摘要
AdaptiveControlbyUsingNeuralNetworks¥(郝继红)(吕强)(段运波)(许耀铭)HAOJihong;LUQiang;DUANYunbo;XUYaoming(Dept.ofPowerEngineering,Harbini...
A kind of model reference adaptive control system has been made up by using Madaline neural networks. The neuron controller NN1 and its adaptation algorithm are proposed in this system.Neural networks identifier NN2 is adopted,it can approximate the plant's input/output mapping accurately, furthermore, as back-propagating path of the output error, it provided valid teacher signal to controller NN1. Applying the present algorithm to the control of linear and nonlinear systems in simulation, the result shows its validity.