izing Switch)的交替動作來達成,這些的開關狀態是可藉由邏輯性質的描述來表示,同時因其最終之目標多為計算亦或是估算之結果,因此在配電規劃的問題可以此之明確的計算來表示成問題的目標函數(Object Function),如此便可以基因演算法(Genetic Algorithm, GA)來搜尋到最佳處理問題的解。因此,本論文以基因演算法作為處理配電系統之配電重構之問題對象之最佳工具,以搜尋出處理配電系統長期規劃(Long-Term Planning)之損失最小化(Loss Minimization)及負載平衡(Load Balance)等問題的最佳策略解。以往的傳統式基因演算法(Tradition GA)在處理此類似之組合性最佳化問題時,大部份以修複策略來達成不合法的解(Illegal Solution)的處理,但是,如此可能會造成搜尋時間的增加。本論文藉由對於不合法解的處置策略,進而衍生”有條件突變方法(Conditional Mutation Method)”,並以此為基礎發展出無性生殖之基因運算(Asexual GA Operator)架構,建構具有較佳收斂速度且保證搜尋到最佳解之改良式基因演算法(Improvement GA)。 摘要(英)鍵入您的論文摘要Distribution system planning is a type of multi-constraint combinatorial optimization pro
facts that several constraints have to be satisfied while the variables are discrete. The solutions of such problems can be achieved by adjusting the locations of tie switches and sectionalizing switches. Also the objective function can be described precisely. Because of these characteristics, Genetic Algorithm (GA) can be applied for solving the problems. Genetic Algorithm is used to solve loss minimization and load balancing problems. The solution proposed by traditional GA may result in illegal solution. This will increase the search time since extra step is needed to ensure that the solutions are all legal. In order to overcome this problem, this thesis proposes conditional mutation method. From this method, asexual GA is proposed. From the simulation, it is found that the proposed asexual GA operator can achieve the optimal solution quicker than traditional GA operators. 指導教授楊宏澤蔡孟伸Click Here 可以新增或移除口試委員資訊 論文檔案9178027-new.pdfClick Here 可以新增或移除檔案