2019 22nd International Conference on Electrical Machines and Systems (ICEMS)

Sliding Mode Speed Regulation of Linear Induction Motors Based on Direct

Thrust Control with Space-Vector Modulation Strategy
Mosaad M. Ali1,2, Wei Xu1, Mahmoud F. Elmorshedy1,3,Yi Liu1, Said M. Allam3, and Minghai Dong4
1 School of Electrical and Electronic Engineering, Huazhong University of Science and Technology, 430074, Wuhan, China
2 Department of Electrical Engineering, Faculty of Engineering, Kafrelsheikh University, Egypt
3 Department of Electric Power and Machines Engineering, Faculty of Engineering, Tanta University, Egypt
4 Foshan Golden Age Motor technology Co., Ltd, Foshan, China

E-mails: [email protected], [email protected], [email protected], [email protected],

[email protected], [email protected]

Abstract — In this paper, one improved speed control behavior of the conventional PI controller in the flexibility to
algorithm of linear induction motors (LIMs) using sliding mode load variations is much poor.
control (SMC) is firstly proposed. Then, an extended sliding On the other hand, the advanced control methods, such as
mode load thrust observer (ESO) has been proposed to
fuzzy control method and the neural network method, they
compensate the sliding mode controller under the load thrust
disturbance. Moreover, the SMC integrated with ESO is need adjustment and optimization in the control procedure.
designed based on direct thrust control (DTC) technique with Thus, the high-performance underlying controller, often,
space-vector modulation (SVM) to regulate speed, flux, and goes beyond reality, which makes it difficult for the control-
thrust of the LIM. The output of the SMC is considered as the system based on them to perceive both hardware and
reference thrust for the DTC based on space-vector modulation software [11]. In other word, to make a high-performance
(DTC-SVM). A comparison between the proposed method and
control system for LIMs, it is difficult to use either of these
the conventional method with PI controllers is presented to
illustrate the superiority of the proposed control method. control strategies.
Comprehensive simulation results have demonstrated that the The sliding mode controller (SMC) is gradually applied in
proposed control scheme has a better transient control the motor control system because of its robustness, easy to
performance with a fast and good speed tracking. realize, and adaptability to load variations. An extended
Index Terms-- Sliding mode control, Direct thrust control, sliding mode load thrust observer (ESO) is proposed to
Speed regulation, Load thrust observer, Linear induction overcome the impacts of the load thrust disturbance on LIM
motor. control system. To the best of authors knowledge, the
I. INTRODUCTION combined approach of SMC-ESO with DTC-SVM for LIM
Recently, LIMs have been paid more attention from both is not presented. Therefore, this paper presents a sliding
academia and industry for their advantages of direct drive, mode speed controller with proposed load thrust
greater acceleration/deceleration, compact volume, etc.[1, 2]. compensation technique based on (DTC-SVM) for the LIM.
The most obvious advantages of the LIMs are that it has The robust control performance is achieved by the proposed
high-starting thrust force, no gears and no mechanical rotary control strategy. The proposed control algorithm applied to
requires between motor and the motion devices, reduction of improve a speed-tracking objective under disturbance of load
mechanical losses, low cost, silence, and so on [3-5]. On the thrust. Simulation results are presented to verify the validity
other hand, the thrust dynamic of LIMs has some drawbacks of the proposed control strategy.
such as higher air gap, higher leakage flux, and the effect of II. DYNAMIC MODEL OF LIM CONSIDERING END
end effect [2, 6]. The driving principles of the LIMs are like EFFECTS
those of a rotary induction motors however, its control The linear induction motor (LIM) equivalent circuit is
characteristics are more complicated because the effect of similar to the rotary induction motor equivalent circuit
end effect. The direct thrust control (DTC) was presented to without considering the effect of end effect [5, 12-14]. The
achieve faster dynamic response and decouple the control of LIM dynamic model is obtained in synchronous reference
primary flux linkage and electromagnetic thrust. DTC frame based on the equivalent circuit that introduced in [1].
strategy has more advantages compared with the field- The LIM model can be illustrated as the following equations:
oriented control (FOC) such as less dependence on machine The dq-axis voltage equations can be calculated from the
parameters, high efficiency and does not require current following equations:
regulation loop or coordinate transformation. The control (1)
uda = Ra ida + pψ da − ωψ 1 qa 
method of DTC strategy is based on hysteresis controllers 
and switching table to determine the suitable voltage vector uqa = Raiqa + pψ qa + ωψ 
1 da 
of the inverter to regulate stator flux and thrust. Because of udb = Rb idb + pψ db − (ω1 − ω2 )ψ qb = 0  (2)
this control method, DTC strategy has some drawbacks such 
as higher thrust ripple, higher primary flux linkage ripple and uqb = Rb iqb + pψ qb + (ω1 − ω2 )ψ db = 0 
variable switching frequency. Several modifications are used where subscripts a and b refer to both primary and secondary
to reduce both primary flux linkage and thrust ripples and variables respectively. While ω1 and ω2 are the synchronous
achieve constant switching frequency as in [7, 8]. In this and rotor speed respectively.
paper, the switching table is replaced by the space vector The dq-axis linkage fluxes can be written as:
modulation SVPWM for the voltage vector selection to ψ da = La ida + Lmeq idb 
 (3)
achieve a constant switching frequency. As regards the speed
ψ qa = La iqa + Lmeq iqb 
loop control, its common to use PI controller for all
researches [9], However, for high performance LIMs system, ψ db = Lb idb + Lmeq idb 
 (4)
the conventional PI controller cannot meet the request of ψ qb = Lb iqb + Lmeq iqb 
multi-variable nonlinear control [10]. Moreover, the

