Neural-Network-Based Stochastic Scheduling Control of Unknown Nonlinear Systems

This article addresses the problem of stability of unknown nonlinear systems with a stochastic scheduling scheme. In order to solve the difficulty resulted by the unknown nonlinearity, the neural-network approximation technique is introduced. Notice that for the controller design of uncertain nonlinear systems, numerous simulation studies and actual industrial implementations show that the neural network is a good candidate to handle the design difficulty resulted by unknown nonlinearities. The feedback control signal in this article is produced based on periodic sampled data. In the stochastic scheduling scheme, both the choices of controllers and their execution time allotted to the scheduler are random. Sufficient conditions based on the probability distribution of almost sure stability are obtained by using a general Lyapunov functional and some stochastic techniques.