Research On The Thread Of Oil Tube Identification System

In the oil extraction process, the pipeline’s quality of crude oil plays an important part in the production and the safety of oil industry. However, due to the poor quality of the pipeline, many pumping string damage and failure, galling, leakage, slip off wells, collapsing, etc. have occurred, which leads to serious pipeline faults, to the production and operation and to great economic losses. According to statistics, more than95%of the oil pipeline failure occurred in the threaded connections, if the screw geometry and deviation exceeds the scope of the provisions of API, such as the taper is too large, bad teeth may lead to fracture-type failure. So tubing thread parameter identification is of great sense.The topics take pipe thread digital images as the research object, based on digital image processing technology, discussing the screw image acquisition and processing system development and research, as well as the technical implementation. According to the American Petroleum Institute (API) standard, we take thread taper, teeth-angle (API known as the thread angle), pitch, threaded depth (API known as the thread height) as the thread pattern recognition analysis of the characteristic parameters, and gives all characteristic parameters of the definition and calculation methods.This paper applies the improved three-layer feed-forward BP neural network to obtain the characteristic parameters and use experimental ways to obtain the best value of the number of network hidden nodes, momentum coefficient, the error level and step size and other network parameters. We use VB.NET and MATLAB Programming to achieve system functions, SQL Server for establishing the database, which builds the basis for the study of image processing and the characteristic parameters of the thread database management.