考虑气象条件下的电力系统短期负荷预测研究
|
摘要
电力系统负荷预测的水平已成为衡量电力系统运行管理现代化的标志之一。尤其是准确的短期负荷预测更是具有重要的意义。负荷预测的影响因素较多，既由负荷本身的历史表现决定，还要受众多非负荷因素的影响。非负荷因素中又以气象因素权重最大。
负荷坏数据辨识是由负荷曲线抗差聚类和坏数据曲线模式分类两个顺序的过程组成的；本文通过对Kohonen网的抗差聚类和BP网模式分类的效果分析，设计由这两种网络组合而成的神经网络模型，较好地完成了坏数据辨识的任务。
在正常日预测中，通过深入研究各个时段负荷和气象之间的变化关系，本文设计了一个由时间序列、模糊推理和一元回归修正模型组成的综合模型，在气温变化剧烈的季节利用一元回归模型对前两种模型进行修正，预测的结果表明了这种修正的有效性。考虑到节假日负荷中商业和居民用电比例高，本文结合相邻近的周末负荷考虑温度因素来对其进行预测。节后工作日预测的预测中则采用灰色预测模型，能用较少的历史数据通过累加生成推导其变化规律。
单一模型所用到的信息毕竟有限，各个模型所用到的信息不会完全相同，所以本文最后对各个单一模型的结果建立了EW组合预测模型，所有的建模和计算都是针对浙江省负荷数据的，据此开发的程序已经在浙江省电力调度中心得到了应用，达到了较高的准确率。
The level of load forecasting is one of the measures of modernization of power system management. So load forecasting, especially accurate short-term load forecasting is of great importance to power system. There are many factors that affect system load, such as history data of load, many non-load factors in which weather factor is the most important.
The outlier identification is divided into two sequential parts: the robust day-load-curves cluster and the bad curve pattern classification. By analyzing the effects of Kohonen network clustering and BP network classification, the dissertation designs an outlier identification model comprising these two kinds of neural network and implements the tasks of bad data identifications and adjustments.
In order to forecast normal day load, this thesis advances a integrated model consisting of time series model, fuzzy model and linear regression model through studying the relations between load and weather; In order to forecast the load of such day which temperature change sharply, this thesis uses the 3r model to adjust the former two models and the forecasting results validate the effectivity of the adjustment; In order to forecast festal day load, this thesis uses the neighboring weekends load data while considering temperature factor and uses gray model to forecast the load of the day after the feast.
A single model can only make use of limited information and the information that the models used can't be the completely same, so a EW integrated forecasting model is proposed in the end of the thesis to improve the accurateness. All modeling and forecasting are based on the data of Zhejiang province load and the program is applied in the power dispatch center of Zhejiang province and it's performance is satisfactory.
负荷坏数据辨识是由负荷曲线抗差聚类和坏数据曲线模式分类两个顺序的过程组成的；本文通过对Kohonen网的抗差聚类和BP网模式分类的效果分析，设计由这两种网络组合而成的神经网络模型，较好地完成了坏数据辨识的任务。
在正常日预测中，通过深入研究各个时段负荷和气象之间的变化关系，本文设计了一个由时间序列、模糊推理和一元回归修正模型组成的综合模型，在气温变化剧烈的季节利用一元回归模型对前两种模型进行修正，预测的结果表明了这种修正的有效性。考虑到节假日负荷中商业和居民用电比例高，本文结合相邻近的周末负荷考虑温度因素来对其进行预测。节后工作日预测的预测中则采用灰色预测模型，能用较少的历史数据通过累加生成推导其变化规律。
单一模型所用到的信息毕竟有限，各个模型所用到的信息不会完全相同，所以本文最后对各个单一模型的结果建立了EW组合预测模型，所有的建模和计算都是针对浙江省负荷数据的，据此开发的程序已经在浙江省电力调度中心得到了应用，达到了较高的准确率。
The level of load forecasting is one of the measures of modernization of power system management. So load forecasting, especially accurate short-term load forecasting is of great importance to power system. There are many factors that affect system load, such as history data of load, many non-load factors in which weather factor is the most important.
The outlier identification is divided into two sequential parts: the robust day-load-curves cluster and the bad curve pattern classification. By analyzing the effects of Kohonen network clustering and BP network classification, the dissertation designs an outlier identification model comprising these two kinds of neural network and implements the tasks of bad data identifications and adjustments.
