RT24 UK | Finite Control Set-Model Predictive and AI-based Power Electronics Control Development using OPAL-RT

This presentation showcases two key applications of OPAL-RT's real-time simulation platform for validating advanced power electronics control systems. The first study introduces an FPGA-based threaded parallel searching algorithm that improves the efficiency of long-horizon finite-control-set model predictive control (FCS-MPC), achieving 40-45% faster calculation times for horizon lengths of 5. The second study demonstrates an AI-based approach using a compressed 1D Convolutional Neural Network (CNN) implemented on FPGA, which reduces hardware DSP resource requirements by 75% compared to conventional methods while maintaining performance, highlighting OPAL-RT's capabilities in simulating and validating complex control strategies for modern power electronics applications.