摘 要

本論文主要討論一些非固定式類神經網路於電力負載預測的應用。有別於傳統固定式類神經網路技術，此論文所採用之神經網路在訓練與測試期間，其架構為非固定的，輸入神經元的個數將會依據欲預測負載的特性而定。修正式的學習理論包括模糊倒傳遞學習理論和隨機倒傳遞學習理論，將個別地應用於吾等所發展的負載預測器中。為選取有效而精確的類神經網路輸入變數，電力負載與天氣溫度間的相互關係分析及負載與負載間的灰關聯分析將被探討及研究。

兩種未來一日至數日之短期性型態的電力負載預測，即尖峰負載與小時性負載預測，為本論文主要研究對象。由台灣電力公司及中央氣象局所提供之小時性電力負載與天氣溫度為論文主要研究資料，其資料範圍包含1992年至1998年。此外，為映證吾等所研究之負載預測器的實用性與優越性，數種傳統式的預測方法，包含具有常數值學習速率與動量係數之固定式類神經網路、遞迴式時間序列模式、及於1995年所提出之人工類神經網路短期負載預測器也一併被模擬以做為比較。
由實驗結果發現，吾等所研究之負載預測器的表現比傳統式方法，無論在網路過度訓練現象的降低和預測精確度以及類神經網路學習速率的提升上，均可達到有效改善之目的。

ABSTRACT

In this thesis, some aspects of the non-fixed neural network for power load forecasting are discussed. Unlike traditional fixed neural network technique, the structure of neural network is non-fixed during its training and testing phases. Based on the characteristic of the desired forecasting day, the number of input node utilized is changeable. The modified learning algorithms, including fuzzy back-propagation learning algorithm and stochastic back-propagation learning algorithm, will be used in the load forecasters we developed. For precise input selection of the neural network model, the analysis of mutual relationship between load and temperature and gray relational analysis between desired forecasting load and the related previous load are studied.

Two types of load forecasting, i.e., peak load forecasting and hourly load forecasting, are investigated. Short term (one-to-several-day-ahead) load forecasting is considered in this research. Hourly loads and relevant temperature data from 1992 to 1998 provided by Taipower Utility and the Central Weather Bureau is implemented for this research. For demonstrating the feasibility and superiority of the forecasters we develop, several forecasting models, including fixed neural network with constant learning rate and momentum, recursive time series model, and artificial neural network short term load forecaster (ANNSTLF) proposed by [Kho.2], are also performed for a comparison.

From the results of the simulation, better performances could be obtained by the methods we proposed. Not only the over-training phenomenon is obviously reduced, the forecasting accuracy and the learning speed of the neural model are also effectively improved.

摘 要 iii

ABSTRACT v

LIST OF FIGURES vii

LIST OF TABLES ix

CHAPTER 1 INTRODUCTION 1
1.1 Research Objective 1
1.2 Dissertation Organization 6

CHAPTER 2 HISTORIC REVIEW OF PREVIOUS TECHNIQUES 7
2.1 Introduction 7
2.2 Time Series Model 7
2.3 Regression Model 8
2.4 Expert System 9
2.5 Neural Network 9

CHAPTER 3 NEURAL NETWORK MODEL 11
3.1 Introduction 11
3.2 Back-Propagation Learning Algorithm 15
3.3 Non-Fixed Neural Network 16
3.4 Modified Learning Algorithm 19
3.4.1 Fuzzy Back-Propagation Learning Algorithm 19
3.4.2 Stochastic Learning Algorithm 23

CHAPTER 4 DATA ANALYSIS 26
4.1 Introduction 26
4.2 Data Normalization 26
4.3 Input Selection 28
4.3.1 Gray Relational Analysis 28
4.3.2 Mutual Relationship Analysis 29
4.4 Measure of Performance 30

CHAPTER 5 INITIAL MODEL CONSTRUCTION 31
5.1 Introduction 31
5.2 Forecasting Model with Fuzzy Learning 31
5.2.1 Neural Network Structure 33
5.2.2 Fuzzy Learning Algorithm 35
5.2.3 Next-Day Peak Load Forecasting 37
5.2.4 One-Day-Ahead Hourly Load Forecasting 41
5.3 Forecasting Model with Stochastic Learning 44
5.3.1 Next-Day Peak Load Forecasting 44
5.3.2 One-to-Two-Day-Ahead Hourly Load
Forecasting 48

