蜂群最佳化之實數問題求解研究

THUIR > College of Engineering > Department of Industrial Engineering and Enterprise Information > Theses and Dissertations > Item 310901/1406

Please use this identifier to cite or link to this item: http://140.128.103.80:8080/handle/310901/1406

Title:	蜂群最佳化之實數問題求解研究
Other Titles:	Study of Real Function Problem Solution of MBO
Authors:	莊惟丞 Chuang, Wei-Chen
Contributors:	曾宗瑤;張炳騰 Tseng, Tsueng-Yao;Chang, Ping-Teng 東海大學工業工程與經營資訊學系
Keywords:	蜂群演算法;複合型啟發式演算法;函數特性;全域搜尋;區域搜尋 Marriage in Honey Bees Optimization;Multi-Heuristic algorithm;Function Characteristic;Global search;Local search
Date:	2009
Issue Date:	2011-02-25T01:55:17Z (UTC)
Abstract:	啟發式演算法發展至今已有許多相當成熟的模型方法，然而尚無一通用的演算法可解決所有的問題函數，每種方法皆有其利弊。故在實際應用上為了截長補短，衍生出許多複合型態的方法。而蜂群最佳化則是藉由模仿蜜蜂繁衍的行為模式所發展的一套演算法則，利用蜜蜂繁衍的機制達到演化最佳。目前被大量應用在最佳化、排程、分群、資料探勘等議題。其結合了退火函數、交配操作，如此全域搜尋的特性，保有求解的複雜度；另與其他單純的搜尋方法結合，在區域搜尋上做競爭，面對不同問題能有更好的適應能力，達到快速收斂的效果。本研究針對不同函數特性與數學環境，探討蜂群最佳化的演化方法，歸納其求解特性，瞭解其中不同啟發式演算法組合邏輯。並提出實數問題的參數設定，經由實驗驗證的結果顯示，蜂群最佳化在實數變數底下也能有相當好的收斂效果。證明如此的結合方法，不論單一或多重區域解，皆能擁有優秀的全域搜尋能力以及穩定的最佳解收斂效率，是一優良的演算法。 Heuristic algorithm develop have a lot of familiar model method already so far, but does not have a common algorithm that can solve all problem functions yet. Each kind of method all has its advantages and disadvantages. In order to draw on the strong points to offset the weaknesses in practical application, derive a lot of methods of compound attitude out. Marriage in Honey Bees Optimization (MBO) is by imitating a set of algorithms of development of manner that the honeybee multiplies. The mechanism of utilizing honeybees to multiply reaches and evolves the best. Apply optimization to, production schedule, clustering, data mining topic that in a large amount at present. It has combined annealing function and mating operation, such characteristic of global searched, keep the complexity solved. Bind with other simple search patterns separately, do the competition in local searching, there can be better adaptive capacity in the face of different problems, get the result of convergence fast. This research directs against different function characteristics and mathematics environment, probe into the evolution method that MBO, sum up it ask characteristic of solving and understand with heuristic to perform algorithm make logic up it. The parameter of putting forward the real number question is set up, the result that proves via the experiment reveals, the MBO in the real number parameter there can be quite good convergence results too. Prove the combination method like MBO, no matter solve in the single or multiple local solutions, there can all be outstanding global search capability and steady solving efficiency. It is a fine algorithm of performing.
Appears in Collections:	[Department of Industrial Engineering and Enterprise Information] Theses and Dissertations

Files in This Item:

File	Size	Format
097THU00030054-001.PDF	1379Kb	Adobe PDF	1488	View/Open

Loading...