基于随机游走法的热特性参数重建研究

With the continuous development of the aerospace industry, the importance of temperature control for spacecraft is becoming increasingly prominent, especially in obtaining external heat flow data during on-orbit flight. Based on the Monte Carlo random walk method, the inverse analysis of thermal characteristic parameters are investigated for a 2-D rectangular domain using the Covariance Matrix Adaptive Evolution Strategy. The inversion of wall temperature and thermal conductivity are investigated by considering various positions, quantities of temperature, and distributions of internal heat sources within the domain. Errors of 0%, 1%, and 3% are individually applied, and the average runtime is subjected to statistical analysis. The results indicate that for a computational domain with 10 measurement points, better inversion results can be achieved when known point temperatures are closer to the boundary and there are three measurement points, and the internal heat source distribution does not affect the inversion results.