简约粒子群优化算法

针对全局版粒子群的早熟和局部版粒子群的最优位置信息利用率低的问题，提出简约粒子群算法．该算法使用速度松弛迭代策略，使粒子不必频繁更新速度，当粒子速度有利于适应度进一步提高时，就在下一个迭代周期内维持该速度，这有利于提高良好速度信息的利用率，减小算法的计算量，加快运算的收敛速度．同时，利用精英集团策略，使多个最优位置信息在种群内充分共享，有效地控制了种群多样性，避免了早熟现象．在典型标准测试函数上进行了全局、局部版惯性因子粒子群和全局、局部版约束因子粒子群测试比较，结果表明简约粒子群算法具有更强的寻优能力和更高的稳定性，且计算量也比较小．

Compact Particle Swarm Optimization Algorithm

In order to avoid premature convergence in the global version of PSO(paruicle swarm optimization),and low information utilization of the best positions in the local version of PSO,the compact(Com-PSO) algorithm is proposed,in which two strategies are employed to improve the performance.Firstly the relaxation-velocity-update strategy is used to improve the utilization of good velocities,where the velocity can maintain unchanged at the next iteration when it benefits to further improving the fitness.The strategy not only enhances the local search ability but also saves the computation.Secondly the elite strategy is adopted to control the diversity of the swarm,where the information of the several best positions is shared.The resulting algorithm effectively utilizes the information of the best positions,and avoids the premature convergence,so it achieves the good performance.In experiments,Com-PSO,linearly decreasing weight particle swarm optimization(LDW-PSO) algorithm,constriction PSO(CPSO),local version of LDW-PSO with von Neumann topology,and local version of CPSO with von Neumann topology are compared on benchmark functions.The experimental results show that Com-PSO has better search ability and higher stability compared to the other algorithms and it also requires less computation time.