Probabilistic control design for continuous-time stochastic nonlinear systems: a PDF-shaping approach

摘要

For the purposes of industrial control, it is often desirable that process variables remain near a set operating point over the long term. This type of long term process performance can be concisely summarized by the stationary probability density function (PDF). In this work, feedback control laws are designed for continuous-time stochastic nonlinear processes so that the closed-loop system is indeed stationary with the PDF chosen in the control design.