数据特征驱动的火电产能过剩分解集成预测模型

有机融合数据特征驱动与多模态信息集成建模思想,构建了中国火电行业产能过剩组合预测方法和模型.首先识别火电产能过剩规模时序数据的本质和模式特征,发现其不仅具有非平稳,非线性特征,还呈现高复杂性和突变性;其次采用与数据特征相配的变分模态分解方法将时序数据分解,得到多个分量;然后识别各分量的数据特征,据此选择三次指数平滑-最小二乘支持向量机模型进行预测;最后集成各分量预测结果,得到火电产能过剩规模的最终预测结果.实证检验表明,所构建模型的预测水平精度,方向精度和稳定性均优于目前广泛使用的单一模型和其他组合预测模型.预测结果显示,2020-2022年中国火电产能过剩规模仍处于较高水平,呈先降后升趋势,且体制扭曲仍将是火电产能过剩的决定性因素.

A data-characteristic-driven decomposition ensemble forecasting model for thermal power overcapacity

By organically integrating data-characteristic-driven modeling idea with multi-modal information ensemble modeling idea,a combination forecasting method and model for overcapacity in China's thermal power industry is constructed.Firstly,the nature and pattern characteristics of thermal power overcapacity scale data are identified and it is found that the data not only has non-stationary and non-linear characteristics,but also has high complexity and mutability characteristics.Secondly,variational mode decomposition method which matches the data characteristics is used to decompose the time series data to obtain multiple components.Then the data characteristics of obtained components are identified,and then a triple exponential smoothing-least square support vector machine model is selected for prediction.Finally the forecasting results of each component are integrated to obtain the final forecasting result of the scale of thermal power overcapacity.Empirical tests show that the forecasting performance of the constructed model is better than the single and other combined prediction models currently widely used in terms of level accuracy,directional accuracy,and stability.The forecast results show that the scale of China's thermal power overcapacity will be still at a relatively high level showing a trend of falling first and then rising.And the institutional distortion will still be the decisive factor for thermal power overcapacity.