Research of Adjusted Smith Predictor Based on Immune Feedback

Abstract

Immune feedback Smith predictor is a model self-modified Smith predictor with immune feedback method. In the immune feedback algorithm, the difference between Smith predictor model and actual object model is seen as antigen, the antigen can produce antibody, and antibody can adjust the model of predictor simultaneously. The immune feedback Smith predictor shows good performance as traditional Smith predictor when the object has significant time delay, and can modify the model according the actual object’s response automatically, so it have better character than traditional Smith predictor when then object have non-linear and time-varying characteristic. The simulation result prove that the Smith predictor with immune algorithm will have better Robust feature, and better adaptive when the object static characteristic change, and it can be suggested to use in real engineering projects.