基于智能优化算法的供水管网漏水点定位

Identification of Leakage Location Based on Modern Optimization Algorithms

摘要:

针对城市供水管网的爆管事故，基于管网水力学模型和管网水压监测点的监测信息，利用3种智能优化算法——杜鹃算法、遗传算法和粒子群算法，建立了爆管定位模型，并在2个算例管网中进行了测试.2个算例管网的运行结果显示，杜鹃算法由于其调节参数少，快速搜寻能力强，在爆管定位的智能优化算法中显示了更优秀的定位能力，在小规模算例管网中可以实现对爆管点的定位，在较大规模的算例管网中也可以实现90%以上的定位寻优效果，显示了强大的算法应用可扩展性.

Abstract:

The paper presents an algorithm for the location of sudden bursts in combination with both continuous monitoring of pressure and hydraulic model computation. The Cuckoo Search, genetic algorithm and particle swarm approaches were employed to identify the location of leakage. Their performances were compared. The results show that the Cuckoo Search has a better performance in terms of searching speed and parameter requirement among the three optimization algorithms. The case study results reveal the potential of the proposed burst location identification technique in a real life water distribution system.