CHAPTER 6 LOAD FORECASTING MODULES 51
6.1 Introduction 51
6.2 Module A 51
6.2.1 Next-Day Peak Load Forecasting 52
6.2.2 One-to-Two-Day-Ahead Load Forecasting 60
6.3 Module B: One-to-Five-Day-Ahead Load
Forecasting 64
6.3.1 Input Selection 64
6.3.2 Learning Algorithm 69
6.3.3 Performance Study 70

CHAPTER 7 CONCLUSIONS AND FUTURE RESEARCH 74
7.1 Conclusions 74
7.2 Future Research 75

REFERENCES 77

REFERENCES

[Bak.1] Bakirtzis A. G., Peprldls V., Klartzis S. J., Alezladls M. C. and
Malssls A. H., “A neural network short term load forecasting model
for the Greek power systems,” IEEE Trans. on Power Systems, Vol. 11,
No. 2, pp. 858-863, 1996.
[Box.1] Box G. E. P. and Jenkins G. M., Time series analysis forecasting and
control, Holden Day, San Francisco, 1970.
[Bun.1] Bunn D. W. and Farmer E. D., Eds., Comparative models for electrical
load forecasting, John Wiley and Sons, New York, 1985.
[Cha.1] Chaturvedi D. K., Mishra R. K. and Agrawal A., “Load forecasting using
genetic algorithms,” Journal of Institution of Engineers, Vol. 76, pp.
161-165, 1995.
[Che.1] Chen S. T., Yu D. C. and Moghaddamjo A. M., “Weather sensitive short
term load forecasting using non-fully connected artificial neural
networks,” IEEE Trans. on Power Systems, Vol. 7, No. 3, pp. 1098-1105,
1992.
[Chen.1] Chen Y. Q., Yin T. and Babri H. A., “A stochastic back-propagation
algorithm for training neural networks,” Proc. of First International
Conf. on Information, Communication and Signal Processing, pp. 703-
707, 1991.
[Cho.1] Chow M. Y., Zhu J. and Tram H., “Application of fuzzy multi-objective
decision making in spatial load forecasting,” IEEE Trans. on Power
Systems, Vol. 13, No. 3, pp. 1185-1190, 1998.
[Chr.1] Christiaanse W. R., “Short term load forecasting using general
exponential smoothing,” IEEE Trans. on Power Applications and
Systems, Vol. PAS-90, No. 2, pp. 900-910, 1972.
[Con.1] Connor J., Atlas L. E. and Martin D., “Recurrent neural networks and
load forecasting,” Proc. of First International Forum on Applications
of Neural Networks to Power System, pp. 211-215, 1991.
[Dan.1] Daneshdoost M., Lotfalian M., Bumroonggit G. and Ngoy J. P., “Neural
network with fuzzy set-based classification for short-term load
forecasting,” IEEE Trans. on Power Systems, Vol. 13, No. 4, pp. 1386-
1391, 1998.
[Das.1] Dash P. K., Satpathy H. P. and Liew A. C., “A real-time short-term
peak and average load forecasting system using a self-organizing fuzzy
neural network,” Engineering Applications of Artificial Intelligence,
Vol. 11, pp. 307-316, 1998.
[Den.1] Deng J. L., “Introduction to gray systems theory,” Journal of Gray
Systems, Vol. 1, No. 2, pp. 1-20, 1989.
[Dil.1] Dillon T. S., Sestito S. and Leuang S., “Short term load forecasting
using an adaptive neural network,” International Journal of
Electrical Power & Energy Systems, Vol. 13, No. 4, pp. 186-192, 1991.
[Gra.1] Grady W. M., Groce L. A., Huebner T. M., Lu Q. C., and Crawford M. M.,
“Enhancement, implementation and performance of an adaptive short-term
load forecasting algorithm,” IEEE-PES Summer Meeting, pp. 1873-1878,
1991.
[Gun.1] Gunst R. F. and Mason R. L., Regression analysis and its applications:
A data-oriented approach, Dekker, New York, 1980.
[Hag.1] Hagan M. and Behr S. M., “The time series approaches to short-term
load forecasting,” IEEE Trans. on Power Systems, PWRS, Vol. 2, pp.
785-791, 1987.