978-1-7281-3398-0/19/$31.00 ©2019 IEEE

 2019 22nd International Conference on Electrical Machines and Systems (ICEMS)

The electromagnetic developed thrust can be determined the appropriate vector the trajectory of the primary flux
by: moves in the direction of the inverter output voltage vector.
3π Furthermore, the LIM developed thrust, illustrated in
Fe = (ψ da iqa −ψ qa ida ) (5) equation (6), can be expressed in another equation based on
2τ
The motion equation is expressed as: the primary flux linkage and secondary flux linkage. Fig. 2
shows the primary flux linkage vector, secondary flux
dv
Fe = Fl + M r + Dvr (6) linkage and the angle between the two fluxes vectors in αβ
dt plane.
where vr is the linear speed of LIM, M is the mass and the
3 π Lmeq
effect of dynamic end effects on LIM parameters is based on Fe = ψ aψ b sin(θ ab ) (12)
f(Q) and its value depends on end effect factor Q as follows: 2 τ σ La Lb
f (Q ) = (1 − exp(−Q)) / Q (7) where
where L2meq
σ = 1− (13)
Q = La Ra / [vr ( Llb + Lm )] (8) La Lb
and La is the primary length. It can be noted that, the procedure to increase/decrease the
The value f(Q) is a coefficient that describes the average thrust rate achieved by selecting the switching voltage vector
attenuation rate of the flux and it is utilized to modify the that increases/decrease the flux angle (θab).
magnetizing inductance as given:
Lmeq = [1 − f (Q)]Lm (9) u3 (010) u2 (110)

where Lm is the mutual inductance at standstill [13, 15]. Then

Sector
the modified primary and the secondary flux linkage are Sector II Sector
given by the following equations: III I

Vref.
La = Lla + Lmeq u4 (011)
u0 (000)
u7 (111) u1 (100)
(10)
Lb = Llb + Lmeq
III. DIRECT THRUST CONTROL WITH SPACE-VECTOR Sector Sector
MODULATION IV
Sector
VI
V
Space vector modulation based direct thrust control
scheme (DTC-SVM) results from the combination between u5 (001) u6 (101)

