Particle Swarm Optimization–Support Vector Regression (PSO-SVR)-Based Rapid Prediction Method for Radiant Heat Transfer for a Spacecraft Vacuum Thermal Test

• DOI: 10.3390/app14209407
• Publication Date: 2024-10-17T08:42:15Z
• Authors: Xinming Su; Hongsen Jiang; Taichun Qin; Guiping Lin
• Keywords: Technology; QH301-705.5; T; Physics; QC1-999; infrared heating cage; rapid prediction method; PSO-SVR model; radiant heat transfer; Engineering (General). Civil engineering (General); Chemistry; TA1-2040; Biology (General); QD1-999

The simulation of external heat flow has a pivotal role in the vacuum thermal test of spacecraft. The key to simulating the external heat flow of a spacecraft through an infrared heating cage lies in the calculation of radiative heat transfer, and existing Monte Carlo simulation methods for simulating the external heat flow of an infrared heating cage have the disadvantages of complicated modeling and slow calculation speed. In this paper, we consider the spacecraft and infrared cage spacing, partition height, partition arc length, curvature, circumferential relationship, radial relationship, and other variables. The particle swarm optimization–support vector regression (PSO-SVR) method is used to establish the angular coefficient relationship model between spacecraft and infrared cages with different shapes, which realizes the rapid prediction of heat flow in the infrared cage. The angular coefficients obtained by the rapid prognostic model are essentially the same as those obtained by Monte Carlo simulation, while the efficiency is improved by 29,750 times. Taking the vacuum thermal test of a small thermal control star as a case study, the prognostic error gradually decreases with the increase of heat flow, and the maximum error is 6.1%.