本文關鍵詞： 粒子群混合優(yōu)化算法 給水管網(wǎng) 動態(tài)調整 自適應懲罰函數(shù) 決策系統(tǒng)　出處：《北京工業(yè)大學》2014年碩士論文　論文類型：學位論文

【摘要】：城市給水管網(wǎng)系統(tǒng)是城市建設和工業(yè)生產的重要基礎設施，，隨著城市規(guī)模的擴大，管網(wǎng)系統(tǒng)的規(guī)模也逐漸呈現(xiàn)出復雜化、大型化的發(fā)展趨勢，管網(wǎng)系統(tǒng)的建設資金投入也隨之增加。一般給水管網(wǎng)的投資占到工程總投資的四分之三，通過優(yōu)化計算，可以節(jié)省工程投資的5%—10%，具有巨大的經濟效益和現(xiàn)實意義。如何對給水管網(wǎng)進行科學的優(yōu)化設計已經成為國內外專家關注的問題。 粒子群優(yōu)化算法機制簡單、可調參數(shù)少、易于實現(xiàn)，具有較強的全局收斂能力，不需要借助問題的特征信息等特點，因此文中采用粒子群優(yōu)化算法對城市給水管網(wǎng)進行優(yōu)化設計。論文的主要研究工作如下： 1.針對標準粒子群優(yōu)化算法在優(yōu)化給水管網(wǎng)問題中存在易陷入局部最小難以尋求最優(yōu)解的問題，文中在分析粒子群優(yōu)化算法的慣性權重對算法性能影響的基礎上，提出了一種基于粒子聚集程度動態(tài)調整慣性權重的策略，該調整策略利用種群在進化過程中粒子分布信息動態(tài)改變慣性權值，充分平衡了算法在優(yōu)化過程中全局探索能力和局部開發(fā)能力，提高了算法的尋優(yōu)精度。利用4個典型的基準測試函數(shù)對改進的算法進行測試。測試結果表明，改進的算法與傳統(tǒng)的ω調整策略相比更能夠適應粒子動態(tài)搜索的性能。 2.為了解決給水管網(wǎng)優(yōu)化設計這樣一類帶有約束條件的、高度非線性的離散組合優(yōu)化問題，提出一種改進型粒子群混合優(yōu)化算法。該算法在動態(tài)調整慣性權重的改進基礎之上，將極值優(yōu)化算法引入到改進粒子群算法中，利用極值優(yōu)化算法精細的局部搜索能力，增加種群多樣性并使算法有效地跳出局優(yōu)。根據(jù)給水管網(wǎng)優(yōu)化設計問題的最優(yōu)解分布的特點，采用自適應懲罰函數(shù)法處理管網(wǎng)優(yōu)化的約束條件，提高算法的搜索效率。將改進型粒子群混合優(yōu)化算法應用到給水管網(wǎng)優(yōu)化設計中，仿真結果表明，算法有較好收斂速度的同時還有效地避免了陷入局部最優(yōu)，并得到了更優(yōu)的工程造價。 3.開發(fā)了單機版給水管網(wǎng)智能計算決策系統(tǒng)軟件。該系統(tǒng)軟件使用VS2010和MATLAB作為開發(fā)工具，嵌入多種智能優(yōu)化算法及新提出的改進算法，對給水管網(wǎng)進行優(yōu)化計算，迅速、準確地得到管網(wǎng)總造價和最優(yōu)管徑的結果，實現(xiàn)在管線的鋪設階段根據(jù)實際情況提供優(yōu)化決策的功能，通過設置管網(wǎng)運行、造價的不同參數(shù)，分析對比數(shù)據(jù)結果為決策者提供更優(yōu)的選擇。通過C#設計界面，使系統(tǒng)軟件的操作具有直觀性和可視性。 論文針對粒子群優(yōu)化算法在解決給水管網(wǎng)優(yōu)化問題所存在的問題，對算法進行了改進研究，提出一種改進型粒子群混合優(yōu)化算法并將其應用在城市給水管網(wǎng)優(yōu)化設計中。在滿足供水要求的前提下，取得了較優(yōu)的結果。并利用Visual Studio2010和MATLAB等工具開發(fā)了給水管網(wǎng)智能計算決策系統(tǒng)，為城市規(guī)劃設計、建設等提供基礎數(shù)據(jù)支撐。
[Abstract]:Urban water supply network system is an important infrastructure for urban construction and industrial production. With the expansion of the scale of the city, the scale of the pipe network system also gradually presents a complex, large-scale development trend. The investment of water supply network accounts for 3/4 of the total investment of the project. Through the optimization calculation, it can save 5- 10% of the project investment. It is of great economic benefit and practical significance. How to optimize the water distribution network scientifically has become a problem that experts at home and abroad pay close attention to. Particle swarm optimization (PSO) has the advantages of simple mechanism, few adjustable parameters, easy implementation, strong global convergence ability and no need for the characteristic information of the problem. Therefore, particle swarm optimization algorithm is used to optimize the design of urban water distribution network. The main research work of this paper is as follows: 1. Aiming at the problem that the standard particle swarm optimization algorithm is easy to fall into the local minimum, it is difficult to find the optimal solution in the optimization of water supply network. On the basis of analyzing the influence of inertia weight of particle swarm optimization algorithm on the performance of the algorithm, a strategy of dynamically adjusting inertia weight based on particle aggregation degree is proposed in this paper. The adjustment strategy uses the information of particle distribution to change the inertia weight dynamically in the process of population evolution, and balances the global exploration ability and local development ability of the algorithm in the process of optimization. The improved algorithm is tested by using four typical benchmark functions. The test results show that the proposed algorithm can be used to improve the accuracy of the algorithm. 