【作者】 杨毅；

【导师】 杨立中；

【作者基本信息】 西南交通大学 ， 环境工程， 2004， 硕士

【摘要】 水污染控制系统规划是水环境管理的核心内容，其实质是运用系统工程的思想和方法，分析和协调水污染控制系统各组成因素之间的关系，以便用较小的代价获得有效的或满意的水质效果，寻求最佳的污染控制实施方案。由此可见，水污染控制系统规划是协调环境、经济和社会所构成的具有多变量、多层次的复杂系统的一种寻优过程，在原理上可以归结为一个有约束非线性优化模型的求解。 遗传算法(GA)是一种基于自然选择原理和自然遗传机制的随机优化方法。它求解优化问题的基本思想是把问题的目标函数解释为生物对环境的适应性，把问题的一个解对应于生物群体中的一个个体。由当前群体开始，通过复制、交叉和变异操作，产生新一代群体。重复这个过程，直到得出满足条件的最优个体，该最优个体就代表了原优化问题的最优解。 在理论上，一定条件下遗传算法可以收敛到全局最优解，但在实际应用中往往很快收敛到局部极值点后算法就停滞不前，即出现所谓的早熟性收敛。针对上述问题，在对涪江绵阳段水环境概况进行分析和评价的基础上，本文引入了自适应算子和局部退化算子，对涪江绵阳城区段的污水控制系统进行了优化布局，并与传统规划方法——试探分解法的规划结果进行了比较。结果表明，通过改进后的遗传算法较好地改善了标准遗传算法的早熟现象，具有更好的全局搜索能力，是一种理论上可行、实践上具有深入研究开发价值的水污染控制系统优化方法。更多还原

【Abstract】 Planning for control system of water pollution, the main content of water environment government, aims at analyzing and coordinating relationship in composition factors of control system of water pollution with idea and method of system engineering in order to gain effective or satisfied water quality with smaller cost and find the best execute program. Therefore, planning of control system of water pollution which is to solve a constrained nonlinear planning issue is a process of seeking the best program which coordinates a multi-variable and multi-level complex system composed of environment, economy and society.Genetic Algorithms (GA) is a random optimization method based on natural selection theory and natural genetic mechanics. Its basic idea to solve optimization issue is to use creature’s adaptability to denote the object function and use individual of population to denote solution of issue. Newer population is generated from current population by copying, crossing and variation. This process is duplicated constantly until the best individual which represents the best solution of optimization issue is gained.In theory, GA can gain global optimum solution in a certain condition but always traps in deadlock after gaining local optimum solution, which is called as early mature. This paper applies self-adapting operator and local degeneracy operator to optimization layout of water pollution control system of Fujiang river in Mianyang city against problem mentioned above. The result that improved GA improves early mature phenomenon of standard GA and owns better global searching ability indicated that improved GA is a feasible optimization method of control system of water pollution in theory with profound researching and development value.更多还原