TCN-DBN模型在油井产油量预测中的应用  

Application of TCN-DBN Model in Oil Production Prediction of Oil Wells

作　　者：潘少伟[1] 宋倩 王树楷 PAN Shaowei;SONG Qian;WANG Shukai(School of Computer Science,Xi’an Shiyou University,Xi’an,China,710065)

机构地区：[1]西安石油大学计算机学院,西安710065

出　　处：《福建电脑》2025年第3期16-19,共4页Journal of Fujian Computer

基　　金：西安石油大学2023年度“立德树人”研究课题“OBE理念下思政元素融入程序设计类课程教学的探索与实践—以《JAVA程序设计》为例”(No.LD202309)的研究成果,并获得它的资助。

摘　　要：针对长短期记忆神经网络中存在的并行性、长时间序列中梯度稳定性等问题,本文提出一种用于油井产量预测的TCN-DBN算法。将时序卷积网络与深度置信网络相结合,算法使用时序卷积网络提取时间序列特征,同时融合自注意力机制以提高模型预测精确度。实验结果显示,TCN-DBN算法在Z25生产井上与LSTM-DBN算法相比,MAE值下降了72.95%,RMSE值下降了68.94%,MAPE值下降了48.47%。实验结果说明了TCN-DBN算法的预测精度较高,可为原油开采计划的制定提供理论和技术支持。This paper proposes a TCN-DBN algorithm for oil well production prediction to address the issues of parallelism and gradient stability in long short-term memory neural networks.Combining temporal convolutional networks with deep belief networks,the algorithm uses temporal convolutional networks to extract temporal series features,while integrating self attention mechanisms to improve model prediction accuracy.The experimental results showed that compared with the LSTM-DBN algorithm,the TCN-DBN algorithm reduced the MAE value by 72.95%,RMSE value by 68.94%,and MAPE value by 48.47%on the Z25 production well.The experimental results demonstrate that the TCN-DBN algorithm has high prediction accuracy and can provide theoretical and technical support for the formulation of crude oil extraction plans.

关 键 词：时序卷积网络 深度置信网络 油井产油量预测 

分 类 号：TE328[石油与天然气工程—油气田开发工程]