基于小波变换的电液伺服试验多变量系统神经网络控制

针对电液伺服试验系统的复杂非线性和不确定性特性．提出一种基于小波变换的“主控制器”结合“参征器”的复合控制策略，并用于电液伺服试验系统的多变量控制。主控制器由一个包含PID控制规则的神经网络构成，在整个系统控制中起着主导作用；“参征器”的作用是抑制干扰，保证系统响应的快速性。仿真结果证明，该方法具有良好的自学习和自适应解耦控制性能，能有效地提高系统的稳态精度，使系统具有较强鲁棒性。并具有响应速度快、超调量小等特点。

Neuron Network Control of Electro-Hydraulic Servo Testing Multi Variable System Based on Wavelet

As for complicated characteristics of Electro-hydraulic Servo testing system,such as non-linearity and undefined,a composite controller algorithm of using a host controller com-bined with join-oscillator based on wavelet transform is present-ed in this paper.The host controller is made up of neuron nets of containing PID control rules,it plays a leading role in the whole system controlling;join-oscillator takes affect of restraining interference and ensures speedy of system responding.Simulation and testing show that this algorithm has good self study and adaptive decoupling control properties.It has good performance,small overshoot and can reduce the static error effectively.