Analysis And Research Of Abnormal Wells In Oil Field Based On Clustering And Association Rules

With the application of the Internet of Things in oil and gas production and the real-time monitoring system in oil and gas fields,the real-time database and relational database will generate a large amount of production data at every moment.How to analyze and utilize these data reasonably and efficiently is related to all aspects of oil and gas production.With the rapid development of data mining technology,oilfield industry has gradually applied this technology to actual production and operation of oilfileds.As an important way to analyze and process data,clustering analyze technology and association rules have attracted more and more attention.The operating state of pumping Wells is an important factor affecting the oil production and how to judge the production status of pumping efficiently is an urgent problem to be solved now.Based on the existing research and the actual production situation of oil field,the article aims to propose an improved k-means algorithm to identify the abnormal well.According to the clustering results,using the apriori algorithm to analyze the production parameters which may affect the operation status of pumping wells.This article first to analysis the research background of the abnormal well in oil field and introduce the applicaion of data mining technology in the oil and gas production process,then analyzed the k-means algorithm and Apriori algorithms in the theory.The article include main research work as follows:1.In view of the k-means algorithm is affected with the abnormal point and the outliers,the article put forward a kind of weighted Euclidean distance to achieve the purpose that reduce the influence of abnormal points and outliers on clustering effect,and then using the Python programming to analysis the UCI IRIS data set,the experiment result shows that the improved k-means algorithm has a better clustering quality and higher accuracy than k-means algorithm.2.Use the improved K-means algorithm to clustering analysis the production parameters of pumping well and production data.Firstly,using the data preprocessing techniques to deal with the source data and using statistical mathematics formula to calculate the volatility of production parameters,then use python programming experiment to clustering analysis the volatility of production parameters of well pumping,last judge the exception types and give the corresponding solutions.3.According to the results of cluster analysis,using the apriori algorithm is to analyze the correlation of production parameters of abnormal Wells,and by setting different support degrees to find the strong correlation between production parameters,achieving the goal of to find the problems quickly and solving problems as soon as possible in the oil and gas production process.