「機組調派」是電力系統經濟運轉的重要課題之一。主要的目的是符合各項必要的限制條件下，能夠使總火力發電成本達到最低成本之要求。
本論文中，提出一個以基因演算法及倒傳遞網路法作短期火力發電機組之調派。基因演算法乃根據自然進化原則所發展出之最佳化求解理論。在尋優的過程中，同時針對一組解而非單一點進行搜尋，某一解尋優至另一解採用隨機方式之移動規則，可避免陷入局部最小值內。類神經網路法具有快速穩定等優點，本論文以倒傳遞網路法完成類神經網路，並以基因演算法所求得最佳機組組合狀態做為目標輸出值。
在考慮固定的電力系統下，可透過類神經網路計算達到即時反應的能力，當系統架構改變時，可透過基因演算法重新評估最佳的機組調派，期能改善傳統方法的缺點。
本論文以六部機組之電力系統為例進行能力評估，結果顯示，基因演算法在求解最佳機組調派的問題上，比傳統的方法更接近整體的最佳解，而類神經網路法不僅能接近基因演算法所得的解，在處理速度上更快於其他方法。

Unit commitment is one of the most important subjects with respect to the economical operation of power systems, which attempts to minimize the total thermal generating cost while satisfying all the necessary restrictive conditions.
　　This thesis proposes a short-term thermal generating unit commitment by genetic algorithm and back propagation network. Genetic algorithm is based on the optimization theory developed from natural evolution principles, and in the optimization process, seeks a set of solutions simultaneously rather than any single one by adopting stochastic movement rule from one solution to another, which prevents restriction to fractional minimal values. Neural networks method outperforms in speed and stability. This thesis uses back propagation network method to complete neural networks and sets the optimal unit combination derived from genetic algorithm as the target output.
　　Under fixed electrical systems, instant responsiveness can be calculated by neural networks. When the systematical architecture changes, genetic algorithm can be applied to re-evaluation of the optimal unit commitment, hoping to improve the pitfalls of traditional methods.
　　This thesis takes the power system of six units for example to conduct performance assessment. The results show that genetic algorithm provides solutions closer to the overall optimal solution than traditional methods in optimizing unit commitment. On the other hand, neural networks method can not only approximate the solution obtained by genetic algorithm but also process faster than any other methods.

目錄
中文摘要∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙I
英文摘要∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙III
目錄∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙V
圖目錄∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙IX
表目錄∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙XI

第一章 緒論
　　1.1 　研究背景∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙1
　　1.2　 研究動機∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙3
　　1.3 　論文架構∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙4
第二章 機組調派問題與傳統解法
　　2.1 　前言∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙5
　　2.2 　目標函數∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙6
　　2.3 　機組選定的限制條件∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙7
　　　　　2.3.1 負載供需平衡∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙7
　　　　　2.3.2 機組發電輸出的上下限∙∙∙∙∙∙∙∙∙∙∙7
　　　　　2.3.3 備轉容量∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙8
　　　　　2.3.4 最少運轉時間∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙8
　　　　　2.3.5 最少停機時間∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙9
　　　　　2.3.6 罰點因數∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙9
　　2.4 　優先次序表列法∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙10
　　2.5 　前向動態規劃法∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙13
第三章　基因演算法求目標輸出值
　　3.1 　前言∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙15
　　3.2 　理論分析∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙16
　3.2.1 變數轉換∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙17
　 3.2.2 初始狀態∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙17
　 3.2.3 計算適合度函數∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙18
　 3.2.4 挑選∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙19
　3.2.5 複製∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙20
　3.2.6 交配∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙20
　 3.2.7 突變∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙21
　　3.3 　基因演算法求目標輸出值∙∙∙∙∙∙∙∙∙∙∙∙∙∙22
　3.3.1 機組狀態編碼與解碼函數∙∙∙∙∙∙∙∙23
　　　　　3.3.2 本文所提適合度函數∙∙∙∙∙∙∙∙∙∙∙∙24
　　　　　3.3.3 模糊系統架構∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙24
　　　　　3.3.4 基因演算法∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙25
第四章　倒傳遞網路結合基因演算法在機組調派上的應用
　　4.1 　前言∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙28
　　4.2 　倒傳遞網路的模型架構∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙29
　　4.3 　倒傳遞網路的學習演算法∙∙∙∙∙∙∙∙∙∙∙∙∙∙31
　　4.4 　網路參數的決定∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙38
　　4.5 　訓練樣本∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙40
　　4.6 　網路訓練的評估指標∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙41
　　　　　4.3.1 訓練(測試)範例總錯率∙∙∙∙∙∙∙∙∙∙41
　　　　　4.3.2 訓練(測試)範例誤差均方根∙∙∙∙∙∙∙42
　　4.7 倒傳遞網路結合基因演算法在機組調派上的應用∙∙∙∙∙∙∙∙∙43
第五章　結果與討論
　　5.1 　前言∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙48
　　5.2 　基因演算法各項參數進一步分析∙∙∙∙∙∙∙∙50
　5.2.1 交配率及突變率∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙50
　　　　　5.2.2 群體數目對計算的影響∙∙∙∙∙∙∙∙∙∙51
5.2.3 不同適合度函數的執行結果∙∙∙∙∙∙∙53
　　　 　5.2.4 比例係數對執行結果的影響∙∙∙∙∙∙∙56
　　5.3　 倒傳遞網路學習速率對訓練的影響∙∙∙∙∙∙∙59
5.4 系統測試的結果∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙61
第六章　結論與未來展望
6.1 結論∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙62
6.2 未來展望∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙63
參考文獻∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙64
附錄A∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙68