the direct thrust control and the field-oriented control. There

Fig.1. Reference primary -voltage vector of space-vector modulation
are many articles proposed DTC-SVM strategy to reduce the
The thrust and primary flux can be controlled
stator flux ripple and torque ripple and achieve constant
simultaneously by choosing the appropriate primary voltage
switching frequency in conventional rotating machines [16-
vector and then realized by SVM technique. The choice of
19]. DTC-SVM can be implemented with different schemes,
the optimum voltage vector to achieve the required value of
such as a) DTC-SVM scheme with closed-loop torque control
flux and thrust is introduced in Fig. 3.
only, b) DTC-SVM scheme with-closed loop flux linkage
β
control only, c) DTC-SVM with closed-loop torque and flux
linkage control.
The first two schemes are sensitive to any disturbance or Ψa-ref.
distortion in the feedback signal which may lead to Ψa-act.
instability problems in the control system [20]. All of these
δΨa
methods can be implemented for LIMs and is used for
regulating the thrust instead of torque.
θΨa Ψb-act.
The method presented in this paper is DTC-SVM with
closed loop thrust and flux linkage control to overcome the θab
α.
problems associated with the other two control schemes.
In order to drive the control gates of the three-phase
inverter, there are eight voltage vectors resulting from the Fig.2. Primary and secondary flux linkage vectors in αβ plane
combination of switching state for each switching device. Based on the field-oriented control concept, the primary
Fig. 1 shown the six active voltage vectors (u1-u6) where two flux linkage is oriented in the d-axis. The primary voltage
voltage vectors are zero (u0, u7). The reference primary equations can be modified as follows:
voltage vector can be rephrased in the vector representation uda = Ra ida + pψ da
form as follows: (14)
uqa = Ra iqa + ωaψ da
→
→ dψ a → Substituting equation (14) into equation (5), the developed
u1 = + Ra I a (11)
dt thrust equation can be expressed by:
The primary resistance, Ra, can be neglected. 3π
Fe = ψ a (uqa − ωsψ a ) (15)
According to the above assumption, the trajectory of the 2τ Ra
primary flux linkage can be directly controlled by controlling It can be observed from equation (14) and (15) that the d-
the output reference voltage vector. Therefore, by selecting axis stator voltage, uda can be used to adjust the value of
primary flux linkage while the q-axis primary voltage, uqa
 2019 22nd International Conference on Electrical Machines and Systems (ICEMS)

has an influence on the developed thrust. The control procedure is to find a control law so that the
LIM speed can track the desired reference vr to achieve this
u3 u2 Ψa Fe Ψa Fe
control objective, the tracking error e is defined as the
u3 u2 difference between the reference and actual linear speed. The
Ψa Fe
u1
sliding mode surface s is defined as the function of speed
u0 & u7
Sector error, represented as e = vr* - vr , where vr* is the linear speed
u6
I
Ψa Fe
reference. The states of LIM speed error can be defined by x1
u4
Vref. and x2 as illustrated in the following equation:
u1
x1 = vr * − vr 
• •*  (20)
x2 = x1 = vr − vr 
The sliding mode variable is defined as:
s = k * x1 + x2 (21)
where k is a positive constant of the SMC surface variable.
u5 u6
The reaching law s· is designed as equation (22) for fast
Fig.3. Choosing the optimum voltage vector based on the required
approaching and reducing chattering phenomena.
value of flux and thrust s⋅ = −ε sgn( s ) − cs (22)
According to above observations, the speed of LIM can
be adjusted by controlling the value of the developed thrust where ε ≥ 0 is switching gain and c > 0 is exponent
coefficients of reaching law. The LIM controlled system
according to the motion equation (6). The complete block
arrived stable state at s= s· =0.
diagram of the proposed SMC system to adjustable speed of
−ε sgn(s ) − cs = k * x 1• + x 2• (23)
LIM is shown in Fig. 4. The reference speed is compared
with the actual linear speed and the error is controlled using Combined with equation (23) and (16), the thrust
a sliding mode controller and the output is considered the reference of sliding mode speed controller is designed as:
* ••

reference thrust. Two PI controllers are used to regulate the Fe = Bn −1 [ k * vr * + vr ∗ + ε sgn( s ) + cs

