Iterative model and trajectory refinement for orbital trajectory optimization. (26th April 2017)

• Record Type: Journal Article
• Authors: Hudson, Jennifer; Gupta, Rohit; Li, Nan; Kolmanovsky, Ilya
• Is Part Of: Optimal control applications and methods. Volume 38:Number 6(2017)
• Page Start: 1132
• Page End: 1147
• Publication Date: 2017-04-26
• Subjects: iterative/numerical methods -- optimal control -- orbital trajectory optimization; Control theory -- Periodicals; Mathematical optimization -- Periodicals; 629.8312
• DOI: 10.1002/oca.2319
• Languages: English
• ISSNs: 0143-2087

Abstract

Summary: An iterative model and trajectory refinement (IMTR) strategy is proposed for trajectory optimization of nonlinear systems. A high‐ and a low‐fidelity models are used. The high‐fidelity model accurately represents the system but is not easily amenable to trajectory optimization, because of degree of nonlinearity, computational cost, or to being of "black‐box" type. The low‐fidelity model is suitable for numerical optimization but approximates the system dynamics with an error. The IMTR method is proposed to systematically iterate between the 2 models and efficiently converge on a control solution. Examples are drawn from orbital mechanics. The IMTR approach is compared to optimal nonlinear quadratic control using Pontryagin maximum principle. A convergence criterion for the IMTR iterations is established.