Comparative analysis of PID and neural network controllers for improving starting torque of wound rotor induction motor

TELKOMNIKA Telecommunication, Computing, Electronics and Control
Vol. 18, No. 6, December 2020, pp. 3142~3154
ISSN: 1693-6930, accredited First Grade by Kemenristekdikti, Decree No: 21/E/KPT/2018
DOI: 10.12928/TELKOMNIKA.v18i6.14571  3142
Journal homepage: http://journal.uad.ac.id/index.php/TELKOMNIKA
Comparative analysis of PID and neural network controllers for
improving starting torque of wound rotor induction motor
Hashmia Sh. Dakheel1
, Zainab B. Abdulla2
, Helen Jasim Jawad3
, Ali Jasim Mohammed4
1,2
Department of Electromechanical, University of Technology Baghdad, Iraq
3
Department of Electrical, University of Technology Baghdad, Iraq
4
Ministry of education/Directorate of vocational education, India
Article Info ABSTRACT
Article history:
Received Nov 11, 2019
Revised Jun 21, 2020
Accepted Jul 9, 2020
Unlike 3-phase squirrel cage induction motor, starting-up of 3-phase wound rotor
counterpart can be improved by adding an external resistance to the rotor circuit.
Thus, leads to reduce starting current and increase starting torque. In this paper
two controllers for 3-phase wound rotor induction motor have been proposed
include conventional proportional integral derivative (PID) controller and the
other based on artificial neural network (NARMA-L2). A comparison between
these controllers has been conducted. It has been shown that starting torque of
the motor has been improved, when utilizing the neural network controller
compared to the conventional counterpart. It should be noted that
MATLAB/SIMULINK has been used to implement both controllers.
Keywords:
External resistance (Rext)
NARMA-L2
PID controller
Starting torque (Ts)
Wound rotor induction motor
This is an open access article under the CC BY-SA license.
Corresponding Author:
Hashmina Sh. Dakheel,
Department of Electromechanical,
University of Technology,
168 Al-Sinna’ street, 10066, Baghdad, Iraq.
Email: hashimia.shr@yahoo.com
1. INTRODUCTION
The induction motors (IM) are greater used than 60% of total electric strength consumed worldwide [1].
IM have unique advantages such as slow maintenance, high robustness, and low cost which makes it to be suitable
for industrial applications [2-5]. Wound-rotor induction motor (WRIM) has some specified applications that
requires a high torque, adjustable speed drives, high inertia loads, so in order to improve the performance of
WRIM use several methods of rotor control which include resistance and impedance [1]. The rotor of IM has
3-phase winding connected in star, and slip rings are connected to the rotor winding, these rings give
the possibility of connecting external resistance (Rext) adding to the rotor that helps in speed control of WRIM [6].
The authors of [7] present a new method for soft starting of IM in rotor circuit that represented in a parallel
combination of resistors, self-inductors and capacitors to obtain soft and higher starting torque (Ts) and starting
current is limited. In [8] this study presents different methods of starting performance for 3-phase wound-rotor
IM that include variable rotor resistance, online direct and double feed starting as well as compare them, this study
achieves high starting torque by using double feed starting. Study [9] refer to different starter connected with IM

