基于改进混合蛙跳算法的给水管网改扩建优化模型研究

发布时间：2018-01-13 20:39

本文关键词：基于改进混合蛙跳算法的给水管网改扩建优化模型研究　出处：《重庆大学》2014年硕士论文　论文类型：学位论文

【摘要】：城市给水管网系统是城市的生命线工程，是连接用户与给水厂的桥梁。然而随着工作年限的增加，，给水管网系统出现了腐蚀严重、爆管频繁、二次污染问题，使得管网系统的改扩建十分必要而迫切。管网改扩建模型的科学与否不仅关乎居民的用水需求，也影响着城市给水管网系统的建设和运行维修费用。本文在前人研究的基础上，提出利用新颖的混合蛙跳算法来对给水管网系统改扩建进行优化。 1）混合蛙跳算法（SFLA）作为新兴的群智能随机搜索优化算法，结合了微粒群算法和模因算法的优点，可维持局部开发能力和全局寻优能力的平衡，具有概念简单、参数少、全局寻优能力强、鲁棒性好的特点。本文针对混合蛙跳算法的局部搜索随机性差，步长大小波动性大的缺陷，分别提出措施来改进混合蛙跳算法，使其更趋完善。 2）城市中长期需水量的准确预测是城市给水管网系统改扩建优化的前提和基础。本文利用SFLA来优化BP神经网络的连接权值和阈值，从而建立SFLA-BP网络，来预测城市的中长期需水量。以西南某市为实例，首先运用SPSS对影响水量的因素进行主成分分析来得到主要影响因素，然后利用SFLA-BP网络来进行需水量预测，较之BP网络和PSO-BP网络，SFLA-BP网络的计算时间更短，预测精度更高，从而验证了SFLA良好的计算性能和全局搜索能力。 3）基于改进的SFLA来建立管网改扩建的优化模型，并在目标函数中除经济性外，引入水力可靠性和熵值可靠性，同时利用分目标乘除法将多目标函数转化为单目标函数，以MATLAB为平台，X市为工程实例，用传统模型作为对照，计算两模型在最不利、消防和事故时的水力状况，进而比较两模型的经济性指标、可靠度指标和综合指标，结果表明，较之传统模型： ①优化模型得到的设计方案的平均年动力费用和新建水厂的年费用折算值，费用高一些，但管网建造维修年费用要低很多，总的来说，优化模型得到的设计方案比传统模型节省约6%的费用。 ②优化模型得到的设计方案的熵值可靠度与传统模型近似，但有更好的水力可靠度。 ③对于综合评价指标—单位可靠度所需的费用，优化模型的计算结果更低，验证了基于改进SFLA的给水管网改扩建优化模型在实际工程应用中的可行性和优越性。
[Abstract]:The urban water supply network system is the lifeline project of the city, it is the bridge between the user and the water supply plant. However, with the increase of the working life, the water supply network system has the problems of serious corrosion, frequent pipe burst and secondary pollution. It is very necessary and urgent to rebuild and extend the pipe network system. The scientific model of the network reconstruction and extension model is not only related to the water demand of the residents. On the basis of previous studies, a novel hybrid leapfrog algorithm is proposed to optimize the reconstruction and expansion of urban water supply network system. 1) the hybrid leapfrog algorithm (SFLAs), as a new swarm intelligence random search optimization algorithm, combines the advantages of particle swarm optimization algorithm and meme algorithm, and can maintain the balance between local development ability and global optimization ability. It has the advantages of simple concept, few parameters, strong global optimization ability and good robustness. In this paper, the local search randomness of the hybrid leapfrog algorithm is poor, and the step size is volatile. Measures are proposed to improve the hybrid leapfrog algorithm to make it more perfect. 2) accurate prediction of medium and long term urban water demand is the premise and foundation of urban water supply network reconstruction and expansion optimization. This paper uses SFLA to optimize the connection weight and threshold of BP neural network. SFLA-BP network is established to predict the medium and long term water demand of the city. Taking a city in southwest China as an example, the principal component analysis of the factors affecting the water quantity is carried out by using SPSS to get the main influencing factors. Then the SFLA-BP network is used to predict the water demand. Compared with BP network and PSO-BP network, the calculation time is shorter and the prediction accuracy is higher. Thus, the good computing performance and global search ability of SFLA are verified. 3) based on the improved SFLA, the optimization model of pipe network reconstruction and extension is established, and the hydraulic reliability and entropy reliability are introduced into the objective function besides economy. At the same time, the multi-objective function is transformed into a single-objective function by the method of sub-objective multiplication and division. Taking MATLAB as the platform and the city of X as an engineering example, the two models are calculated in the most disadvantageous way, compared with the traditional model. The hydraulic status of fire protection and accident is compared, and the economic index, reliability index and comprehensive index of the two models are compared. The results show that compared with the traditional model: 1 the average annual power cost of the design scheme obtained by the optimization model and the annual cost conversion value of the new water plant are higher, but the annual cost of pipe network construction and maintenance is much lower. The cost of the optimized model is about 6% less than that of the traditional model. (2) the entropy reliability of the optimization model is similar to that of the traditional model, but it has better hydraulic reliability. (3) for the cost of unit reliability, the result of optimization model is lower. The feasibility and superiority of the water supply network reconstruction and extension optimization model based on improved SFLA are verified.
【学位授予单位】：重庆大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TU991.33

【参考文献】