Short-term power load forecasting method based on ICEEMDAN-DCN-Transformer
To address the problem that traditional load forecasting methods are susceptible to complex environmental factors,a combined short-term electric load forecasting model based on ICEEMDAN-DCN-Transformer was proposed,the original power load data was decomposed into several IMF and a Res by ICEEMDAN method.Taking into account the influence of complex environmental factors,the decomposed components and environmental characteristics were input into DCN-Transformer in parallel for prediction.Finally,the sets of prediction data were linearly summed to obtain the complete prediction results.Experiments were conducted according to the historical data of power load in Quanzhou City,four single prediction models and three combined prediction models were established as comparison models to predict the 10-day 240 h power load sequence at the location.The results show that the as-proposed algorithm can significantly improve the accuracy of load forecasting and effectively reduce the value of the error evaluation index,compared with the traditional algorithm,providing a theoretical basis for the safe operation and planning of power system.
power load forecastingICEEMDANDCNprediction accuracyshort-term loadcombined forecasting modelerror evaluation
万方数据
维普
