Artificial intelligence and PI gain optimisation for sensorless indirect vector control of induction motor-based electric vehicle drives

Abstract

This paper aims to develop indirect vector-controlled induction motor (IM) drives for EV applications. In this paper, controllers are used in two places. The first is the speed controller, used to reduce the error between reference speed and actual EV speed. And the second one is the speed estimator used to reduce the error between the adaptive model and reference model during sensorless operation. The performance of these controllers is compared by using ANFIS, ANN and PI controllers. The real-time performance of the IM-based EV drive system is dependent on the PI gain value. Hence in this paper, seven human and nature-inspired metaheuristic optimisation techniques such as JFO, GWO, ABC, PSO, GA, TLBO and DE are used to obtain the optimised value of the PI controller.

KEYWORDS:

• Artificial neuro-fuzzy inference system (ANFIS)
• artificial neural network (ANN)
• electric vehicle (EV)
• optimisation technique
• PI gains
• induction motor

Disclosure statement

No potential conflict of interest was reported by the author(s).