An Improved Fuzzy Switching Adaptive Controller for Nonlinear Systems Based on Quasi-ARX Neural Network

In this paper, we offer the fuzzy switching con-troller based on Quasi-ARX neural network model using Lyapunov learning algorithm to control nonlinear dynamical system. This work exploits the idea to use Lyapunov function to train multi-input multi-output neural network on the core-part sub-model.We can get a linear controller, and a controller based on the characteristic of the fuzzy switching algorithm designed between the two controllers. Based on the result of simulation, the Lyapunov learning algorithms have better performancecompared with the propagation learning algorithms. The improving performances of the Lyapunov learning algorithm are stable in the learning process, fast convergence of error, and able to increase the accuracy of the controller.