TELKOMNIKA Telecommun Comput El Control 
Comparative analysis of PID and neural network controllers for improving... (Hashmina Sh. Dakheel)
3143
which represented in auto-transformer starter, and direct-on-line at starting by using MATLAB SIMULINK to
analysis starting current, torque and speed also compare them. This research used dynamic simulation of
speed-control and starting for wound-rotor of IM based on differential equations for angular velocity and flux
linkages [10] and comparison between developed model with experimental results in laboratory. In study [11]
show using chopper control resistance and feedback circuit to obtain constant speed and constant torque, the speed
for a given load torque may be varied by varying the rotor resistance. There are many types of controllers can be
utilized with the motors such as proportional integral derivative (PID) controller, these controllers are widely used
in industrial applications for control which have a reasonable performance and different plants.
Hence, the gain parameters of PID can be obtained by evolutionary algorithms [12]. Because of
parameters and non-linear behavior of IM, the process of control convergence has many problems at different
conditions of operating, thus chooses artificial intelligent controller (AIC) that represents the best one for IM
control, the use of artificial neural networks (ANN) for nonlinear system modeling and control has demonstrated
to be extremely successful because of their ability to learn the dynamics of the plant, robustness, inherent
approximation capability and a high degree of tolerance [13]. With the development in artificial intelligence
applications, neural network has been used in identification and control of linear and non-linear systems.
The main advantage of ANN based techniques over conventional techniques is non-algorithmic
parallel-distributed architecture for information processing that allows it to learn any complex input-output
mapping, ANN is extremely useful in the area of learning control and capability of learning by training data under
diverse operations conditions [14], therefore this paper includes study the effect of adding Rext to the rotor circuit
of WRIM in order to improve Ts that is represented one of the most characteristics of IM at starting operation
condition, as well as comparison between PID controller and ANN (NARMA-L2) at different values of these
resistance in order to improve and obtain the maximum and best Ts to develop performance of WRIM.
2. STARTING TORQUE
The starting torque is one of great importance in case of driving high-inertia loads that is proportional
to the rotor resistance. It's necessary to have an excessive resistance at some points of starting period, and
reduce the value of this resistance until put off from rotor circuit while motor reaches to steady state case so
that the starting current is usually limited in drivers in order to protect the winding from the over load current
and low voltages at starting and breaking cases of the motor [7]. Therefore, it is required to a starter in
the circuit of motor and reduced the applied voltage to the motor at starting condition [9].
3. STARTING WITH EXTERNAL RESISTANCE
In wound-rotor induction motor (WRIM) is additionally Rext to the rotor circuit for the duration at starting
as a result as shown in Figure 1 shows WRIM with rotor external resistance [15]. When Ts increase and rotor
current decrease, that is very important to obtain a maximum torque (Tmax) at starting condition, this resistance
proportional to the slip, when the resistance value is high, the slip will be increase instantly, for that reason it is
viable to attain "pull-out" torque even with low speeds [15, 16]. Whilst the motor reaches its rated speed as the Rext
is removed and when the motor operates at steady state condition, it behaves in the identical way as squirrel-cage
induction motor [16]. The value of rotor resistance specifies the torque consequently; when the rotor resistance is
increasing at a regular torquereasonsa proportionate increase in themotor slip with lower in speed ofrotor, because
of that, the rate for a given load torque can be varied by various rotor resistance, its illustrates as show in (1) and
to achieve high Ts, to obtain Tmax at stand still should be chosen the value of R2 is appropriately as shown in
Figure 2 indicate to mechanical characteristic at variable values of Rext [17].
Figure 1. WRIM with external resistance Figure 2. Mechanical characteristic of WRIM

 ISSN: 1693-6930
TELKOMNIKA Telecommun Comput El Control, Vol. 18, No. 6, December 2020: 3142 - 3154
3144
𝑇 =
𝐾𝑠 𝐸2
2 𝑅2
𝑅2
2+𝑆2 𝑋2
2 (1)
4. PID CONTROLLER
PID controllers consist of 3 fundamental control modes, which are proportional, integral and
derivative modes, these modes are extensively utilized in control system especially in industrial applications,
they are simple to implement and provide good performance. The process of this controller includes set points
and feedback to perform, as show in (2) refers to this process in domain time [18].
𝑈(𝑇) = 𝐾𝑝 𝑒(𝑡) + 𝐾𝑖 ∫ 𝑒(𝑡)𝑑𝑡 + 𝐾𝑑
𝑡
0
𝑑𝑒(𝑡)
𝑑𝑡
(2)
PID controller (proportional integral derivative controller) is widely used in industrial control system,
it is calculates an “error” value as the difference between the measured process variable and the desired set
point [19-21], a proportional controller may not give steady state error performance which is needed in
the system, an integral controller may give steady state error performance, but it slows a system down. So,
the addition of a derivative term helps to cure both of these problems, Figure 3 illustrates a basic block of PID
controller [22-24]. Figure 4 shows simulink model of WRIM that is proposed in this work without any
controllers, and Figure 5 refers to the model with PID controller and subsystem include the simulink model
of WRIM.
Figure 3. Basic block of PID controller
Figure 4. Simulink model of WRIM without controller

