Fault-tolerant control design for stochastic distribution control systems based on k-step fault estimation

A novel scheme of fault-tolerant control (FTC) and fault estimation (FE) for the stochastic distribution control (SDC) system with disturbance is investigated. The output probability density function (PDF) is modelled by the weight vector of the B-spline neural network with the given basis function, and control of output probability density function can be converted to control of the weight vector of the B-spline neural network. A k-step fault estimation algorithm aiming at enhancing the accuracy of fault estimation is first applied to the stochastic distribution control system. Subsequently, a dynamic output feedback fault-tolerant control algorithm is constructed by using the k-step fault estimation information to guarantee that the target weight vector can be still approximated by the output weight vector in case of actuator fault. A simulation example is used to confirm the availability of the presented method.