Multi-Load Forecasting of Integrated Energy Systems Based on STL-Crossformer
Multi-load forecasting in integrated energy systems is crucial for the operation and scheduling of the system.Traditional forecasting models have not fully captured the long-term dependencies in time series or considered the coupling relationships between multiple loads,limiting improvements in forecasting accuracy.To address the challenges of multi-load forecasting in integrated energy systems,this paper proposes a forecasting model that integrates Seasonal Trend Decomposition and Crossformer.Initially,the original load data is decomposed into three sub-sequences using seasonal trend decomposition.Then,by employing a dimension-segmented embedding method and a two-stage attention mechanism,the model extracts cross-time and cross-dimensional correlations of multi-load data.Finally,a hierarchical encoder-decoder structure is utilized to generate forecasting results.Comparative experiments on real load datasets demonstrate that the model proposed in this paper has higher accuracy compared to other comparison models.
integrated energy systemsmulti-load forecastingseasonal trend decompositionattention mechanismcoupling relationships
NETL
NSTL
万方数据
