remote sensing

Article
An Improved Adaptive Kalman Filter for a Single Frequency
GNSS/MEMS-IMU/Odometer Integrated Navigation Module
Peihui Yan , Jinguang Jiang * , Fangning Zhang, Dongpeng Xie, Jiaji Wu, Chao Zhang, Yanan Tang
and Jingnan Liu

GNSS Research Center, Wuhan University, Wuhan 430079, China; [email protected] (P.Y.);
[email protected] (F.Z.); [email protected] (D.X.); [email protected] (J.W.);
[email protected] (C.Z.); [email protected] (Y.T.); [email protected] (J.L.)
* Correspondence: [email protected]

Abstract: Aiming at the GNSS receiver vulnerability in challenging urban environments and low
power consumption of integrated navigation systems, an improved robust adaptive Kalman filter
(IRAKF) algorithm with real-time performance and low computation complexity for single-frequency
GNSS/MEMS-IMU/odometer integrated navigation module is proposed. The algorithm obtains
the scale factor by the prediction residual, and uses it to adjust the artificially set covariance matrix
of the observation vector under different GNSS solution states, so that the covariance matrix of
the observation vector changes continuously with the complex scene. Then, the adaptive factor is
calculated by the Mahalanobis distance to inflate the state prediction covariance matrix. In addition,
the one-step prediction Kalman filter is introduced to reduce the computational complexity of the





 algorithm. The performance of the algorithm is verified by vehicle experiments in the challenging
urban environments. Experiments show that the algorithm can effectively weaken the effects of
Citation: Yan, P.; Jiang, J.; Zhang, F.;
Xie, D.; Wu, J.; Zhang, C.; Tang, Y.;
abnormal model deviations and outliers in the measurements and improve the positioning accuracy
Liu, J. An Improved Adaptive of real-time integrated navigation. It can meet the requirements of low power consumption real-time
Kalman Filter for a Single Frequency vehicle navigation applications in the complex urban environment.
GNSS/MEMS-IMU/Odometer
Integrated Navigation Module. Keywords: integrated navigation; robust adaptive Kalman filter; low computation complexity; real-
Remote Sens. 2021, 13, 4317. https:// time performance; single-frequency GNSS/MEMS-IMU/odometer module; low power consumption;
doi.org/10.3390/rs13214317 challenging urban environment

Academic Editors: Juliette Marais,

Li-Ta Hsu and Yanlei Gu

1. Introduction
Received: 6 September 2021
Accepted: 23 October 2021
Global navigation satellite system (GNSS) has high positioning accuracy and no cu-
Published: 27 October 2021
mulative error. However, the GNSS receiver needs at least four satellites to be able to
locate [1]. In the environment of avenues, viaducts, urban canyons, and tunnels, the satel-
Publisher’s Note: MDPI stays neutral
lite signal is vulnerable to occlusion and multipath effects [2]. The GNSS receiver with the
with regard to jurisdictional claims in
real-time kinematic (RTK) algorithm is prone to losing lock in the complex environment. It
published maps and institutional affil- takes several seconds to fix the integer ambiguity when reentering the open environment,
iations. which makes the continuity and availability of GNSS positioning not guaranteed. Inertial
navigation has high short-term positioning accuracy and is not restricted by the environ-
ment. However, the INS positioning accuracy will diverge exponentially with time due
to the errors of device noise, zero offset, scale factor, axis deviation, and nonlinearity [3].
Copyright: © 2021 by the authors.
Taking advantage of each other, combining GNSS and INS can provide more accurate [4],
Licensee MDPI, Basel, Switzerland.
continuous and reliable navigation information (including position, speed, and attitude).
This article is an open access article
In other words, through the information fusion of multi-source heterogeneous sensors,
distributed under the terms and complementary advantages can be achieved.
conditions of the Creative Commons Kalman filter is one of the most commonly used optimal estimation methods in inte-
Attribution (CC BY) license (https:// grated navigation [5,6]. However, its performance will be degraded by GNSS measurement
creativecommons.org/licenses/by/ outliers or non-Gaussian distribution noise [7]. In response to this problem, various filter-
4.0/). ing algorithms have been proposed to control their impact on the accuracy of integrated

Remote Sens. 2021, 13, 4317. https://doi.org/10.3390/rs13214317 https://www.mdpi.com/journal/remotesensing

Remote Sens. 2021, 13, 4317 2 of 28

navigation positioning, such as robust Kalman filter, H filter [8,9], DIA methods [10], robust
interval Kalman filter [11], adaptive robust filter based on Mahalanobis distance [12–14],
adaptive Kalman filter [15–17], and so forth. Robust filtering can control the influence of
measurement outliers on motion state estimation [18], but it cannot effectively manage the
effect of model deviation. H filter can solve the interference of the stochastic model [19],
but it cannot do anything about GNSS outliers. In addition, the calculation amount of
this method is enormous. DIA methods are detection, identification, and adaptation. This
method can eliminate the influence of outliers on integrated navigation, but it is difficult
to identify outliers when the measurement is unreliable [20]. A robust interval Kalman
filter combines robust estimation with an interval Kalman filter. It can deal with the model
parameters’ uncertainty and weaken the effects of model errors. The adaptive robust filter
based on Mahalanobis distance defines the judgment index based on Mahalanobis distance.
When the observation value is reliable, the adaptive Kalman filter is executed, otherwise,
the robust estimation method is used for the robust filtering. This method can weaken the
influence of abnormal model deviation and abnormal measurement. Adaptive Kalman
filter can control the influence of model error [21,22]. However, adaptive Kalman filter
based on prediction residual needs reliable observation information, and its performance
is significantly affected by outliers. Different from the adaptive filtering of prediction
residual strategy, Yang et al. proposed an adaptive robust filter technology combining
robust estimation and adaptive factors [23,24]. This technology can effectively suppress the
influence of observation anomaly and carrier state disturbance anomaly on the parameter
estimation of dynamic systems, and has been successfully applied in many fields [24–26].
Currently, deep neural networks have been widely used and are very effective [27–30]. De-
spite having a good performance, these methods cannot be used when power consumption
is a primary concern.
The above methods solve the problem of poor integrated navigation positioning
accuracy caused by abnormal GNSS observations and model deviations under dynamic
conditions. However, their effect on single-frequency GNSS that is vulnerable and does
not output the covariance of the observation vector is limited. (GNSS vulnerability means
that GNSS signals easily interfere in a complex urban environment. When it is suddenly
blocked or blocked in a short time, the positioning error will increase instantaneously,
and it is difficult to converge from the RTK float solution to a fixed solution.) Besides,
these methods cannot operate on low-cost and low power consumption embedded chips.
A real-time low computation complexity IRAKF algorithm is proposed to solve these
problems. The algorithm obtains the scale factor by the single epoch prediction residual
estimation method and adjusts the artificially set observation vector covariance matrix
under different GNSS solution states. Subsequently, the Mahalanobis distance based on
the prediction residual is used to detect disturbance anomalies and obtain the adaptive
factor to inflate the state prediction covariance matrix. In addition, the one-step prediction
Kalman filter is combined to reduce the amount of calculation. The combination of these
three methods realizes low power consumption chip-level single-frequency GNSS/MEMS-
IMU/odometer robust adaptive integrated navigation. The algorithm is verified on a small
volume embedded module, and a real-time vehicle experiment is carried out in Wuhan,
China. This paper focuses on improving positioning accuracy and reducing the power
consumption of low-cost integrated navigation and positioning modules in complex scenes
to be widely used in vehicle navigation. The main contributions of this paper are as follows:
First, an improved robust adaptive algorithm is proposed for single-frequency GNSS that is
vulnerable and does not output the covariance of the observation vector; Second, a one-step
predictive Kalman filter is adopted to reduce power consumption.
The paper is organized as follows: Section 2 introduces the conventional robust
adaptive extended Kalman filter algorithm. Next, the improved robust adaptive Kalman
filter with real-time performance and low computation complexity for the single-frequency
GNSS/MEMS-IMU/odometer integrated navigation is presented in Section 3. In Section 4,
the experimental route and hardware platform are described. Next, the corresponding
Remote Sens. 2021, 13, 4317 3 of 28

results are presented and analyzed in Section 5. Finally, the discussions about the results
and the conclusions are given in Sections 6 and 7, respectively.

2. Conventional Robust Adaptive Extended Kalman Filter

The adaptive filter has been widely applied and investigated in the field of dynamic
navigation and positioning. In order to demonstrate the performance of the real-time low
computation complexity IRAKF algorithm more clearly, the conventional robust adaptive
Kalman filter (CRAKF) is first briefly introduced. The CRAKF mainly adjusts the observa-
tion vector covariance matrix and the state prediction vector covariance matrix [15,25,26].
Take the calculation window as m, and the predicted residual is used to obtain the
estimation value of the observation vector covariance matrix:

1 m
m i∑
T
R̂k = Vk−i V k−i − Hk Pk,k−1 Hk T , (1)
=0

where Vk−i is the predicted residual vector, Hk denotes the design matrix, Pk,k−1 represents
the state prediction covariance matrix. The prediction residual is defined by:

Vk = Zk − Hk Xk,k−1 , (2)

where Z represent the observation vector. The prediction residual is used to construct a
two-segment adaptive factor, which can be written as
  
1 Vk  ≤ c
e 
αk =   , (3)
 c Vk  > c
e 
|Ve k |
q
where Ve k = kVk k/ trace(P ), P is the prediction residual covariance matrix,
Vk Vk
PVk = Hk Pk,k−1 Hk + Rk . kk and trace() represents the least-squares norm and the trace of
the matrix, respectively. c is the threshold by the experience, and 1 ≤ c ≤ 2.5. The adaptive
factor is used to inflate the state prediction covariance matrix.
The estimated value of the observation vector covariance matrix R̂k and adaptive factor
αk are substituted into extended Kalman filter [6], and the gain matrix, state estimation
vector, and state estimation vector covariance matrix of the CRAKF can be obtained
as follows:
1 1
K̂k = Pk,k−1 Hk T ( Hk Pk,k−1 Hk T + R̂k ) (4)
αk αk
X̂k = Xk,k−1 + K̂k (Zk − Hk Xk,k−1 ) (5)
Pk = (I − K̂k )Pk,k−1 , (6)
where X is the error state vector; K is the Kalman filter gain matrix, which determines the
weights of measurement information when the error state is updated; I denotes the identity
matrix. All the subscripts indicate the state change. For example, Xk,k−1 means the error
state vector from epoch k − 1 to epoch k.

