基于相似日的96点短期负荷预测实用方法及其应用

Method and application of ninety-six points short-term load forecasting based on similar historical day

短期负荷预测是各供电分公司调度部门的基础性工作,96点日负荷预测采用模式识别中的聚类分析方法,对影响负荷的不同相关因素建立历史映射数据库,在不同日之间计算相关度而得到相似日,并在相似日的基础上采用线性外推法得到预测日的对应负荷。文章对奉贤地区三月份某一周的负荷数据做了预测验证,通过比较分析,表明该方法具有较高的精确度和较强的工程实用价值,不但减少了负荷预测对人的经验的依赖,而且应用简便、有效。

Ninety-six points short-term load forecasting is the basic work for all Power Supply Companies. The paper presents a new method for short-term load forecasting, which applies cluster analysis technique of pattern recognition principle to calculate the similar coefficients between the future day and past days, then chooses the most similar days according to these coefficients and uses linear extrapolation method to predict the load of future day. With this method, the index-mapping database is also designed. A week load data in Fengxian district on March is forecasted in the paper. Analysis results show that this method has good performance in engineering application. It not only reduces the dependence of load forecasting upon the experience, bus also behaves as simple and accurate.