The shape regulation of probability density function for stochastic systems

Abstract

For nonlinear stochastic systems which are excited by Gaussian white noise, an innovational regulation method is proposed to control the shape of the probability density function of state response to track a desired shape. Firstly, a polynomial feedback scheme is established, and the nonlinear part is replaced by polynomials expansion. Then the recursive equations of the moments which are related to control gain are derived under Fokker-Planck-Kolmogorov theory framework. Meanwhile, regarding the tracking requirement, an optimization problem about the moment approximation is constructed, and the gain of regulation function is obtained by solving this optimization problem using the gradient method. Furthermore, the probability density function of state response is reconstructed from the relationship of the Fourier transform pairs between the characteristic function and probability density function. Finally, two examples are given to demonstrate the effectiveness of the method developed in this paper.