Development of Efficient Energy Management Strategy to Mitigate Speed and Torque Ripples in SR Motor Through Adaptive Supervisory Self-Learning Technique for Electric Vehicles

Abstract

Switched reluctance motors (SRM) are receiving extensive attention from the industry because of their simple construction, high reliability and efficiency. Antithetically, controlling SRM is challenging due to its nonlinearities and variable parameters such as speed, torque, flux, etc. In this study, a novel approach is used to control the speed and torque of SRM with various controllers. This methodology integrates model-in-loop (MIL) and hardware-in-loop (HIL) simulations to evaluate the efficiency of SRM in electric vehicles (Evs). To reduce the speed variations in the SRM, PID, intelligent, hybrid, and adaptive supervisory self-learning control approaches (ASSC) are used. The mathematical models of the aforementioned controllers were created in MATLAB/Simulink, and these control approaches are used to reduce speed, torque, and current fluctuations under various load and speed conditions. In this paper, both simulation and experimental work are used to investigate the various dynamic characteristics of the SRM in Evs. From the simulation findings, the proposed ASSC controller exhibits less overshoot (1.05%), settling time (0.02s) and risetime (0.01s) than PID, intelligent and hybrid control approaches. The proposed ASSC method combines numerical data and real-time rules knowledge with ANN to predict error responses in the SR motor. Also, it is significantly reducing the speed, torque, current and flux ripples of the SRM from low to high speed with different load conditions. Further, to verify the simulation results experimental work is conducted and similar results are measured under different load and speed conditions. From experimentation, efficiency maps are developed for PID, intelligent, hybrid, and ASSCs across the entire operating range. The maps show that the controller's maximum efficiency is, respectively, 85%, 88%, 91%, and 95%. The supervisory controller is 10% more efficient than PID, intelligent, and hybrid controllers in terms of various dynamic characteristics of SRM. From the observations, both experimental and simulation results corroborated that the recommended ASSC improves the SRM efficiency.

Evaluation of Electric Vehicle Performance Characteristics for Adaptive Supervisory Self-Learning-Based SR Motor Energy Management Controller under Real-Time Driving Conditions

“…In the proposed work, using the Ziegler Nichols approach, the PID parameters are selected. When the SR motor dynamics are unknown or not readily available, the P, I, PI, and PID controllers are fine-tuned using the Ziegler Nichols approach [35]. Based on the different transient responses of the EV SR motor, the PID controller gains are estimated using the Ziegler Nichols rules.…”

Critical Review on Torque Ripple Sources and Mitigation Control Strategies of BLDC Motors in Electric Vehicle Applications

“…Various BLDC motor control strategy discussions explore different control schemes in terms of sensors or sensorless controllers to improve motor performance and address different power optimization issues in real-time [18,19]. Furthermore, the optimal control strategy to regulate torque ripples in various control methods [20][21][22][23][24][25][26][27][28][29][30] and the investigation of various commutation techniques and PWM schemes to control stator current in BLDC motors have been discussed [31,32], as mentioned in Fig 1 . According to the review, some control strategies have been developed to improve BLDC motor performance to meet vehicle performance requirements, such as driving range and energy optimization in real-time operating conditions.…”

“…are described by, Ea=Kb fa(λ) ω(21) Eb=Kb fb(λ) ω(22) Ec=Kb fc(λ) ω(23) Where, λ is rotor angle (electrical degree), ω is angular speed (rad/sec), Kb is the back EMF constant and fa,fb & fc are the trapezoidal signal functions. From the following back EMF equations, the trapezoidal signal functions can be expressed by the following equations (24,25& 26), ) = fa(λ-120 ▫ )…”

Prediction of Electric Vehicle Driving Range and Performance Characteristics: A Review on Analytical Modeling Strategies With Its Influential Factors and Improvisation Techniques

“…The simulation tools play a vital role in evaluating the EV performance in different scenarios by reducing time consumption, cost and human effort associated with realtime testing. However, the accuracy of simulation results is unfailing on the data quality of powertrain components and vehicle dynamics mathematical model [47][48]. The primary outputs of EV powertrain simulation are EC and DR. Other than this, the size, operating region and dynamic behaviours of powertrain components can be evaluated for selected DCs to know the limit of powertrain components.…”