一种带数据整数规划的新型并行自平衡PSO算法

为有效解决标准粒子群(PSO)算法在进化后期缺乏多样性且精度不高的问题,利用多核系统及实际高校地理数据,给出一种高校数据的整数规划方法及并行自平衡PSO算法模型来并行求解高校路网问题,同时体现算法性能。将自平衡机制采用多核系统并行处理方式生成相互独立的子群体,每个子群体间并行求解,最终生成主群体最优路径即高校路网。在Visual Studio2005.NET环境下用C++编程实现仿真。实验结果表明,此算法从求解精度及计算时间两个重要方面综合改善了算法性能。

A New Parallel Self-balancing PSO Algorithm with Data Integer Programming

To effectively solve the problem of the standard particle swarm （PSO） algorithm, which is the lack of diversity and not high accuracy, this paper uses multi-core systems and actual universities geographic coordinates. Then it gives an integer pro- gramming and parallel self-balancing PSO algorithm to solving university path. Self-balancing mechanism uses parallel processing of multi-core systems to generate independent sub-groups. Every sub-group can do parallel computation and finally generate the optimal path of the main group. It uses C ＋＋ programming in Visual Studio 2005 . NET to realize simulation. The results show that this algorithm improves the performance of the PSO algorithm from the solution accuracy and computation time which are two important aspects.