基于粒子群算法的水资源优化配置

Rational Allocation of Water Resources Based on Particle Swarm Optimization

本文基于可持续发展理论,以社会、经济和环境的综合效益最大为目标,建立了区域水资源优化配置模型。根据模型的特点,采用粒子群算法(PSO)对模型进行求解。针对粒子群算法的迭代原理,通过对粒子编码方法、适应度函数构造和约束条件处理等环节的改进,构成了用于多目标有约束条件模型求解的粒子群优化算法。不仅拓展了粒子群优化算法的应用领域,同时也为复杂多目标模型的求解提供了一种新途径。本文以北京市为例,借助本文提出的模型,得到了该市2010、2020和2030年三个水平年在50%保证率下的水量配置方案。优化结果表明,该算法应用于水资源优化配置中是合适的。

On the basis of sustainable development theory, a regional water resources allocation model with a maximum social, economic and environment benefit was developed in this study. The Particle Swarm Optimization （PSO） was used to solve the model. According to the iteration principle of PSO, an improved PSO algorithm for multi-objective model was proposed by particle coding, fitness function, and constraint condition. This method does not only expand the application of PSO, but also provide a new way of solving the complex and multi-objective model. Beijing was taken as a case study in this paper. The results of water resources allocation in the study were analyzed by scenario simulation for three benchmark years （2010, 2020 and 2030） at reliability of 50%. The results show that both the model and method are practicable and suitable.