[Hag.2] Hagan M. and Klein R., “Identification techniques of Box and Jenkins
applied to the problem of short-term load forecasting,” Proc. IEEE
Summer Power Meeting, pp. A77 618-2, 1977.
[Hag.3] Hagan M. and Klein R., “Off-line and adaptive Box and Jenkins models
for load forecasting,” Proceedings of Lawrence Systems and Decision
Science, Berkeley, 1977.
[Hag.4] Hagan M. and R. Klein, “On-line maximum likelihood estimation for
load forecasting,” IEEE Trans. on Systems, Man, and Cybernetics, Vol.
SMC-8, No. 9, pp. 711-715, 1978.
[Hay.1] Haykin S., Neural networks: A comprehensive foundation, Macmillan, New
York, 1994.
[Hey.1] Heydt G., Khotanzad A. and Farahbakhshian N., “A method for the
forecasting of the probability density function of power system
loads,” IEEE Trans. on Power Applications and Systems, Vol. PAS-100,
No. 2, pp. 5002-5010, 1981.
[Ho.1] Ho K. L., Hsu Y. Y., Chen C. F., Lee T. E., Liang C. C., Lai T. S.,
and Chen K. K., “Short term load forecasting of Taiwan power system
using a knowledge-based expert system,” IEEE Trans. on Power Systems,
Vol. 5, pp. 1214-1221, 1990.
[Hsi.1] Hsia K. H. and Wu J. H., “A study on the data preprocessing in gray
relation analysis,” Journal of the Chinese Gray System Association,
Vol. 1, No. 1, pp. 47-53, 1998.
[Hua.1] Huang H. C., Hwang R. C. and Hsieh J. G., “Short-term power load
forecasting by non-fixed neural network with fuzzy BP learning
algorithm,” accepted by International Journal of Power and Energy
Systems (IASTED), 2000.
[Hua.2] Huang H. C., Hwang R. C. and Hsieh J. G., “A new artificial
intelligent peak power load forecaster based on non-fixed neural
networks,” accepted by International Journal of Electrical Power &
Energy Systems, 2001.
[Huan.1] Huang S. R., “Short-term load forecasting using threshold
autoregressive models,” IEE Proc. Electric Power Application, Vol.
144, No. 5, pp. 477-481, 1997.
[Hwa.1] Hwang R. C., Hourly load forecasting by neural networks, Ph.D. Thesis,
Southern Methodist University, Texas, 1993.
[Hwa.2] Hwang R. C., Huang H. C. and Hsieh J. G., “Short-term power load
forecasting by neural network with modified back-propagation learning
algorithm,” IEEE Winter Power Meeting, Singapore, 2000.
[Kho.1] Khotanzad A. and Lu J. H., “Nonparametric prediction of AR processes
using neural networks,” Proc. of IEEE International Conf. Acoustics,
Speech and Signal Processing, ICASSP-90, Albuquerque, New Mexico, pp.
2551-2554, 1990.
[Kho.2] Khotanzad A., Hwang R. C., Abaye A. and Maratukulam D., “An adaptive
modular artificial neural network hourly load forecaster and its
application at electric utilities,” IEEE Trans. on Power Systems,
Vol. 10, No. 3, pp. 1716-1722, 1995.
[Kho.3] Khotanzad A., Afkhami-Rohani A., Lu T. L., Abaye A., Davis M. and
Maratukulam D., “ANNSTLF-A neural network based electric load
forecasting system,” IEEE Trans. on Neural Networks, Vol. 8, No. 4,
pp. 835-845, 1997.
[Kho.4] Khotanzad A., Abaye A. and Maratukulam D., “An adaptive recurrent
neural network system for multi-step-ahead hourly prediction of power
system loads,” Proc. IEEE International Conf. on Neural Networks, pp.
3393-3397, 1994.
[Lee.1] Lee K. Y., Chen Y. T. and Park J. H., “Artificial neural network
methodology for short-term load forecasting,” Proc. of the Workshop
on Applications of Artificial Neural Network Methodology in Power
Systems Engineering, pp. 109-112, 1990.
[Lu.1] Lu C. N., Yu H. T. and Vemuri H., “Neural network based short term
