PID神经元网络多变量控制系统分析

PID神经元网络是一种新的多层前向神经元网络,其隐含层单元分别为比例(P)、积分(I)、微分(D)单元,各层神经元个数、连接方式、连接权初值是按PID控制规律的基本原则确定的,它可以用于多变量系统的解耦控制.给出了PID神经元网络的结构形式和计算方法,从理论上证明了PID神经元网络多变量控制系统的收敛件和稳定性,通过计算机仿真证明了PID神经元网络具有良好的自学习和自适应解耦控制性能.

ANALYSIS OF PID NEURAL NETWORK MULTIVARIABLE CONTROL SYSTEMS

PID neural network is a new kind of feedforward multi layer network. Its hidden layer neurons are proportional neuron (P), integral neuron (I) and derivative(D) neuron. The numbers of the neurons, the connective forms and primary value of the weights are based on the rules of the PID control. The stabilization and convergence of the PID neural network multivariable control system are theoretically proved. Computer simulation displays the perfect self study and adaptive decoupling control properties of the PID neural network.