Real time implementation for robust adaptive sliding mode predictive control of PMSM

Model Predictive Control is recognized as a promising approach for electric drives, with particular interest in the development of robust and high-performance predictive models. In this work, we propose an Adaptive Integral Sliding Mode Predictive Control for Permanent Magnet Synchronous Motors, combining a predictive model with an adaptive law based on the integral sliding mode. This method automatically adjusts the sliding function limit, reducing the chattering phenomenon while enhancing the system dynamics. To ensure compatibility with industrial electronic boards, the Adaptive Integral Sliding Mode Predictive Control was implemented following the V-cycle development process, including Model-in-the-Loop, Software-in-the-Loop, and Processor-in-the-Loop validation. This framework facilitates the deployment of embedded control software in the automotive sector and provides a cost-effective evaluation of the hardware implementation. Furthermore, real-time simulations of control, Invariant Sliding Mode Predictive Control, and Sliding Mode Control configurations were carried out on the dSPACE DS1104 platform, showing excellent correlation with MATLAB/Simulink results. Experimental validation on the STM32F4 board confirms that the proposed approach offers faster response to load torque disturbances and better performance over a wide speed range, demonstrating its reliability, robustness, and effectiveness. • Proposes an Adaptive Integral Sliding Mode Predictive Control (AISMPC) for speed and current control of PMSM drives. • Develops a robust continuous predictive model derived directly from the integral sliding mode surface, improving performance and robustness. • Introduces a self-adjustment mechanism for the sliding mode boundary to reduce chattering and enhance system dynamics. • Applies the V-model development process, including MIL, SIL, and PIL validation, targeting automotive and aerospace applications. • Demonstrates successful embedded implementation on an STM32F4 platform, with experimental results confirming accurate tracking and high control performance.