CAJ | 학술논문

针对基本粒子群算法（ PSO）寻优过程中存在收敛速度慢、易陷入局部最优和计算精度差等缺陷，采用分簇思想和碰撞策略，提出了一种改进的粒子群算法（C-PSO），在该算法中，粒子通过分簇并行搜索，有效避免了群体过度集中现象，极大地增强粒子全局搜索能力。将C-PSO算法应用于混凝土面板堆石坝断面优化设计中，优化结果表明，该算法对解决复杂的多变量多约束非线性问题具有较好的适应性，为复杂的混凝土面板堆石坝断面优化设计问题提供了新的解决思路。
침대기본입자군산법（ PSO）심우과정중존재수렴속도만、역함입국부최우화계산정도차등결함，채용분족사상화팽당책략，제출료일충개진적입자군산법（C-PSO），재해산법중，입자통과분족병행수색，유효피면료군체과도집중현상，겁대지증강입자전국수색능력。장C-PSO산법응용우혼응토면판퇴석패단면우화설계중，우화결과표명，해산법대해결복잡적다변량다약속비선성문제구유교호적괄응성，위복잡적혼응토면판퇴석패단면우화설계문제제공료신적해결사로。
In the optimization process with particle swarm optimization (PSO) algorithm, problems such as slow conver-gence rate, local optimization, computational inaccuracy are commonly found .Based on the clustering and particle colli-sion strategy , a modified algorithm , namely the clustering particle swarm optimization ( C-PSO) , is proposed .With the clustering of particles , the searching process is carried out in parallel .Excessive concentration of particles is then avoi-ded, which greatly improves its global searching ability .Then, the proposed C-PSO algorithm is applied to the optimiza-tion design of the section of concrete face rockfill dam .It shows an good adaptability to solve the complex nonlinear problems with multi-constraints and multi-variables , providing a new solution for the optimization design of the section of concrete face rockfill dam .