In order to forecast normal day load, this thesis advances a integrated model consisting of time series model, fuzzy model and linear regression model through studying the relations between load and weather; In order to forecast the load of such day which temperature change sharply, this thesis uses the 3r model to adjust the former two models and the forecasting results validate the effectivity of the adjustment; In order to forecast festal day load, this thesis uses the neighboring weekends load data while considering temperature factor and uses gray model to forecast the load of the day after the feast.
A single model can only make use of limited information and the information that the models used can't be the completely same, so a EW integrated forecasting model is proposed in the end of the thesis to improve the accurateness. All modeling and forecasting are based on the data of Zhejiang province load and the program is applied in the power dispatch center of Zhejiang province and it's performance is satisfactory.
引文
[1] 边肇祺，张学工 等。模式识别。清华大学出版社。2001年5月。
[2] 阎平凡．张长水。人工神经网络与模拟进化算法。1999年12月
[3] Jiawei Han, Micheline Kambr,数据挖掘：概念与技术，北京：高等教育出版社，2001年5月。
[4] 张智星，孙春在，(日)水谷英二 著，张平安，高春华 等 译，神经、模糊和软计算，西安：西安交通大学出版社，2000年2月。
[5] 方开泰，潘恩沛 著，聚类分析，北京：地质出版社，1982年
[6] 杨曾武 统计预测原理，中国财政经济出版社。
[7] 徐秉铮，张百灵，韦岗 编著，神经网络理论与应用，华南理工大学出版社，1994年12月第一版。
[8] 杨行峻，郑君里，人工神经网络，北京：高等教育出版社，1992．9。
[9] 袁曾任．人工神经网络原理及其应用[M]．北京：清华大学出版社，1999．
[10] 孙增圻，智能控制理论与技术。北京：清华大学出版社，1997。
[11] 袁亚湘，孙文瑜，最优化理论与方法。北京：科学出版社，1999。
[12] 张国江，邱家驹，李继红，基于模糊推理系统的多因素电力负荷预测，电力系统自动化，Vol．26，No．5，March 2002。
[13] 牛东晓．陈志业．邢棉等．具有二重趋势性季节型电力负荷预测组合优化灰色神经网络模型．中国电机工程学报．Vo．22 No．1 Jan 2002
[14] 牛东晓．曹树华等．电力负荷预测技术及其应用。北京：中国电力出版社。1998年
[15] 鞫平．姜巍等．96点短期负荷预测方法及其应用。电力系统及其自动化。Vol．25 No．22 Nov，25，2001。
[16] 周江文，等．抗差最小二乘法[M]．武汉：华中理工大学出版社，1997．
[17] 吕安民 等，数据挖掘和知识发现的技术方法，测绘科学 2000 v25 No4．
[18] 沈清，汤霖，模式识别导论，国防科技大学出版社，1991年5月。
[19] 候志检，吴际舜，张绮雨。电力系统短期负荷预报的几种改进手段，电力系统及其自动化，Vol．20，No．7．1996
[20] 灰色预测与决议。武汉：华中理工大学出版社，1992
[21] 软计算方法和数据挖掘理论在电力系统负荷预测中的应用。电力系统及其自动化学报。Vol．15，No．1，Feb．2003。
[22] T. Takagi and M.Sugeno. Fuzzy identification of systems and its applications to modeling and control. IEEE Transactions on Systems, Man, and Cysbernetics, 15:116-132, 1985.
[23] Shi Qing-yun, K.S.Fu, A Method for the Design of Binary Tree Classifiers, Pattern Recognition, v16, No.6, pp 593-603, 1983.
[24] Neuro-fuzzy approaches to short-term electrical load forecasting, Bartkiewicz, W. , Neural
Networks, 2000. IJCNN 2000, Proceedings of the IEEE-INNS-ENNS International Joint Conference on, Volume: 6,2000, Page(s): 229 -234 vol.6.
[25] P.A. Mastorocostas, J.B. Theocharis, A.G. Bakirtzis, Fuzzy modeling for short term load forecasting using the orthogonal least squares method, IEEE Trans. On power systems, Vol.14, No.l, February 1999.
[26] T. Takagi and M.Sugeno, Fuzzy identification of systems and its applications to modeling and control. IEEE Trans. On Systems, Man, and Cysbernetics, 15:116-132, 1985.
