面向安全服务重组多目标优化的粒子群遗传算法

Multi-objective Optimization GAPSO Arithmetic for Security Service Reconfiguration

遗传算法具有快速随机的全局搜索能力,但局部搜索能力差,易陷入早熟收敛,迭代效率低.粒子群算法采用速度——位置模型,可以较快收敛到指定精度.将粒子群算法与遗传算法融合,采用多目标遗传算法得出初步的优化结果,并将其作为粒子,利用粒子群算法强化局部搜索,加快收敛速度,仿真结果证明了该算法的优越性.在CSSM对底层安全服务的重组时利用粒子群和遗传算法的结合(GAPSO),能够提高效率.

Genetic algorithm can perform global searching rapidly and stochastically but for local searching it is not apt because it is easy to converge prematurely, thus leading to a low efficiency of iteration. Particle swarm algorithm adopts a velocity-position model and converges quickly to the designated precision. This algorithm is a combination of genetic algorithm and particle swarm optimization （GAPS, O）, in which multiple-objective genetic algorithm is adopted to get primary optimized results, treated as particles. It is then applied to the enhancing of local searching and speed convergence. Its superiority is demonstrated by the simulation results. For application, GAPSO is employed in security service reconfiguration and the results prove that it can improve efficiency.