基于深度残差网络的电力负荷预测策略

Power Load Forecasting Strategy Based on Deep Residual Network

在电力能源使用中,多重外界因素影响着电力负荷总耗能量。当前已有技术仅限于对机组内部因素及数字型外界因素对电力负荷的影响进行测算,无法综合处理较复杂环境、人文等数字化处理较困难的影响因素,测算精准度受到制约。在预测电力负荷总量问题上,基于深度学习理论,搭建一种更精准的预测模型,利用数据降维(PCA)、数据清洗等方法使模型具有处理复杂影响因素的能力,并采用仿真分析技术将新模型与现存模型进行预测精准度对比。研究表明,新模型的预测水平较其余模型更良好。

In the use of power energy,multiple external factors affect the total energy consumption of power load.At present,the existing technology is limited to measuring the influence of unit internal factors and digital external factors on power load.However,it is unable to comprehensively deal with the difficult influencing factors of digital processing such as complex environment and humanities,and the measurement accuracy is restricted.On the problem of forecasting the total power load,a more accurate forecasting model is built based on the deep learning theory.Data dimensionality reduction(PCA)and data cleaning are used to make the model have the ability to deal with complex influencing factors,and simulation analysis technology is used to compare the prediction accuracy between the new model and the existing model.The research shows that the prediction level of the new model is better than that of other models.