Innovative Application of Deep Learning Driven Active Disturbance Rejection Control Technology in Motor Speed Regulation

Abstract

With the increasing complexity of motor speed control system, the traditional active disturbance rejection control method has performance limitations when dealing with nonlinear, time-varying and strong coupling systems. This paper studies the combination of deep learning and active disturbance rejection control technology, and proposes an active disturbance rejection control algorithm based on deep learning, which can adaptively optimize system parameters and improve the suppression ability of complex disturbances. The design and implementation of the algorithm in motor speed regulation combines the nonlinear feature extraction ability of deep learning and the fast response characteristics of active disturbance rejection control. The experimental results show that the active disturbance rejection control system based on deep learning has improved the motor speed regulation accuracy, dynamic response speed and anti-interference ability, and shows better robustness and stability especially in complex working conditions.