Hits:

Indexed by:期刊论文

Date of Publication:2017-07-01

Journal:COMMUNICATIONS IN NONLINEAR SCIENCE AND NUMERICAL SIMULATION

Included Journals:SCIE、EI、Scopus

Volume:48

Page Number:95-114

ISSN No.:1007-5704

Key Words:Symplectic Pseudospectral method; State-delayed system; Nonlinear optimal control; Dual variational principle; Hamiltonian system

Abstract:Nonlinear state-delayed optimal control problems have complex nonlinear characters. To solve this complex nonlinear problem, an iterative symplectic pseudospectral method based on quasilinearization techniques, the dual variational principle and pseudospectral methods is proposed in this paper. First, the proposed method transforms the original nonlinear optimal control problem into a series of linear quadratic optimal control problems. Then, a symplectic pseudospectral method is developed to solve these converted linear quadratic state-delayed optimal control problems. Coefficient matrices in the proposed method are sparse and symmetric since the dual variational principle is used, which makes the proposed method highly efficient. Converged numerical solutions with high precision can be obtained after a few iterations due to the benefit of the local pseudospectral method and quasilinearization techniques. In the numerical simulations, other numerical methods were used for comparisons. The numerical simulation results show that the proposed method is highly accurate, efficient and robust.(C) 2016 Elsevier B.V. All rights reserved.