车用电机定子电流新型优化控制方法  

A New Optimal Control Method for Stator Current of Vehicle Motor

作　　者：周雅夫 王翰涛 连静 ZHOU Ya-fu;WANG Han-tao;LIAN Jing(School of Faculty of Vehicle Engineering and Mechanics Automotive Engineering Dalian University of Technology,Liaoning Dalian 116024,China)

机构地区：[1]大连理工大学运载工程与力学学部汽车工程学院,辽宁大连116024

出　　处：《机械设计与制造》2023年第3期230-233,238,共5页Machinery Design & Manufacture

基　　金：大连市科技创新基金项目(2018J12GX061)。

摘　　要：为了实现交直轴电流最优匹配控制,提出了一种基于深度确定性策略梯度的车用电机定子电流优化控制方法。在不对交直轴耦合项电压进行补偿解耦的情况下实现IPMSM交直轴电流的最优匹配控制。实验分析表明基于DDPG的定子电流优化控制方法在优化交直轴电流匹配和最大转矩电流比控制方面优势显著。相比较传统的前馈补偿解耦(FFCDC)方法实现MTPA控制,同等定子电流产生转矩输出平均高出最高14%,实验验证进一步证明了DDPG具有优化交直轴电流最佳匹配和产生最大转矩输出的能力,验证了该方法的有效性。In order to realize optimal matching control of d-q axis current,a novelNon-Decoupling Control approach based on Deep Deterministic Policy Gradient(DDPG)is proposed. The optimal matching control of the IPMSM d-q axis current is achieved without compensating and decoupling the d-q axis coupling term voltage. The experiment results show that the DDPG-based stator current optimization control method has significant advantages in optimizing the d-q axis current matching and improve MTPA output. Compared with feed-forward compensation decoupling control(FFCDC)to achieve MTPA control during equivalent stator current command,the proposed scheme generated torque output about 14% higher than the FFCDC. The experimental verification further proves that DDPG can optimize the d-q axis current matching and the maximum torque output,which validates the effectiveness of the proposed method.

关 键 词：内置式永磁同步电机 最大转矩电流比 深度确定性策略梯度 交直轴电流最优匹配 定子电流优化控制 

分 类 号：TH16[机械工程—机械制造及自动化] TP29[自动化与计算机技术—检测技术与自动化装置] U463.233[自动化与计算机技术—控制科学与工程]