Research on integrated design method of wide-range hypersonic vehicle/ engine based on dynamic multi-objective optimization

A hypersonic vehicle powered by an air-breathing engine enables efficient long-range delivery and high-speed flexible access to space. The key to achieving high-performance operation of hypersonic vehicles lies in high- efficiency and well-matched hypersonic vehicle/engine integration configuration design. While machine learning-assisted intelligent optimization has shown initial success in hypersonic vehicle/engine integration design, over-reliance on basic and simplistic intelligent methods has led to a significant dependency on sample size and a tendency to easily converge to local optima. This study addresses the need for wide-speed-range, small- sample, and multi-criteria hypersonic vehicle/engine integration design by developing a parametric model for the hypersonic vehicle/engine configuration. Leveraging computational fluid dynamics (CFD) technology, the study uses the Deep Active Subspace (DAS) model along with the Improved Multi-Objective Coati Optimization Algorithm (IMOCOA). This approach is applied to small-sample dynamic multi-point and multi-objective optimization design with the objective of achieving an optimal hypersonic vehicle/engine configuration design characterized by low drag, a high lift-drag ratio, and a high total pressure recovery coefficient across various operating conditions. The results indicate that the Mean Absolute Percentage Error (MAPE) for predicting hypersonic vehicle/engine integration performance using the DAS model is <2 %. Validation of the Pareto solution set from multi-objective optimization shows that dynamic multi-objective optimization enhances performance by >3 % compared to static multi-objective optimization. In comparison to the pre-optimization configuration, the optimized configuration demonstrates a 12.97 % reduction in total drag, with a 9.77 % improvement in lift-drag ratio and a 10.27 % enhancement in total pressure recovery coefficient, highlighting rapid and efficient hypersonic vehicle/engine integration configuration design and performance improvement.