BLDC Motor Control For EVs Using Cuckoo Search Algorithm In PI

Controller Tunning

Arul P

1 a

, Ananda M.H

2 b

, Dhamodharan Selvaraj

2 c

and Dhanalakshmi R

3 d

Electrical and Electronics Engineering Department, Kongunadu College of Engineering and Technology , Tamil Nadu,

India.

School of Electrical and Electronics Engineering, REVA University, India

Department of Electrical and Electronics Engineering, Dayananda sagar college of engineering, Bangalore, India

Keywords: CSA, PI tuning, EV, BLDC motor

Abstract: electric vehicles (EV) are the present and future technology to overcome the environment problems created

by the traditional fossil fuel-based engines. Compact, high torque machines are needed for changing the future

of vehicles. The Brushless DC Motors (BLDC) are ones which satisfies the requirement. The BLDC motor

speed control is creating more importance in EV industry as it is used in many applications now-a-days.

Optimal control of BLDC motor is much need to make the EV more efficient and less consuming ones. DTC

control has a PI controller to produce proportional torque reference. If the PI controller parameters like K

and K

values are changed arbitrarily the torque ripple and settling time of speed are changing. So, there is

numerous combinations of K

and K

parameters are available. In this paper the problem is defined with multi-

objective. Identification of Kp and Ki parameters is done by minimizing toque ripple and settling time. The

objective function is solved by cuckoo search algorithm and results are discussed and compared with manual

tuning of PI controller, PSO based PI tuning and CSA based tuning.

1 INTRODUCTION

Many concepts are discussing about torque ripple

minimization of Direct Torque Control (DTC) of

Brushless DC Motor (BLDC) by changing the

switching pattern of pulse width. In 2004, ripple in

the torque is identified due to BLDC motor power

circuit commutation (Song & Choy, 2004). In 2005

torque ripple is reduced due to implementation of

space vector change (Liu et al., 2005). In 2010 stator

current improvement and torque ripple reduction is

implemented using model predictive control (Li &

Cheng, 2010). In 2015, by implementing DTC control

to BLDC motor torque ripple is reduced (Mahalingam

& Ramji, 2022). In 2012, a new PWM scheme is

proposed to eliminated the torque ripple caused due

to commutation by using the non-ideal back EMF

(Devi et al., 2017) is implemented. Using

multilevel inverter and a current controller an attempt

https://orcid.org/0000-0002-5927-6023

https://orcid.org/0000-0001-6026-3708

https://orcid.org/0000-0001-5946-2427

https://orcid.org/0000-0002-3317-6673

made to minimize the torque ripple (Mahalingam &

Ramji, 2022) In 2017 repetitivecontrol is used to

minimize the torque ripple (Devi et al., 2017) and by

using adaptive input-output feedback linearization

also a literature is proposed (Fang et al., 2012)

Speed response improvement of BLDC motor

is made in many attempts. In 2016, fuzzy PID

controller is used to improve the speed response

(Varshney et al., 2017). And with only fuzzy

implementation also done for improving speed

response (Geetha & Thangavel, 2016)

Fig. 1 presents an idea for the control of speed

in electric vehicles. The direct torque control (DTC),

which was initially created for the control of

induction motor drives in which direct control of flux

and electromagnetic torque was attempted, evolved as

a solution to these challenges. DTC was initially

developed for the control of induction motor drives.

It made use of the estimated flux and electromagnetic

torque to determine optimal inverter switching, which

enabled it to acquire quick response times. Because

of the non-sinusoidal back-EMF, the DTC of a BLDC

motor is different from that of an induction motor and

P, A., M H, A., Selvaraj, D. and R, D.

BLDC Motor Control For EVs Using Cuckoo Search Algorithm In PI Controller Tunning.

DOI: 10.5220/0012508600003808

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Intelligent and Sustainable Power and Energy Systems (ISPES 2023), pages 80-83

ISBN: 978-989-758-689-7

a PMSM motor. In order to lessen the effects of

torque ripple, the hybrid conduction mode was

Figure.1 Proposed Block Diagram of the System.

developed on the basis of the features of BLDC. It is

essential that the estimated torque be correct in order

to achieve direct control of the torque. In this article,

we will discuss the direct self-controlled approach for

BLDC that makes use of stator flux linkage reference

with three phase conduction. The rounding effect in

phase back-EMF is one of the causes for the creation

of torque ripple, and the back-EMF is derived by

making use of shape functions in order to estimate

torque.

