摘要
目前成熟且最具实用性的抽油机故障诊断方法为示功图分析法,而示功图识别算法又是示功图分析法的基础,所以示功图识别算法的精确度对抽油机故障诊断的准确度有极大的影响。利用油井实测数据,对灰度矩阵识别算法、矢量法识别算法和标准误差法识别算法等对示功图识别精确度较高的算法进行了实验研究。提出了基于3种识别算法的相似度计算方法;通过对实验结果的对比分析,阐述了3种算法各自的优缺点以及对示功图波动的敏感度,为示功图识别技术的可靠性分析提供了一定的理论依据。
The three recognition algorithms which have higher accuracy to identify the dynamometer card have been studied by using the measured data of oil wells,including the gray-level matrix recognition algorithm,the vector recognition algorithm and the standard error recognition algorithm.The similarity calculation method based on the three recognition algorithms is proposed.By contrasting and analyzing the experiment results,the advantages and disadvantages of the three algorithms and their sensitivity to dynamometer cards fluctuation are demonstrated in this paper.
出处
《工业控制计算机》
2018年第2期110-111,114,共3页
Industrial Control Computer
基金
陕西省工业科技攻关项目(2015GY110)资助
关键词
抽油机
示功图
模式识别
故障诊断
识别算法
pumping unit
dynamometer card
pattern recognition
fault diagnosis
recognition algorithm
