工业用电市场用电量组合预测方法研究

Research on Combination Forecasting Method of Electricity Consumption in Industrial Electricity Market

在总结现有用电量预测的主流方法基础上,对工业用电量数据特征进行分析,提出一种新的工业市场用电量预测方法。文章通过分析文献,总结了当前主流预测模型和方法的优势和劣势,从而提出灰色预测和梯度提升回归的组合模型。针对文章建立的模型,使用某地区四年实际用电量数据,与传统灰色预测模型、神经网络模型、单一梯度提升回归模型的预测结果进行对比,发现该组合模型在数据量较少或数据量比较充足的情况下,预测精度和稳定性很高,证明了所建立模型可靠性和有效性。

On the basis of summarizing the current mainstream methods of electricity consumption forecasting, the paper analyzes the characteristics of industrial electricity consumption data, and proposes a new industrial market electricity consumption forecasting method. By analyzing the literature, the paper summarizes the advantages and disadvantages of current mainstream prediction models and methods, then it proposes a combined model of grey prediction and gradient elevation regression. According to the model established in the article, using the four-year actual electricity consumption data of a certain area, compared with the prediction results of the traditional gray prediction model, neural network model and single gradient lifting regression model, it is found that the combined model has less data volume or data volume. In the case of sufficient conditions, the prediction accuracy and stability are high, which proves the reliability and effectiveness of the established model.