### 博碩士論文 105521075 完整後設資料紀錄

 DC 欄位 值 語言 DC.contributor 電機工程學系 zh_TW DC.creator 馬冠群 zh_TW DC.creator Guan-Chun Ma en_US dc.date.accessioned 2018-7-4T07:39:07Z dc.date.available 2018-7-4T07:39:07Z dc.date.issued 2018 dc.identifier.uri http://ir.lib.ncu.edu.tw:88/thesis/view_etd.asp?URN=105521075 dc.contributor.department 電機工程學系 zh_TW DC.description 國立中央大學 zh_TW DC.description National Central University en_US dc.description.abstract 傳統的粒子群演算法的學習策略常被認為是讓粒子失去多樣性的原因，也因為多樣性的喪失而引發過早的收斂。因此本論文中提出了兩個方法，來改善上述提到的問題。第一個方法是「單維度模仿」(Single Dimension Imitation, SDI)，此方法是用來提高粒子的收斂速度以及結果的準確度，並且將它運用在傳統的粒子群演算法上。第二個方法是「區間搜索」(Interval Search, IS)，此方法是用來提高粒子群的多樣性，接著再提出「混合經驗解」(Hybrid Experience,H_(i,best) )來增強IS結果的準確性。接著我們做一些實驗，來證明這兩個方法的可行性。然後我們發現了在處理比較簡單的問題時第一個方法比第二個方法優秀，而在處理比較複雜的問題時第二個方法比第一個方法優秀，因此我們決定將這兩個方法結合在一起，形成一個新的演算法。最後我們將使用一些測試函數與其它已提出的演算法進行比較。從實驗結果得知，本論文提出的新的改良方法，不論在結果還是在速度上面，都比其他演算法還要優秀許多。 zh_TW dc.description.abstract The learning strategy in the canonical particle swarm optimization (PSO) algorithm is generally considered to be the reasons for the loss of diversity, and this leads to premature convergence. In this thesis, two methods are proposed to improve the above mentioned problems. The first method is called single dimension imitation. This method is used to improve the convergence speed of the particles and the accuracy of the results, and it is used in the canonical particle swarm optimization (PSO) algorithm. The second method is the Interval Search, IS, which is utilized to improve the diversity of the particle swarm. In order to enhance the accuracy of the results of IS, we propose a new strategy which called Hybrid Experience,H_(i,best), and add it to the IS. Since we found that the first method was better at dealing with simpler problems than the second, while the second method was superior to the first in dealing with more complex problems, so we decided to combine the two methods to form a new algorithm. Finally, we will use some test functions to compare with other proposed algorithms. According to the experimental results, the new improved method proposed in this thesis is much better than other algorithms in both the results and the speed. en_US DC.subject 粒子群演算法 zh_TW DC.subject 區間搜索 zh_TW DC.title 混合式區間搜索粒子群演算法 zh_TW dc.language.iso zh-TW zh-TW DC.title A Particle Swarm Optimization With Hybrid Interval Search en_US DC.type 博碩士論文 zh_TW DC.type thesis en_US DC.publisher National Central University en_US