Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Oil pumping unit fault diagnosis method based on multiscale convolutional neural network

A technology of convolutional neural network and diagnosis method, which is applied in the field of fault diagnosis of oil pumping units based on multi-scale convolutional neural network, can solve the problems of single filter setting, limited parameter flexibility, etc., and achieves the improvement of fault diagnosis accuracy. Effect

Inactive Publication Date: 2019-08-16
SOUTHWEST PETROLEUM UNIV
View PDF6 Cites 15 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] However, the existing deep neural networks applied to fault diagnosis of dynamometer diagrams are all completed in a single channel, and the size of the filter is set in each layer, which limits the flexibility of parameters

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Oil pumping unit fault diagnosis method based on multiscale convolutional neural network
  • Oil pumping unit fault diagnosis method based on multiscale convolutional neural network
  • Oil pumping unit fault diagnosis method based on multiscale convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0027] The present invention will be described in detail below in conjunction with the accompanying drawings. figure 1 It is an overall flowchart of the present invention. Concrete implementation steps of the present invention are as follows:

[0028] Step 1: Collect the suspension point displacement and load data, and preprocess the original data. The specific preprocessing method is:

[0029] (1) Smooth and normalize the data;

[0030] (2) Draw a two-dimensional dynamometer diagram with displacement and load data. The input of the network model is 32*32, so the size of the dynamometer diagram is also processed as 32*32 pixels.

[0031] Step 2: Make labels for the dynamometer data and divide them into training set and test set according to a certain ratio.

[0032] Convert the dynamometer label into one-hot encoding form, for example: there are 8 categories of dynamometer data, the gas-affected working condition is the 4th category, the original label is 3, and its one-hot...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses an oil pumping unit fault diagnosis method based on a multiscale convolutional neural network. The conventional fault diagnosis method has problems of dependence on artificialselection of features, complicated calculation and poor accuracy rate, and the existing deep neural networks applied to the oil pumping unit fault diagnosis are all completed in a single path, size ofa filter is set singly at each layer, and flexibility of parameters is limited. A multiscale convolutional block is taken as a core structure, and the oil pumping unit fault diagnosis method based ona multiscale convolutional neural network is improved. The method avoids influence of complicated feature engineering and uncertainty of feature selection on fault identification accuracy rate in theconventional fault diagnosis method, meanwhile, the method can extract global and local features with more abundant and effect indicator diagrams, and can improve fault diagnosis accuracy rate.

Description

technical field [0001] The invention relates to a fault diagnosis technology for an oil pumping unit, in particular to a fault diagnosis method for an oil pumping unit based on a multi-scale convolutional neural network. Background technique [0002] Rod pumping units are widely used in my country's petroleum industry. At present, the most widely used method for fault diagnosis of rod well pumps is to use the dynamometer data for analysis. Now the main diagnostic method of the dynamometer diagram is to extract the features of the dynamometer diagram, analyze the dynamometer diagram by using invariant moments, Freeman chain code, gray matrix and other methods, and extract the eigenvectors that can effectively reflect the working conditions of the oil well pump. Combined with BP neural network, support vector machine (SVM) and other diagnostic models for diagnosis. These fault diagnosis methods all rely on manual selection of features, the calculation is complex, and the acc...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G01M99/00G06N3/04
CPCG01M99/007G06N3/045
Inventor 罗仁泽袁杉杉苏赋马磊吕沁王瑞杰张可
Owner SOUTHWEST PETROLEUM UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com