3145
Figure 5. Simulink model with PID controller
5. NEURAL NETWORK CONTROLLER
In the recent years, ANN controllers has given performance effectively to improve nonlinear systems.
ANN is implemented in many applications such as electric power and there are promising in dentification/control
system, the self-adapting and notable-rapid computing functions of ANN cause them to well desirable to deal
with nonlinearities [25, 26].This model consists of two nets, the first net implement is a controller and the other
net simulates a version of the plant, NARMA-L2 model can be used to model the plant previously cited, using
two distinct neural networks, it makes use of a nonlinear identification tool. The neuro controller is noted by
special names: nonlinear auto-regressive moving average and feedback linearization control, while the plant
model has a selected method (associate formulation), the plant model can be estimated by NARMA-l2 controller
when the same formula, therefore the object of NARMA-L2 is to convert the system from nonlinear to linear
dynamics system [27], when using the NARMA-L2 controller, the first step is identify the plant model, that is
included of two functions f and g [28], these functions are representing the past values for each the output (y)
and the control effort (u) determine by tapped delay lines (TDL), while the second step includes simply
the rearrangement ofthe two sub networks f and g trained offline, at the result the computation time is decrease
[29]. Figure 6 can be shown NARMA-L2 control structure, when using the NARMA-L2 controller, the first step
is identify the plant model, that is included of two functions f and g [28], these functions are representing
the past values for each the output (y) and the control effort (u) determine by tapped delay lines (TDL), while
the second step includes simply the rearrangement of the two sub networks f and g trained offline, at the result
the computation time is decrease [29]. Figure 7 shows the model which consists of WRIM with NARMA-L2
controller that includes two input and one output represents starting torque. This model consists of multi hidden
layers variation with the value of external resistance.
Figure 6. NARMA-L2 control structure
Figure 7. The model of WRIM with NARMA-L2

 ISSN: 1693-6930
3146
6. SIMULATION RESULT AND DISCUSSIONS
This part of paper discuss simulation results of WRIM using MATHLAB SIMULINK environment
which achieves real-time observation to obtain best torque at starting by adding (Rext) to rotor circuit with
values 0.05Ω, 0.1Ω and 0.15Ω because this resistance led to decrease starting current and therefore (Ts)
increase but up of this value of Rext the current began to increase and Ts decrease, also this paper use two
different control techniques such as conventional controller (PID) controller as well as intelligent controller
(ANN (NARMA-L2)) to improve Ts of motor that led to develop the performance of the motor. The parameters
of WRIM considered in this paper are summarized in the Table 1.
Table 1. Parameters of WRIM
Parameters Spesification
Phase 3-Phase
Frequency 50 HZ
Power 2.2 KW/3HP
Pole 6
Efficency % 81%
Speed 1000RPM
Voltage of stator 415 V
Winding Y-connection
Stator resistance 0.7384 Ω
Stator inductance 3 mH
Rotor resistance 0.7402 Ω
Rotor inductance 3 Mh
Mutual inductance 0.1241 H
Torque 14.85 N-m
Rotor Inertia 0.0343 Kg-m2
6.1. Simulation result with Rext only
This part of simulation results shows using external resistance adding to rotor circuit (star- connecting)
by using MathLab/simulink model in order to increase Ts of motor which leads to better performance, this part
takes three values of external resistance which are Rext (0.05Ω, 0.1Ω and 0.15 Ω), in this case starting torque
values increase with adding external resistance and equal to (722N.m, 1140N.m, and 1232 N.m) respectively,
Figure 8 shows simulation results of model without any controller also Table 2 indicate the values of Ts at these
different values of Rext.
(a) (b)
(c)
Figure 8. Simulation result without controller; (a) Ts at Rext = 0.05 Ω, (b) Ts at Rext = 0.1 Ω, (c) Ts at Rext = 0.15 Ω

3147
Table 2. Simulation result without controller
Rext (Ω) Ts (N.m)
0.05 722
0.1 1140
0.15 1232
6.2. Simulation results by using PID controller
This part of paper deals with using PID controller to improve (Ts) than the first case above, by using
this controller Ts increase to 1754 N-m When Rext = 0.05 Ω while the values of constant gains of PID controller
which represented in proportional gain (KP=3), integral gain (Ki = 1) and derivative gain (Kd = 0.001) also
the value of Ts equal to 1840N.m when Rext = 0.1 Ω and 2080N.m at Rext = 0.15 Ω, Figure 9 shows these results
by using PID controller. Table 3 indicates to the values of constant gains for PID controller and Ts for each
value of Rext.
(a) (b)
(c)
Figure 9. Simulation results with PID controller, (a) Ts at Rext = 0.05 Ω, (b) Ts at Rex t = 0.1 Ω, (c) Ts at Rext = 0.15 Ω
Table 3. Constant gains of PID and Ts
Rext (Ω) Gains of PID Ts (N.m)
0.05 KP=3, Ki=1, Kd=0.001 1754
0.1 KP=5, Ki=1, Kd=0.001 1840
0.15 K P=3, Ki=2, Kd=0.001 2080
6.3. Simulation result by using NARMA-L2
This part of paper discuss using artificial neural network (NARMA-L2), this intellegrnt technique
achieves improving Ts than two other previous cases by obtaining higher Ts in order to develop performance
of WRIM, this network is controlled by network architecture which represented by number of delayed plant
input, number of delayed plant output and size of hidden layers, in addition to training data to get simulation
results. Also this part consists of three states include three different values of Rext (0.05Ω, 0.1Ω and 0.15 Ω)
and NARAMA-L2 is trained at each external resistance value to obtain highest starting torque.

