基于BP神经网络PID算法的多电机同步控制研究

发布时间：2019-05-15 15:50

【摘要】：近年来随着工业自动化程度的不断提高,传动系统也提出了更加严苛的要求。传动控制系统在工农业生产中得到越来越广泛的应用,多电机控制的研究显得尤为重要。多电机同步控制精度的提高将直接带动工农业生产效率提升,带来极大的经济效应,因而多电机同步控制的研究越来越受到人们的关注。本文主要从多电机控制策略和多电机同步控制方式两个方面进行研究。首先,在多电机同步控制系统中需要单台电机具有很好的速度跟随特性,因而单台电机控制效果的好坏将直接制约着多电机同步控制系统的精度。工农业生产中主要是以传统PID控制器来提升控制器响应速率。传统PID控制器简单易实现,但是PID参数整定比较困难,对于多电机同步控制系统这样一个多变量、非线性、强耦合的控制对象而言显得捉襟见肘。因而,本文采用BP神经网络PID替代传统PID控制器,在深入研究过程中发现传统BP神经网络算法具有收敛速度慢、易陷入局部极小值等缺陷,提出引入惯性项、引入动量项、改进搜索方向、改进学习速率四点改进策略,重新设计基于改进BP神经网络的PID控制器,改善BP神经网络PID控制器性能。其次,在多电机同步控制系统中,多电机同步控制方式与多电机的同步控制精度也是关系密切。本文将并行控制、主从控制、交叉耦合控制和偏差耦合控制进行比较研究,得出偏差耦合控制能够较好的解决多电机同步控制问题,但是传统偏差耦合控制速度补偿器对于各台电机之间的同步误差修正不够快,反应不够灵敏。因而本文提出了三点改进速度补偿器的策略:引入速度信号补偿增益、引入误差因子、添加BP神经网络PID控制器。重新设计了偏差耦合控制速度补偿器,改善速度补偿器的不足。最后,在Matlab/Simulink环境下搭建多电机同步控制系统的仿真控制平台,进行仿真实验分析,从实验结果可以看出本文所做研究明显提高了多电机同步控制精度。
[Abstract]:In recent years, with the continuous improvement of the degree of industrial automation, the transmission system has made more demanding requirements. The transmission control system is more and more widely used in industrial and agricultural production, and the research of multi-motor control is particularly important. The improvement of multi-motor synchronous control precision will directly drive the efficiency of industrial and agricultural production and bring great economic effect, so the research of multi-motor synchronous control is more and more concerned. In this paper, the two aspects of multi-motor control strategy and multi-motor synchronous control are studied. First, in the multi-motor synchronous control system, a single motor is required to have a good speed-following characteristic, so that the control effect of a single motor will directly restrict the accuracy of the multi-motor synchronous control system. In the industrial and agricultural production, the response rate of the controller is improved by using the traditional PID controller. The traditional PID controller is simple and easy to implement, but the PID parameter setting is difficult. For a multi-variable, non-linear, strong-coupled control object of the multi-motor synchronous control system, it is difficult to see. In this paper, the BP neural network PID is used to replace the traditional PID controller. In the course of the in-depth study, it is found that the traditional BP neural network algorithm has the defects of slow convergence speed, easy to fall into local minimum, etc. The inertia term is introduced, the momentum term is introduced, and the search direction is improved. Improve that four-point improvement strategy of the learning rate, and re-design the PID controller based on the improved BP neural network to improve the performance of the BP neural network PID controller. Secondly, in the multi-motor synchronous control system, the synchronous control accuracy of the multi-motor synchronous control system and the multi-motor is also closely related. In this paper, the parallel control, the master-slave control, the cross-coupling control and the bias coupling control are compared and studied, and the problem of synchronous control of the multi-motor can be solved well by the deviation coupling control. But the conventional deviation coupling control speed compensator is not fast enough to correct the synchronous error between the motors, and the reaction is not sensitive enough. The strategy of three-point improved speed compensator is proposed in this paper. The speed signal compensation gain is introduced, the error factor is introduced, and the BP neural network PID controller is added. The deviation coupling control speed compensator is redesigned to improve the speed compensator. Finally, the simulation control platform of the multi-motor synchronous control system is set up in the environment of Matlab/ Simulink, and the simulation experiment is carried out.
【学位授予单位】：沈阳工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM301.2

【参考文献】