摘要
智能电网是电网企业发展的必然趋势,针对智能电网建成后带来的新能源、分布式能源的并网问题,以及智能电网建成后的高度集成信息化的环境变化,需要建立新环境下的负荷预测数据仓库,数据仓库中不但包含传统预测所考虑的定量性数据,而且包括新的定性数据。基于数据仓库根据数据挖掘技术计算与预测日具有高度相似特征的相似度,提取相应的数据,形成相应的知识库和推理规则库。研究具有一定自适应性的能够自己确定模型结构的智能预测方法,并研究相应的后干预校正算法,使预测精度得到突破性的提高,形成自适互动的智能负荷预测体系方法。研究将提高智能电网乃至整个电力工业的经济效益;增强我国的节能减排、可持续发展能力;完善智能电网的智能性和高效性以及拓展电力负荷预测研究理论的研究范围。
Smart grid is the inevitable trend of electric grid enterprises.After the completion of the smart grid,with the new energy and distributed energy accessed into electric grid,and a highly integrated information system completed,the electric grid enterprises＇ environment will be changed.A data warehouse of the load forecasting will be established,it contains not only quantitative data for traditional forecasting,but also new qualitative data.Based on the data warehouse,firstly,data mining techniques is used to calculate a high degree of similarity days of the forecasting in order to extract the corresponding data,and then a corresponding knowledge database and inference database are created.Secondly,a group of self-adaptive intelligent models which can determine the structure of themselves is applied for load forecasting.Finally,the corresponding intervention correction algorithm will be applied to improve the accuracy.An adaptive interactive intelligent load forecasting system is formed.The study will improve the economic benefits of smart gird and the whole electric power industry,enhance China＇s ability of energy-saving emission reduction and sustainable development,and will be conducive to the theory and applications of electric load forecasting study.
出处
《陕西电力》
2010年第5期11-15,共5页
Shanxi Electric Power
基金
国家自然科学基金(70671039)
中央高校基本科研业务费专项资金