圖目錄
圖2.1 前向動態規劃法∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙13
圖3.1 基因演算法執行程序∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙16
圖3.2 交配過程∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙20
圖3.3 突變過程∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙21
圖3.4 基因演算法流程∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙22
圖3.5 編碼與解碼∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙23
圖3.6 模糊系統架構∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙25
圖3.7 機組調派問題的交配運算∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙26
圖4.1 典型三層倒傳遞網路∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙29
圖4.2 類神經網路訓練階段∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙32
圖4.3 S函數曲線∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙34
圖4.4 機組選定階段∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙38
圖4.5 訓練樣本選取∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙40
圖4.6 總錯率∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙41
圖4.7 誤差均方根∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙42
圖4.8 倒傳遞網路處理機組調派的流程∙∙∙∙∙∙∙∙∙∙∙44
圖5.1 群體數目對計算的影響
(a) 群體數目=10∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙51
(b) 群體數目=20∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙52
(c) 群體數目=30∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙52
圖5.2 不同適合度函數的執行結果
(a) 線性等級∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙53
(b) 非線性等級∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙54
(c) 比例因數∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙54
(d) 本論文所提適合度函數∙∙∙∙∙∙∙∙∙∙∙∙∙∙55
圖5.3 比例係數對執行結果的影響
(a) k=100∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙56
(b) K=500∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙57
(c) K=1000∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙57
(d) K=2000∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙58
圖5.4 不同學習速率對訓練的影響
(a) η=0.9∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙ 59
(b) η=0.7∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙59
(c) η=0.5∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙60

表目錄
表2.1 不同機組的特性∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙10
表2.2 滿載平均成本∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙11
表2.3 優先次序表列∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙11
表2.4 所有組合狀態∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙14
表5.1 各發電機組之特性∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙48
表5.2 各時段之負載需求量∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙49
表5.3 交配率演算法則∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙50
表5.4 突變率演算法則∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙50
表5.5 各種不同方法的執行結果∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙61

[1] C.K. Pang, G.B. Sheble, and F. Albuyeh,“Evaluation of Dynamic Programming
Based Methods and Multiple Area Representation for Thermal Unit
Commitments,”IEEE Trans. on Power Appratus and System. Vol. PAS-100,
No.3, March 1981, pp.1212-1218.

[2] R.M. Burns and C.A. Gibson,“Optimization of Priority List for a
Unit Commitment Program,”IEEE Power Engineering Society Summer Metting,
Paper No. A75453-1, 1975.

[3] A. Merlin and P. Sandrin,“A New Method for Unit Commitment at Electricite
de France,”IEEE Trans. on Power Appratus and Systems, Vol. PAS-102, No.
5, May 1983, pp.1218-1225.

[4] W.L. Snyder, H.D. Powell, and J.C. Rayburn,“Dynamic Programming Approach
to Unit Commitment,”IEEE Trans. On Power Systems,Vol. PWRS-2, No.2, May
1987, pp.339-350.

[5] H. Sasaki, M. Watanable, J. Kabokawa, N. Yorino and R. Yokoyoma, “A
Solution method of Unit Commitment by Artificial Neural Networks,” IEEE
Trans. on Power Systems, Vol. PWRS-7, No. 3, August 1992, pp.974-981.

[6] F. Zhuang and F.D. Galian,“Unit Commitment by Simulated Annealing,”IEEE
Trans. on Power Systems, Vol.5, No.1, February 1990, pp. 311-317.

[7] X. Yin and N. Germay,“Investigation on Solving the Load Flow Problem by
Genetic Algorithms,”Electrical Power System Research, Vol.22, 1991,
pp.151-163.

[8] 葉怡成, ”類神經網路模式應用與實作,”台北儒林圖書有限公司，民國八十二年，第
267~312頁，第451~470頁。

[9] J.M. Zurada,“Introduction to Artificial Neural Systems,”Info Access
Distribution Pte Ltd, 19992, pp. 163-218.

[10] C. Wang and S.M. Shahidehpour,“Effects of Ramp-Rate Limits on Unit
Commitment and Economic Dispatch,”IEEE Trans. on Power Systems, Vol.8,
No.3,1993, pp.1341-1357.

[11] H. Sasaki, M. Watanabe, and R. Yokoyama,“A Solution Method of Unit
Commitment by Artificial Neural Networks,”IEEE Trans. on Power Systems,
Vol.7, No.3,1992, pp.974-981.