4 typical benchmark functions are used to test the improved algorithm. Compared with the traditional 蠅 adjustment strategy, the improved algorithm is more suitable for particle dynamic search. 2. In order to solve this kind of highly nonlinear discrete combinatorial optimization problem with constraints, the optimal design of water supply network is solved. An improved particle swarm optimization (PSO) hybrid optimization algorithm is proposed, which is based on the dynamic adjustment of inertia weight, and the extremum optimization algorithm is introduced into the improved PSO algorithm. By using the fine local search ability of the extremum optimization algorithm, the population diversity is increased and the algorithm effectively jumps out. According to the characteristics of the optimal solution distribution of the water supply network optimization design problem. Adaptive penalty function method is used to deal with the constraint conditions of pipe network optimization, and the search efficiency of the algorithm is improved. The improved particle swarm optimization algorithm is applied to the optimization design of water supply network. The algorithm has better convergence speed and effectively avoids falling into local optimum and gets better project cost. 3. The software of intelligent calculation and decision system for water supply network is developed, which uses VS2010 and MATLAB as developing tools. A variety of intelligent optimization algorithms and new improved algorithms are embedded to optimize the water supply network. The results of the total cost and the optimal diameter of the pipe network are obtained quickly and accurately. In the pipeline laying stage according to the actual situation to provide the function of optimal decision-making, through setting up the pipeline network operation, the cost of different parameters. Through the C # design interface, the operation of the system software is intuitive and visible. Aiming at the problem of particle swarm optimization algorithm in solving the problem of water supply network optimization, the improvement of the algorithm is studied in this paper. An improved particle swarm optimization algorithm is proposed and applied to the optimal design of urban water supply network. The intelligent calculation decision system of water supply network is developed by using Visual Studio2010 and MATLAB tools, which can be used for urban planning and design. Construction to provide basic data support.
【學位授予單位】：北京工業(yè)大學
【學位級別】：碩士
【學位授予年份】：2014
【分類號】：TU991.33;TP18

【參考文獻】

相關期刊論文 前5條

1 張頂學;關治洪;劉新芝;;一種動態(tài)改變慣性權重的自適應粒子群算法[J];控制與決策;2008年11期

2 李祥立 ,鄒平華;基于模擬退火算法的供熱管網(wǎng)優(yōu)化設計[J];暖通空調;2005年04期

3 喬俊飛;佟磊;潘廣源;;基于改進蟻群算法的Hanoi給水管網(wǎng)優(yōu)化設計[J];控制工程;2013年03期

4 朱家松,龔健雅,鄭皓;遺傳算法在管網(wǎng)優(yōu)化設計中的應用[J];武漢大學學報(信息科學版);2003年03期

5 杜繼永;張鳳鳴;李建文;楊驥;;一種具有初始化功能的自適應慣性權重粒子群算法[J];信息與控制;2012年02期

本文編號：1469864