The hybrid PSO-IC algorithm for grid

connected PV power system with EV battery is

introduced for charging the EV vehicle (Ahmed et al.,

2020). And for roof top photovoltaic controlled with

new hybrid optimization technique also proposed for

battery charging purposes (Selvaraj & Rangasamy,

2022),.

This paper made an attempt to reduce the

torque ripple and minimization of speed settling time.

Multi-objective problem is formulated in this paper

and the PI controller output is controlled by changing

the K

and K

parameter. Here cuckoo search

algorithm is used to compare the performance after

and before optimization.

2 PROBLEM IDENTIFICATION

In order to manage the torque that the motor

produces, the DTC control utilizes a speed control

loop that is equipped with a PI controller. This loop

produces a proportional torque value that is used as

an electromagnetic torque reference. It can be

understood by considering the following:









  (1)

When applied to PI controller it can be defined as



























 



(2)

Therefore, the outcomes of the torque

measurements are entirely dependent on the Kp and

Ki values that the PI controller uses. The ripples in

the torque can be adjusted by making adjustments to

the values of the PI controller. Therefore, it is

formulated as an equation for discrete optimization,

with settling time functioning as an additional target.

Therefore, the problem is stated as the minimizing of

torque and settling time of speed of the DTC control,

and this is accomplished by taking into consideration

an arbitrary limit for the values of Kp and Ki.

3 OBJECTIVE FUNCTION

Torque ripple is a key factor that can easily be

the cause of vibration in permanent magnet motors as

well as mechanical noise. Torque ripple has a

negative impact on the control precision of BLDC

speed control systems since it is a big factor. The

dependability of the motor may be jeopardized as a

direct outcome of the situation, particularly in the

event that the issue is severe. Continuous application

of the motor torque is required in virtually every

circumstance. Suppressing ripple in the torque output

is a necessary step in the development of a high-

precision permanent magnet motor, which is why this

step is included in the design process. The oscillations

in the motor's torque, as well as the amount of time it

takes for those oscillations to settle, are shown as,

























 





















 



 





 

 





Constraints:













….(4)















BLDC Motor Control For EVs Using Cuckoo Search Algorithm In PI Controller Tunning

4 CUCKOO SEARCH ALGORITHM

BASED PI TUNING

Xin-She Yang is responsible for the

development of the mathematical model of the

cuckoo algorithm (CS). There is a class of algorithms

known as "nature inspired algorithms," and the

cuckoo method is a member of that class. The

approach was devised as a result of the observation of

the cuckoo's mating behavior, which served as its

inspiration.

Here the host nests are the Kp and Ki

parameters. The objective is the torque ripple and

speed settling time.

Procedure of CSA is given below

Step 1. Initialize N host nests (K

& K

) Xi (i=

1, 2,..n) and maximum number of iteration.

Step 2. (minimization of eq(3)) or cost

function (Fi) is evaluated. Cuckoo is selected random

basis with levy flights algorithm.

Step 3. Choose a nest among N nests and name

it as (j).

Step 4. Check old solution is less than new one

and replace j by new solution.

Step 5. Best nest(solutions) are saved.

Step 6. Rank the solutions and find the current

best.

Step 7. Do this for all the iterations

Step 8. Display the results.

Figure.2 Cuckoo search algorithm convergence graph

Figure.3 Speed graph of PSO-PI, CSA-PI and manual tuned

PI controller (PSO settling time is .12 sec; CSA settling

time is 0.0454sec; PI setting time is .25 sec)

Figure.4 Error torque of PSO-PI, CSA-PI and Manual tuned

PI controller (PSO-PI maximum torque ripple is 68 Nm;

CSA-PI maximum torque ripple is 35Nm; PI maximum

torque ripple is 68Nm)

Figure.5 Flux wave of PSO-PI, CSA-PI and manual tuned

PI (saffron is PI controller and blue is CSA-PI – where

saffron is visible and Blue is less visible)

5 DISCUSSIONS

Fig.2 shows the Cuckoo search algorithm

convergence graph where the fitness function is

reducing n every iteration. Fig.3 shows Speed graph

of PSO-PI, CSA-PI and manual tuned PI controller.