value of the reference primary flux linkage and the reference ∧ (24)
thrust. The output of theses PI controllers is the command + c * Fl − c * An vr − An vr  ]
values of dq-axis primary voltage components in The coefficients of the sliding mode variable states k, ε,
synchronous reference frame. The command value of the and c will meet to zero in finite time.
primary voltage is transformed to the αβ-coordinates (uαa* Although the switching gain ε has a strong influence on
the control performance, it is difficult to determine the gain ε
and uβa*) using the primary flux angle, θψ. These command
to reduce external disturbance and the undesirable chattering
values (uαa* and uβa*) are used to generate the required effect [9]. For more improving the function sgn (s) can be
inverter pulses aiding with SVM scheme. designed as:
Vdc
s
Ψa* PI
uda*
uαa* sat ( s ) = (25)
Ψa Converting
Pulses
VSI
s +γ
Sliding Fe* SVPWM
vr* uqa* from dq to αβ uβa*
mode PI Where γ is a small positive constant
controller Fe
vr
Fl
θΨ
V. EXTENDED THRUST OBSERVER AND COMPENSATION
Load thrust Flux & Thrust
iabc BASED ON SLIDING MODE CONTROL
observer Fe vabc
vr
calculation From the thrust equations of LIM motor, the extended
state space equation can be obtained from (13). With
electrical linear speed vr and load thrust Fl as the observed
Sensoer
SLIM objects.
  vr •   − DM −1 − M −1  vr   M −1 
Fig.4. Block diagram of the proposed control strategy for LIM.  •  =   +  Fe
  Fl   0 0   Fl   0 
IV. DESIGN OF SLIDING MODE SPEED CONTROL  (26)
Based on LIM mechanical equation, the design of the  vr 
 vr = [1 0]  F 
proposed SMC of LIM speed involves the following steps:   l
The LIM mechanical equation can be represented A desired sliding variable s is designed as a function of
considering the nominal parameters as: speed error, ev= vr*- vr^, where vr^ is the estimated speed, Fl^
vr • = Bn Fe + Dn Fl + An vr (16) the estimated thrust load, h feedback gain of the observer
where M and D is the total mass and viscous coefficient of and cv, εv, kv, are positive gain. The proposed load
LIM respectively and Bn = − Dn = 1 / M , An = − D / M are disturbance sliding mode observer ESO is obtained by:
the values of the parameters in the nominal condition.  ∧•  −1  
∧
v
 = − DM −1
−M   r  + M  F + 1  f (v − v ) (27)
v −1 ∧
If an external load disturbance occurs, the mechanical r
     
∧ 0 0  ∧   0 
e
h
r r
equation of the controlled LIM system can be modified as:  Fl •    Fl 
vr • = ( Bn + ΔB) Fe + ( Dn + ΔD ) Fl + ( An + ΔA)vr (17)
Where, f (vr - vr^) is the sliding mode control function of
vr • = Bn Fe + Δz + An vr (18) speed observation errors which depends on its sliding
The lumped uncertainty Δz is given by: variable and reaching law.
Δz = ΔBFe + ( D + ΔD ) Fl + ΔAvr Subtracting equation (26) from equation (27) yields, the
(19)
observed dynamic error can be written as:
 2019 22nd International Conference on Electrical Machines and Systems (ICEMS)

