基于改进局部搜索遗传算法的目标分配决策

Target allocation decision making based on improved genetic algorithms with local search

为满足舰载武器目标分配需求,对传统的局部搜索遗传算法进行了改进,并用其求解目标分配问题的最优解。构造了适合于目标分配问题的染色体;设计了搜索性能较好且能够保留优秀基因的交叉操作方法;将局部搜索机制引入标准遗传算法,提高了目标分配算法的收敛速度;把模拟退火算法引入局部搜索问题,在一定程度上避免了局部最优问题;将贪婪算法应用于局部搜索提高了最优分配方案的搜索效率。仿真计算表明,改进局部搜索遗传算法的目标分配性能优于已有算法。

To make proper target allocation （TA） for ship borne weapons, an improved genetic algorithm （GA） with local search is proposed. Suitable chromosomes are constructed for the TA problem. A new cross- over operator is proposed to keep good genes and enhance the search performance. Local search mechanism is added into GA to find solutions for TA problems with a more rapid convergence speed, Simulated annealing （SA） is used to overcome the problem of local optimum. A heuristically greedy local search is also proposed to improve the search efficiency. The simulation results show that the improved algorithm can improve the search efficiency of GA. Several examples are also exploited to explicit the superiority of the proposed approach over other existing algorithms,