基于MPSO方法的边坡可靠度计算研究  

Study for Calculation of Slope Reliability Based on MPSO Method

作　　者：张玉娇[1] 朱晓丽[1] 贺子光[2,3] 党亚倩 陈磊[2] 

机构地区：[1]济源职业技术学院建筑工程系,河南济源454000 [2]长安大学地质工程与测绘学院,西安710054 [3]西部地质资源与地质工程教育部重点实验室,西安710054

出　　处：《安阳工学院学报》2015年第6期51-54,109,共5页Journal of Anyang Institute of Technology

基　　金：陕西省科技计划项目:社会发展公关计划(S2012SF3082);陕西省财政厅(陕财办建(2012)398号)

摘　　要：借鉴遗传算法中的杂交概念,将其引入标准粒子群方法(Particle Swarm Optimization,PSO),形成混合粒子群算法(MPSO)。该方法改善了PSO方法的全局搜索能力,提高了算法的收敛速度和计算精度,并用其计算可靠度指标及相应的验算点。以典型的边坡为例,通过工程算例并与其他方法对比,表明了MPSO方法较标准PSO方法的计算精度高,收敛速度快;分析了算法中各控制参数对可靠度指标的影响;算例结果表明:MPSO方法对求解功能函数呈高度非线性的边坡可靠性问题具有很好的适应性,是科学可行的,具有很好的应用前景。Drawing on the concept of hybrid genetic algorithms, it is introduced to standard Particle Swarm methods（Particle Swarm Optimization, PSO） to form Mixed Particle Swarm Optimization（MPSO）. This method not only improves global searching ability, but also advance the rate of convergence and computational accuracy, and isused to calculated reliability index and corresponding check points. Taking one typical slope for examples, comparing to other methods by first example, it is showed that MPSO method has higher precision, more fast convergence speed than standard PSO, while analysis the effect caused by each control parameters in algorithm on reliability index; The results show this method is good to solve the reliability analysis problem with highly nonlinearperformance function, is scientific and feasible and has good application prospect.

关 键 词：混合粒子群算法 边坡 可靠度 

分 类 号：TU311.2[建筑科学—结构工程]