[27] B.Cestnik, I. Kononenko, and I.Bratko, ASSISTANT 86: A knowledge-elicitation tool for sophisticated users. In Progress in Machine Learning, I.Bratko and N.Lavrac. Eds. Wilmslow, UK: Sigma Press, 1987.
[28] P. Clark, and T. Ninlett, The CN2 induction algorithm, Machine Learing, vol.3, no.4, pp.261-284, 1989.
[29] Nima Amjady, Short-term hourly load forecasting using time-series modeling with peak load estimation capability. IEEE Trans. On power systems, Vol. 16, No.3 Aug. 2001
[30] Tomonobu Senjyu, Hitoshi Takara, Katsumi Uezato, One-hour-ahead load forecasting using neural network. IEEE Trans. On power systems,Vol. 17,No. 1,Feb. 2002
[31] Jyh-Shing Roger Jang. ANFIS: Adaptive-Network-Based Fuzzy Inference System. IEEE Transactions on Systems, Man, and Cybernetics. 1996, 23(3): 665～684.
[32] Juberias, et al. New arima model for hourly load forecasting. Proceedings of the IEEE Power Engineering Society Transmission and Distribution Conference, 1999, 1: 314～319.
[33] ZhengTongxin, et al. Hybrid wavelet-Kalman filter method for load forecasting. Electric Power Systems Research, 2000, 54(1): 11-17.
[34] S.E.Papadkis, et al. A novel approach to short-term load forecasting using fuzzy neural networks. IEEE Transactions on Power Systems, 1998, 13(2): 480～489.
[35] H. Mori, H.kobayashi. Optimal Fuzzy Inference for Short-Term Load Forecasting. IEEE Transactions on Power Systems, 1996,11(1): 390～396.
[36] Thomas G. Dietterich, An Experimental Comparison of Three Methods for Constructing Ensembles of Decision Trees: Bagging, Boosting, and Randomization, Machine Learning, 40, 139-157, 2000.
[37] Eric Bauer, Ron Kohavi, An Empirical Comparison of Voting Classification Algorithms: Bagging, Boosting, and Variants, Machine Learning, 36, 105-139, 1999.
[38] Sreerama K.Murthy, Automatic construction of decision trees from data: a multi-disciplinary survey, Data mining and knowledge discovery,2,345-389,1998.
[39] J.Mui and K.S.Fu, Automated classification of nucleated blood cells using a binary tree classifier, IEEE Trans. Patt. Anal. Mach. Intell., vol.PAMI-2,pp.429-443, 1980.
[2] 阎平凡．张长水。人工神经网络与模拟进化算法。1999年12月
[3] Jiawei Han, Micheline Kambr,数据挖掘：概念与技术，北京：高等教育出版社，2001年5月。
[4] 张智星，孙春在，(日)水谷英二 著，张平安，高春华 等 译，神经、模糊和软计算，西安：西安交通大学出版社，2000年2月。
[5] 方开泰，潘恩沛 著，聚类分析，北京：地质出版社，1982年
[6] 杨曾武 统计预测原理，中国财政经济出版社。
[7] 徐秉铮，张百灵，韦岗 编著，神经网络理论与应用，华南理工大学出版社，1994年12月第一版。
[8] 杨行峻，郑君里，人工神经网络，北京：高等教育出版社，1992．9。
[9] 袁曾任．人工神经网络原理及其应用[M]．北京：清华大学出版社，1999．
[10] 孙增圻，智能控制理论与技术。北京：清华大学出版社，1997。
[11] 袁亚湘，孙文瑜，最优化理论与方法。北京：科学出版社，1999。
[12] 张国江，邱家驹，李继红，基于模糊推理系统的多因素电力负荷预测，电力系统自动化，Vol．26，No．5，March 2002。
[13] 牛东晓．陈志业．邢棉等．具有二重趋势性季节型电力负荷预测组合优化灰色神经网络模型．中国电机工程学报．Vo．22 No．1 Jan 2002
[14] 牛东晓．曹树华等．电力负荷预测技术及其应用。北京：中国电力出版社。1998年
[15] 鞫平．姜巍等．96点短期负荷预测方法及其应用。电力系统及其自动化。Vol．25 No．22 Nov，25，2001。
[16] 周江文，等．抗差最小二乘法[M]．武汉：华中理工大学出版社，1997．
[17] 吕安民 等，数据挖掘和知识发现的技术方法，测绘科学 2000 v25 No4．
[18] 沈清，汤霖，模式识别导论，国防科技大学出版社，1991年5月。
[19] 候志检，吴际舜，张绮雨。电力系统短期负荷预报的几种改进手段，电力系统及其自动化，Vol．20，No．7．1996
[20] 灰色预测与决议。武汉：华中理工大学出版社，1992
[21] 软计算方法和数据挖掘理论在电力系统负荷预测中的应用。电力系统及其自动化学报。Vol．15，No．1，Feb．2003。
[22] T. Takagi and M.Sugeno. Fuzzy identification of systems and its applications to modeling and control. IEEE Transactions on Systems, Man, and Cysbernetics, 15:116-132, 1985.
