查詢結果分析
來源資料
頁籤選單縮合
題 名 | 以衍生型均值退火神經網路作短期水火力發電調度=Short-Term Hydrothermal Scheduling Using an Extended Mean Field Annealing Neural Network |
---|---|
作 者 | 梁瑞勳; 康芳菖; | 書刊名 | 科技學刊 |
卷 期 | 8:2 1999.04[民88.04] |
頁 次 | 頁117-123 |
分類號 | 448.14 |
關鍵詞 | 衍生型均值退火神經網路; 水火力發電調度; Extended mean field annealing neural network; Hydrothermal scheduling; |
語 文 | 中文(Chinese) |
中文摘要 | 在本論文中,提出以衍生型均值退火神經網路法作短期水火力發電調度。在電力 系統中,水火力發電調度的主要目的是如何利用有限的水資源條件下,滿足必要的限制條件 ,所求出各個時段的水力及火力機組的發電量,使得火力機組的總發電成本為最低。也就是 說,在未來的 K 小時中去決定電力系統中水力、火力發電機組的最佳發電調度。 類神經網 路的主要優點是其知識分佈於神經元中並以平行方法作資訊處理,而衍生型均值退火神經網 路又共同包含有模擬退火法之近似最佳解或整體最佳解的品質與類神經網路法之快速收斂的 性質之優點。故在本論文中,我們將應用衍生型均值退火神經網路法來求出最佳化水火力 發電調度的結果,並由兩個測試範例所得到的結果證實此法於水火力發電調度問題上的有效 性。 |
英文摘要 | In this paper, an extended mean field annealing neural network approach is used for the short-term hydrothermal scheduling. In power systems, the major goal of hydrothermal scheduling is to determine the optimal amounts of generated powers for the hydro units and thermal units to minimize the total fuel cost of thermal units and subject to some practical constraints. That is easy to say, we want to find the optimal hydrothermal scheduling in the power system for the next K hours in the future. A major advantage of the artificial neural network is that the domain knowledge is distributed in the neurons and information processing is corroborated out in a paralleldistributed manner. Therefore, it is rather efficient for the artificial neural network to reach the desired solutions. The extended mean field annealing neural network has the advantages which combine both good solution quality for simulated annealing and rapid convergence for artificial neural network. So we use the extended mean field annealing neural network to find the solution of short-term hydrothermal scheduling. From the results of two test examples, it is found that the extended mean field an nealing neural network has the ability to find out the optimal solution of the hydrothermal scheduling. |
本系統中英文摘要資訊取自各篇刊載內容。