摘要翻譯--改進的蟻群算法在車輛路徑優(yōu)化中的研究與應(yīng)用

1、<p>　　改進的蟻群算法在車輛路徑優(yōu)化中的研究與應(yīng)用</p><p><b>　　摘要</b></p><p>　　車輛路徑問題是物流領(lǐng)域所關(guān)注的熱點問題，因為車輛路徑問題（Vehicle Routing Problem，VRP）具有復(fù)雜性與多樣性，如何合理高效的規(guī)劃車輛路徑以最低的成本完成貨物運輸，這是一個極富挑戰(zhàn)性的問題。因此，研究物流作業(yè)中配送車輛的

2、路徑優(yōu)化問題，不僅具有重要的理論意義，而且還有很高的實用價值。</p><p>　　在綜述了國內(nèi)外有關(guān)車輛路徑優(yōu)化問題的研究基礎(chǔ)上，選取蟻群算法，并對其加以改進，來實現(xiàn)車輛路徑的優(yōu)化功能。</p><p>　　蟻群算法（Ant Colony Optimization, ACO）是由意大利M. Dorigo等學(xué)者提出，通過模擬蟻群的覓食行為，構(gòu)造出一種基于種群的進化算法。蟻群算法選用分布式并

3、行的計算機機制，并具有很強的魯棒性，易于與其他方法相結(jié)合。所以，對其進行改進，并應(yīng)用到路徑優(yōu)化問題中，這是一種很好的嘗試。</p><p>　　本文的主要工作包括蟻群算法的改進及其在車輛路徑優(yōu)化問題中的應(yīng)用，主要研究內(nèi)容如下：</p><p>　?。?）對基本蟻群算法的原理與模型進行了學(xué)習(xí)，了解其實現(xiàn)步驟、算法特點，并建立起相應(yīng)模型；對于基本蟻群算法中的多個關(guān)鍵參數(shù)，進行了實驗分析。<

4、;/p><p>　?。?）基于基本蟻群算法的兩項缺點——收斂速度緩慢、易陷入局部最優(yōu)解，提出了一種改進方案。該改進算法借鑒了最大最小蟻群算法中限制信息素范圍的思想，可以有效抑制由于最長路徑與最短路徑所含信息量的巨大差距而引起的停滯現(xiàn)象。與此同時，引入了局部搜索、局部信息素更新策略，加快了蟻群算法的運算速度。在此基礎(chǔ)上，繼續(xù)改良信息素的全局更新機制，來實現(xiàn)算法更快搜索到全局最優(yōu)解的設(shè)想。</p><

5、p>　?。?）將改進后的蟻群算法應(yīng)用到基本車輛路徑問題與有時間窗的車輛路徑問題中，對比于使用原始蟻群算法所得結(jié)果，改良后算法的搜索性能有了顯著提升，對車輛路徑實現(xiàn)了優(yōu)化作用。</p><p>　　本文基于MATLAB.R2011a平臺，實現(xiàn)了蟻群算法的改進及其在車輛路徑優(yōu)化問題中的應(yīng)用。</p><p>　　The research and application of the im

6、proved Ant Colony Optimization in the Vehicle Routing optimization problem</p><p><b>　　Abstract</b></p><p>　　Vehicle routing problem has been a hot issue of logistics industry. Owing

7、 to the complexity and diversity of vehicle routing problem ,how rational and efficient planning of the vehicle at the lowest cost path to complete the transport of goods becomes very challenging problems. Therefore, the

8、 study of logistics operations in the distribution vehicle routing problem, not only has important theoretical significance, but also a high practical value.</p><p>　　On the basis of the research of the vehi

9、cle routing problem at home and abroad ,we select Ant Colony Optimization and improved to achieve the optimization of the vehicle path.</p><p>　　Ant Colony Algorithm (Ant Colony Optimization, ACO) was propos

10、ed by the Italian scholar M. Dorigo, etc., by simulating the foraging behavior of ant colonies constructed an evolutionary algorithm based on population. Ant colony Optimization select distributed parallel computer mecha

11、nism and has strong robustness, so combined with other methods easily. Therefore, it turns out to be an excellent attempt to improve and apply to the path optimization problem.</p><p>　　The main study includ

12、es the improvement and application of Ant Colony Algorithm in the vehicle routing problem.And the main research content is as follows:</p><p>　　(1) The study of principles and model of basic ant colony optim

13、ization， their implementation steps, the algorithm characteristics, and establish a corresponding model. Carrying out experimental analysis on several key parameters of basic ant colony optimization.. </p><p&g

14、t;　　(2)To solve the disadvantages based on two basic ant colony algorithm - slow convergence, easy to fall into local optimal solution an improved scheme is proposed. The improved algorithm draws on the max-min ant colon

15、y optimization the idea of ??limiting the scope of the pheromone.It can effectively suppress stagnation caused by the huge amount of information gap including longest path and the shortest path. At the same time, the int

16、roduction of local search, local pheromone update strategy to sp</p><p>　　(3) The improved algorithm which applied to the basic and with time windows vehicle routing problem compared to the results with the

17、original ant colony algorithm proceeds has been significantly improved and achieves the optimal vehicle routing effect.</p><p>　　This paper based on MATLAB.R2011a platform to achieve the improvement of ant c