Model Free Predictive Current Control Employing Neural Network with Jaya—Optimized Speed control Gain Parameter Tuning for Enhanced Drive Performance

This paper presents a data-driven current control strategy that eliminates the requirement for explicit plant models while efficiently reducing parameter mismatches. The proposed data-driven current control solution employs fewer input features, lowering the computational cost during training while maintaining control network performance. By utilising the inherent adaptability of data-driven methodologies, the given control solution improves overall control system performance by increasing robustness to changes in system parameters. Furthermore, the Jaya optimisation algorithm simplifies the optimisation of gain parameters for the speed control loop, resulting in improved dynamic performance and responsiveness. The combination of the model-free current control methodology and the Jaya optimisation algorithm produces a very effective control solution that exceeds existing methods. To verify the closed-loop control performance, the test results are presented compared with the conventional control design to show the superiority of the proposed appraoch.