Adaptive Backstepping Neural Control for Switched Nonlinear Stochastic System with Time-Delay Based on Extreme Learning Machine

In this paper, for a class of switched stochastic nonlinear systems with time-varying delays, the output feedback stabilization problem is addressed based on single hidden layer feed-forward network (SLFN) and backstepping technique. Furthermore, an adaptive backstepping neural switching control scheme is presented for the above problem. In the scheme, only a SLFN is employed to compensate for all known system nonlinear terms depending on the delayed output. The output weights and control laws are updated based on the Lyapunov synthesis approach and backstepping technique to guarantee the stability of the overall system. Then a special switching law is given based on attenuation speed of each subsystem. Different from the existing techniques, the parameters of the SLFN are adjusted based on a new neural networks learning algorithm named as extreme learning machine (ELM), where all the hidden node parameters randomly be generated. Finally, the proposed control scheme is applied to an example and the simulation results demonstrate good performance.