摘要： 提出一种异构分布式系统的负载均衡调度算法。对异构系统进行建模，使用染色体建立任务集合调度模型，根据该模型制定适应度函数，将其作为衡量负载均衡的标准，利用该标准对异构系统进行任务调度，并动态设定最大进化代数，以此改进动态遗传算法。实验结果表明，该算法具有较好的负载均衡性能。

关键词: 异构系统, 变异操作, 负载均衡, 遗传算法

Abstract: The classic Genetic Algorithm(GA) limits the evolution because the next generation cannot inherit the most adaptable chromosome. To improve the algorithm, this paper proposes a dynamic genetic algorithm. It creates a model for the heterogeneous system, and formulates criterion for measuring load balance according to the model, then uses the formulated criterion in scheduling jobs on the heterogeneous system. The algorithm allows configuring the maximum evolution generation dynamically. Experimental results show that the improved algorithm has better load balance performance.

Key words: heterogeneous system, mutation operation, load balance, Genetic Algorithm(GA)

中图分类号: 

• TP301.6