基于CNN-BiLSTM-Attention的城镇燃气用气量预测

Urban Gas Consumption Prediction Based on CNN-BiLSTM-Attention Network Model

在线阅读下载PDF

导出

摘要为解决传统用气量预测方法在处理复杂时间序列数据时的局限性,提出一种结合卷积神经网络、双向长短期记忆网络和注意力机制的组合模型(CNN-BiLSTM-Attention)用于城镇燃气用气量预测,并对西部某市的实际用气情况进行实证分析。结果显示,该模型的均方根误差、平均绝对误差和决定系数分别为19.14、17.53和0.9666,预测效果显著优于其他模型。CNN-BiLSTM-Attention网络模型为城镇燃气用气量预测提供了一种有效的解决方案,为城市能源管理和决策提供了科学依据。 To address the limitations of traditional gas consumption prediction methods in dealing with complex time series data,a combined model(CNN-BiLSTM-Attention)integrating convolutional neural networks,bidirectional long short-term memory networks,and attention mechanisms is proposed for urban gas consumption prediction.An empirical analysis was conducted on the actual gas consumption situation of a city in the western region.The results show that the root mean square error,mean absolute error,and coefficient of determination of this model are 19.14,17.53 and 0.9666 respectively,and its prediction effect is significantly better than that of other models.The research indicates that the CNN-BiLSTM-Attention network model provides an effective solution for urban gas consumption prediction and offers a scientific basis for urban energy management and decision-making.

作者丁磊邓小毅马玺郭乐华龙海红廖春梅黎国鑫徐玲 DING Lei;DENG Xiaoyi;MA Xi;GUO Lehua;LONG Haihong;LIAO Chunmei;LI Guoxin;XU Ling(Neijiang China Resources Gas Co.,Ltd.,Neijiang 641000,Sichuan,China;China Petroleum and Chemical Corporation Southwest Oil and Gas Branch Gas Production Plant 2,Langzhong 637400,Sichuan,China)

机构地区内江华润燃气有限公司中国石油化工股份有限公司西南油气分公司采气二厂

出处《科技和产业》 2025年第15期66-73,共8页 Science Technology and Industry

关键词城镇燃气用气量预测卷积神经网络双向长短期记忆神经网络注意力机制 urban gas gas consumption prediction convolutional neural network bidirectional long short term memory attention mechanisms

分类号 TU996.3 [建筑科学—供热、供燃气、通风及空调工程]

作者简介丁磊(1995-),男,四川自贡人,硕士,初级工程师,研究方向为智慧城镇燃气;邓小毅(1975-),男,四川内江人,工程师,研究方向为燃气安全管理;马玺(1981-),男,四川内江人,硕士,高级工程师,研究方向为信息项目管理;郭乐华(1986-),男,四川资阳人,工程师,研究方向为燃气供销差管理;龙海红(1982-),男,四川广安人,工程师,研究方向为燃气信息化技术;廖春梅(1989-),女,四川遂宁人,工程师,研究方向为燃气输配技术;黎国鑫(1995-),男,四川资阳人,初级工程师,研究方向为燃气信息化技术;徐玲(1995-),女,四川南充人,硕士,初级工程师,研究方向为天然气管道集输防腐。