针对标准的遗传算法在任务分配中收敛速度慢的问题,对多agent系统中的任务分配进行形式化描述的基础上,融合模拟退火算法的优化思想,提出了一种基于模拟退火遗传算法的任务分配方法,详细阐述了该算法的基本思想和关键步骤,并通过仿真实验进行验证。仿真实验结果表明,基于模拟退火遗传算法比标准的遗传算法具有更快的收敛速度和寻优效果。
Aiming at the shortcomings of normal genetic algorithm that its convergence speed is slow in task allocation,based on giving the formal specification of task allocation in multi-agent system,this paper proposed a simulated annealing genetic algorithm(SAGA) by integrating simulated annealing,presented the basic thought and pivotal steps of SAGA in detail,and validated the algorithm by simulation experiment.The simulation results illustrate that SAGA has better convergence speed and optimal results than normal genetic algorithm.