STOCHASTIC CONTROL FOR SYSTEMS WITH NON-GAUSSIAN NOISE

Abstract

In this proposal some stochastic control problems and related topics are described for stochastic equations that are driven by some non-Gaussian noise processes. These noise models include Rosenblatt processes. These noise models have a long range dependence property that has been empirically determined in a wide variety of physical phenomena. Since these processes are not Markov optimal controls cannot be determined by solving partial differential equations such as Hamilton-Jacobi-Bellman equations. Furthermore stochastic maximum principles with forward-backward stochastic differential equations are not available for these problems. The proposers plan to obtain explicit optimal stochastic controls as they have done for an ergodic linear-quadratic control problem for a scalar system. It is also planned to study the problems of parameter estimation and adaptive control for these stochastic systems with non-Gaussian, long range dependent noise.

Document Details

Document Type
DoD Grant Award
Publication Date
Aug 12, 2021
Source ID
FA95502010336

Entities

People

  • Tyrone Edward Duncan

Organizations

  • Air Force Office of Scientific Research
  • United States Air Force
  • University of Kansas

Tags

Fields of Study

  • Mathematics

Readers

  • Adaptive Control and Estimation with Uncertainty in Dynamic Systems.
  • Finite Element Method (FEM) for solving Partial Differential Equations (PDEs)
  • Statistical inference.