本論文旨在利用蒙地卡羅模擬法估計未來電力系統發電成本與輸電
線損，藉此模擬，能模擬真實系統操作，並較現有方法接近真實值。首先
假設機組故障及修復的時間是指數分配，利用此分配產生數個隨機觀察值
，以模擬發電機組在運轉過程中，何時發生故障，一旦故障又何時修復等
隨機特性。因機組狀態組合為數眾多，另採用統計抽樣與估計方法以減少
計算時間。另一方面，在估算發電成本與輸電線損需進行大量的機組約定
與經濟調度，為提高計算效率與收斂性，本論文提出新經濟調度法應用在
蒙地卡羅模儗。因此，本論文較現有蒙地卡羅模擬法在降低估計發電成本
誤差有效率有四項理由：第一，模擬機組修復的隨機特性。第二，結合統
計抽樣理論。第三，提出新的分層抽樣方法。第四，提出新的統計估計二
方法。　　　蒙地卡羅方式的發電成本估計法主要在抽取適當的機組狀態
組合之少數樣本，再針對各樣本計算發電成本，最後利用少數樣本的成本
值估計母體的成本均值。同時本論文亦改善傳統旋積運算法中最常使用的
混合常態近似法較傳統方法在計算效率與精確度上較佳，並與蒙地卡羅法
在理論與數值比較。　　　在輸電線損估計方面，應用蒙地卡羅法研擬合
理的管制目標，作為未來調度計劃的重要參考依據。此蒙地卡羅法能考慮
未來網路結構變動與增建機組，能準確推估線損。

CONTENTS
ABSTRACT
ABSTRACT
LIST OF FIGURES
LIST OF TABLES
ABBREVIATIONS
CHAPTER 1: INTRODUCTION
1.1 Background of The Study
1.2 Scope of The Dissertation
1.3 Contributions
1.4 Outline of The Dissertation
CHAPTER 2: CONVOLUTION APPROACH TO PRODUCTION SIMULATION
2.1 Introduction
2.2 Basic Concept of Convolutional Production Simulation
2.3 Brief Review of Mixture of Normal Approximation
2.4 Proposed Load Categorization Algorithm
2.5 Numerical Tests
2.6 Summary
CHAPTER 3: GENERATION SCHEDULING ALGORITHM FOR USE IN MONTE CARLO PRODUCTION SIMULATION
3.1 Introduction
3.2 Simplified Scheduling Algorithm
3.3 Unit Commitment
3.4 Proposed Economic Dispatch Algorithm
CHAPTER 4: VARIANCE REDUCTION TECHNIQUES FOR MONTE CARLO SIMULATION
4.1 Introduction
4.2 Monte Carlo Simulation
4.3 Crude Monte Carlo Approach
4.4 Stratified Sampling
4.5 Control Variate Method
4.6 Antithetic Variate Method
4.7 Linear Regression Estimation
4.8 Ratio Estimation
4.9 Estimation by Sample''s Rank Statistics
4.10 Summary
CHAPTER 5: PROPOSED MONTE CARLO PRODUCTION COST ESTIMATION
5.1 Introduction
5.2 Problem Description
5.3 Proposed Stratification Process
5.4 Optimal Sampling
5.5 Control Variate Sampling and Linear Regression Estimator-Conventional Approach
5.6 Extension to Combine with Antithetic Sampling
5.7 Proposed Estimation by Sample''s Rank Statistics
5.8 Computer Implementation
5.9 Numerical Tests
5.10 Summary
CHAPTER 6: COMPARISON OF MONTE CARLO WITH CONVOLUTIONAL PRODUCTION SIMULATION
6.1 Introduction
6.2 Computing Efficiency
6.3 Simulation Accuracy
6.4 Numerical Test
6.5 Suggested Application Area
CHAPTER 7: EXTENSION OF MONTE CARLO PRODUCTION SIMULATION TO TRANSMISSION LOSS ESTIMATION
7.1 Introduction
7.2 Structure of Proposed Algorithm
7.3 Numerical Tests
7.4 Summary
CHAPTER 8: CONCLUSIONS
APPENDIX A: PRODUCTION SIMULATION OF HYDRO GENERATION SYSTEM
REFERENCES
BIOGRAPHICAL NOTE