Internal Model Decoupling Based on Single-neuron Adaptive PID

In order to solve the coupling problem in large delay system with more variables,a new decoupling algorithm with single neuron self–adaptive PID internal model is recommended,which is based on internal model control theory and the on-line self-learning ability of single neuron.Furthermore,the decoupling theory in large delay system with more variables is analyzed.Internal decoupling can tranfer a more in and more out system into a single in and single out system with more sub-systems by using decoupling predictive compensator,and turn the object model into diagonal predominance.After doing all the work,the principal diagonal elements can be regarded as the predictive model of internal control.Experimental results show that the decoupling capability of this new internal model algorithm is rather good.It also shows fast property and better anti-disturbance capability.