GA和PSO在煤矿进度计划优化中的比较应用  

Comparison application of GA and PSO to optimization of mine schedule plan

作　　者：王德银[1,2] 刘玲[1] 郭海湘[1] 

机构地区：[1]中国地质大学(武汉)经济管理学院,湖北武汉430074 [2]平顶山煤业集团第五矿,河南平顶山467000

出　　处：《煤炭工程》2009年第12期118-121,共4页Coal Engineering

基　　金：国家自然科学基金项目资助(70573101);高等学校博士学科点专项科研基金资助(20070491011)

摘　　要：运用遗传算法(GA)和粒子群算法(PSO)在网络图优化的基础上对平煤天安五矿井巷与安装工程的施工进度计划进行二次优化控制。首先通过计划评审法(PERT)得到己二采区各个工序的时间参数和相应的网络图,接着在网络图的基础上,以净现值NPV最大化作为进度安排目标,以各工序的开工日期为决策变量,以各工序之间的先后顺序和时间关系以及系统外资金控制为约束,采用GA和PSO进行二次优化,比较结果表明PSO要优于GA。The genetic algorithm （GA） and particle swarm optimization （PSO） were applied to optimize the project scheduling of the roadway and installation project in Pingmei Coal. The process of optimization contains two parts in this paper ： the first part is obtaining the time parameters of each operation and the network graph of the roadway and installation project by PERT method which are based on the raw data; another part based on the network graph is the second optimization of which the objective is the maximal NPV （Net Present Value） and the starting dates of all operations are decision - making variables and operations order and time and cost restriction out of system are the constraints by GA and PSO. Then the optimizing result of PSO is compared with the result of GA. The optimizing results show that PSO is better than GA and the NPV based on optimized operations is more than original plan 1497.4 ten thousand RMB.

关 键 词：煤矿进度计划 工序管理 净现值 PSO GA 

分 类 号：N945.22[自然科学总论—系统科学]