Robust tracking controller design of output stochastic distribution system based on instant performance index

In this paper, robust tracking control problems are considered for a class of output stochastic distribution system with model uncertainty and random disturbance. Firstly, the output PDF tracking error dynamic system is established by using B-spline neural network to approximate the PDFs of the system output variables. Afterwards, an instant performance index is proposed to instead of the conventional ones with infinite integral. The effect caused by the model uncertainty and disturbance are also estimated by a virtual system rather than the worst condition of the stochastic systems was chosen for the traditional robust control. Then a robust controller is designed based on this estimate results and the expectation performance. A simulation case study is included to show the effectiveness of the proposed algorithm and encouraging results have been obtained.