考虑不同发展情境的江苏省中长期电力需求预测

传统的中长期组合预测模型通常没有考虑不同发展阶段各个指标的不同权重,也没有在中长期的预测中考虑不同的发展情形。依据经济发展的周期特性及与电力的内在联系,将电力数据划分为不同阶段,利用电量误差指标和趋势误差指标计算各单一预测模型权重,利用客观熵权法得到各单个预测方法的综合权重。充分考虑未来经济发展的不同可能以及电动汽车、需求响应等用电技术改变因素,设定低速发展、高速发展和考虑其他因素的发展情境,并分别进行组合预测。将该组合预测方法应用于江苏省中长期负荷预测中,得到不同情境下的预测结果。

The Medium- and Long-Term Power Demand Forecasting for Jiangsu Province with Consideration of Different Developing Situations

Different weights of various indicators in different development stages are not considered in the conventional medium- and long-term combined forecasting model, and different development scenarios are not considered either in the medium- and long-term forecasting. Based on the periodicity of economy and its relationship with electric power, the electricity data are grouped according to different development stages, and two indicators including electricity error and trend error are used to calculate the weight of each single model. The entropy method is adopted to get the comprehensive weight of each single forecasting method. With a full consideration of different scenarios of future economic development and changes of electricity technology such as electric vehicle and demand response, a combined forecasting is made respectively for different development scenarios including low-speed development, high-speed development and other factors. The combined forecasting model has been applied to the medium- and long-term power load forecasting of Jiangsu Province and satisfactory results are obtained under different scenarios.