Robust PI Tracking Strategy for Output Probability Distributions Based on Uncertain B-Spline Neural Networks

This paper considers the robust tracking control problem for output stochastic distributions of dynamic non-Gaussian systems. By using the square root B-spline approximations with modelling errors, a robust constrained tracking control strategy with proportional-integral (PI) structure is investigated for a nonlinear weighting system in the presence of exogenous disturbances. The main objective is to make the output probability density functions (PDFs) to follow a target PDF. An LMI-based PI control algorithm is proposed to track the desired weight dynamics, where the robust peak-to-peak measure is applied to optimize the tracking performance and the state constraints system related to the B-spline expansion can be guaranteed. Rigorous stability and performance analysis is provided for the constrained weight tracking control problem