Smith online identification predictive control and its application in STM

In this paper, a Smith online identification predictive control method was proposed for the system with delay. As we know, the Smith predictor, put forward by Smith (1959), resolves how to control the object with delay theoretically. However, due to the existing modeling error, a good control effect cannot be be acquired. It is even impossible to design a proper controller in order to stabilize the system. This paper incorporates system identification with the Smith predictor to-realize Smith online identification prediction control to compensate the delay. The identification module models the object dynamically. Based on the result, the delay is compensated by the Smith predictor. Experiments are presented to demonstrate that the method is of strong self-adaptability. Furthermore, it was successfully applied in the compensation control system of the AJ-I scanning tunneling microscope (STM) manufactured by Shanghai Aijian Nanometer science & technology development Ltd.