Chaotic optimization of smith predictor

A controller may be failure if the parameters of a Smith predictor are not precisely selected. Therefore, a method is proposed by which the parameters can be chaotically optimized on-line. The transfer function of the plant is identified first by chaotic mapping of the step response curve, then, the PID parameters are optimized off-line by using the identified model. When the model is changed, the parameters of the controller can be adjusted adaptively. The plant can be automatically identified, and rolling optimization process is performed in the method. The control system simulation for the plant is carried out, and the results show that the tracking performance and robustness of the controller are satisfactory. ©, 2015, Northeastern University. All right reserved.