混合变异算子的自适应粒子群优化算法安晓会.pdfVIP

第 28卷 计算机应用 Vo.l 28 2008年 6月 ComputerApplications June 2008 : 1001- 9081( 2008) S1- 0028- 03 1 2 安晓会 , 高岳林 ( 1. , 750021; 2. , 750021) ( anx iaohui_2007@ 163. com) :针对惯性权重线性递减粒子群算法( DW )不能适应复杂的非线性优化搜 过程的问题, 提出了 一种非 线性递减的惯性权重策略,使算法很快地进入局部搜 , 并在算法中引入混合变异算子, 克服算法易早熟收敛的缺 陷对几种典型函数的测试结果表明, 本文算法的收敛速度和收敛精度都明显优于 DW 算法 :粒子群算法; 自适应惯性权重; 变异算子; 全局优化 : TP182 : A A dap tive particle swarm algorithm w ith hybrid m u tation operator 1 2 AN X iao-hui, GAO Yue-lin (1. Sh ool of M a th em atics and C omp uter, N ingx ia Un iv ersity, Yinchuan N ingx ia 750021, Ch ina; 2. R esearch In stitu te of Inf orma tion and Sy stem Comp u ta tion S cience, N or th N ationa l Un iversity, Yinchuan N ingx ia 750021, China ) A bstract: A new inertia weightw ith nonlinearly descend ing strategy w as presented to solve the problem that the inearly DecreasingW eight ( DW ) of particle swarm algorithm cou ld not adapt to the com plex and nonlinear optmi ization process, which could quickly pu t the algortthm into local search ing. Further more, mutation operator w as introduced to overcom e the problem of the prem ature and low precision of the standard PSO. The algorithm of DW-PSO and ourm ethod were testedw ith fivew ell-know n benchm ark functions. The expermi ents show that the convergence speed and accuracy of our m ethod are sign ificantly superior to that of DW-PSO. K ey w ords: Standard Particle Sw arm Optmi ization (SPSO); adaptive inertia weight; m utation operator; global optmi ization