e v = − DM ev − M eF − f (ev )
• −1 −1 It can be noticed from Fig. 5 that the actual speed tracks
 • (28) the reference speed which ensure the effectiveness of the
 e F = −hf (ev ) proposed method. Being the comparison in Fig. 5 the
∧ ∧ tracking performances are gradually deteriorated at changing
where ev = vr * − vr and eF = Fl − Fl speed for the PI controller system. In contrast, the proposed
In this study, an integral sliding variable is designed to SMC controlled system demonstrates its robustness in the
reduce the speed observation error. The integral sliding tracking performance during speed variation, as illustrate in
variable can be expressed as: Fig. 5.
 ev + cv * ev ev < ev (lim) It can be observed that a high level of the actual speed
sv =   (29) rise-time is due to the high mass of the employed LIM in
ev ev > ev (lim) addition to that the motor is started with a 100 N load.
On the other hand, Fig. 6 shows that the actual flux
where cv and ev(lim) are the integral coefficient and the linkage tracks the reference flux linkage with acceptable
separation threshold of the ESO integral sliding surfaces, small variation during speed changes. The flux linkage ripple
respectively. has also a small reduction using the proposed SMC system
The exponent reaching law with respect to the speed compared to PI controller. Fig. 6 verifies the effectiveness of
observation error can be expressed as the proposed SMC system based on DTC-SVM strategy.
sv • = ev • + cv * ev = −ε v sgn( sv ) − kv sv (30) Moreover, the profile of the thrust load is shown in Fig. 7.
where εv is the switching gain and kv is the exponent
coefficient of the ESO reaching law.
By substituting equation (28) into equation (30), the
sliding mode control law of the ESO can be obtained as
follows:
∧
f (vr − vr ) = (cv − DM −1 )ev + ε v sgn( sv ) + kv sv ] (31)
VI. STABILITY ANALYSIS
Lyapunov stability theorem is used to guarantee that the
controlled system arrives the sliding surface and stays on it.
The stability proof of the controlled LIM system is expresses
as follow:
1
v( x ) = s 2
2 Fig.5. Speed profile of the PI - proposed sliding controller.
v• = s * s• < 0
v •1 = s * s •
= (k * x1 + x2 ) * (−ε sgn( s ) − cs )
= (k * x1 + x2 ) *[−ε sgn(k * x1 + x2 ) − c(k * x1 + x2 )]
= −ε (k * x1 + x2 ) *sgn(k * x1 + x2 ) − c(k * x1 + x2 ) 2
= −ε k * x1 + x2 − c (k * x1 + x2 ) 2 ≤ 0
The stability proof of ESO, it obtains
v • 2 = sv * s • v
= sv (cv ev + ev • )
∧
= sv (cv ev − DM −1ev − M −1eF − hf (vr − vr ))
= sv (−ε v sgn( sv ) − kv sv ) Fig.6. LIM flux linkage of PI - proposed sliding controller during
speed variation.
= −ε v sv − kv s 2 v ) ≤ 0
This can ensure that the proposed SMC system is stable
and any tracking error trajectory will reach to zero in a finite
time.
VII. SIMULATION RESULTS
To verify the validity of the proposed SMC system,
several simulation results of speed tracking are conducted.
The parameters of the LIM are listed in Table 1 (appendix
section). The proposed SMC system studied and discussed
under two different operating conditions.
A. Variable Speed under Constant Thrust Load
Sliding mode speed regulation of linear induction motors
based DTC-SVM control strategy is tested under constant
thrust load equals 100 N. In this case, two different levels of
reference speed are adopted, Where the reference linear Fig.7. Load and developed thrust response of the PI - proposed
speed is decreased from 10 m/s to 8 m/s at time t=10 s. sliding controller under speed Variation.
 2019 22nd International Conference on Electrical Machines and Systems (ICEMS)

The effect of speed variation, shown in Fig. 5, on the

developed thrust is illustrated in Fig. 7. It can be observed
that the developed thrust varies directly with speed change to
keep the actual speed tracks the reference value.
The input three-phase current of a LIM controlled with
the proposed sliding controller during speed changes are
illustrated in Fig. 8. It can be noticed that the three-phase
current are affected by the speed change due to the variation
of the magnetizing inductance.

Fig.10. Load and developed thrust Profile of the PI - proposed

sliding mode controller.

Fig.8. The input three-phase current of a LIM controlled with the

proposed sliding controller during speed changes

B. Variable Thrust Load under Constant Speed

In order to proof the effectiveness and validity of the
proposed sliding mode control strategy under different Fig.11. LIM flux linkage of PI - proposed sliding controller
operating conditions, the system performance is tested under under thrust load Variation.
variable thrust loading condition. In this case, it is assumed
that the LIM is loaded, with 80 N for the first 10 s after that
the load is increased to 150 N for 5 s. The profile of the
thrust load is shown in Fig. 10. The capability and
effectiveness of the proposed SMC method over PI
controller for speed control are shown in Fig. 9 where good
speed tracking can be achieved.
While, Fig. 9, Fig. 10 and Fig. 11 illustrate the capability
i abc (A)

of the SMC strategy of DTC-SVM for regulating the speed at

the required value, achieving the load requirements and
keeping the primary flux linkage constant at the desired
value. Furthermore, the input three-phase current of a LIM
controlled with the proposed sliding mode controller during
load thrust changes are shown in Fig. 12.

