基于体感温度的电力系统负荷分类及负荷预测

Load Classification and Load Forecasting of Power System Based on Apparent Temperature

针对体感温度与负荷之间的变化关系进行了深入研究,研究表明体感温度在不同范围内变化时将对地区负荷影响表现出截然不同特征。将负荷分为对体感温度敏感和不敏感2种类别,并提出2种负荷预测方法。2种负荷预测方法均以径向基神经网络为基础,并针对不同类型待预测负荷采取差异化样本选取和处理方法,有效提高了该负荷预测模型适用性和负荷预测精度。将该方法运用到某市总负荷预测中,预测结果表明该方法具有较高精度和较好实用性,是一种有效的短期负荷预测新方法。

In-depth research on the relationships between the load and the apparent temperature suggests that, in different ranges, the apparent temperature responds to the regional load in totally different ways. Two kinds of loads are classified accordingly asinto apparent temperature-sensitive one and the apparent temperature non-sensitive one, based on which two load forecasting methods are proposed in this paper. Both of the two methods are based on Radial Basis Function （RBF） Neural Network with the differentiated sample selection and handling applied to improve the applicability of the load forecasting model effectively. The method recommended in this paper is applied in the load forecasting of a certain city, and the results show that it the method, with its high precision and good practicability, is an effective one forin short-term load forecasting.