遗传算法在供水管网优化设计中的应用研究

Application of Genetic Algorithms in Optimizing Design of Water Supply Pipe Network

鉴于给水管网优化设计在给水工程中占有重要地位,国内外学者对其进行了广泛而深入的研究,提出了多种优化方法,诸如经典优化法、线性规划法、动态规划法、广义简约梯度法以及标准遗传算法。但这些方法在实际应用中均存在一定的局限性。该文着重研究了遗传算法作为一种强有力的搜索寻优工具在给水管网管径组合优化设计中的应用。遗传算法具有思路简单、适用广泛、全局寻优能力强等特点,目前在各个研究领域日益为人们关注。通过对可能的管径组合进行编码、适应度评价、选择、交叉、变异等一系列步骤,经过一段时间(一定代数)的进化后,可以找到最优的管径组合方案,从而避开了线性、非线性规划中的复杂理论。

The optimizing design of water supply pipe network has been broadly and deeply studied by many local and foreign scholars because of its important status in water supply engineering. The scholars advanced many kinds of optimizing methods, such as classic optimal method, linear programming method, dynamic programming method, generalized reduction gradient (GRG) method and simple genetic algorithms (SGA).These methods all have some limitation in the practical application. The application of genetic algorithm, as a powerful optimizing tool in the optimizing design of water supply networks is studied in this paper. As GA method is easy to understand and can solute varies of problems, moreover, it can effectively find the global solution, and it is drawing more and more attentions in most fields. The optimizing combination method of pipe diameters can be found after some time (some generations) through a series of steps including coding of possible pipe diameters combination, evaluation of fitness, selection, crossover and mutation. As a result, the complexity of linear programming and nonlinear programming can be eluded.