3. The Proposed Improved Robust Adaptive Kalman Filter

3.1. System Overview
Due to power consumption, the GNSS receiver adopts a simplified RTK algorithm,
which leads to poor positioning accuracy and vulnerability. It cannot be compared with
other similar receivers on the market in terms of positioning accuracy. Besides, because the
receiver signal is susceptible to interference from a complex environment, the observation
vector covariance matrix is prone to error values, so the receiver does not output the
covariance matrix of the observation vector. As a result, the positioning accuracy of the
integrated navigation system with the GNSS receiver is poor. Based on the disadvantage
of the GNSS receiver, this paper proposes an IRAKF algorithm for a single frequency
Remote Sens. 2021, 13, 4317 4 of 28

GNSS/MEMS-IMU/odometer low power consumption real-time loosely coupled inte-

grated navigation. The algorithm combines the robust adaptive algorithm and the one-step
prediction Kalman filter, which improves the robustness of the integrated navigation sys-
tem and dramatically reduces the algorithm’s computational complexity. It meets the
requirements of real-time and low-power consumption of integrated navigation modules
in challenging urban environments. The flow chart of the proposed algorithm is shown in
Figure 1. The red dotted rectangle indicates the calculation of the scale factor and adaptive
factor, and the blue dotted rectangle indicates the one-step prediction Kalman filter. The
specific implementation process is as follows: firstly, the prediction residual is calculated
by GNSS observation information, and the estimated observation vector covariance matrix
is obtained according to the prediction residual. Then it is compared with the artificially
set covariance matrix of the observation vector. If the estimated covariance matrix of the
observation vector is larger than the set covariance matrix of the observation vector, the
outlier GNSS exists at this time. Subsequently, the scale factor adjusts the set covariance
matrix of the observation vector to reduce the weight of GNSS in the Kalman filter. On
the contrary, the current GNSS observation quality is consistent with the estimated value,
and there is no need to adjust the R matrix. With the outlying observations, the estimated
value of the observation vector covariance matrix calculated by the scale factor becomes
unreliable. The Mahalanobis distance based on the prediction residual is introduced to
detect disturbance anomalies and obtain the adaptive factor to inflate the state prediction
covariance matrix.
In addition, a particular case is also considered. GNSS RTK is susceptible to interfer-
ence and turns into a pseudo-fixed solution when the vehicle enters the tunnel, which will
disturb the system state after it participates in the integrated navigation, resulting in poor
positioning accuracy in the tunnel. The PDOP value and the number of satellites tracked
reflect the influence of weather and environment on GNSS positioning accuracy. With
these two parameters, the pseudo-fixed solution of GNSS can be identified and eliminated.
Finally, to reduce the computation complexity of the integrated navigation algorithm, the
one-step prediction Kalman filter is adopted. In the one-step prediction Kalman filter, the
Kalman prediction process only calculates the state transition matrix, and the prediction
state covariance matrix is executed in the Kalman filter update process. In this way, the
computing load of the P-matrix is reduced. The clock frequency of MCU in the module
is reduced by decreasing the algorithm’s calculation amount to realize the low-power
consumption of the module. The relationship between power consumption and the clock
frequency is as follows [31]:

Ptotal = PMCU + PGNSS + PI MU + Pothers (7)

PMCU = Pclock f requency + Pperipherals + PIO port + Pothers . (8)

Remote Sens. 2021, 13, 4317 5 of 28
Remote Sens. 2021, 13, x FOR PEER REVIEW 6 of 30

Kalman filter prediction

NO

YES

Calculate scaling factor

NO

YES

Calculate adaptive factor

YES

NO

Kalman filter update

Figure 1. The algorithm flow chart of improved robust adaptive Kalman filter algorithm applied to GNSS/INS/odometer
Figure 1. The algorithm flow chart of improved robust adaptive Kalman filter algorithm applied to GNSS/INS/odometer
integrated navigation module.
integrated navigation module.
Remote Sens. 2021, 13, 4317 6 of 28

From Equation (7), the power consumption of the single-frequency GNSS/MEMS-

IMU/odometer integrated navigation module mainly includes MCU, GNSS, IMU, and
other sensors. Since the power consumption of GNSS, IMU, and other sensors are fixed
values and cannot be changed. So, only MCU can be used to reduce power consumption.
However, as shown in Equation (8), MCU power consumption comprises clock frequency,
internal peripherals, and IO ports. The peripherals used inside the MCU include SPI, serial,
encoder, timer, and so forth. These peripherals and IO ports are connected to external
sensors and cannot be reduced. Therefore, the module can only reduce the system power
consumption by lowering the clock frequency of the MCU. In this paper, the computational
complexity is reduced by decreasing the algorithm complexity, which in turn can reduce
the clock frequency of the MCU. In the experiment, the operating frequencies of MCU are
400 MHz, 200 MHz, 100 MHz, and 50 MHz, respectively. It is concluded that the lower the
frequency, the lower the power consumption of the integrated navigation system. Through
the real-time vehicle navigation experiment, the clock frequency of the MCU is reduced to
50 MHz in this research, and the average power consumption of the module is 143 mW.
In addition, during the initialization of the integrated navigation, we have adopted
a dynamic alignment method to calibrate the INS, that is, using satellites for dynamic
alignment. In dynamic alignment, the position and speed information of IMU can be
obtained by using GNSS information through lever arm vector correction. Since the
vehicle is on a horizontal plane, it can be considered that the pitch and roll are zero, and the
heading can be solved according to the following formula through GNSS speed information
ψ = tan−1 (v E /v N ). At this time, the attitude information of IMU is still rough. In the
subsequent process, the attitude information uses the position information provided by
GNSS to converge through the Kalman filter to realize fine alignment.

3.2. Improvement Algorithm

This research uses the IRAKF with real-time performance and low computation
complexity to implement the single-frequency GNSS/MEMS-IMU/odometer integrated
navigation [32]. The IRAKF fuses the single-frequency GNSS, MEMS-IMU, and an odome-
ter to achieve optimal integrated navigation system state estimates. In order to show the
IRAKF algorithm, a scaling factor based on the GNSS solution state, an adaptive factor
based on the Mahalanobis distance, and a one-step prediction Kalman filter model are
introduced in this section.

3.2.1. Scaling Factor Based on GNSS Solution State

Single-frequency GNSS signal is vulnerable to serious interference in the complex envi-
ronment such as boulevards, tunnels, urban canyons, and viaducts. The positioning result
of single-frequency GNSS is easy to change from the RTK fixed solution to a float solution
or a differential solution. Since the commercial receiver does not output the observation
vector covariance matrix, the matrix Rk can be set according to the GNSS solution.
2

 diag(σ f ix )
 ; Fixed solution state
2
diag(σ f loat ) ; Float solution state
Rk = (9)
 diag(σ2

) ; di f f erential solution state
di f f erential

kσ2f ix k < kσ2f loat k < kσdi

2
f f erential k, (10)

where diag() function converts a vector into a diagonal matrix, and the elements not on
the diagonal are all zero. σ2f ix , σ2f loat and σdi
2
f f erential are the position variances of fixed,
float, and differential solutions of the single-frequency RTK, respectively. Considering
the different hardware characteristics of other receivers, this value needs to be obtained
through many experimental data statistical calculations.
The Rk set by prior knowledge is discontinuous, but it changes continuously in the
challenging urban environment. It is inconsistent with the value of GNSS receiver output
Remote Sens. 2021, 13, 4317 7 of 28

in the complex urban environment. In addition, when the positioning error of the GNSS
RTK changes from large to small, the vehicle may not move from an obscured environment
to an open environment. In this case, Rk given by prior knowledge is unreliable. If this
value is used as the covariance matrix of the observation vector, integrated navigation
will be significantly affected. Therefore, a robust adaptive algorithm based on the GNSS
solution state is designed.
In vehicle navigation applications with medium and low dynamics, the state pre-
diction vector Xk and the state prediction vector’s covariance matrix Pk,k−1 are generally
considered reliable within one second. Therefore, the reliability of the measurement
information Zk can be inferred from the predicted residual Vk = Zk − Hk Xk,k−1 . That
is, if Vk is small, Zk and Xk,k−1 are more consistent, and Zk is reliable; otherwise, Zk is
considered unreliable.
According to the propagation law of covariance matrix, the theoretical covariance of
prediction residual is as follows:

PVk = Hk Pk,k−1 HTk + Rk . (11)

However, there are many methods for estimating covariance of prediction residuals,
including the Sage windowing estimation method, single epoch prediction residual es-
timation method, recursive estimation method, and so on [25,26]. In these methods, the
Sage windowing estimation method sometimes cannot reflect the size of the model error of
the current epoch. In contrast, the recursive estimation method uses the estimation model
based on the information of the previous epoch. It cannot also reflect the state of the current
epoch in time. Therefore, the single-epoch prediction residual estimation method with a
small amount of calculation is selected according to the hardware conditions to estimate
the covariance of the prediction residuals. The covariance matrix of Vk can be obtained
T
by E(Vk Vk )
T T
P̂Vk = E(Vk Vk ) = Vk Vk . (12)
Since the error of the dynamic model is small within one second, the error of the
observation vector can be seen by comparing the theoretical covariance and estimated
covariance of the prediction residual. In the robust theory, if the observation vector error is
large, the covariance matrix of the observation vector Rk should be αk Rk (αk > 1), which
means that the weight of the observation vector is weakened. However, we hope that the
theoretical and estimated values of the predicted residual covariance after applying the
scaling factor are the same. So, suppose that the two formulas of (11) and (12) are equal,
that is:
T
Vk Vk = Hk Pk,k−1 HTk + Rk . (13)
According to the above formula, the estimated value of the observation vector covari-
ance matrix is as follows:
T
R̂k = Vk Vk − Hk Pk,k−1 HTk . (14)
After R̂k and Rk are traced, the scaling factor is

