摘要
设计了一种基于电参数和示功图信息融合的综合工况诊断算法,实现了抽油井工况的准确、高效诊断。基于数据字典技术、机器深度学习和信息融合技术,首先使用抽油机电参数法进行快速匹配,无法精准诊断工况时引入示功图方法,充分发挥抽油机电参数法的快速性和抽油井示功图法的准确性优势,融合多种信息进行综合判断;在对示功图进行深度学习过程中,改进了网络模型使其轻量化,使其适用于嵌入式硬件环境。新算法识别判断过程快速、稳定,识别结果准确、可靠,大大提高了抽油井工作效率。
A comprehensive condition diagnosis algorithm based on the fusion of electrical parameters and indicator diagram information is designed,which realizes the accurate and efficient diagnosis of pumping well conditions.Based on data dictionary technology,machine deep learning and information fusion technology,the electrical parameter method of the pumping unit is used for fast matching in the first place,and the indicator diagram method is introduced when the conditions cannot be accurately diagnosed.Give full play to the rapidity of the electric parameter method of the pumping unit and the accuracy of the indicator diagram method of the pumping well,and integrate various information for comprehensive judgment.In the process of deep learning the indicator diagram,the network model is improved to make it lightweight and suitable for embedded hardware environment.
出处
《工业控制计算机》
2023年第6期16-18,共3页
Industrial Control Computer
基金
南京市工业和信息化发展专项资金项目
江苏省研究生科研与实践创新计划项目支持。
关键词
抽油井
电参数
示功图
机器学习
信息融合
oil well
electrical parameter
indicator diagram
machine learning
information fusion