摘要
文中结合克隆选择算法,模拟退火算法和遗传算法的优点,提出了一种改进的混合克隆退火遗传算法,并将该算法应用于网格计算任务调度问题的求解之中.该算法先通过克隆,退火交叉和高斯变异等操作来产生一组新的抗体,然后再对所产生的抗体进行模拟退火,直到退火温度不能再降低为止,从而求得问题的最优解.理论分析和实验结果表明这种任务调度算法优于其他调度算法,并可以成功地应用于网格环境下的任务调度问题.
This paper combined with the advantages of clonal selection algorithm, simulated annealing and genetic algo-rithm, brings forward an improved hybrid clonal annealing genetic algorithm and applied to solve grid computing task scheduling problem. It first generates a new group of antibodies through operations such as clone, annealing crossover, gauss mutation, etc, and than simulated anneals independently all the generated antibodies respectively. When the temperature in the process of cooling no longer falls, the result is the optimal solution on the whole. From the experiment result, it concludes that this algorithm was superior to other algorithm, and can be applied to the optimization of task scheduling successfully.
出处
《微电子学与计算机》
CSCD
北大核心
2009年第10期154-157,共4页
Microelectronics & Computer
关键词
网格计算
任务调度
模拟退火
遗传算法
grid computing
task scheduling
simulated annealing
genetic algorithm
