摘要
为降低徐深气田新建产能管网系统投资,考虑集气支线的连接方式和障碍对整体建设费用的影响,将集气支线连接方式作为优化变量之一,采用R函数法和分层求凸包法,推导并建立任意障碍多边形的数学表达式和可行布局约束条件,以管网建设费用最小为目标,建立拓扑布局优化数学模型。根据数学模型的结构层次,构建改进的混合遗传算法求解策略,设计几何位置实数编码和拓扑关系整数编码的多参数级联编码方式,调整选择复制遗传算子的操作方式,建立考虑集气支线气量均匀性的自适应种群进化的适应度函数;结合K—中心聚类法、叉积法、贪心算法和Prim算法给出初始种群的建立方法,验证模型和算法的有效性。结果表明:集气支线的连接方式和障碍是管网布局优化的重要影响因素,基于合理初值的多参数级联编码遗传算法比常规遗传算法的寻优效果和速度更好。
In order to reduce the investment of new capacity network system in Xushen gas field,the mathematical model of topology layout optimization is established where the minimum pipe network construction cost are taken as the objective function.In the model,the influence of connection mode of gas gathering branch pipe and obstacle on the whole construction cost are considered,thus the gas gathering branch pipe connection mode is taken as one of the optimization variables of model,and at the same time,by taking in the R function method and layered convex hull method,the mathematical expressions and feasible layout constraints are derived and established.According to the structure level of the model,the improved hybrid genetic algorithm solving strategy is constructed.The multi-parameter cascade coding mode of the geometric position based on real number coding and topological connection mode based on integer coding is designed.And the operation to select and copy is adjusted and the adaptive population evolution function which considering gas volume uniformity of gas gathering branch pipe is established.Besides,the K-center clustering method,the cross product method,the greedy algorithm and the Prim algorithm are combined to give the establishment method of initial population.Finally,the validity of the model and algorithm is verified by an example.The results show that the connection mode of gas gathering branch pipe and obstacle are important factors for the optimization of pipe network layout,and the improved hybrid genetic algorithm based on reasonable initial value is faster and better than the conventional genetic algorithm.
出处
《东北石油大学学报》
CAS
北大核心
2016年第4期96-105,共10页
Journal of Northeast Petroleum University
基金
国家科技支撑计划项目(2012BAH28F03)
国家自然科学基金面上项目(51674086)
东北石油大学创新科研项目(YJSCX2015-012NEPU)
关键词
集气支线
拓扑优化
R函数
障碍
混合遗传算法
徐深气田
gas gathering branch pipe
topological optimization
R function
obstacle
hybrid genetic algorithm
Xushen gas field