load forecasting,” IEEE Trans. on Power Systems, Vol. 8, No. 1, pp.
336-342, 1993.
[Luq.1] Lu Q. C., Grady W. M., Crawford M. M. and Anderson G. M., “An
adaptive nonlinear predictor with orthogonal escalator structure for
short-term load forecasting,” IEEE Trans. on Power Systems, Vol. PWRS-
4, No. 1, pp. 158-165, 1989.
[Mba.1] Mbamalu G., El-Hawary F. and M. El-Hawary E., “Decomposition approach
to forecasting electric power system load using an artificial neural
network,” Electric Machine and Power Systems, Vol. 25, pp. 875-883,
1997.
[Mes.1] Meslier F., “New advances in short-term load forecasting using Box
and Jenkins approach,” IEEE Winter Power Meeting, paper no. A78 051-
5, 1978.
[Mog.1] Moghram I. and Rahman S., “Analysis and evaluation of five short-term
load forecasting techniques,” IEEE Trans. on Power Systems, PWRS,
Vol. 4, No. 4, pp. 1484-1491, 1989.
[Mor.1] Mori H. and Kobayashi H., “Optimal fuzzy inference for short term
load forecasting,” IEEE Trans. on Power Systems, Vol. 11, No. 2, pp.
390-396, 1996.
[Pap.1] Papadakis S. E., Theocharis J. B., Kiartzis S. J. and Bakirtzis A. G.,
“A novel approach to short-term load forecasting using fuzzy neural
networks,” IEEE Trans. on Power Systems, Vol. 13, No. 2, pp. 480-489,
1998.
[Papa.1] Papalexopoulos A. D. and Hesterberg T. C., “A regression-based
approach to short-term system load forecasting,” IEEE Trans. on Power
Systems, PWRS, Vol. 5, No. 4, pp. 1535-1544, 1990.
[Par.1] Park D. C., El-Sharkawi M. A., Marks R. J. Atlas L. E. and Damborg M.
J., “Electric load forecasting using artificial neural networks,”
IEEE Trans. on Power Systems, Vol. 6, No. 2, pp. 442-449, 1991.
[Pen.1] Peng T. M., Hubele N. F. and Karady G. G., “Advancement in the
application of neural networks for short-term load forecasting,” IEEE
Trans. on Power Systems, Vol. 7, No. 1, pp. 250-257, 1992.
[Pir.1] Piras A., Germond A., Buchenel B., Emhof K. and Jaccard Y.,
“Hetrogenous artificial neural networks for short-term load
forecasting,” IEEE Trans. on Power Systems, Vol. 11, No. 1, pp. 397-
402, 1996.
[R.1] IEEE Committee Report, “Load forecasting bibliography, Phase I,”
IEEE Trans. on Power Applications and Systems, Vol. 1, PAS-100, No. 1,
pp. 53-58, 1980.
[R.2] IEEE Committee Report, “Load forecasting bibliography, Phase II,”
IEEE Trans. on Power Applications and Systems, Vol. 1, PAS-100, No. 7,
pp. 3217-3220, 1981.
[Rah.1] Rahman S. and Bhatnagar R., “An expert system based algorithm for
short term load forecast,” IEEE Trans. on Power Systems, PWRS, Vol.
3, No. 2, pp. 392-399, 1989.
[Rah.2] Rahman S. and Hazim O., “A generalized knowledge-based short term
load forecasting,” IEEE Trans. on Power Systems, Vol. PWRS-8, No. 2,
pp. 508-514, 1993.
[Rum.1] Rumelhart D. E. and McClelland J. L., Parallel distributed processing,
Vol. 1, MIT Press, Cambridge, 1986.
[Sri.1] Srinivasan D., Liow A. C. and Chang C. S., “Forecasting daily load
curves using a hybrid fuzzy-neural approach,” IEE Proc. Electric
Power Application, Vol. 141, No. 6, pp. 561-567, 1994.
[Toy.1] Toyoda J., Chen M. and Inouye Y., “An application of state estimation
to short-term load forecasting, Part I and Part II,” IEEE Trans. on
Power Applications and Systems, Vol. PAS-89, No. 7, pp. 1678-1688,
1970.
[Ver.1] Vermaak J. and Botha E. C., “Recurrent neural networks for short term
load forecasting,” IEEE Trans. on Power Systems, Vol. 13, No. 1, pp.
126-132, 1998.