 ISSN: 1693-6930
3148
6.3.1. Simulation result at Rext = 0.05 Ω
This part refers to simulation results of the model by using NARMA-L2, the value of starting torque
which equal to 1940 N.m and number of delayed plant input and output of NARMA-L2 that equal to 4 and 3
respectively while size of hidden layers = 13, sampling interval = 0.001, training samples = 3000, maximum
plant input = 4, minimum plant input = -1, minimum plant output = 0.2, maximum interval value =1, minimum
interval value = 0.1, Figure 10 shows these simulation results include neural network, plant identification-
NARMA-L2 in (a), (b) while (c) indicate to Ts , as well as (d), (e), and (f) show each of training, testing and
validation data for NARMA-L2, wheras (g), (h), and (i) show neural network training for each performance,
state and regression.
(a) (b)
(c) (d)
Figure 10. Simulation results by using NARMA-L2 at R ext = 0.05 Ω; (a) neural network,
(b) plant identification-NARMA-L2, (c) Ts = 1940 N.m, (d) training data

3149
(e) (f)
(g) (h)
(i)
Figure 10. Simulation results by using NARMA-L2 at R ext = 0.05 Ω; (e) testing data, (f) validation data,
(g) training performance, (h) training state, (i) training regression (continue)

 ISSN: 1693-6930
3150
6.3.2. Simulation results at Rext = 0.1 Ω
In this part simulation results show when Rext increases to 0.1 Ω and by training NARMA-L2, we get
higher starting torque compared to previous case when R ext = 0.05 Ω and the value of starting torque reaches
2150 N.m, Figure 11 shows simulation result of model, (a) refers to plant identification-NARMA-L2 with
number of both delayed plant input and output of NARMA-L2 which equal to 2,3 respectively, whereas size
of hidden layers = 8, (b) shows the value of Ts, also (c), (d), (e) refer to data for NARMA-L2, as well as (f),
(g), and (h) show network training neural for each performance, state and regression.
6.3.3. Simulation results at Rext = 0.15 Ω
Figure 12 refers to simulation results which indicate to increase Ts to 2200N.m and number of delayed
plant input and output = 4.3 respectively while size of hidden layer = 9, in this part we obtain higher and best
value of Ts than two other cases. Table 4 shows all simulation results by using NARMA-L2 include Ts, number
of delayed plant input and output as well as size of hidden layer for each case. Table 5 shows the comparison
of simulation results by using (Rext) adding to rotor circuit of WRIM at three cases without controller and with
two techniques include conventional PID controller and artificial neural network (NARMA-L2), these results
indicate to NARMA-L2 controller gives higher value of Ts than two other cases under the same value of Rext.
From above we conclude one of the ways to increase Ts in WRIM is adding Rext to rotor circuit by decrease
stating current in order to improve the performance of the motor.
(a) (b)
(c) (d)
Figure 11. Simulation results by using NARMA-L2 at Rext = 0.1 Ω; (a) plant of NARMA,
(b) Ts = 1940 N.m, (c) training data, (d) testing data

3151
(e) (f)
(g) (h)
Figure 11. Simulation results by using NARMA-L2 at Rext = 0.1 Ω; (e) validation data,
(f) training performance, (g) training state, (h) training regression (continue)
(a) (b)
Figure 12. Simulation results by using NARMA-L2 at Rext = 0.15 Ω;
(a) plant of NARMA, (b) Ts = 2150 N.m

