【摘要】 负荷与气象是密切相关的,尤其在夏冬两季。以省网负荷预报为例,在传统的电力负荷预测中,所采用的气象信息一般是全省、全天的气象信息,比如全省最高温度、最低温度等等。当前,气象预报技术已经可以做到分区、逐时预报。该文总体思路是将分区逐时气象预测数据应用于负荷预报,以进一步提高负荷预测精度。指出了按照行政区域和气候区域相结合来分区,定义了基于负荷的综合气象因素,提出了根据网供电与综合气象因素进行预测的策略,给出了将相似日方法与支持向量机相结合的负荷预测方法。最后给出在河南电网应用的实际结果,对比表明采用逐时气象信息的预测精度比采用全天性气象指标的预测精度更高。更多还原

【Abstract】 It is well known that electric load is closely related to weather especially in summer and winter.Taking the load forecasting in a provincial power grid as example,the weather data used in traditional load forecasting are globally and whole date,such as maximum and minimum temperature in the province.The weather forecasting has been developed to be able to predict hourly in area of a province.This paper uses the hourly area weather data for load foresting.The issues discussed include: dividing areas,weather data synthesis,load forecasting strategies and so on.Finally,the practical application results in Henan Power Grid are given to validate the methods.更多还原