摘要
针对星载电子系统硬件演化电路的应用需求,提出一种求解可重构资源替换问题的两阶段寻优演化算法.该算法将遗传算法与局部最优的启发式算法相结合,在任务运行时出现故障节点时,利用局部最优算法找到与工作节点距离最近的节点集合作为备选节点,并利用有约束条件的遗传方法,在集合中选择全局路径长度最短的最优节点来替换故障节点,该算法兼顾了局部通信和全局遍历的应用需求,又提高了求解效率.在搭建的测试环境中,对所提出的算法进行了验证,仿真结果表明两阶段寻优演化算法可以高效地寻找到最优替换节点.
The two phase optimization evolution algorithm is proposed to solve the problem of reconfigurable resource replacement, which is applied to the hardware evolution of satellite electronic system. In this algorithm, the basic genetic algorithm and local optimi- zation algorithm are used to solve the problem of failure node selection when the hardware evolution chip is running. The local optimi- zation algorithm is used to find the idle nodes, which has the shortest distance between the work nodes, and the genetic algorithm with constraints is used to select the effective node. The HGA algorithm provides the efficiency of selection,in the real test environment, the proposed algorithm is verified. The simulation results show that the two phase optimization evolution algorithm can effectively find the optimal replacement node.
出处
《小型微型计算机系统》
CSCD
北大核心
2016年第10期2380-2384,共5页
Journal of Chinese Computer Systems
基金
国防科工局民用航天基金项目(2014537)资助
北京市自然基金项目(4312016)资助
关键词
星载电子系统
硬件演化
局部最优
遗传算法
satellite electronic system
hardware evolution
local optimal
genetic algorithm