Modified ADRC Design of Permanent Magnet Synchronous Motor Based on Improved Memetic Algorithm

Area of Science:

• Electrical Engineering
• Control Systems
• Artificial Intelligence

Background:

• Traditional auto-disturbance rejection control (ADRC) suffers from system ripple due to non-differentiable control functions.
• Optimizing ADRC parameters is crucial for effective permanent magnet synchronous motor (PMSM) control.

Purpose of the Study:

• To propose a novel modified ADRC design for PMSM speed control.
• To develop an improved memetic algorithm (IMA) for optimizing ADRC parameters.
• To reduce system ripple and enhance control precision.

Main Methods:

• A differentiable and smooth optimal control function was developed to modify the ADRC design, mitigating system ripple.
• An improved memetic algorithm (IMA) was proposed, incorporating adaptive nonlinear convergence factor, Gaussian mutation, enhanced learning mechanisms (competitive and opposition-based learning), and elite set maintenance.
• The proposed IMA was used to optimize the parameters of the modified ADRC.

Main Results:

• The modified ADRC with IMA demonstrated a significantly better optimization effect compared to fuzzy PI, traditional ADRC with genetic algorithm, and improved ADRC with moth-flame optimization.
• Experimental results on PMSM speed control cases validated the effectiveness of the proposed method.
• The novel control function effectively weakened system ripple.

Conclusions:

• The proposed modified ADRC based on IMA offers superior performance for PMSM speed control.
• The developed IMA provides high-quality parameter optimization with improved precision.
• This approach effectively addresses the limitations of traditional ADRC and other optimization algorithms.