αk = trace(R̂k )/trace(Rk ). (15)

Thus, the scaling factor for the GNSS RTK fixed solution can be expressed as

 1 trace(R̂k ) ≤ trace(Rk, f ix )
αk = trace(R̂k ) , (16)
 trace(R ) trace(R̂k ) > trace(Rk, f ix )
k, f ix

where Rk, f ix = diag(σ2f ix ) is the observation vector covariance matrix of the fixed solution.
Remote Sens. 2021, 13, 4317 8 of 28

The scaling factor for the GNSS RTK float solution can be expressed as

 1 trace(R̂k ) ≤ trace(Rk, f loat )
αk = trace(R̂k ) , (17)
 trace(R )
trace(R̂k ) > trace(Rk, f loat )
k, f loat

where Rk, f loat = diag(σ2f loat ) is the observation vector covariance matrix of the float solution.
The scaling factor for the GNSS differential solution can be expressed as

 1 trace(R̂k ) ≤ trace(Rk,di f f erential )
αk = trace(R̂k ) , (18)
 trace(R )
trace(R̂k ) > trace(Rk,di f f erential )
k,di f f erential

2
where Rk,di f f erential = diag(σdi f f erential ) is the observation vector covariance matrix of the
differential solution.
Comparing trace(R̂k ) and trace(Rk, ), if the former is greater than the latter, then
αk Rk > Rk, . Obviously, the influence of the observation vector on the state estimation is
weakened. However, suppose the former is less than the latter. In that case, it means that
the observation vector error is small, and there is no need to adjust the covariance matrix
of the observation vector by the scale factor. At this time, αk is equal to 1. Therefore, the
estimated observation vector covariance can be written as follows:

R̂k = αRk . (19)

Additionally, there is another situation that quickly disturbs the state of the integrated
navigation system in a complex environment, leading to increased positioning errors. That
is, the GNSS RTK state is a fixed solution, but the positioning accuracy is poor. In this
case, the observation covariance matrix Rk is minimal. It means that the weight of the
observation vector is strengthened, and the Kalman filter trusts the observation information
more. However, after the observation information with significant errors participates in
the integrated navigation, it is easy to disturb the system state, resulting in extensive
positioning errors for an extended period. Therefore, before the measurement update of
the Kalman filter, the horizontal dilution of precision (HDOP) and the number of tracked
satellites of the receiver are combined to eliminate the wrong GNSS RTK fixed solution.
The judgment conditions are as follows:

(HDOP > C1 ||Nsat < C2 )&& state == Fix, (20)

where HDOP is the horizontal dilution of precision at the current time; Nsat is the number
of satellites tracked by the receiver; C1 and C2 are the threshold of HDOP and the number
of satellites, respectively.

3.2.2. Adaptive Factor Based on the Mahalanobis Distance

The scaling factor obtained by the prediction residual single-epoch method in the
above section may not optimize the state estimation value of the Kalman filter. In the
CRAKF, based on the estimation value of the observation vector covariance matrix, the
adaptive factor is used to adjust the state prediction covariance matrix. However, with the
outlying observations, the prediction residual becomes unreliable, and the adaptive factor
may degrade the performance of the CRAKF. Therefore, it is necessary to find a reliable
way to obtain the adaptive factor. Based on the Mahalanobis distance of the prediction,
residual is adopted to detect the disturbance anomaly and uses it to obtain the adaptive
factor [7,12,14]. The specific implementation process is as follows.
Consider the state-space model of a stochastic system,

Xk = Φk,k−1 Xk−1 + Wk−1
, (21)
Zk = Hk Xk + Vk
Remote Sens. 2021, 13, 4317 9 of 28

where Φk,k−1 is the state transition matrix; Wk−1 and Vk are the system noise vector and
the observation noise vector, respectively.
For the above formula, the observation vector should be Gaussian distributed with
the mean Zk and covariance PZk , and the probability density function ρ(Zk ) can be written
as [7,12]:
 
T
exp − 21 (Zk − Zk ) (PZk )−1 (Zk − Zk )
ρ(Zk ) = N (Zk ; Zk , PZk ) = r   , (22)
m
(2π ) PZk 


where m represents the dimension of an observation vector. The probability density func-
tion will no longer hold if the outlying observations or observation noise is contaminated.
Therefore, it can be judged whether there is a gross error in the observation according to
this characteristic.
The test statistic is constructed according to formula (22) to detect modeling errors.
Consequently, the square of the Mahalanobis distance from observation Zk to mean Zk is
used as the test statistic.
T
Mk2 = (Zk − Zk ) (PZk )−1 (Zk − Zk ), (23)
q
T
where Mk = (Zk − Zk ) (PZk )−1 (Zk − Zk ) is the Mahalanobis distance. The prediction
residual between the observation vector and the predicted state vector is just the Maha-
lanobis distance. Thus, the Mahalanobis distance of the prediction residual can be written
as follows:
γ k = V k T ( PV k ) − 1 V k . (24)

The prediction residual Vk follows a standard normal distribution, so the Mahalanobis

distance between the observation vector and predicted state vector should be Chi-square
distributed with the degree of freedom m.
n o
P γk > χ2α (m) = α, (25)

where α is the significance level, it is a small value, and 0.01 is used in this contribution; m
is the dimension of the prediction residual; P{} denotes the probability of γk being more
significant than χ2α , and the probability value is α. However, it is a small probability event
that the calculated value of γk is more remarkable than χ2α (m). If this small probability
event holds, it indicates some outlying measurements in the observation data. Otherwise,
there are no outliers. According to this feature, the adaptive factor can be solved. An
adaptive factor β k is introduced to inflate the state predicted vector covariance matrix,

P̂k,k−1 = β k Pk,k−1 . (26)

According to the Formula (25), the following equation can be satisfied

γk = Vk T (PVk )−1 Vk = χ2α . (27)

Change the above formula into the following form,

f ( β k ) = Vk T (PVk )−1 Vk − χ2α = 0. (28)

The adaptive factor β k is solved according to Gauss–Newton iterative method [14],

f ( β k (i − 1))
β k ( i ) = β k ( i − 1) − , (29)
f 0 ( β k (i − 1))
Remote Sens. 2021, 13, 4317 10 of 28

where i represents the i-th iteration. In Equation (29), the derivative of an inverse matrix is
as follows [33]:

f 0 (βk ) = (Vk T (PVk )−1 Vk − χ2α )0 = (Vk T (PVk )−1 Vk )0

h i
= Vk T ((PVk )−1 )0 Vk = Vk T −(PVk )−1 Hk Pk,k−1 Hk (PVk )−1 Vk . (30)
−1 −1
= − V k T ( PV k ) Hk Pk,k−1 Hk (PVk ) Vk

Substituting (28) and (30) into (29) gives

Vk T (PVk )−1 Vk − χ2α

β k ( i ) = β k ( i − 1) + . (31)
Vk T (PVk )−1 Hk Pk,k−1 Hk (PVk )−1 Vk

The initial value β k (0) = 1.

3.2.3. One-Step Prediction Kalman Filter Model

In traditional integrated navigation Kalman filter, the state prediction vector covari-
ance matrix P and state prediction vector X are performed together, significantly increasing
the computation load and unsuitable for systems with real-time and low power consump-
tion requirements [5,6,34,35]. In our research, the Kalman filter is improved, and the state
prediction vector covariance matrix P is performed together with the measurement update,
which dramatically reduces the computation load. The improved algorithm is as follows:
Prediction:
Xk,k−1 = Φk,k−1 Xk , (32)
Φ̂k,k−1 = Φk,k−1 Φ̂k−1,k−2 (33)
Update:

T 1 T
Pk,k−1 = Φ̂k,k−1 Pk Φ̂k,k −1 + Qk−1 + 2 ( Φ̂k,k−1 Qk−1 + Qk−1 Φ̂k,k−1 )(n − 1) ∆t, (34)
−1
Kk = Pk,k−1 HTk (Hk Pk,k−1 HTk + Rk ) , (35)
Xk = Xk,k−1 + Kk (Zk − Hk Xk,k−1 ), (36)
Pk = (I − Kk Hk )Pk,k−1 , (37)
where Φ represents the state transition matrix; Φ̂ denotes the calculated state transition
matrix, which is obtained by multiplying the current Φ by the previous Φ̂. For example,
Φk,k−1 means the state transition matrix from epoch k − 1 to epoch k.
In order to put the state predicted covariance matrix P into the measurement update
process, it is necessary to use the state transition matrix Φ to achieve this goal. Therefore,
theoretical reasoning is needed to prove the feasibility of the formula.
According to the theoretical derivation, Q and P of the k + 3 epoch in the extended
Kalman filter are as follows [36]:

1
Qk +2 = (Φ Q + Qk+2 ΦkT+3,k+2 )∆t (38)
2 k+3,k+2 k+2

Pk+3,k+2 = Φk+3,k+2 Pk+2 ΦkT+3,k+2 + Qk+2

= (Φk+3,k+2 Φk+2,k+1 Φk+1,k )Pk (Φk+3,k+2 Φk+2,k+1 Φk+1,k )T + Qk+2 + , (39)
Φk+3,k+2 Qk+1 ΦkT+3,k+2 + (Φk+3,k+2 Φk+2,k+1 )Qk (Φk+3,k+2 Φk+2,k+1 )T
where Q is the constant spectral density matrix; ∆t is the sampling interval of the
navigation system.
Remote Sens. 2021, 13, 4317 11 of 28

Expand the Formula (39):

Pk+3,k+2 = Φk+3,k+2 Pk+2 ΦkT+3,k+2 + Qk+2

= (Φk+3,k+2 Φk+2,k+1 Φk+1,k )Pk (Φk+3,k+2 Φk+2,k+1 Φk+1,k )T + Qk+2 +
T , (40)
2 [(Φk +3,k+2 Φk+2,k+1 )Qk +1 Φk +3,k +2 + Φk +3,k +2 Qk +1 ( Φk +3,k +2 Φk +2,k +1 ) ] ∆t +
1 T

1 T T
2 [(Φk +3,k+2 Φk+2,k+1 Φk +1,k )Qk ( Φk +3,k +2 Φk +2,k+1 ) + ( Φk+3,k+2 Φk+2,k +1 )Qk ( Φk +3,k +2 Φk +2,k +1 Φk+1,k ) ] ∆t

where Qk ≈ Qk+1 ≈ Qk+2 ≈ · · · ≈ Qk+n is the system noise covariance matrix.