[12] Z. Ouyang, and S.M. Shahidehpour,“A Hybrid Artificial Neural Network-
Dynamic Programming Approach to Unit Commitment,”IEEE Trans. on Power
Systems, Vol.7, No.1, 1992, pp.236-242.

[13] Li.S. Shahidehpour, S.M. and C. Wang, "Promoting the Application of
Expert Systems in Short-Term Unit Commitment,”IEEE Trans. on Applied
Supercondutivity, Vol.3, 1993, No.1, pp.286-292.

[14] G.B. Sheble,“Solution of the Unit Commitment Problem by the Method of
Unit Periods,”IEEE Trans. on Power Systems, Vol.5, No.1,1990 pp.257-260.

[15] E. Handschin, and H. Slomski,“Unit Commitment in Thermal Power Systems
with Long-Term Energy Constraints,”IEEE Trans. on Power Systems, Vol.5,
No.4,1990 pp.1470-1477.

[16] S. Tong, K. Shahidehpour, S.M. and Z. Ouyang,“A Heuristic Short-Term
Unit Commitment,”IEEE Trans. on Power Systems, Vol.6, No.3,1991 pp.1210-
1216.

[17] A.I. Cohen and M. Yoshimura,“A Branch-and-Bound Algorithm for Unit
Commitment,”IEEE Trans. on Power Appratus and Systems, Vol. PAS-No.2,
February 1983, pp.444-451.

[18] F. Zhuang and F.D. Galian,“Towards a More Rigorous and Practical Unit
Commitment by Lagrangian Relaxation,”IEEE Trans. on Power System, Vol.
PWRS-3, No. 2, May 1988, pp.1218-773.

[19] K.Nara, A. Shiose, M.Kitagawa, and T. Ishihara,“Implementation of
Genetic Algorithm for Distribution Systems Loss Minimun Re-
Configuration,”IEEE Trans. Power System, Vol. PWRS-7, No.3, August 1992,
pp.1044-1051.
[20] R.N. Wu,“A Practical Method for Short-Term Hydro-Thermal-Interchange
Scheduling,”Ph.D. Thesis, Department of Electrical Engineering, The
University of New Brunswick,1989, pp.44-52.

[21] A.J. Wood and B.F. Wollenberg,“Power Generation Operation and Control,”
John Wiley and Sons, New York, 1984, pp.111-153.

[22] S.K. Tong, S.M. Shahidehpour, and Z. Ouyang,“A Heuristic Short-Term Unit
Commitment,”IEEE Trans. on Power Systems, Vol.6, No.3,1991 pp.1210-1216.

[23] Z. Ouyang and S.M. Shahidehpour,“An Inteligent Dynamic Programming for
Unit Commitment Application,”IEEE Trans. on Power Systems, Vol.6,
No.3, 1991, pp.1203-1209.

[24] F. Zhuang and F.D. Galiana,“Towards a More Rigorous and Practical Unit
Commitment by Lagrangian Relaxation,”IEEE Trans. on Power Systems,
Vol.3, No.2, 1988, pp.763-773.

[25] S. Virmani, E.C Adrian,. K. Imhof, and S. Mukherjee,“Implementation of a
Lagrangian Relaxation Based Unit Commitment Problem,”IEEE Trans. on
Power Systems, Vol.4, No.4, 1989, pp.1373-1380.

[26] F. Zhuang and F.D. Galiana,“Unit Commitment by Simulated Annealing,”
IEEE Trans. on Power Systems, Vol.5, No.1, 1990, pp.311-318.

[27] C.C. Su, and Y.Y. Hsu,“Fuzzy Dynamic Programming : An Application to
Unit Commitment,”IEEE Trans. on Power Systems, Vol.6,1991, No.3, pp.1231-
1237.

[28] Z. Ouyang and S.M. Shalidelpour,“A Hybrid Artificial Neural Network-
Dynamic Programming Approach to Unit Commitment,”IEEE Trans. on Power
Systems, Vol.7, No.1, 1992, pp.236-242.

[29] C. Wang and S.M. Shahidehpour,“Effects of Ramp-Rate Limits on Unit
Commitment and Economic Dispatch,”IEEE Trans. on Power Systems, Vol.8,
No.3, 1993, pp.1341-1357.

[30] S. Vemuri and L. Lemonidis,“Fuel Constrained Unit Commitment,”IEEE
Trans. on Power Systems, Vol.7, No.1, 1992, pp410-415.

[31] 盧炳勳,曹登發譯, ”類神經網路理論與應用,”台北, 全華科技圖書公司, 民國八十
一年,第105~127頁.

[32] S.A. Kazarlis, A.G. Bakirtzis, and V. Petridis,“A Genetic Algorithm
Solution To The Unit Commitment Problem,”IEEE Trans. on Power Systems,
Vol.11, No.1,1996, pp83-92.

[33] 柯志諭,“類神經網路在火力機組選定之應用,”國立台灣工業技術學院, 電機工程
研究所碩士論文, 民國八十三年.

[34] 廖國清, 曹大鵬,“以混合基因演算法及模糊系統法作短期火力機組排程,”第二十
屆電力研討會, 1999, pp. 957-961.