 ISSN: 1693-6930
3152
(c) (d)
(e) (f)
(g) (h)
Figure 12. Simulation results by using NARMA-L2 at Rex t = 0.15 Ω; (c) training data, (d) testing data
(e) validation data, (f) training performance, (g) training state, (h) training regression (continue)

3153
Table 4. Simulation results with NARMA-L2
Rext (Ω) Parameters of NARMA-L2 plant Ts (N.m)
0.05 I/P = 4, O/P = 3
Hidden layers = 13
1940
0.1 I/P = 2, O/P = 3
Hidden layers = 8
2150
0.15 I/P = 4, O/P = 3
Hidden layers = 9,
2200
Table 5. Comparison simulation results
Rext(Ω) TS (N.m) without controller. Ts (N.m) by using PID Ts (N.m) by using NARMA-L2
0.05 722 1754 1940
0.1 1140 1840 2150
0.15 1232 2080 2200
7. CONCLUSION
This paper proposes adding external resistance to the rotor circuit of 3-phase wound rotor induction
motor with three different values (0.05 Ω, 0.1 Ω, 0.15 Ω), this model uses conventional PID controller and
artificial neural netwok (NARMA-L2) and training this network to obtain the best starting torque at each value
of external resistance. A comparative study between PID controller and (NARMA-L2) has presented to develop
performance of the motor at starting. The simulation results has demonstrated a successful implementation of
adaptive neural network (NARMA-L2) which gives better results than PID controller by obtaining a highest
starting torque and achive improving performance of WRIM.
REFERENCES
[1] K. Ranjith Kumar, S. Palaniswami and G. Lourds Shiny, “Performance enhancement of wound rotor induction motor
by rotor capacitive reactance control using fuzzy controller,” International Journal of Electrical Engineering. vol. 3,
no.9, pp.197-207, July 2012.
[2] S. Chandran, “Rotor resistance adaptation scheme using neural learning algorithm for fuzzy logic based sensor less
vector control of induction motor,” IOSR Journal of Electrical and Electronics Engineering, vol. 8, no. 1,
pp. 13-21, 2013.
[3] Hanan Mikhael D., et al., “Speed control of induction motor using PI and V/F scalar vector controllers,” International
Journal of Computer Applications, vol. 151, no.7, pp. 36-43, October 2016.
[4] Z. Li, et al., “Rotor speed and stator resistance identification scheme for senseless induction motor drive,”
TELKOMNIKA Telecommunication Computing Electronics and Control, vol 11, no.1, pp. 503-512, 2013.
[5] C. T. Pham, A. W. Shen, “A comparative study of control methods for induction motor and high performance
Z-source inverter,” TELKOMNIKA Telecommunication Computing Electronics and Control, vol 11, no.6,
pp. 2912-2925, June 2013.
[6] V. V. Maruti Prasad, W. Prasanna Narasimham, G Sai Raghu Ram, “Rotor Side control of high power slip ring
induction motor using single thyristor,” IOSR Journal of Electrical and Electronics Engineering, vol.9, no. 2,
pp. 49-53, March-April 2014.
[7] M. B. Bannae Sharifian, et al., “A new soft starting method for wound-rotor induction motor,” Journal of Electrical
Engineering, vol.62, no.1, pp.31-36, January 2011.
[8] R. M. Hamouda, A. I. Alolah, M. A. Bader and M. A. Abdel-Halim, “A Comparative Study on the Starting Methods
of Three Phase Wound-Rotor Induction Motors,” IEEE Transactions on Energy Conversion, vol.14, no. 4,
December 1999.
[9] D. D. Dhawale, et al., “Analysis of different starting methods of induction motor,” IJARIIE, vol.3, no. 2,
pp. 2484-2494, 2017.
[10] S. AL-Jufout, Kamal Khandakji, “Dynamic simulation of starting and chopper speed control of wound-rotor
induction motor,” International Journal of Simulation Systems Science & Technology, vol. 8, no 2, July 2007.
[11] B. M. Frederic, et al.,"The control of induction electromotor by the resistance change in rotor circuit,” American
Journal of Engineering Research (AJER), vol.8, no. 7, pp-169-175, 2019.
[12] R. Essakiraj, et al., “Speed control of induction machine using GA based PID controller,” Middle- East Journal of
Scientific Research 23(Sensing, Signal Processing and Secuirty), pp.164-169, 2015.
[13] M. George, K. P. Basu, “NARMA-L2 controlled variable frequency three-phase induction motor drive,” European
Journal of Scientific Research, vol. 70, no. 1, pp. 98-111, February 2012.
[14] H. Sh. Dakheel, “Speed control of separately DC motor using artificial neural network,” ournal of Engineering and
Development, Vol.16, No.4, pp.349-362, Dec. 2012.
[15] R. Kumar, R. Dogra and P. Aggarwal, “Rotor side speed control methods using MATLAB/Simulink for
wound induction motor,” International Journal of Mechanical and Mechatronics Engineering, vol.11, no. 7,
pp. 1378-1386, 2017.
[16] M. A. Abdel-Halim, et al., “Smooth starting of slip ring induction motors,” IEEE Transactions on Energy
Conversion, Vol.12, No.4, December 1997.