By observing Equation (40), it can be found that the first and second terms on the right
side of the equation are easier to calculate, while the third and fourth terms are relatively
complicated. If we want to calculate the third and fourth terms, we must save the previous
value of Φ. However, as time grows, the number of Φ that also needs to be saved increases,
which seriously wastes hardware memory resources. Therefore, in order to facilitate the
calculation, we should construct the third and fourth terms into the form of the first term,
as shown below:
1
s3 = [(Φk+3,k+2 Φk+2,k+1 Φk+1,k )Qk+1 + Qk+1 (Φk+3,k+2 Φk+2,k+1 Φk+1,k )T ]∆t (41)
2
1
[(Φk+3,k+2 Φk+2,k+1 Φk+1,k )Qk + Qk (Φk+3,k+2 Φk+2,k+1 Φk+1,k )T ]∆t.
s4 = (42)
2
Consequently, substituting (41) and (42) for the last two items in (40),
Pk+3,k+2 = Φk+3,k+2 Pk+2 ΦkT+3,k+2 + Qk+2
= (Φk+3,k+2 Φk+2,k+1 Φk+1,k )Pk (Φk+3,k+2 Φk+2,k+1 Φk+1,k )T + Qk+2 +
T . (43)
2 [( Φk +3,k+2 Φk +2,k +1 Φk +1,k )Qk +1 + Qk +1 ( Φk +3,k +2 Φk +2,k +1 Φk +1,k ) ] ∆t +
1
T
2 [( Φk +3,k+2 Φk +2,k +1 Φk +1,k )Qk + Qk ( Φk +3,k +2 Φk +2,k +1 Φk +1,k ) ] ∆t
1

In order to verify the feasibility of replacing the last two items of Formula (40) with
the constructed Formulas (41) and (42), we use Formula (43) to subtract Formula (40). If the
result approaches zero, it indicates that the construction method is feasible. The procedures
(43)–(40) are as follows:
T
2 [( Φk +3,k +2 Φk +2,k+1 Φk+1,k )Qk+1 + Qk+1 ( Φk+3,k+2 Φk+2,k +1 Φk +1,k ) ] ∆t −
1
(43) − (40) =
T
2 [( Φk +3,k +2 Φk +2,k+1 )Qk+1 Φk+3,k+2 + Φk+3,k+2 Qk +1 ( Φk+3,k +2 Φk +2,k+1 ) ] ∆t +
1 T

T
. (44)
2 [( Φk +3,k +2 Φk +2,k+1 Φk+1,k )Qk + Qk ( Φk+3,k+2 Φk+2,k +1 Φk +1,k ) ] ∆t −
1

T T
2 [( Φk +3,k +2 Φk +2,k+1 Φk+1,k )Qk ( Φk+3,k+2 Φk+2,k +1 ) + ( Φk +3,k +2 Φk +2,k+1 )Qk ( Φk+3,k+2 Φk+2,k +1 Φk+1,k ) ] ∆t
1

According to the theoretical formula of the Kalman filter state transition matrix, it
can be known that there is ∆t in the matrix, so it is assumed that Φk ≈ I + Fk ∆t. In the
Kalman filter, ∆t is equal to 0.01s. Putting it into Equation (44), it can be seen that (43)–(40)
are high-order small quantities, and the result approaches zero. The details are as follows:
1
2 ( ∆t + F2 ∆t + F3 ∆t + F2 F3 ∆t )( F1 − F3 )Qk + 2 Qk ( F1 − F3 )( ∆t + F2 ∆t + F3 ∆t + F2 F3 ∆t )−
2 3 3 4 T 1 T 2 3 3 4
(43) − (40) =
T
2 ( ∆t + F1 ∆t + F2 ∆t + F3 ∆t + F2 F1 ∆t + F3 F1 ∆t + F3 F2 ∆t + F3 F2 F1 ∆t )Qk ( F2 + F3 + F3 F2 ∆t ) −
1 2 3 3 3 4 4 4 5
. (45)
5 T
2 ( F2 + F3 + F3 F2 ∆t )Qk ( ∆t + F1 ∆t + F2 ∆t + F3 ∆t + F2 F1 ∆t + F3 F1 ∆t + F3 F2 ∆t + F3 F2 F1 ∆t )
1 2 3 3 3 4 4 4

Therefore, the third and fourth terms of Formula (40) can be replaced by (41) and (42).
The P matrix of the epoch k + 3 can be rewritten as follows:

Pk+3,k = (Φk+3,k+2 Φk+2,k+1 Φk+1,k )Pk (Φk+3,k+2 Φk+2,k+1 Φk+1,k )T + Qk+2 +

, (46)
2 · 21 [(Φk+3,k+2 Φk+2,k+1 Φk+1,k )Qk + Qk (Φk+3,k+2 Φk+2,k+1 Φk+1,k )T ]∆t

where Pk+3,k is the state predicted covariance matrix from epoch k to epoch k + 3. It can be
seen from the above formula that the calculation of the state prediction covariance matrix
Remote Sens. 2021, 13, 4317 12 of 28

from epoch k to k + 3 can be realized in one step. Then, the state transition matrix from
epoch k to epoch k + 3 can be written as:

Φk+3,k = Φk+3,k+2 Φk+2,k+1 Φk+1,k . (47)

Similarly, the matrix of P and Φ are extended to n epochs, namely from k epochs to
k + n epochs, as follows:

Pk+n,k = Φk+n,k Pk ΦkT+n,k + 12 (Φk+n,k+n−1 Qk+n−1 + Qk+n−1 ΦkT+n,k+n−1 )∆t

(48)
+ 12 (Φk+n,k Qk+n−1 + Qk+n−1 ΦkT+n,k )(n − 1)∆t

Φk+n,k = Φk+n,k+n−1 Φk+n−1,k+n−2 · · · Φk+2,k+1 Φk+1,k (49)

Therefore, the P can be executed in the measurement update process, and in the
Kalman filter prediction process, a new Φ can be obtained by multiplying each previous
Φ. The state transition matrix Φ and the predicted state vector covariance matrix P of the
code running in the chip can be written in the following form:

Φ̂k,k−1 = Φk,k−1 Φ̂k−1,k−2 (50)

T 1 T
Pk,k−1 = Φ̂k,k−1 Pk Φ̂k,k −1 + Qk−1 + 2 ( Φ̂k,k−1 Qk−1 + Qk−1 Φ̂k,k−1 )(n − 1) ∆t. (51)

In Kalman filter prediction, the amount of calculation of the Φ matrix is much less
than that of P matrix, which significantly reduces the computational complexity.

4. Experimental Route and Hardware Platform

This study selected a section of complex scene road in Wuhan, Hubei Province as the
test route, as shown in Figure 2. The route includes scenes such as Boulevard 1, Tunnel 1,
Boulevard 2, and Tunnel 2. They constitute a continuous GNSS high occlusion scene,
and there are continuous s-turns and up and down ramps, and multipath interference of
lake reflected signals. The vehicle-mounted experimental platform is shown in Figure 3.
The testing equipment and reference equipment are installed on the vehicle’s roof, and
the odometer is attached to the wheel. The experimental equipment is a multi-sensor
fusion navigation and positioning module, which incorporates single-frequency GNSS,
MEMS-IMU, magnetometer, barometer, temperature and humidity sensor, NB-IoT, and
e-sim to form an integrated navigation and positioning system. The reference equipment
is NovAtel SPAN CPT6, a navigation-level device with dual-frequency GNSS and fiber
optic gyroscope tight integrated navigation technology. Table 1 shows the performance
indicators of the two devices.
Remote Sens. 2021, 13, x FOR PEER REVIEW 14 of 30
Remote Sens. 2021, 13, x FOR PEER REVIEW 14 of 30

Remote Sens. 2021, 13, 4317 13 of 28

gyroscope tight integrated navigation technology. Table 1 shows the performance indica-
gyroscope
tors tight
of the two integrated navigation technology. Table 1 shows the performance indica-
devices.
tors of the two devices.

Figure 2. Real-time performance verification experiment trajectory and scenes of vehicle navigation.
Figure2.2.Real-time
Figure Real-timeperformance
performanceverification
verificationexperiment
experimenttrajectory
trajectoryand
andscenes
scenesofofvehicle
vehiclenavigation.
navigation.

Figure 3. Vehicle experimental platform.

Figure3.3.Vehicle
Figure Vehicleexperimental
experimentalplatform.
platform.
ote Sens. 2021, 13, x FOR PEER REVIEW 15 of 30
Remote Sens. 2021, 13, 4317 14 of 28

Table 1. Performance specifications of the IMU in the experiment.

