Scribd
Upload
75 views19 pages

Kalman Filters: CS 344R: Robotics Benjamin Kuipers

The lecture covered Kalman filters for linear and nonlinear state space models in higher dimensions. It explained the predictor and corrector steps of the discrete Kalman filter, including computing the Kalman gain and updating the state estimate and error covariance. For nonlinear models, it described linearizing around the current estimate and using the Jacobian matrix in the Kalman filter equations to handle nonlinearities.

Uploaded by

adarshsasidharan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
75 views19 pages

Kalman Filters: CS 344R: Robotics Benjamin Kuipers

The lecture covered Kalman filters for linear and nonlinear state space models in higher dimensions. It explained the predictor and corrector steps of the discrete Kalman filter, including computing the Kalman gain and updating the state estimate and error covariance. For nonlinear models, it described linearizing around the current estimate and using the Jacobian matrix in the Kalman filter equations to handle nonlinearities.

Uploaded by

adarshsasidharan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 19

Lecture 11:

Kalman Filters
CS 344R: Robotics
Benjamin Kuipers
Up To Higher Dimensions
• Our previous Kalman Filter discussion was
of a simple one-dimensional model.
• Now we go up to higher dimensions:
n
– State vector: x ∈ℜ
m
– Sense vector: z ∈ℜ
l
– Motor vector: u ∈ ℜ
• First, a little statistics.
Expectations
• Let x be a random variable.
• The expected value E[x] is the mean:
N
1
E[x] = ∫ x p(x) dx ≈ x = ∑ x i
N 1
– The probability-weighted mean of all possible
values. The sample mean approaches it.
• Expected value of a vector x is by
€ T
component.
E[x] = x = [x1,L x n ]
Variance and Covariance
• The variance is E[ (x-E[x])2 ]
N
1
σ = E[(x − x ) ] = ∑ (x i − x )
2 2 2

N 1
• Covariance matrix is E[ (x-E[x])(x-E[x])T ]
N
1
Cij =
N
∑(x ik − x i )(x jk − x j )
k =1

– Divide by N1 to make the sample variance an


unbiased estimator for the population variance.
Biased and Unbiased Estimators
• Strictly speaking, the sample variance
N
1
σ = E[(x − x ) ] = ∑ (x i − x )
2 2 2

N 1
is a biased estimate of the population
variance. An unbiased estimator is:
N
1
2
s = ∑
N −1 1
(x i − x ) 2

• But: “If the difference between N and N1


ever matters to you, then you are probably
up to no good anyway …” [Press, et al]
€
Covariance Matrix
• Along the diagonal, Cii are variances.
• Off-diagonal Cij are essentially correlations.

⎡C1,1 = σ 12 C1,2 C1,N ⎤


⎢ 2 ⎥
⎢ C2,1 C2,2 = σ 2 ⎥
⎢ O M ⎥
⎢ 2⎥
⎣ CN ,1 L CN ,N = σ N ⎦
Independent Variation
• x and y are
Gaussian random
variables (N=100)
• Generated with
x=1 y=3
• Covariance matrix:
⎡0.90 0.44⎤
Cxy = ⎢ ⎥
⎣0.44 8.82⎦
Dependent Variation
• c and d are random
variables.
• Generated with
c=x+y d=x-y
• Covariance matrix:

⎡10.62 −7.93⎤
Ccd = ⎢ ⎥
⎣−7.93 8.84 ⎦
Discrete Kalman Filter
n
• Estimate the state x ∈ℜ of a linear
stochastic difference equation
x k = Axk −1 + Buk + wk −1
– process noise w is drawn from N(0,Q), with
covariance matrix Q.
• with a measurement z ∈ℜ m
z k = Hx k + vk
– measurement noise v is drawn from N(0,R), with
covariance matrix R.
• A, Q are nxn. B is nxl. R is mxm. H is mxn.
Estimates and Errors
n
ˆ
• x k ∈ℜ is the estimated state at time-step k.
− n
ˆ
• x k ∈ℜ after prediction, before observation.
• Errors: e −
k = x k − xˆ −
k

e k = x k − xˆ k
• Error covariance matrices:
T
− − −
P = E[e e ]
k k k
T
Pk = E[e k e ]k

• Kalman Filter’s task is to update xˆ k Pk


Time Update (Predictor)
• Update expected value of x
ˆx−k = Axˆ k −1 + Bu k
• Update error covariance matrix P
− T
Pk = APk −1A + Q

• Previous statements were simplified


versions of the same idea:
−
xˆ (t ) = xˆ (t2 ) + u[t3 − t 2 ]
3
2 − 2 2
σ (t ) = σ (t2 ) + σ [t3 − t2 ]
3 ε
Measurement Update (Corrector)
• Update expected value
− −
xˆ k = xˆ + K k (z k − Hxˆ )
k k
−
– innovation is z k − Hx ˆk
• Update error covariance matrix
−
Pk = (I − K kH) P k

• Compare with previous form


− −
xˆ (t 3 ) = xˆ(t ) + K(t3 )(z3 − xˆ (t ))
3 3
2 2 −
σ (t 3 ) = (1− K(t 3 ))σ (t3 )
The Kalman Gain
• The optimal Kalman gain Kk is
− T − T −1
K k = Pk H (HPk H + R)
− T
PH k
= − T
HPk H + R

• Compare with previous form


2 −
σ (t3 )
K(t 3 ) = 2 − 2
σ (t3 ) + σ 3
Extended Kalman Filter
• Suppose the state-evolution and
measurement equations are non-linear:
x k = f (x k −1 ,uk ) + wk −1
z k = h(x k ) + vk
– process noise w is drawn from N(0,Q), with
covariance matrix Q.
– measurement noise v is drawn from N(0,R),
with covariance matrix R.
The Jacobian Matrix
• For a scalar function y=f(x),
Δy = f ′ (x)Δx
• For a vector function y=f(x),
⎡∂f1 ∂f1 ⎤
⎡Δy1 ⎤ ⎢ (x) L (x)⎥ ⎡Δx ⎤
⎢ ⎥ ⎢∂x1 ∂x n ⎥ ⎢ 1⎥
Δy = JΔx = ⎢ M⎥ = ⎢ M M ⎥⋅ ⎢ M ⎥
⎢ ⎥ ⎢∂f n ∂f n ⎥ ⎢ ⎥
⎣Δy n ⎦ ⎢ (x) L (x)⎥ ⎣Δx n ⎦
⎢
⎣∂x1 ∂x n ⎥
⎦
Linearize the Non-Linear
• Let A be the Jacobian of f with respect to x.
∂f i
A ij = (x k−1,uk )
∂x j
• Let H be the Jacobian of h with respect to x.
∂hi
H ij = (x k )
∂x j
• Then the Kalman Filter equations are
almost the same as before!
EKF Update Equations
• Predictor step: ˆx−k = f (xˆ k −1 ,uk )
− T
P = APk −1A + Q
k

− T − T −1
• Kalman gain: K k = P H (HP H + R)
k k

− −
• Corrector step: xˆ k = xˆ + K k (z k − h(xˆ ))
k k
−
Pk = (I − K kH) P k
“Catch The Ball” Assignment
• State evolution is linear (almost).
– What is A?
– B=0.
• Sensor equation is non-linear.
– What is y=h(x)?
– What is the Jacobian H(x) of h with respect to x?
• Errors are treated as additive. Is this OK?
– What are the covariance matrices Q and R?
TTD

• Intuitive explanations for APAT and HPHT


in the update equations.

You might also like