一种改进的带有演化规则的元胞遗传算法

An Improved Cellular Genetic Algorithm with Evolutionary Rules

传统的遗传算法(GA)在求解云计算环境下的资源调度问题时存在早熟收敛、易陷入局部最优和鲁棒性差等问题,在GA算法的基础上融入了元胞自动机的作用机理,通过制定相应的演化规则,提出了一种改进的具有演化规则的元胞遗传算法(CGAER)。CGAER算法在种群进化过程中能够根据所制定的演化规则,不断地更新每个个体及其周围邻居的状态,确保整个种群在搜索空间的均匀性和多样性,这样可以有效避免算法过早收敛、易陷入局部最优和寻优能力弱等问题。通过一系列的仿真并与传统的遗传算法(GA)和元胞遗传算法(CGA)相比较,CGAER算法展现了良好的寻优性能和鲁棒性。在解决云资源调度问题时具备全局解空间搜寻能力,是一种更为有效的算法。

When dealing with resource scheduling under cloud computing environment, the traditional Genetic Algorithm (GA ) has some disadvantages such as early convergence, poor efficiency and being easy to fall into local optimum. In order to overcome these shortcomings of the traditional GA, this paper proposed an improved cellular genetic algorithm with evolutionary rules ( CGAER). According to the evolutional rules established, the status of each individual and its neighbors can be updated periodically, which can ensure uniformity and diversity of the entire population in the search space and avoid problems such as premature convergence of the algorithm, being easy to fall into local optimum and poor search ability. Therefore, the CGAER algorithm has a strong capability to find out the global optimum value for cloud resources scheduling problem. The experimental simulation shows that it has a good performance over the GA and CGA algorithm, which verifies that the CGAER algorithm is more effective and suitable to deal with resource scheduling under cloud computing environment.