Here PSO settling time reaches 0.12 sec. CSA settling

time is 0.0454sec; PI setting time is .25 sec. Fig.4

shows Torque of PSO-PI, CSA-PI and Manual tuned

PI controller. Here PSO-PI maximum torque ripple is

65 Nm. CSA-PI maximum torque ripple is 35Nm; PI

maximum torque ripple is 68Nm. Fig.5 shows Flux

wave of PSO-PI, CSA-PI and manual tuned PI. Here

ISPES 2023 - International Conference on Intelligent and Sustainable Power and Energy Systems

also saffron is PI controller and blue is CSA-PI –

where yellow is visible and Blue (PI) magenta (PSO-

PI) is less visible. The identified in PSO is Kp=18 and

Ki = 49. CSA is K

= 2 and K

= 17.

Table 2: Comparison table

Torque

Ripple in Nm

Setting time

in sec

PSO-PI

0.12

CSA-PI

0.0454

Manual

tuning of PI

0.25

% of

improved

48.52

81.84

Table 2: BLDC motor parameters

6 CONCLUSIONS

The BLDC motor is modeled and the DTC control is

applied to control the speed and torque. Then the PI

controller parameters like K

and K

are optimized by

multi-objective for minimization of torque ripple and

settling time. Solution algorithm proves the better

results. It is tabulated in table I. The CSA based PI

controller performs better in both speed settling time

and torque ripple minimization.

REFERENCES

Ahmed, C. C., Cherkaoui, M., & Mokhlis, M. (2020).

PSO-SMC controller based GMPPT technique

for photovoltaic panel under partial shading

effect. International Journal of Intelligent

Engineering and Systems, 13(2).

https://doi.org/10.22266/ijies2020.0430.30

Devi, K. S., Dhanasekaran, R., & Muthulakshmi, S.

(2017). Improvement of speed control

performance in BLDC motor using fuzzy PID

controller. Proceedings of 2016 International

Conference on Advanced Communication

Control and Computing Technologies,

ICACCCT 2016.

https://doi.org/10.1109/ICACCCT.2016.78316

Fang, J., Li, H., & Han, B. (2012). Torque ripple

reduction in BLDC torque motor with nonideal

back EMF. IEEE Transactions on Power

Electronics, 27(11).

https://doi.org/10.1109/TPEL.2011.2176143

GeethaV & Thangavel, S. (2016). Performance

analysis of direct torque-controlled BLDC

motor using fuzzy logic. International Journal

of Power Electronics and Drive Systems, 7(1).

https://doi.org/10.11591/ijpeds.v7.i1.pp144-

151

Li, Z., & Cheng, S. (2010). Torque ripple reduction

in brushless DC motors based on model

predictive control. Proceedings - International

Conference on Electrical and Control

Engineering, ICECE 2010.

https://doi.org/10.1109/iCECE.2010.1095

Liu, Y., Zhu, Z. Q., & Howe, D. (2005). Direct torque

control of brushless DC drives with reduced

torque ripple. IEEE Transactions on Industry

Applications, 41(2).

https://doi.org/10.1109/TIA.2005.844853

Mahalingam, K., & Ramji, N. K. C. (2022). A

comparative analysis of torque ripple reduction

techniques for sensor BLDC drive.

International Journal of Power Electronics and

Drive Systems, 13(1).

https://doi.org/10.11591/ijpeds.v13.i1.pp122-

131

Selvaraj, D., & Rangasamy, D. (2022). Electric

vehicle charging using roof top photovoltaic

controlled with new hybrid optimization

technique. Indonesian Journal of Electrical

Engineering and Computer Science, 26(3).

https://doi.org/10.11591/ijeecs.v26.i3.pp1227-

1234

Song, J. H., & Choy, I. (2004). Commutation torque

ripple reduction in brushless DC motor drives

using a single DC current sensor. IEEE

Transactions on Power Electronics, 19(2).

https://doi.org/10.1109/TPEL.2003.823177

Varshney, A., Gupta, D., & Dwivedi, B. (2017).

Speed response of brushless DC motor using

fuzzy PID controller under varying load

condition. Journal of Electrical Systems and

Information Technology, 4(2).

https://doi.org/10.1016/j.jesit.2016.12.014

BLDC Motor Control For EVs Using Cuckoo Search Algorithm In PI Controller Tunning