Parameter Self-tuning Of N-PID Controller

PID controller has been widely applied in all kinds of industrial process controldue to advantages of simple structure, easy implementation and stronger robustness. Asclassical PID frequently can't meet the control performance requirements for complexnon-linear system. N-PID is an effective way to solve the problem. But it has manydifficult adjustment of parameters. Therefore to explore an effective parameter tuningmethod so as to give full play to the advantage of N-PID has an important applicationvalue.This thesis first introduce the present commonly used several kind of parameterself-tuning methods and summarize the research on parameter self-tuning of domesticand foreign systems development; Next, build mathematical model of N-PID controller,take high order nonlinear system as the research object, on the basis of familiar withSVM principle and usage, control system's overshoots, adjusting time, rise time areconsidered as input samples, nine parameters as output samples, to train appropriatesupport vector machine (SVM) model, so as to realize the N-PID controller parametersoffline tuning. At last, compare relay feedback self-tuning to SVM self-tuning, resultsshow the method of parameter self-tuning of N-PID controller based on SVM has agood ability of self-learning, improves the precision and rate of algorithm, avoidsmanual tune, time saving and enhances system's stability and fast responsiveness.