Sei sulla pagina 1di 3

Artificial Neural Network Short-Term

Electrical Load Forecasting Techniques

by Xu Le Yan

Call_No: MST(E)011.XU
TABLE OF CONTETNS
Chapter 1 INTRODUCTION 1
1.1 General Background of STLF 1
1.2 Historical Development of Methods For STLF 2
1.3 Problems Discussed In This Thesis 4
Chapter 2 FUNDAMENTALS OF ARTIFICIAL NEURAL NETWORKS 7
2.1 Historical Development of Artificial Neural Networks 7
2.1.1 Initial Development Stage 7
Emergence of Neural Networks 7
Rosenblatt's Perceptron 8
Adaptive Linear Element (Adaline) 9
Madaline 10
2.1.2 Dilemma Stage 10
2.1.3 Resurgence Stage 11
2.1.4 Further Development 12
2.2 Biological Basis of Neural Networks 12
2.3 Artificial Neurons 13
2.4 Feedforward Neural Networks 14
2.5 Feathers of Artificial Neural Networks 16
2.6 Back-Propagation Algorithm 17
2.6.1 Introduction 17
2.6.2 Learning Process 18

1
2.6.3 Derivation of the Back-Propagation Algorithm 19
2.6.4 Training procedure of BP algorithm 23
Chapter 3 BASIC BP MODEL WITH TEMPERATURE FOR STLF 25
3.1 Introduction 25
3.2 Architecture of Basic BP Model With Temperature 25
3.3 Algorithm Description 27
3.4 Experiment Results for Basic BP Model with Temp 30
3.5 Problems of the Basic BP Model with Temp 31
Chapter 4 LOCAL MINIMUM PROBLEM 33
4.1 Introduction 33
4.2 Method To Treat Local Minimum Problem 33
4.3 Experiment Results Regarding Local Minimum 35
Chapter 5 FAST CONVERGENCE TECHNIQUES 37
5.1 Variable Learning Rate Method 37
5.1.1 Introduction to Variable Learning Rate 37
5.1.2 Momentum Term 38
5.2 Quasi-Newton Method 39
5.2.1 Introduction 39
5.2.2 Classical Newton Method 40
5.2.3 Quasi-Newton Method 41
5.3 Experiment Results for VLR and Quasi-Newton 42
Chapter 6 HOLIDAY AND WEEKEND PROBLEM AND TREATMENTS 44
6.1 Introduction 44
6.2 Intelligent Treatment of Holidays and Weekends 45
6.2.1 Intelligent Treatment 1 46
6.2.2 Intelligent Treatment 2 47
6.3 Experiment Results for Intelligent Treatment 49
Chapter 7 ABNORMAL DATA PROBLEM 51
7.1 Introduction 51
7.2 Intelligent Preprocessing Data 52
7.3 Results for Intelligent Preprocessing Data 53
Chapter 8 CHINESE NEW YEAR PROBLEM 56

2
8.1 Introduction to Chinese New Year (CNY) Problem 56
8.2 Methods for Solving Chinese New Year Problem 57
8.2.1 Intelligent Treatment of CNY problem 58
8.2.2 Method of Papalexopoulos 58
8.2.3 Method of Yearly Displacement Compensation 58
8.3 Experiment Results for CNY Problem 59
Chapter 9 CONCLUSION 62
REFERENCE 66
Appendix I Holidays of Macao in 1995-1996 69
Appendix II Source Code Files 70
Source code of the basic BP based model 70
Source code of Yearly Displacement Compensation for CNY 73
Source code of the final model 75

Potrebbero piacerti anche