Constrained PI tracking control for the output pdfs based on T-S fuzzy mode

This paper presents a new proportional-integral (PI) tracking control strategy for general non-Gaussian stochastic systems based on neural network approximation and T-S fuzzy model identification. The objective is to control the conditional probability density function (PDF) of system output to follow a desired PDF. Following the B-spline approximation on the measured output PDFs, the PDF tracking is transformed to a constrained dynamic tracking control problem for weighting vectors. Different from previous related works, the time delay T-S fuzzy model is applied to identify the nonlinear weighting dynamics. Meanwhile, an improved PI controller design procedure based on LMIs is proposed which can guarantee the required tracking convergence. Furthermore, the robust peak-to-peak measure is applied to optimize the tracking performance.