A Stochastic Controller Design Using an Approximate Solution of FPK Equation

Abstract

The Fokker-Planck-Kolmogorov (FPK) equation is a powerful tool with Markovian response for analyzing dynamic systems. In this study we consider the design of a stochastic controller using FPK for a system subjected to a wide band random disturbance with constant power spectral density. The time-domain dynamic system equation is transformed into a general dynamic statistical moment equation in the stochastic domain via the FPK equation. The system transformation cannot be realized in real time. In order to improve stochastic controller design in real time, a new method of designing a stochastic controller and its implementation are proposed. The practical efficiency of the proposed method using the approximate probability density function is verified via numerical simulation.