Table 1. Performance specifications of the IMU in the experiment.
Module SPAN CPT6
Module SPAN CPT6
Bias (deg/h) 10 0.027
Gyro Bias (deg/h) 10 0.027
ARW (deg/sqrt(h))
Gyro 0.27 0.0667
ARW (deg/sqrt(h)) 0.27 0.0667
Bias (mGal) Bias (mGal) 1800 1800
50 50
Accelerometer
Accelerometer
VRW (m/s/sqrt(h))
VRW (m/s/sqrt(h))0.042 0.042 0.03 0.03

5. Results 5. Results
5.1. Operands5.1.
Analysis and Analysis
Operands Comparison
and Comparison
In order to verify the effectiveness
In order to verify theofeffectiveness
the simplified ofone-step prediction
the simplified Kalman
one-step filter
prediction Kalman
in the improved
filterrobust adaptive algorithm,
in the improved the calculation
robust adaptive algorithm, amount and positioning
the calculation amount per- and positioning
formance areperformance
analyzed and arecompared
analyzed with the extended
and compared with Kalman filter. Kalman filter.
the extended
In the IRAKFInalgorithm, the simplified one-step prediction
the IRAKF algorithm, the simplified one-step prediction Kalman filter is used.
Kalman Thatfilter is used.
is, the prediction
That is,state
thecovariance
prediction matrix is only calculated
state covariance matrix isonceonlyin the measurement
calculated once in the up- measurement
date cycle. Consequently, the algorithm the
update cycle. Consequently, dramatically
algorithm reduces calculation
dramatically reducesand can run and
calculation in can run in
real-time in real-time
a low-costinMCU with aMCU
a low-cost low operating
with a lowfrequency.
operatingItfrequency.
ensures that the system
It ensures that the system
works with lowworks power
with consumption. In order to In
low power consumption. evaluate
order tothe computational
evaluate complexity
the computational complexity of
of the one-step
the prediction Kalman filter
one-step prediction Kalmanandfilter
extended Kalman Kalman
and extended filter in filter
the IRAKF algo- algorithm,
in the IRAKF
the operation
rithm, the operation number number of multiplication
of multiplication (M), addition
(M), addition and subtraction
and subtraction (A&S),(A&S),
divi-division (D),
square
sion (D), square rootroot (SR),
(SR), andand trigonometric
trigonometric (T) (T) function
function are are counted.
counted. Figure
Figure 4 shows
4 shows the the number
number of eachof each operation
operation in in
thethe
twotwo Kalmanfilters.
Kalman filters.InInthe
theextended
extendedKalman
Kalman filter,
filter, the
the operands of
operands of M, M, A&S,
A&S,D, D,SR,
SR,and
andTTwithin
withinone onesecond
second areare215,959, 203,912,
215,959, 203,912,247,247,
132,132,
andand35, respectively.
35, respectively. In the one-step prediction Kalman filter, the operands of M, A&S, D, SR,T are 52,747,
In the one-step prediction Kalman filter, the operands of M, A&S, D, SR, and
39,216,
and T are 52,747, 143, 132,
39,216, 143,and
132,35,
andrespectively. Compared
35, respectively. with the
Compared extended
with KalmanKal-
the extended filter algorithm,
the one-step prediction Kalman filter reduces the number
man filter algorithm, the one-step prediction Kalman filter reduces the number of opera- of operations of M by 75.6%,
tions of M byA&S by 80.8%,
75.6%, A&S byand D byand
80.8%, 42.1%.
D by 42.1%.

Figure 4.number
Figure 4. The operand The operand
of the number
operationoftype
the operation type in
in the extended the extended
Kalman Kalman
filter and filter and
the one-step the one-step
prediction Kalman filter.
prediction Kalman filter.
Remote
Remote Sens.
Sens. 2021,
2021, 13,
13, x4317
FOR PEER REVIEW 16
15 of
of 30
28

In order to verify whether the simplified one-step prediction Kalman filter in the
In order to verify whether the simplified one-step prediction Kalman filter in the
IRAKF algorithm will affect the positioning performance of the integrated navigation sys-
IRAKF algorithm will affect the positioning performance of the integrated navigation
tem, the real-time vehicle-mounted experiments are carried out in a challenging urban
system, the real-time vehicle-mounted experiments are carried out in a challenging urban
environment. Meanwhile, the
environment. Meanwhile, thesimplified
simplifiedone-step
one-step prediction
prediction Kalman
Kalman filterfilter is replaced
is replaced with
with
the extended Kalman filter in the algorithms, and the positioning accuracy of of
the extended Kalman filter in the algorithms, and the positioning accuracy thethe two
two is
is compared. The positioning errors of the two in the North, East, and
compared. The positioning errors of the two in the North, East, and Down directions areDown directions
are shown
shown in Figure
in Figure 5. 5.
It It
cancan
bebeseen
seenfrom
fromthe
thefigure
figurethat
thatalthough
although thethe simplified
simplified one-step
one-step
prediction
prediction Kalman filter reduces the number of operations of the algorithm, its
Kalman filter reduces the number of operations of the algorithm, its positioning
positioning
accuracy
accuracy hardly
hardly deteriorates.
deteriorates. The
The overlap
overlap curve
curve in
in Figure
Figure 5a5a is
is expanded
expanded to to the
the separation
separation
curve in Figure 5b to make the figures clearer.
curve in Figure 5b to make the figures clearer.

Figure
Figure 5.
5. The positioning errors
The positioning errorsof
ofthe
theextended
extendedKalman
Kalmanfilter
filter and
and thethe one-step
one-step prediction
prediction Kalman
Kalman filter
filter in North,
in North, East,East,
and
and
DownDown directions:
directions: (a) curve
(a) curve overlap;
overlap; (b) curve
(b) curve separation.
separation.

5.2.
5.2. Performance
Performance Verification
Verification
The
The accuracy
accuracy ofof the
the positioning,
positioning, velocity,
velocity, and attitude
attitude are essential indicators
indicators that
reflect
reflect the
the performance
performance of of the
the integrated
integrated navigation
navigation system.
system. In
Inorder
orderto
toverify
verifythe
theperfor-
perfor-
mance
mance of of the
the IRAKF
IRAKF algorithm
algorithm in in complex
complex scenarios,
scenarios, real-time
real-time vehicle-mounted
vehicle-mounted experi-
experi-
ments
ments are
are carried
carried out.
out. This
Thissection
sectionanalyzes
analyzesthe
thehorizontal
horizontalpositioning,
positioning,forward
forwardvelocity,
velocity,
and heading errors of the single-frequency GNSS/MEMS-IMU/odometer integrated nav-
Remote Sens. 2021, 13, 4317 16 of 28

igation module with the IRAKF algorithm. In addition, the positioning performance of
challenging scenes, namely boulevard and tunnel, is also analyzed and compared.
The horizontal position, forward velocity, and heading errors of the whole test route
estimated by the methods of the CRAKF, Sage filter, extended Kalman filter, and IRAKF are
shown in Figure 6. It can be noted that the horizontal position error of the extended Kalman
filter exhibits large values between 281,989 s (280,000 + 1989) and 282,344 s (280,000 + 2344).
Compared with the CRAKF, Sage filter, extended Kalman filter, the IRAKF algorithm
proposed in this paper has the highest positioning accuracy and the best robustness. But it
is only limited to most road scenes. The main reason is that the GNSS receiver cannot output
the observation vector covariance matrix, and the manually set value cannot accurately
represent the current positioning error. In terms of horizontal velocity error, it can be seen
from Figure 6 that the proposed IRAKF can provide better performance than the other
three methods. From the heading error in Figure 6, it can be seen that the heading error
Remote Sens. 2021, 13, x FOR PEER REVIEW
of IRAKF is smaller than the other three methods in most periods and the error curve 18 of 30
is
smoother. To make the figures clearer, the subfigures of horizontal position error, velocity
error, and heading error are shown in Figure 7, showing the curves in each figure.

Figure 6. Cont.
Remote Sens. 2021, 13, 4317 17 of 28

Figure 6. The real‐time horizontal position error, velocity error,
Figure 6. The real-time horizontal position error, velocity and heading error of the conven‐
error, and heading error of the conventional
Remote Sens. 2021, 13, x FOR PEER REVIEW 19 of 30
tional robust adaptive Kalman filter, Sage filter, extended Kalman filter, and improved robust adap‐
robust adaptive Kalman filter, Sage filter, extended Kalman filter, and improved robust adaptive
tive Kalman filter. 
Kalman filter.

Figure7.
Figure 7. The
Thereal-time
real-timehorizontal
horizontalposition error,
position error,velocity error,
velocity andand
error, heading error
heading subfigures
error of conventional
subfigures robustrobust
of conventional adap-
tive Kalman filter, Sage filter, extended Kalman filter, and improved robust adaptive Kalman filter.
adaptive Kalman filter, Sage filter, extended Kalman filter, and improved robust adaptive Kalman filter.

Table 2. The RMS and maximum errors

Theof the horizontal
statistical position,
values forward
of the rootvelocity, and heading
mean square (RMS)of and
the CRAKF, Sage filter,
maximum errorsextended
of the
Kalman filter, and IRAKF. horizontal position, forward velocity, and heading of the CRAKF, Sage filter, extended
CRAKF Kalman filter, and IRAKF are shown in Table 2. It can be seen that the RMS and maximum
Sage Filter Extended Kalman Filter IRAKF
errors of the proposed IRAKF are minor than those of the other three methods. The specific
Pos Vel Heading Pos Vel Heading Pos Vel Heading Pos Vel Heading
comparative analysis is as follows. Compared with the CRAKF, the RMS errors of the
(m) (m/s) (deg) (m) (m/s) (deg) (m) (m/s) (deg) (m) (m/s) (deg)
horizontal position, forward speed, and heading of the IRAKF are increased by 3.8%, 7.9%
RMS 0.26 0.038 0.44 0.28 0.039 0.46 1.3 0.22 0.77 0.25 0.035 0.43and
and 2.3%, respectively. Compared with Sage filtering, it has increased by 10.7%, 10.3%
MAX 1.67 0.18 4.06 1.7 0.18 4.1 17.1 3.4 7.8 1.65 0.14
6.5% respectively. However, compared with extended Kalman filter, it has increased 4.06 by
80.7%, 84.1%, and 44.2%, respectively. In terms of maximum error, the horizontal position,
Thespeed
forward single-frequency
and headingGNSS signalare
of IRAKF is easy to block
improved byon the boulevard,
1.2%, resulting
22.2% and 0%, in poor
respectively
positioning accuracy. As shown in Figure 8, the GNSS signal was blocked at the
compared with the CRAKF. Compared with Sage filtering, it is improved by 2.9%, 22.2% epoch of
282 126 s, and there was an abnormal observation. At this time, the GNSS positioning
error reached 4.12m. The single-frequency GNSS/MEMS-IMU/odometer integrated navi-
gation positioning error with extended Kalman filter algorithm is also 3.8m. Although the
CRAKF can suppress abnormal GNSS positioning errors, it is still inferior in reducing er-
Remote Sens. 2021, 13, 4317 18 of 28

and 1% respectively. However, compared with extended Kalman filter, it has increased by
90.4%, 95.9% and 47.9%, respectively.

