Cascaded generalized predictive control for induction drives under constraints using particle swarm optimization

Abstract

This paper presents a cascaded generalized predictive control (CGPC) strategy for induction motor drives under operational constraints, optimized through particle swarm optimization (PSO). In the proposed scheme, the outer loop regulates the motor speed, while the inner loop controls torque and flux, ensuring accurate multi-level regulation. PSO is employed to optimally tune the prediction horizon and weighting factors, enhancing robustness, transient response, and disturbance rejection. Unlike conventional GPC–PSO approaches that neglect explicit constraint handling, and linear matrix inequalities (LMI)-based model predictive controller (MPC) methods that impose high computational costs, the proposed CGPC–PSO achieves both constraint management and real-time efficiency. Moreover, compared with Neural-MPC strategies that require retraining for each system, the proposed method provides generalizable and adaptive control without sacrificing computational performance. Simulation results validate the effectiveness of the approach, demonstrating superior trajectory tracking, robustness against parameter variations, and improved dynamic performance compared with RST, LMI, and neural-MPC controllers. These findings position CGPC–PSO as a promising candidate for advanced induction motor drive applications.