Lyapunov Exponent and Stability of Optimally Controlled Quasi Partially Integrable Hamiltonian Systems

w q zhu,z l huang
DOI: https://doi.org/10.1007/978-94-010-0179-3_24
2003-01-01
Abstract:A controlled quasi partially integrable Hamiltonian system with pure parametric stochastic excitations is first reduced to averaged Ito equations by using the stochastic averaging method for quasi-Hamiltonian systems. A dynamical programming equation is then established for the ergodic control problem and solved to yield the optimal control law. The expression for the largest Lyapunov exponent is derived and the necessary and sufficient condition for the asymptotically Lyapunov stability with probability one of optimally controlled system is obtained. The system is stabilized by feedback control with proper choosing of the cost function and weighting matrix in the performance index for ergodic control problem.
What problem does this paper attempt to address?