Table 2. The RMS and maximum errors of the horizontal position, forward velocity, and heading of the CRAKF, Sage filter,
extended Kalman filter, and IRAKF.

CRAKF Sage Filter Extended Kalman Filter IRAKF

Pos Vel Heading Pos Vel Heading Pos Vel Heading Pos Vel Heading
(m) (m/s) (deg) (m) (m/s) (deg) (m) (m/s) (deg) (m) (m/s) (deg)
RMS 0.26 0.038 0.44 0.28 0.039 0.46 1.3 0.22 0.77 0.25 0.035 0.43
MAX 1.67 0.18 4.06 1.7 0.18 4.1 17.1 3.4 7.8 1.65 0.14 4.06

The single-frequency GNSS signal is easy to block on the boulevard, resulting in poor
positioning accuracy. As shown in Figure 8, the GNSS signal was blocked at the epoch
of 282,126 s, and there was an abnormal observation. At this time, the GNSS positioning
error reached 4.12 m. The single-frequency GNSS/MEMS-IMU/odometer integrated
navigation positioning error with extended Kalman filter algorithm is also 3.8 m. Although
the CRAKF can suppress abnormal GNSS positioning errors, it is still inferior in reducing
error divergence. It can be seen from the green line in Figure 8 that the proposed IARKF
Remote Sens. 2021, 13, x FOR PEER has
REVIEW
the best accuracy, preventing abnormal GNSS observations and effectively slowing 20 of 30
down the divergence of positioning errors.

Figure
Figure 8. 8.The
Thepositioning
positioningerrors
errorsofofconventional
conventionalrobust
robustadaptive
adaptiveKalman
Kalmanfilter,
filter, Sage
Sage filter,
filter, extended
extended
Kalman filter, improved robust adaptive Kalman filter, and GNSS in the boulevard
Kalman filter, improved robust adaptive Kalman filter, and GNSS in the boulevard scene. scene.

InInthe
theboulevard
boulevardscene,
scene,the
thehorizontal
horizontalposition
positionerrors
errorsofofthe
theCRAKF,
CRAKF,SageSagefilter,
filter,ex-
ex-
tended
tended Kalman
Kalman filter,
filter, andand
IRAKFIRAKF
at theatmoment
the moment
of GNSS of abnormal
GNSS abnormal observations
observations are shown are
inshown
Table 3.inItTable
can be3.seen that
It can bethe integrated
seen that the navigation
integrated system withsystem
navigation the IRAKF
witheliminates
the IRAKF
GNSS gross errors
eliminates GNSS andgrossimproves
errors andpositioning
improvesaccuracy.
positioningCompared
accuracy.with the otherwith
Compared threethe
methods, its positioning
other three methods, its accuracy at theaccuracy
positioning time of GNSS
at the gross
time errors
of GNSSare gross
improved byare
errors 20%,im-
14.3%,
provedandby96.8%, respectively.
20%, 14.3%, and 96.8%, respectively.

Table 3. The horizontal position error of CRAKF, Sage filter, extended Kalman filter, and IRAKF
at the moment of gross error.

Extended Improvement
CRAKF Sage filer IRAKF
KF (%)
horizontal
0.15 0.14 3.8 0.12 20 14.3 96.8
position (m)
Remote Sens. 2021, 13, 4317 19 of 28

Table 3. The horizontal position error of CRAKF, Sage filter, extended Kalman filter, and IRAKF at
the moment of gross error.

CRAKF Sage Filter Extended KF IRAKF Improvement (%)

horizontal
0.15 0.14 3.8 0.12 20 14.3 96.8
position (m)

The tunnel scene significantly influences the positioning accuracy, and if it is not
processed, a significant error will occur. The GNSS signal is blocked when the vehicle
enters the tunnel, and only the odometer assists the INS. However, the GNSS RTK is a
pseudo-fixed solution when the vehicle enters the tunnel, as shown at 282,327 s in Figure 9.
If sufficient measures are not taken, the GNSS pseudo-fixed solution with large errors will
disturb the stable integrated navigation system, leading to significant errors in subsequent
positioning results. The red line in the figure shows the impact of the integrated navigation
with the extended Kalman filter algorithm. At the time of the GNSS outliers, the fixed
solution positioning error reaches 2.1 m. Nevertheless, the positioning error with the
IRAKF is only 0.14 m. This value is also better than the 0.165 m and 0.16 m of Sage filter
and CRAKF. Besides, the divergence speed of the INS in the tunnel is slower than the other
three methods. Therefore, the IRAKF eliminates the RTK pseudo-fixed solution, improves
Remote Sens. 2021, 13, x FOR PEER REVIEW 21 of 30
the stability of the system state and reduces the positioning error.

Figure9.9. The
Figure The positioning
positioning errors
errors of
of conventional
conventional robust
robust adaptive
adaptive Kalman
Kalman filter,
filter,Sage
Sagefilter,
filter,extended
extended
Kalman filter, improved robust adaptive Kalman filter, and GNSS in the tunnel
Kalman filter, improved robust adaptive Kalman filter, and GNSS in the tunnel scene. scene.

In
Inthe
thetunnel
tunnelscene,
scene,the
the horizontal
horizontal position
position errors
errorsof
of the
the CRAKF,
CRAKF,Sage
Sagefilter,
filter,extended
extended
Kalman filter, and IRAKF at the moment of GNSS RTK pseudo-fixed
Kalman filter, and IRAKF at the moment of GNSS RTK pseudo-fixed solution solution areare
shown
shownin
Table 4. It can be seen from the table that the integrated navigation system with
in Table 4. It can be seen from the table that the integrated navigation system with IRAKF IRAKF
eliminates
eliminates GNSSGNSS abnormal
abnormal observation
observation and
andimproves
improvespositioning
positioningaccuracy.
accuracy.Compared
Compared
with
withthe
theother
otherthree
threemethods,
methods,its itspositioning
positioningaccuracy
accuracyat atthe
theRTK
RTKpseudo-fixed
pseudo-fixedsolution
solution
epoch is improved by 12.5%, 15.2%, and 93.3%, respectively.
epoch is improved by 12.5%, 15.2%, and 93.3%, respectively.

Table 4. The horizontal position errors of the CRAKF, Sage filter, extended Kalman filter, and
IRAKF at the moment of GNSS RTK pseudo-fixed solution in the tunnel scene.

Sage Extended
CRAKF IRAKF Improvement (%)
Filter KF
horizontal
position 0.16 0.165 2.1 0.14 12.5 15.2 93.3
(m)
Remote Sens. 2021, 13, 4317 20 of 28

Table 4. The horizontal position errors of the CRAKF, Sage filter, extended Kalman filter, and IRAKF
at the moment of GNSS RTK pseudo-fixed solution in the tunnel scene.

CRAKF Sage Filter Extended KF IRAKF Improvement (%)

horizontal
0.16 0.165 2.1 0.14 12.5 15.2 93.3
position (m)

5.3. Comparison with Other Filtering Algorithms

The IRAKF algorithm is verified in the single-frequency GNSS/MEMS-IMU/odometer
integrated navigation module, and compared with the CRAKF, Sage filter, and extended
Kalman filter. The real-time errors of the position, velocity and attitude of the four algo-
rithms in the entire experimental route are shown in Figures 10–12, respectively. In addition,
to make the graph clearer, the subfigures of the position, velocity, and attitude errors are
shown in Figures 13–15, respectively, which show each curve in the figure. It can be seen
from the figures that, in complex scenes such as Boulevard 1, Tunnel 1, Boulevard 2, and
Tunnel 2, the position, velocity, and attitude errors are relatively large. In these scenarios,
the integrated navigation system with the extended Kalman filter algorithm is difficult
to reduce the positioning error caused by the GNSS gross error. Although the integrated
navigation with the Sage filter and CRAKF algorithms can achieve a certain effect, the
performance is still unsatisfactory in some harsh environments. However, the proposed
IRAKF can effectively eliminate GNSS gross errors, slow down system divergence errors
Remote Sens. 2021, 13, x FOR PEER REVIEW 22 of 30
in complex environments, and improve the position, velocity and attitude accuracy of
single-frequency GNSS/MEMS-IMU/odometer integrated navigation.

Figure 10. Cont.

Remote Sens. 2021, 13, 4317 21 of 28

Remote Sens. 2021, 13, x FOR PEER REVIEW 23 of 30

Figure
Figure10.
10.The
Thereal-time
real-timeposition errors
position ofof
errors thethe
CRAKF,
CRAKF,Sage filter,
Sage extended
filter, Kalman
extended filter,
Kalman and
filter, and IRAKF
IRAKF in the North,
in the North, East,Down
East, and and Down directions
directions in theinwhole
the whole experimental
experimental route.
route.

Figure 11. Cont.

Remote Sens. 2021, 13, 4317 22 of 28

Remote Sens. 2021, 13, x FOR PEER Figure

REVIEW11. The real-time velocity errors of the CRAKF, Sage filter, extended Kalman filter,24 of 30
and
Figure 11. The real-time velocity errors of the CRAKF, Sage filter, extended Kalman filter, and IRAKF
IRAKF in the North, East, and Down directions in the whole experimental route.
in the North, East, and Down directions in the whole experimental route.

Figure 12. Cont.

Remote Sens. 2021, 13, 4317 23 of 28

Remote Sens. 2021, 13, x FOR PEER REVIEW

Figure
Figure 12. The
The real-time
real-timeattitude
attitudeerrors
errors
of of
thethe CRAKF,
CRAKF, SageSage
filter,filter, extended
extended Kalman
Kalman 25and
filter,
filter, and of 30
IRAKF
IRAKF in the
in the Roll, Roll,and
Pitch, Pitch,
Yawand
in Yaw in theexperimental
the whole whole experimental
route. route.

