Research Of Electric Power Load Forecasting Based On Data Mining Technology

Power short-term load forecasting(STLF) is an important and integral component in the operation of any electric utility whose accuracy directly influence power system's security, profit and quality. STLF is characterized by massive data for forecasting, noisy-contained sample data, influenced by weather condition, and randomicity. Based on various data mining technologies, the author aims at each stage of STLF and has done deep research on the pre-process of historical load data, classification of load samples, process of weather condition, establishment of forecasting model and its input parameters mining. All these work had laid a solid foundation for hi-accuracy STLF.The load of power systems is an unsteady stochastic process. Among those observed values there may exist some "dirty data" due to the effect of various factors. These dirty data, participating the training of neural networks intermingled with normal data, badly affect the accuracy of load forecasting. In order to purge the historical load, this paper brings forward an intelligent model which comprising the effects of ART network clustering and CC network classification.Classification of load samples is one must take into consideration when carry out short-term load forecasting. Input sample neither less nor more, too less will not achieve training accuracy, too more will lead to not only meaningless study but even can't converge.. Therefore, this paper adopt Coonan network to train samples based on weather, day type, actual historical load and so on who influence the accuracy of load forecasting, in turn raises choose out the samples similar to forecasting day. Forecasting model using the selected samples can decrease training-time and increase forecasting accuracy effectively.There are so many factors that influenced STLF, how to justify and select the correlative factors is the key to improve the performance of load forecasting. The reduction algorithmbased on rough set theory introduced to mine more correlative attributes in the pending forecasting components, insures the rationality of input parameters of forecasting model. A reduction algorithm through classification reliability algorithm which with certain noise and having very good cover ability and generalizable ability through set classification reliability-β is introduced to overcome the large computational complexity of conventional reduction algorithm.Lastly, construct the short-term load forecasting based on data mining. the author aims at each stage of STLF and has done deep research on the pre-process of historical load data, classification of load samples, process of weather condition, establishment of forecasting model and its input parameters mining. All these work had laid a solid foundation for hi-accuracy STLF. The forecasting results show that the proposed method possesses better forecasting accuracy and the forecasting is satisfactory.