用微粒群算法实现天然气管网运行最优化

Optimizing the Operation of Natural Gas Pipeline Network by Particle Swarm Optimization

以天然气公司的收益最大化为目标,建立了天然气管网优化运行的数学模型。传统的直接搜索法有网格法和复合形法,但这两种方法收敛速度较慢,工作量较大。采用以自适应惩罚函数为目标函数的微粒群算法,并对算法的收缩因子与自适应性加以修正。结合管网稳态分析的节点压力法,编制计算程序对模型进行求解,结果表明,该方法能够高效获得高性能的优化运行调度结果。

Taking the gas company's income as a goal, an optimal running mathematical model of natural gas pipeline network is established. Conventional direct searching method includes network method and compound method. But these methods involve large amount of workload which results in slowly convergent speed. The Particle Swarm Optimization (PSO) adopts a self-adapting penalty function as a target function and a shrink factor and self-adaptation of the algorithm is amended. Combined with node pressure technique, which is used for steady-state analysis of pipeline network, an applicable program is developed for the purpose of solving the model. Practical example of calculations shows that this method can get a highly optimized performance result effectively in pipeline network operation.