Figure 13. The real-time position errors subfigures of the CRAKF, Sage filter, extended Kalman filter, and IRAKF in the
Figure 13. The real-time position errors subfigures of the CRAKF, Sage filter, extended Kalman filter, and IRAKF in the
North, East, and Down directions.
North, East, and Down directions.
Remote Sens. 2021, 13, 4317 24 of 28
Figure 13. The real-time position errors subfigures of the CRAKF, Sage filter, extended Kalman filter, and IRAKF in the
North, East, and Down directions.

Figure
RemoteFigure 14. 13,
Sens. 2021, The real-time
x real-time
FOR velocity
PEERvelocity errors subfigures of the CRAKF, Sage filter, extended Kalman filter, and IRAKF in the
REVIEW errors 26 of 30
14. The subfigures of the CRAKF, Sage filter, extended Kalman filter, and IRAKF in the
North, East, and Down directions.
North, East, and Down directions.

Figure 15. The real-time attitude errors subfigures of the CRAKF, Sage filter, extended Kalman filter, and IRAKF in the
Figure 15. The real-time attitude errors subfigures of the CRAKF, Sage filter, extended Kalman filter, and IRAKF in the Roll,
Roll, Pitch, and Yaw.
Pitch, and Yaw.
In
In the
the entire
entire experimental
experimental route,
route, the
the RMS
RMS errors
errors ofof the
the position,
position, velocity,
velocity, and
and attitude
attitude
of the CRAKF, Sage filter, extended Kalman filter, and IRAKF are shown
of the CRAKF, Sage filter, extended Kalman filter, and IRAKF are shown in Table 5. in Table 5. Since
Since
the extended Kalman filter has weak robustness, its RMS errors of the position,
the extended Kalman filter has weak robustness, its RMS errors of the position, velocity velocity
and
and attitude
attitude are
are relatively
relatively large.
large. The
The CRAKF,
CRAKF, Sage
Sage filter, and IRAKF
filter, and IRAKF can
can eliminate
eliminate gross
gross
errors, and their performance indexes are better than the extended Kalman filter. How-
ever, according to the data in Table 5, the proposed IRAKF has the best performance. In
terms of positioning accuracy, compared with the CRAKF algorithm, IRAKF has im-
proved by 14.3%, 12.5%, and 24.3% in the north, east, and down directions, respectively.
Compared with the Sage filter, it has improved by 10%, 26.3%, and 17.6%, respectively.
Remote Sens. 2021, 13, 4317 25 of 28

errors, and their performance indexes are better than the extended Kalman filter. However,
according to the data in Table 5, the proposed IRAKF has the best performance. In terms of
positioning accuracy, compared with the CRAKF algorithm, IRAKF has improved by 14.3%,
12.5%, and 24.3% in the north, east, and down directions, respectively. Compared with the
Sage filter, it has improved by 10%, 26.3%, and 17.6%, respectively. However, compared
with the extended Kalman filter, it is improved by 83.7%, 78.7%, and 45.1%, respectively.

Table 5. The RMS errors of the position, velocity, and attitude of the CRAKF, Sage filter, extended Kalman filter, and IRAKF
in the whole experimental route.

Position (m) Velocity (m/s) Attitude (deg)

North East Down North East Down Roll Pitch Yaw
CRAKF 0.21 0.16 0.74 0.035 0.034 0.045 0.53 0.214 0.432
Sage filter 0.2 0.19 0.68 0.036 0.036 0.044 0.55 0.22 0.45
Extended Kalman filter 1.11 0.66 1.02 0.19 0.2 0.23 0.65 0.60 0.774
IRAKF 0.18 0.14 0.56 0.033 0.032 0.043 0.51 0.2 0.43

5.4. Real-Time Power Consumption Verification and Comparision

In the process of real-time vehicle navigation, the power meter and power supply are
used to test the power consumption of the multi-sensor fusion navigation and positioning
module, and the data are saved. The real-time power consumption curve is shown in
Figure 16. The MCU operating frequency in the module with the IRAKF algorithm can
Remote Sens. 2021, 13, x FOR PEER REVIEW
be reduced to 50 MHz, and its average power consumption is 143 mW.27Because of 30 CRAKF,
Sage filter, and extended Kalman filter do not adopt Kalman one-step prediction strategy,
the working frequency of MCU can only be reduced to 200 MHz. The average power con-
working frequency
sumption of these of MCU can
methods is only be reduced
relatively to 200MHz.
high—296 The mW,
mW, 300 average
andpower con- respectively.
304 mW,
sumption of these methods is relatively high—296mW, 300mW, and 304mW, respectively.
According to the above analysis, IRAKF reduces the module’s power consumption by
According to the above analysis, IRAKF reduces the module's power consumption by re-
reducing
ducing thethe computational
computational complexity.
complexity.

Figure 16.
Figure 16.The
Thereal-time power
real-time consumption
power of IRAKF,
consumption CRAKF, extended
of IRAKF, CRAKF,Kalman filter,
extended and Sagefilter, and Sage
Kalman
filter in the experimental route.
filter in the experimental route.
6. Discussion
In this paper, robustness and low computation complexity are the topics. According
to the presented results, the IRAKF applied to single-frequency GNSS/MEMS-IMU/odom-
eter integrated navigation in the challenging urban environment can improve the robust-
ness ability of the system and significantly reduce the computational complexity of the
algorithm. Compared with the CRAKF, Sage filter, and extended Kalman filter algo-
Remote Sens. 2021, 13, 4317 26 of 28

6. Discussion
In this paper, robustness and low computation complexity are the topics. According to
the presented results, the IRAKF applied to single-frequency GNSS/MEMS-IMU/odometer
integrated navigation in the challenging urban environment can improve the robustness
ability of the system and significantly reduce the computational complexity of the algorithm.
Compared with the CRAKF, Sage filter, and extended Kalman filter algorithms, the IRAKF
with the one-step prediction Kalman filter dramatically reduces computation. As shown
in Figure 4, the number of M, A&S, and D operations decreased by 75.6%, 80.8%, and
42.3%, respectively. The IRAKF not only reduces the amount of calculation but also slows
down the error divergence. As shown in Figure 6, the maximum errors of horizontal
position, forward speed, and the heading of CRAKF are 1.67 m, 0.18 m/s, and 4.06 degrees,
respectively. For the Sage filter they are 1.7 m, 0.18 m/s, and 4.1 degrees, respectively,
and for the extended Kalman filter they are 17.1 m, 3.4 m/s, and 7.8 degrees, respectively.
However, for the IRAKF they are only 1.65 m, 0.14 m/s, and 4.06 degrees, respectively.
Regarding the RMS errors, the horizontal position, forward speed, and heading of the
IRAKF are 0.29 m, 0.04m/s, and 0.72 degrees, respectively.
Compared with the CRAKF, Sage filter, and extended Kalman filter algorithm, the
IRAKF improves robustness. In the boulevard, when the GNSS signal is blocked suddenly,
the positioning result becomes worse. Since the covariance matrix of the observation vector
is a fixed value, the gain matrix of the extended Kalman filter cannot change with the GNSS
positioning accuracy, so the influence of these outliers cannot be eliminated. However,
the IRAKF algorithm can use the scale and adaptive factors to adjust the measurement
vector and state prediction covariance matrices. The gain matrix can change with the GNSS
positioning accuracy and improve the robust performance of the integrated navigation
system. As shown in Figure 8, the IRAKF suppresses the abnormal GNSS and improves
the positioning accuracy of integrated navigation. In addition, it also has a specific effect in
terms of improving the accuracy of the tunnel scene. When the vehicle enters the tunnel,
GNSS RTK is a pseudo-fixed solution. The IRAKF can eliminate the RTK pseudo-fixed
solution and reduce the system state error. However, sometimes the positioning error of the
IRAKF is greater than the CRAKF, Sage filter, and extended Kalman filter. The main reason
is that the observation vector covariance matrix is an artificially set fixed value, which
cannot accurately represent the current GNSS positioning accuracy. If the GNSS receiver
can output the observation vector covariance matrix, it is believed that the positioning
accuracy of the integrated navigation system with the IARKF will be further improved.
In addition, since the calculation amount of the IRAKF algorithm is less than that of
the CRAKF, Sage filter, and extended Kalman filter, the MCU with the IRAKF algorithm
can run at a lower operating frequency. As can be seen from Figure 16, compared with the
other three algorithms, IRAKF has the lowest power consumption.

7. Conclusions
This paper presents an improved robust adaptive Kalman filter algorithm with real-
time performance and low computation complexity for single-frequency GNSS/MEMS-
IMU/odometer integrated navigation in the challenging environment. It is implemented
in multi-sensor fusion navigation and positioning module. The algorithm obtains the
scale factor by the prediction residual to adjust the artificially set covariance matrix of
the observation vector under different GNSS solution states. It can make the observation
vector covariance matrix constantly change with complex scenes. Since the Mahalanobis
distance can better reflect the disturbance anomaly, the Mahalanobis distance based on the
prediction residual is introduced to detect disturbance anomalies and obtain the adaptive
factor to inflate the state prediction covariance matrix. Furthermore, the one-step prediction
Kalman filter is adopted to reduce the computation complexity of the algorithm and realize
the low power consumption real-time integrated navigation.
The robust adaptive algorithm based on the GNSS state suppresses the GNSS gross
errors in the avenue scene, eliminates the GNSS-RTK pseudo-fixed solution in the tunnel
Remote Sens. 2021, 13, 4317 27 of 28

scene, and improves the positioning accuracy of integrated navigation. In addition, com-
pared with the CRAKF, Sage filter, and extended Kalman filter, the IRAKF reduces the
amount of calculation and realizes the low power consumption of the integrated navigation
module. The average power consumption in the whole test route is only 143 mW. It is of
great significance for real-time vehicle navigation applications in complex environments.
With measurements from the multi-sensor fusion navigation and positioning module, the
proposed filtering method can provide real-time medium and low dynamic horizontal
positioning error of less than 0.25 m, a forward velocity error of less than 0.035m/s, and a
heading error of less than 0.43 degrees, which are far better than the results of the same
level commercial integrated navigation module or other existing filtering methods.

