并行仿真的粒子群优化算法异步模式研究

粒子群优化算法，起源于鸟群行为的研究，是一种基于群智能的进化计算技术，通过粒子之间的协作与竞争以实现对多维复杂空间的高效搜索。该文研究了粒子群优化算法的生物特征，提出粒子群优化算法的异步模式，使进化中的粒子个体充分表现出独立性，种群表现出异步性。异步模式的程序实现通过MFC多线程并行仿真实现。最后，采用经典测试函数验证异步模式的有效性，测试结果表明：与同步模式(经典PSO算法)比较分析，异步模式的收敛速度显著提高，同时刻的寻优效果更好。

Research on the Parallel Simulation of Asynchronous Pattern of Particle Swarm Optimization

Particle swarm optimization(PSO) ,rooting from simulation of swarm of birds, is a new branch of evolution algorithms based on swarm intelligence, realizing effective search on multi - dimension complex spaces through cooperation and competition between particles. An asynchronous pattern of the particle swarm optimization is proposed based on analyzing its biologic characters, each particle acted as an independent individual and the searching of population performed asynchronously in the pattern. Then the asynchronous pattern is simulated with the MFC multi - thread method. The experiment results demonstrate that the asynchronous pattern has bigger speed of convergence and better optimizing result comparing with the synchronous pattern.