【摘要】 韧性度是衡量网络拓扑结构脆弱性的有效指标,它描述了网络在节点失效或遭遇外来攻击时被断裂的难易程度、断裂后剩余子网的连通性及重新连通断裂子网的难易程度,采用韧性度判据来衡量网络结构的抗毁性,可以在一定程度上避免连通度的不足。为解决韧性度约束下的最小网络拓扑规划问题,文章提出了一种基于准并行遗传算法的拓扑规划方法,该方法通过在初始种群中加入特定个体来扩大其覆盖率,利用不同结构种群独立进化的方法来防止早熟。仿真实验表明了算法的有效性。更多还原

【Abstract】 Using tenacity as a criterion of survivable network design can avoid the limitation of connectivity to a certain extent,but the complexity of computation is a great obstacle.A quasi-parallel genetic algorithm is presented to accelerate the computing speed by using more computers.The algorithm focuses on the problem of mini-network topology planning under the restraint of tenacity.In this algorithm, several sub-populations evolve independently,each sub-population can evolve in different modes with different parameters,and the self-governed sub-population schema can avoid precocity.On the other hand,the optimization of the initial population is used to increase the efficiency of evolution,and special individuals are added into the initial population to improve the coverage of the solution space.The result of simulation suggests that this algorithm can converge to the optimal solution efficiently,and the efficiency of the quasi-parallel genetic algorithm is better than the simple genetic algorithm.In addition,the quasi-parallel genetic algorithm can be implemented in one computer with a sequence mode.更多还原