Task-Oriented Enhanced Self-Learning Speed Synchronization Control of Flexible Coupled Dual-Motor Systems

Authors:
Publisher website
Google Scholar
Abstract
In this article, a task-oriented enhanced self-learning speed synchronization control method is proposed for flexible coupled dual-motor (FCDM) systems with unknown model parameters. Firstly, the two-time-scale characteristics of the FCDM system are analyzed. Using singular perturbation theory (SPT), the original system is turned into a reduced-order model. Secondly, aiming at the challenge that the state of the reduced-order system cannot be directly measured, an innovative state reconstruction mechanism is proposed, and the HJB equation is solved by reinforcement learning (RL) algorithm. The actor-critic structure is introduced to approximate the control policy and cost function respectively. Thirdly, in order to enhance the quality of training data and ensure the efficient execution of the algorithm, a variational auto encoder (VAE) is introduced to enrich the sparse data points close to the reference rotational speed. Finally, experiments are carried out on the semi-physical simulation platform based on dSPACE to verify the effectiveness of the proposed method.
Funding Information
  • Key Research and Development Projects in Shanxi Province (202202100401002)
  • National Natural Science Foundation of China (62273350, 62303468, 62303469)
  • Natural Science Foundation of Jiangsu Province (BK20221112, BK20221116)

This publication has 18 references indexed in Scilit: