本論文針對電力系統進行最佳化短期火力機組排程,一般而言機組排程主要是在排定時間內,滿足發電機組及系統上各項限制條件下,有效地調度各台發電機使總發電成本最小;求解機組排程必須有效且迅速,才可做為調度人員參考依據,因此機組排程的求解方法已成為一種重要的課題。 機組排程演算法方面,本文採用二進制粒子群演算法,此演算法是由傳統粒子群演算法進行延伸,其特色是利用二位元的變數進行求解,用以解決二進制的最佳化問題;此外二進制粒子群演算法採用多點搜尋方式,且對於搜尋過的最佳解賦予記憶性,粒子也會相互學習,因此較容易針對全域最佳解進行搜尋。 最後本論文以二進制粒子群演算法進行兩個案例的最佳化機組排程模擬,以IEEE 30-bus與IEEE 57-bus的電力系統做24小時負載的機組排程,並與基因演算法、量子基因演算法做發電成本、輸電線路損失以及運算時間的比較,模擬結果可驗證二進制粒子群演算法適合用於求解最佳化短期火力機組排程。
The thesis has developed a new method which can obtain the better solution for short term unit commitment of power system. Under various constrains, the economic dispatch and unit commitment have to be optimized in order to saving the production costs. The method to process these targets has to be fast and effectively to give the best decision making in time for the power system operators. Therefore, to search the best solution method has become an important topic. In terms of set scheduling algorithm, this study used binary Particle Swarm Optimization, which is developed from traditional Particle Swarm Optimization. It is characterized by binary variable-based solving for binary optimization problem. In addition, it adopts multipoint search pattern, and endues the searched optimal solution with memory. The particle learn from each other, thus, it is capable to search for global optimal solution. Finally, this study used binary Particle Swarm Optimization to simulate optimal set scheduling of two cases, and used IEEE 30-bus and IEEE 57-bus power system to schedule 24-hour load set. The generating cost, electric transmission line loss and computing time were compared with that of genetic algorithm and quantum genetic algorithm. The simulation results have proved that the binary Particle Swarm Optimization is applicable to solving optimal short-term thermal power set scheduling more effectively.
為了持續優化網站功能與使用者體驗,本網站將Cookies分析技術用於網站營運、分析和個人化服務之目的。
若您繼續瀏覽本網站,即表示您同意本網站使用Cookies。