Optimal design of thermal control system for space optical remote sensor based on NSGA-II and opto-mechanical-thermal integration analysis

Abstract

The thermal control system (TCS) is a key technology for ensuring good imaging quality of a space optical remote sensor (SORS) in orbit, and its research is of practical significance. In this paper, a multidisciplinary optimization method was proposed for the TCS of a SORS. A particle-swarm-optimized back-propagation neural network was used as the surrogate model to reduce the computational cost of the opto-mechanical-thermal integrated simulation model. Subsequently, the mean and variance of the modulation transfer function (MTF) of 12 fields of view of the SORS at the Nyquist frequency were considered as objective functions, multi-objective optimization of the design parameters of the TCS was performed using the non-dominated sorting genetic algorithm Ⅱ (NSGA-Ⅱ), and the Pareto front composed of a series of Pareto solutions was obtained. A compromise solution was selected as the optimal TCS design. The results showed that, upon comparison with the initial design, the optimal design scheme increased the mean of the MTFs of the SORS by 34.4%, reduced the variance by 31.3%, and significantly improved the comprehensive optical performance of the SORS.

Keywords

• Space optical remote sensor
• Thermal control system
• Opto-mechanical-thermal integration
• Surrogate model
• Multi-objective optimization (MOO)
• Non-dominated sorting genetic algorithm Ⅱ