首页|期刊导航|煤气与热力|基于LiBi-Transformer模型的城市燃气日负荷预测

基于LiBi-Transformer模型的城市燃气日负荷预测OA

Urban Gas Daily Load Forecasting Based on LiBi-Transformer Model

中文摘要英文摘要

提出一种基于LiBi-Transformer算法的供暖期燃气日负荷预测模型.该模型的输入特征参数为日最高温度、日最低温度、日平均温度、风力等级、天气类型、星期类型、节假日类型,输出参数为日用气量.与BP、LSTM、GRU和Transformer模型相比,LiBi-Transformer模型预测效果更好,表明LiBi-Transformer模型具有更高的准确性,可用于供暖期城市燃气日负荷的预测.

A daily gas load forecasting model for the heating season based on the LiBi-Transformer algo-rithm is proposed.The input feature parameters of the model include daily maximum temperature,daily mini-mum temperature,daily average temperature,wind scale,weather type,day-of-week type and holiday type,while the output parameter is the daily gas consump-tion.Compared with BP,LSTM,GRU and Transformer models,the LiBi-Transformer model achieves better prediction performance.The results demonstrate that the LiBi-Transformer model possesses higher accuracy and can be applied to forecast urban daily gas load dur-ing the heating season.

郝学军;许佳雯

北京建筑大学 环境与能源工程学院,北京 102616北京建筑大学 环境与能源工程学院,北京 102616

建筑与水利

燃气日负荷预测注意力机制LiBi-Transformer模型

daily gas load forecastingattention mecha-nismLiBi-Transformer model

《煤气与热力》 2026 (9)

43-46,4

评论