Converters for induction motors enhancing fault tolerance in matrix: a hybrid EOO–RERNN approach

This research presents a hybrid technique named EOO–RERNN, integrating the Eurasian oystercatcher optimizer (EOO) and Recalling enhanced recurrent neural network (RERNN), to enhance fault tolerance in Matrix converters (MCs) for Induction Motors (IMs). The proposed method assesses fault impacts, reconstructs healthy phases, manages switching frequency with Space vector modulation (SVM), and diagnoses faults to optimize switching states. Comparative analysis using MATLAB/Simulink shows a 1.1% reduction in torque ripple compared to existing methods like the Cuckoo Search Algorithm and Particle Swarm Optimization, demonstrating superior performance and improved motor reliability.