Fig.12. The input three-phase current of a LIM controlled with

the proposed sliding controller during load thrust changes

The simulation results proved the speed tracking of the

vr (m/s)

proposed sliding mode controller is much better than that PI

controller. Moreover, the results illustrate the capability of
the proposed SMC method for regulating the speed at the
required value with good dynamic and static performance.
With achieving the load requirements and keeping the stator
flux linkage constant at the desired value.
VIII. CONCLUSION
In this paper, a sliding mode control law for regulating
the speed of LIM is proposed. Moreover, an extended sliding
Fig.9. LIM Speed response of the PI - proposed sliding mode mode load thrust observer has been proposed and
controller under thrust load Variation implemented to compensate the proposed controller. The
effectiveness and validity of the presented control strategy
 2019 22nd International Conference on Electrical Machines and Systems (ICEMS)

have been verified using appropriate simulation tools for [12] Duncan, J. “Linear induction motor-equivalent-circuit model,” in IEE
LIM controlled using DTC-SVM. The comparative study Proceedings B - Electric Power Applications, vol. 130, no. 1, pp. 51-
57, January 1983.
supported with simulation results proof that the proposed [13] J. Zou, W. Xu and C. Ye, “Improved deadbeat control strategy for
SMC-ESO has the best dynamic performance with the linear induction machine,” IEEE Trans. Magn., vol. 53, no. 6, pp. 1-4,
minimum tracking error and settling time under speed and June 2017, Art no. 8106804.
load variation conditions. [14] J. Zou, W. Xu, J. Zhu and Y. Liu, “Low-complexity finite control set
model predictive control with current limit for linear induction
VI. ACKNOWLEDGEMENT machines,” IEEE Trans. Ind. Electron., vol. 65, no. 12, pp. 9243-9254,
This work has been partly supported by National Natural Dec. 2018.
[15] M. Cirrincione and M. Pucci, “An MRAS-based sensorless high-
Science Foundation of China (NSFC51877093, 51707079, performance induction motor drive with a predictive adaptive model,”
and 51807075), National Key Research and Development IEEE Trans. Ind. Electron., vol. 52, no. 2, pp. 532-551, April 2005.
Program of China (Project ID: YS2018YFGH000200), and [16] Z. Koutsogiannis, G. Adamis and A. Fyntanakis, “Direct Torque
Fundamental Research Funds for the Central Universities Control using Space Vector Modulation and Dynamic Performance of
(2019kfyXMBZ031). the Drive, via a Fuzzy Logic Controller for Speed Regulation, in
Torque Control,” 2007 European Conf. on Power Electronics and
APPENDIX Applications, Aalborg, 2007, pp. 1-10.
TABLE I. MAIN PARAMETERS OF LIM [17] Zhang, Y., Q. Wang, and W. Liu. “Direct torque control strategy of
Pareamter Value induction motors based on predictive control and synthetic vector duty
Rated thrust, FN 280 N ratio control,” 2010 International Conf.on Artificial Intelligence and
Rated power, PN 3 kW Computational Intelligence, Sanya, 2010, pp. 96-101.
Rated speed, vN 11 m/s [18] A. Ammar, A. Benakcha, and A. Bourek, “Closed loop torque SVM-
Rated voltage, UN 180 V DTC based on robust super twisting speed controller for induction
motor drive with efficiency optimization,” Int. J. Hydrogen Energy.,
Rated current , IN 22 A
vol. 42, no. 28, pp.17940-17952,July 2017.
