基于正交遗传算法和灵敏度分析的体系仿真优化方法

体系对抗已成为当今战争的主要形式,因而急需对体系优化问题进行深入细致地研究.鉴于此,提出了一种基于正交遗传算法和灵敏度分析的体系仿真优化方法.该方法采用正交遗传算法在可行域内快速地搜索一些较优方案(解);利用灵敏度分析方法从已评估方案中得到待研究体系输入、输出之间的灵敏度关系;应用这种灵敏度关系来指导正交遗传算法的后续搜索.最后采用一些数据实例来验证该方法,实验结果表明,该方法是可行的、正确的和有效的.该方法能在较少次数的仿真后,得到待研究体系的满意解或最优解;同时,该方法可推广到其它复杂优化问题的求解中.

The Simulation Optimization Approach for System-of-Systems based on Orthogonal Genetic Algorithm and Sensitivity Analysis

system-of-systems antagonism has already become the dominating modality of modern warfare,so it is urgent to study the optimization problem of system-of-systems.For this purpose,it presents a simulation optimization approach for system-of-systems based on orthogonal genetic algorithm and sensitivity analysis.In the proposed approach,it applies the orthogonal genetic algorithm to find some better solutions from the whole feasible space at first;it uses the sensitivity analysis method to obtain the input-output relationship of system-ofsystems by analyzing the evaluated solutions at second;it adopts the sensitivity relations to guide the subsequent heuristic searching of the orthogonal genetic algorithm at third.Many numerical examples were applied to validate the proposed approach.The optimization result suggests that,the proposed approach is feasible,correct and valid.The proposed approach can achieve the optimum solution(or near-optimum solution) of system-of-systems by executing a few times simulation.This approach can be popularized to solve other complex optimization problems.