基于粒子群算法的高拱坝仓面排序多目标优化研究  

Study on multi-objective particle swarm optimization of concrete-placement sequencing for high arch dam

作　　者：钟登华[1] 程普[1] 任炳昱[1] 关涛[1] 刘肖军[1] 

机构地区：[1]天津大学水利工程仿真与安全国家重点实验室,天津300072

出　　处：《水力发电学报》2015年第8期7-17,共11页Journal of Hydroelectric Engineering

基　　金：国家自然科学基金创新群体基金项目51321065;国家自然科学基金资助项目51339003;国家重点基础研究发展计划(973计划)资助项目(2013CB035904)

摘　　要：混凝土高拱坝浇筑施工中,仓面排序是重要环节,如何通过施工进度和施工均衡性的优化获得合理的仓面排序方案是需要解决的重要问题。目前的排序方法大都是根据已有的经验或基于对坝块属性值的数学分析来制定跳仓排序规则,无法有效地解决仓面排序的多目标优化问题。本文综合考虑影响混凝土浇筑的各项因素,建立了以高拱坝跳仓排序规则为变量,以优化施工工期、月浇筑强度、浇筑机械利用率为目标的多目标优化模型,应用粒子群算法对该多目标优化问题进行求解,从而获得多目标综合最优的仓面排序方案。实例表明,本文采用粒子群算法实现了对仓面排序方案的优化调整,计算得到的Pareto解集能够为决策者提供多个方案以便更好地决策,利用逼近理想解法优选的排序方案与传统仓面排序方法的施工仿真结果相比,能够获得更短的工期和更均衡的浇筑过程,对于指导现场施工具有重要意义。Concrete-placement sequencing method is a key link in the construction of high concrete arch dams. How to find a reasonable placement sequencing scheme through optimizing the progress and proportionality of concrete-placement construction is an interesting issue, but the current methods cannot effectively solve such a multi-objective problem. In this work, we have developed a multi-objective optimization model that incorporates the factors of concrete pouring and the sequencing schemes as variables and optimizes the objectives of time limit, month pouring strength, and pouring machine utilization. It adopts a particle swarm optimization （PSO） algorithm to solve the multi-objective problem. Application of our model in a case study shows that the placing sequence can be optimized and optimal schemes can be found by using Pareto optimal solutions and an optimization technique of approximating ideal solution （AIS）. The results are better than those calculated by traditional simulation method and the new method presented would make a significant improvement on on-site construction of concrete dams.

关 键 词：混凝土高拱坝 施工仿真 仓面排序 粒子群算法 多目标优化 

分 类 号：TV544.91[水利工程—水利水电工程]