The Study Of Model-Free Adaptive Controller Based On DSPACE

In order to solve the problems related to the dependence with the model of the system and the unmodeling system dynamics, model-free adaptive control method was proposed. Model-free adaptive control can implement the control only by the input data and output data without knowing the system model. This paper deeply explores Model-free Adaptive control (MFAC) applied to industrial processes characteristic, explains the concept of pan-model and analyzes the model-free control law and the identification of the characteristic parameters, as well as stability and convergence of model-free control methods in detail.Researchers hope to apply a control algorithm which is developed by using design software into practice in a rapid and convenient way when researching in control engineering. Moreover, they hope to revise and apply an algorithm repeated rapidly so that they can apply a best control way in a control application. It is significant to develop and construct a real-time testing platform.dSPACE real-time simulation system is an excellent platform for development and test developed by dSPACE Corporation based on MATLAB/Simulink software. This system can realize a perfect link with MATLAB/Simulink software. dSPACE real-time simulation system owns a hardware system with high speed calculation capacity, furthermore, it includes a high speed processor, input/output interface, convenient code generator/downloading and experiment/debugging software environment. This paper constructed a real-time environment to verify control algorithm based on dSPACE software and hardware platforms.This paper researches the model-free adaptive controller based on dSPACE real-time simulation, and makes use of simulink control system to set up system model for simulation study, which used a format based on the compact linear method of the model-free adaptive control algorithm. In addition, profound analyses and comparison have been made on model free control and conventional PID control. Simulation shows that model free control is obviously better than PID control on the adaptability, anti-interference and against large time-lag.