Author Contributions: Conceptualization, J.J., J.L. and P.Y.; methodology, J.J. and P.Y.; software, P.Y.;
validation, J.J. and P.Y.; formal analysis, J.J., P.Y. and F.Z.; investigation, J.J. and P.Y.; resources, J.J.
and P.Y.; writing—original draft preparation, J.J. and P.Y.; writing—review and editing, J.J., P.Y., D.X.,
J.W., C.Z. and Y.T.; visualization, D.X. and J.W.; supervision, J.J. and P.Y.; project administration, D.X.;
funding acquisition, J.J. All authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by the National Key Research and Development Program of
China (2018YFB0505200 and 2018YFB0505201); the Fundamental Research Funds for the Central
Universities(2042018kf0253).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Acknowledgments: Thank you very much for the technical support of Fangning Zhang, Yanan Tang,
Dongpeng Xie, Jiaji Wu and Chao Zhang.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Grewal, M.; Andrews, A.; Bartone, C. Global Navigation Satellite Systems, Inertial Navigation, and Integration; John Wiley & Sons:
Hoboken, NJ, USA, 2020. [CrossRef]
2. Kaplan, E.D.; Hegarty, C. Understanding GPS Principles and Applications; Artech House Publishers: Boston, MA, USA, 2005.
3. Gallon, E.; Joerger, M.; Pervan, B. Development of Stochastic IMU Error Models for INS/GNSS Integration. In Proceedings of
Proceedings of the 34th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2021),
St. Louis, MI, USA, 20–24 September 2021; pp. 2565–2577.
4. Shin, E.H. Estimation Techniques for Low-Cost Inertial Navigation. Ph.D. Thesis, The University of Calgary, Calgary, AB,
Canada, 2005.
5. Hu, S.; Xu, S.; Wang, D.; Zhang, A. Optimization Algorithm for Kalman Filter Exploiting the Numerical Characteristics of
SINS/GPS Integrated Navigation Systems. Sensors 2015, 15, 28402–28420. [CrossRef]
6. Niu, X.; Zhang, Q.; Gong, L.; Liu, C.; Zhang, H.; Shi, C.; Wang, J.; Coleman, M. Development and evaluation of GNSS/INS data
processing software for position and orientation systems. Surv. Rev. 2015, 47, 87–98. [CrossRef]
7. Chang, G. Robust Kalman filtering based on Mahalanobis distance as outlier judging criterion. J. Geod. 2014, 88, 391–401.
[CrossRef]
8. Zhao, L.; Qiu, H.; Feng, Y. Analysis of a robust Kalman filter in loosely coupled GPS/INS navigation system. Measurement 2016,
80, 138–147. [CrossRef]
9. Jiang, C.; Zhang, S.B.; Zhang, Q.Z. A New Adaptive H-Infinity Filtering Algorithm for the GPS/INS Integrated Navigation.
Sensors 2016, 16, 2127. [CrossRef]
10. Teunissen, P.J.G. Quality control in integrated navigation systems. In Proceedings of the IEEE Symposium on Position Location
and Navigation. A Decade of Excellence in the Navigation Sciences, Las Vegas, NV, USA, 20–23 March 1990; pp. 158–165.
11. Jiang, C.; Zhang, S.-b.; Zhang, Q.-z. A Novel Robust Interval Kalman Filter Algorithm for GPS/INS Integrated Navigation. J.
Sens. 2016, 2016, 1–7. [CrossRef]
12. Jiang, C.; Zhang, S.B. A Novel Adaptively-Robust Strategy Based on the Mahalanobis Distance for GPS/INS Integrated Navigation
Systems. Sensors 2018, 18, 695. [CrossRef]
13. Gao, B.; Hu, G.; Zhu, X.; Zhong, Y. A Robust Cubature Kalman Filter with Abnormal Observations Identification Using the
Mahalanobis Distance Criterion for Vehicular INS/GNSS Integration. Sensors 2019, 19, 5149. [CrossRef]
14. Yang, Z.; Li, Z.; Liu, Z.; Wang, C.; Sun, Y.; Shao, K. Improved robust and adaptive filter based on non-holonomic constraints for
RTK/INS integrated navigation. Meas. Sci. Technol. 2021, 32, 105110. [CrossRef]
Remote Sens. 2021, 13, 4317 28 of 28

15. Yang, Y.; Xu, T. An Adaptive Kalman Filter Based on Sage Windowing Weights and Variance Components. J. Navig. 2003, 56,
231–240. [CrossRef]
16. Mohamed, A.H.; Schwarz, K.P. Adaptive Kalman Filtering for INS/GPS. J. Geod. 1999, 73, 193–203. [CrossRef]
17. Wang, J.; Stewart, M.P.; Tsakiri, M. Adaptive Kalman Filtering for Integration of GPS with GLONASS and INS; Springer:
Berlin/Heidelberg, Germany, 2000; Volume 121, pp. 325–330.
18. Gandhi, M.A.; Mili, L. Robust Kalman Filter Based on a Generalized Maximum-Likelihood-Type Estimator. IEEE Trans. Signal
Process. 2010, 58, 2509–2520. [CrossRef]
19. Duan, Z.; Zhang, J.; Zhang, C.; Mosca, E. Robust H2 and H∞ filtering for uncertain linear systems. Automatica 2006, 42, 1919–1926.
[CrossRef]
20. Zhang, L.; Viktorovich, P.A.; Selezneva, M.S.; Neusypin, K.A. Adaptive Estimation Algorithm for Correcting Low-Cost MEMS-
SINS Errors of Unmanned Vehicles under the Conditions of Abnormal Measurements. Sensors 2021, 21, 623. [CrossRef]
21. Zhao, L.; Liu, J. An Improved Adaptive Filtering Algorithm with Applications in Integrated Navigation. In Proceedings of the
2012 Third International Conference on Digital Manufacturing & Automation, Guilin, China, 31 July–2 August 2012; pp. 182–185.
22. Shi, E. An Improved Real-Time Adaptive Kalman Filter for Low-Cost Integrated GPS/INS Navigation. In Proceedings of the 2012
International Conference on Measurement, Information and Control, Harbin, China, 18–20 May 2012; Volume 2, pp. 1093–1098.
23. Yang, Y.; He, H.B.; Xu, G. Adaptively robust filtering for kinematic geodetic positioning. J. Geod. 2001, 75, 109–116. [CrossRef]
24. Yang, Y.; Cui, X. Adaptively robust filter with multi adaptive factors. Surv. Rev. 2013, 40, 260–270. [CrossRef]
25. Yang, Y. Adaptively Robust Kalman Filters with Applications in Navigation. In Sciences of Geodesy-I; Springer: Berlin/Heidelberg,
Germany, 2010; pp. 49–82.
26. Yang, Y.; Xu, T.; Xu, J. Principles and Comparisons of Various Adaptively Robust Filters with Applications in Geodetic Positioning.
In Proceedings of the 1st International Workshop on the Quality of Geodetic Observation and Monitoring Systems (QuGOMS’11),
Munich, Germany, 13–15 April 2011; pp. 101–105.
27. Jung, S.; Hwang, S.; Shin, H.; Shim, D.H. Perception, Guidance, and Navigation for Indoor Autonomous Drone Racing Using
Deep Learning. IEEE Robot. Autom. Lett. 2018, 3, 2539–2544. [CrossRef]
28. Rostami, M.; Kolouri, S.; Eaton, E.; Kim, K. SAR Image Classification Using Few-Shot Cross-Domain Transfer Learning. In
Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), Long Beach,
CA, USA, 16–17 June 2019; pp. 907–915.
29. Abdallah, A.; Kassas, Z. Deep Learning-Aided Spatial Discrimination for Multipath Mitigation; IEEE: Piscataway, NJ, USA, 2020;
pp. 1324–1335.
30. Fayjie, A.; Hossain, S.; Doukhi, O.; Lee, D. Driverless Car: Autonomous Driving Using Deep Reinforcement Learning in Urban
Environment. In Proceedings of the 2018 15th International Conference on Ubiquitous Robots (UR), Jeju, Korea, 28 June–1 July
2018; pp. 896–901.
31. Yan, P.; Jiang, J.; Tang, Y.; Zhang, F.; Xie, D.; Wu, J.; Liu, J.; Liu, J. Dynamic Adaptive Low Power Adjustment Scheme for
Single-Frequency GNSS/MEMS-IMU/Odometer Integrated Navigation in the Complex Urban Environment. Remote. Sens. 2021,
13, 3236. [CrossRef]
32. Zhou, Y.; Chen, Q.; Niu, X. Kinematic Measurement of the Railway Track Centerline Position by GNSS/INS/Odometer Integration.
IEEE Access 2019, 7, 157241–157253. [CrossRef]
33. Simon, D. Optimal State Estimation (Kalman, H∞, and Nonlinear Approaches)||Optimal Smoothing; John Wiley and Sons: Hoboken,
NJ, USA, 2006; pp. 263–296.
34. Zhu, Q.J.; Yan, G.M.; Yang, P.X.; Qin, Y.Y. A Rapid Computation Method for Kalman Filtering in Vehicular SINS/GPS Integrated
System. Appl. Mech. Mater. 2012, 182–183, 541–545. [CrossRef]
35. Emami, M.; Taban, M. A Low Complexity Integrated Navigation System for Underwater Vehicles. J. Navig. 2018, 71, 1–17.
[CrossRef]
36. Li, Q.; Ban, Y.; Niu, X.; Zhang, Q.; Gong, L.; Liu, J. Efficiency Improvement of Kalman Filter for GNSS/INS through One-Step
Prediction of P Matrix. Math. Probl. Eng. 2015, 2015, 1–13. [CrossRef]