HIL simulation of AI-based sensorless control of induction motors with dual-field orientation

This paper presents real-time hardware-in-the-loop (HIL) testing of an artificial-intelligence- based sensor less control strategy for an induction motor using dual-field orientation. The experimental setup includes a PLECS RT Box I, used as the HIL simulator for the three-phase PWM inverter and the induction machine, and an NI PXIe-1071 system serving as the real-time target computer. The control algorithms are developed and implemented in the MATLAB/Simulink environment. Rotor-speed estimation is carried out using classical recurrent neural networks (RNNs) as well as a hybrid RNN-PI approach that combines RNNs with a conventional PI controller. Both estimation structures operate as deep learning models designed to improve the robustness and accuracy of rotor-speed estimation across a wide speed range and under various load conditions.