 ISSN: 1693-6930
3154
[17] C. Sooji, et al., “MATHLAB/simulink based design and simulation of speed control of three phase slip ring induction
motor using chopper,” International Journal of science, Engineering and Technology Research (IJSETR), vol. 5,
no. 1, pp.267-274, January 2016.
[18] K. Devi, S. Gautam and D. Nagaria, “Speed control of 3-phase induction motor using self tuning fuzzy PID controller
and conventional PID controller,” International Journal of Information and Computation Technology, vol.4, no.12,
pp.1185-1193, 2014.
[19] S. W. Shneen, et al., “Biogeography based optimization tuned fuzzy logic controller to adjust speed
of electric vehicle,” Indonesian Journal of Electrical Engineering and ComputerScience (IJEECS), vol.16, no. 3,
pp. 509-519, December 2015.
[20] M. L. Mhaisgawali and S. P. Muley, “Speed control of induction motor using PI and PID controller,” IOSR Journal
of Engineering (IOSRJEN), vol.3, no. 5, pp25-30, May 2013.
[21] M. M. Abdullah Alqadasi, et al., “Optimization of PID for industrial electro-hydraulic actuator using PSOGSA,”
TELKOMNIKA Telecommunication Computing Electronics and Control, vol.17, no.5, pp. 2625-2635. October 2019.
[22] A. J. Attiya, et al., “Compared with PI, Fuzzy _PI & PSO_PI controllers of robotic grinding force servo system,”
TELKOMNIKA Telecommunication Computing Electronics and Control, vol.16, no.1, pp. 65-74, October 2015.
[23] A. Budianto, et al., “PID control design for biofuel furnace using Arduino,” TELKOMNIKA Telecommunication
Computing Electronics and Control, vol. 18, no.6, pp.3016-3023, December 2018.
[24] B. Ahmed Obaid, et al., “Resolving of optimal fractional PID controller for DC motor drive based on
anti-windup by invasive weed optimization technique,” Indonesian Journal of Electrical Engineering and Computer
Science, vol. 15, no.1, pp. 95-103, July 2019.
[25] S. W. Shneen and Chengxiong Mao, “Artificial optimal fuzzy control strategy for elevator drive system by using
permanent magnet synchronous motor,” TELKOMNIKA Telecommunication Computing Electronics and Control,
vol.14, no.3, pp. 470- 480, June 2015.
[26] S. Arun Balaji and K. Baskaran, “Design and development of artificial neural networking (ANN) system using
sigmoid activation function to predict annual rice production in tamilnadu,” International Journal of Computer
Science, Engineering and Information Technology (IJCSEIT), vol.3, no.1, February 2013.
[27] Abdulrahim T. Humod, et al., “Direct torque control for permanent magnet synchronous,” Eng. &Tech. Journal,
vol. 34, no.3, pp.464-482, 2016.
[28] I. Iulia and A. Mihail, “Design of NARMA-L2 neural controller for an induction motors speed control,” Journal of
Computer Science and Control Systems, vol.6, no.2, pp. 18-21, October 2013.
[29] K. Srakaew, et al., “Design of NARMA-L2 neurocontroller for nonlinear dynamical system,” 29th International
Conference on Modeling Identification, and Control, pp. 210-215, February 2010.

Comparative analysis of PID and neural network controllers for improving starting torque of wound rotor induction motor

More Related Content

What's hot (20)

Similar to Comparative analysis of PID and neural network controllers for improving starting torque of wound rotor induction motor (20)

More from TELKOMNIKA JOURNAL (20)

Recently uploaded (20)

Comparative analysis of PID and neural network controllers for improving starting torque of wound rotor induction motor