高超声速再入飞行器轨迹优化算法评估策略
An Evaluation Strategy of Trajectory Optimization Algorithms for Hypersonic Reentry Vehicle

针对高超声速再入飞行器轨迹优化算法种类多样、决策者难以优选的问题,提出了一种轨迹优化算法综合评估策略。该策略首先对算法本身的稳定性、实时性以及优化结果的可行性、安全性、过渡性、最优性等性能进行指标建模,建立优化算法综合评估指标体系;然后基于层次分析法中的两两比较判断矩阵,以欧氏距离与违例次数为精度衡量依据,将指标权重的计算转化为一类优化问题并利用遗传算法求解;之后计算加权规范化决策矩阵,综合算法数据与理想对照组在空间距离与曲线形状两方面的贴近度,提出一种灰色逼近理想解排序加权法,最终完成评估分析过程。仿真结果表明:该策略以客观数据的形式实现了轨迹优化算法的性能量化,形成了不同算法之间统一的对比评估标准,能够帮助决策者快速遴选出适合任务需求的最优算法,避免了过程中的盲目性与主观性。
A comprehensive evaluation strategy is proposed to deal with the evaluation of various trajectory optimization algorithms for the hypersonic reentry vehicle. An integrated index system is established by taking the stability and real？？time performance of each algorithm itself as well as the feasibility, security, transitivity and optimality of its optimization result into account. And the calculation of index weights is transformed into an optimization problem based on the pairwise judgement matrix of the analytic hierarchy process, and the Euclidean distance and the number of violations are used as precision measures. The genetic algorithm is used to solve the optimization problem. Then, the normalized decision matrix is obtained, and a weighted grey technique for order preference by similarity to ideal solution (TOPSIS) is proposed by synthesizing the nearness degree in both the space and curve between the algorithm data and the ideal control groups to achieve the final evaluation results. Simulation results show that the proposed strategy uses objective data to quantify the performance of trajectory optimization algorithms, and forms a comparative evaluation standard for various algorithms. The strategy can help designers quickly select an optimal algorithm according to different mission requirements and avoid blindness and subjectivity