[23] Shi Qing-yun, K.S.Fu, A Method for the Design of Binary Tree Classifiers, Pattern Recognition, v16, No.6, pp 593-603, 1983.
[24] Neuro-fuzzy approaches to short-term electrical load forecasting, Bartkiewicz, W. , Neural
Networks, 2000. IJCNN 2000, Proceedings of the IEEE-INNS-ENNS International Joint Conference on, Volume: 6,2000, Page(s): 229 -234 vol.6.
[25] P.A. Mastorocostas, J.B. Theocharis, A.G. Bakirtzis, Fuzzy modeling for short term load forecasting using the orthogonal least squares method, IEEE Trans. On power systems, Vol.14, No.l, February 1999.
[26] T. Takagi and M.Sugeno, Fuzzy identification of systems and its applications to modeling and control. IEEE Trans. On Systems, Man, and Cysbernetics, 15:116-132, 1985.
[27] B.Cestnik, I. Kononenko, and I.Bratko, ASSISTANT 86: A knowledge-elicitation tool for sophisticated users. In Progress in Machine Learning, I.Bratko and N.Lavrac. Eds. Wilmslow, UK: Sigma Press, 1987.
[28] P. Clark, and T. Ninlett, The CN2 induction algorithm, Machine Learing, vol.3, no.4, pp.261-284, 1989.
[29] Nima Amjady, Short-term hourly load forecasting using time-series modeling with peak load estimation capability. IEEE Trans. On power systems, Vol. 16, No.3 Aug. 2001
[30] Tomonobu Senjyu, Hitoshi Takara, Katsumi Uezato, One-hour-ahead load forecasting using neural network. IEEE Trans. On power systems,Vol. 17,No. 1,Feb. 2002
[31] Jyh-Shing Roger Jang. ANFIS: Adaptive-Network-Based Fuzzy Inference System. IEEE Transactions on Systems, Man, and Cybernetics. 1996, 23(3): 665～684.
[32] Juberias, et al. New arima model for hourly load forecasting. Proceedings of the IEEE Power Engineering Society Transmission and Distribution Conference, 1999, 1: 314～319.
[33] ZhengTongxin, et al. Hybrid wavelet-Kalman filter method for load forecasting. Electric Power Systems Research, 2000, 54(1): 11-17.
[34] S.E.Papadkis, et al. A novel approach to short-term load forecasting using fuzzy neural networks. IEEE Transactions on Power Systems, 1998, 13(2): 480～489.
[35] H. Mori, H.kobayashi. Optimal Fuzzy Inference for Short-Term Load Forecasting. IEEE Transactions on Power Systems, 1996,11(1): 390～396.
[36] Thomas G. Dietterich, An Experimental Comparison of Three Methods for Constructing Ensembles of Decision Trees: Bagging, Boosting, and Randomization, Machine Learning, 40, 139-157, 2000.
[37] Eric Bauer, Ron Kohavi, An Empirical Comparison of Voting Classification Algorithms: Bagging, Boosting, and Variants, Machine Learning, 36, 105-139, 1999.
[38] Sreerama K.Murthy, Automatic construction of decision trees from data: a multi-disciplinary survey, Data mining and knowledge discovery,2,345-389,1998.
[39] J.Mui and K.S.Fu, Automated classification of nucleated blood cells using a binary tree classifier, IEEE Trans. Patt. Anal. Mach. Intell., vol.PAMI-2,pp.429-443, 1980.