Pole pitch, τ 0.1485 m
[19] A. B. Jidin, N. R. B. N. Idris, A. H. B. M. Yatim, M. E. Elbuluk and T.
Primary length, ls 1.3087 m
Sutikno, “A wide-speed high torque capability utilizing
Secondary resistance, Rb 2.4 Ω
overmodulation strategy in DTC of induction machines with constant
Secondary leakage inductance, Llb 0.0043 H
switching frequency controller,” IEEE Trans. Power Electron., vol.
Primary leakage inductance, Lla 0.0114 H 27, no. 5, pp. 2566-2575, May 2012.
Primary resistance, Ra 1Ω [20] Zou, J. and W. Xu. “Optimal current reference with maximum thrust
criterion for linear induction machine based deadbeat control,” IECON
REFERENCES 2017 - 43rd Annual Conf. of the IEEE Industrial Electronics Society,
Beijing, 2017, pp. 3682-3687.
[1] W. Xu, J. G. Zhu, Y. Zang, Z. Li, Y. Li, Y. Wang, Y. Guo, and Y. Li,
“Equivalent circuits for single-sided linear induction motors,” IEEE
Trans. Ind. Appl., vol. 46, no. 6, pp. 2410–2423, Nov./Dec. 2010.
[2] I. Boldea, L. N. Tutelea, W. Xu and M. Pucci, “Linear electric
machines, drives, and MAGLEVs: an overview,” IEEE Trans. Ind.
Electron., vol. 65, no. 9, pp. 7504-7515, Sept. 2018.
[3] M. F. Elmorshedy, W. Xu and Y. Liu, “Speed control of linear
induction motor with thrust and stator flux ripple reduction,” 2018
21st International Conf. on Electrical Machines and Systems
(ICEMS), Jeju, 2018, pp. 1789-1794.
[4] W. Xu, M. F. Elmorshedy, Y. Liu, M. R. Islam and S. Allam, “Finite-
set model predictive control based thrust maximization of linear
induction motors used in linear metros,” IEEE Trans. Veh. Technol.,
early access 2019.
[5] W. Xu, G. Sun, G. Wen, Z. Wu and P. K. Chu, “Equivalent circuit
derivation and performance analysis of a single-sided linear induction
motor based on the winding function theory,” IEEE Trans. Veh.
Technol., vol. 61, no. 4, pp. 1515-1525, May 2012.
[6] M. M. Ali, S. M. Allam and T. H. Abdel-Moneim, “Dynamic
characteristics of an isolated self-excited synchronous reluctance
generator driven by a wind turbine,” Turk J Elec Eng & Comp Sci.,
vol. 24, no. 6, pp. 5238-5250, 2016.
[7] Zou, J., et al., “Multistep model predictive control with current and
voltage constraints for linear induction machine based urban
transportation,” IEEE Trans. Veh. Technol., vol. 66, no. 12, pp. 10817-
10829, Dec. 2017.
[8] N. Abdelfatah, and G. Brahim, “A Novel Lithium Ion Battery
Autonomous Strategy Improvement Based on SVM-DTC for Urban
Electric Vehicle under Several Speeds Tests,” Majlesi Journal of
Electrical Engineering vol. 5, no. 4, Dec. 2011.
[9] M. A. M. Cheema, J. E. Fletcher, M. Farshadnia and M. F. Rahman,
“Sliding mode based combined speed and direct thrust force control of
linear permanent magnet synchronous motors with first-order plus
integral sliding condition,” IEEE Trans. Power Electron., vol. 34, no.
3, pp. 2526-2538, March 2019.
[10] X. Zhang, L. Sun, K. Zhao and L. Sun, “Nonlinear speed control for
PMSM system using sliding-mode control and disturbance
compensation techniques,” IEEE Trans. Power Electron., vol. 28, no.
3, pp. 1358-1365, March 2013.
[11] A. Boucheta, I. K. Bousserhane, A. Hazzab, P. Sicard, and M. K.
Fellah, “Speed Control of Linear Induction Motor using Sliding Mode
Controller Considering the End Effects,” Journal of Electrical
Engineering and Technology, vol. 7, no. 1, pp. 34 - 45, Jan. 2012.