摘要
基于梯度学习算法，提出一种具有自适应功能的ＰＩＤ控制器．它采用一种改进型神经网络，在线学习相控整流直流传动系统的动力学特性，继而给控制器提供必要的灵敏度信息以实现自适应补偿控制．从而提高整个控制系统在参数变化及受到负载扰动时的鲁棒性．改进型神经网络由两个子神经网络综合构成：一个是应用递推最小二乘算法的线性神经网络；一个是应用动态ＢＰ算法的动态递归网络．对基于相控整流直流传动系统的自适应控制实现策略的仿真结果表明，上述自适应控制器性能好于传统的ＰＩＤ控制器．
A PIDlike controller based on gradient descent learning algorithm is preserted. A PIDlike cost function is proposed, and it can be proved that the PID algorithm is the gradient descent method as long as the cost function is selected as ours. The controller's structure and the learning algorithm are very simple and easy to be realized. A modified Neural network (MNN) is presented to learn the characteristics of the dynamic system for the online solution of the socalled sensitivity. The proposed controller is then applied to a speed control system that uses a phasecontrolled bridge converter and a separately excited DC (Direct Current) machine, instead of the conventional proportionalintegral(PI)control method. The simulation study indicates the superiority of the novel controller over the conventional PI controller, and shows that Neural Network control technique seems to have a lot of promise in the applications of power electronics.
出处
《东北大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
1998年第1期36-39,共4页
Journal of Northeastern University(Natural Science)
基金
辽宁省自然科学基金
关键词
直流电机
调速系统
神经网络
DC drive control,neural network control,intelligent control.
