Application of RBF neural network in the model-free adaptive control

To solve the impact of the unmodelled dynamics of the model process, model-free adaptive control based on RBF neural network is proposed. In this algorithm nonlinear system is linearized by linearization of tight format. Then the system parameters are identified by the RBF neural network algorithm. The parameters are used to directly recursively compute model-free adaptive control input. The controller is designed only by using I/O data of the controlled system, and no structural information or external testing signals are needed. Simulation result shows that the proposed algorithm is an effective strategy with excellent tracking ability and strong robustness.