CN-121999262-A - Method and device for determining type of corrosion outside steel pipe based on computer vision

CN121999262ACN 121999262 ACN121999262 ACN 121999262ACN-121999262-A

Abstract

The invention discloses a method and a device for determining the type of corrosion outside a steel pipe based on computer vision. The method comprises the steps of obtaining outer corrosion failure image data and outer corrosion working condition data of a pipeline to be identified, identifying that the outer corrosion appearance type of the pipeline to be identified is local corrosion or comprehensive corrosion by using a trained machine learning model according to the outer corrosion failure image data of the pipeline to be identified, determining the corrosion type of the pipeline to be identified based on a pre-built multi-mode identification model according to a first set of specified parameters in the outer corrosion working condition data of the pipeline to be identified if the outer corrosion appearance type of the pipeline to be identified is local corrosion, and determining the corrosion type of the pipeline to be identified based on the pre-built multi-mode identification model if the outer corrosion appearance type of the pipeline to be identified is comprehensive corrosion according to a second set of specified parameters in the outer corrosion working condition data of the pipeline to be identified. The method can rapidly, accurately and effectively identify the type of corrosion outside the steel pipe, has high identification efficiency and low cost, and provides accurate data support for the management and control of the risk of the steel pipe.

Inventors

TANG DEZHI
LIU JIE
HU MIN
GU TAN
ZHAO FEI
GUO DACHENG

Assignees

中国石油天然气股份有限公司

Dates

Publication Date: 20260508
Application Date: 20241108

Claims (11)

1. The method for determining the type of the steel pipe external corrosion based on computer vision is characterized by comprising the following steps of: Acquiring outer corrosion failure image data and outer corrosion working condition data of a pipeline to be identified; According to the external corrosion failure image data of the pipeline to be identified, using a trained machine learning model to identify that the external corrosion morphology type of the pipeline to be identified is local corrosion or comprehensive corrosion; If the corrosion is local corrosion, determining the corrosion type of the pipeline to be identified based on a multi-mode identification model constructed in advance according to a first set of specified parameters in the pipeline external corrosion working condition data to be identified, wherein the first set of specified parameters comprise at least one of alternating current density, cathodic protection current density, self-corrosion potential, cathodic protection setting condition, outage potential, lap joint condition and heat preservation setting condition; And if the corrosion is comprehensive corrosion, determining the corrosion type of the pipeline to be identified based on a pre-constructed multi-mode identification model according to a second set of specified parameters in the pipeline external corrosion working condition data to be identified, wherein the second set of specified parameters comprises at least one of corrosion failure position, heat preservation layer setting condition and cathode protection setting condition.
2. The method of claim 1, wherein determining the corrosion type of the pipe to be identified based on the pre-constructed multi-modal identification model based on a first set of specified parameters in the pipe-out corrosion condition data to be identified comprises: If the pipeline to be identified meets at least one preset AC stray current corrosion index, determining that the corrosion type is AC stray current corrosion, wherein the AC stray current corrosion index comprises at least one of an AC current density index, a power-off potential index and a cathode protection current density index; If the pipeline to be identified does not meet any preset AC stray current corrosion index, but meets at least one preset DC stray current corrosion index, determining that the corrosion type is DC stray current corrosion, wherein the DC stray current corrosion index comprises a cathode protection setting condition, a corrosion potential index, a self-corrosion potential and a power-off potential index; if the pipeline to be identified does not meet any preset AC stray current corrosion index, does not meet any preset DC stray current corrosion index, and the lap joint condition is lap joint, determining that the corrosion type is galvanic corrosion; If the pipeline to be identified does not meet any preset alternating current stray current corrosion index, and does not meet any preset direct current stray current corrosion index, the lap joint condition is that lap joint is not performed, the heat preservation layer is arranged, and the corrosion type is determined to be corrosion under the heat preservation layer; If the pipeline to be identified does not meet any preset alternating current stray current corrosion index, any preset direct current stray current corrosion index is not met, the lap joint condition is that lap joint is not achieved, the heat preservation layer is not arranged, and the corrosion type is determined to be natural soil corrosion.
3. The method of claim 2, wherein, The alternating current stray current corrosion index comprises: The alternating current density is larger than a preset first current density threshold value; the alternating current density is smaller than a preset first current density threshold value and is larger than a preset second current density threshold value, and the power-off potential is more positive than the first potential threshold value or more negative than the second potential threshold value; The alternating current density is smaller than a preset first current density threshold value and is larger than a preset second current density threshold value, the power-off potential is more positive than the first potential threshold value, and the cathodic protection current density is larger than the cathodic protection current density threshold value; the direct current stray current corrosion index comprises: No cathodic protection is applied and the corrosion potential is more positive than the self-corrosion potential + set point; Cathodic protection is applied and the power-off potential is more positive than the third potential threshold; The corrosion potential of the pipeline periodically fluctuates; the power-off potential fluctuates periodically.
4. The method of claim 1, wherein determining the corrosion type of the pipe to be identified based on the pre-constructed multi-modal identification model based on a second set of specified parameters in the pipe-out corrosion condition data to be identified comprises: if the corrosion failure position in the pipeline external corrosion working condition data to be identified is the joint coating position, determining that the corrosion type is joint coating corrosion; if the corrosion failure position in the pipeline external corrosion working condition data to be identified is not the joint coating position and the heat preservation layer is arranged, determining that the corrosion type is corrosion under the heat preservation layer; If the corrosion failure position in the pipeline external corrosion working condition data to be identified is not the joint position, the heat preservation layer is not arranged, and the cathode protection is applied, determining that the corrosion type is corrosion caused by cathode protection failure; If the corrosion failure position in the pipeline external corrosion working condition data to be identified is not the joint coating position, the heat preservation layer is not arranged, and the cathode protection is not applied, the corrosion type is determined to be natural soil corrosion.
5. The method of claim 1, wherein the identifying the type of the external corrosion profile of the pipe to be identified as localized corrosion or general corrosion using a trained machine learning model based on the pipe external corrosion failure image data to be identified comprises: Inputting the external corrosion failure image data of the pipeline to be identified into a trained convolutional neural network model, analyzing the external corrosion failure image data by the convolutional neural network model, obtaining external corrosion morphology feature data of the steel pipe, and determining that the external corrosion morphology type of the pipeline to be identified is local corrosion or comprehensive corrosion according to the external corrosion morphology feature data.
6. The method as recited in claim 5, further comprising: Acquiring pipeline external corrosion failure sample data, wherein the sample data comprises pipeline external corrosion failure image data and an external corrosion morphology type tag; Training the constructed machine learning model by using the pipeline outer corrosion failure sample data to obtain a trained machine learning model, wherein the machine learning model is an outer corrosion morphology recognition model based on a neural network.
7. The method of claim 6, wherein training the constructed machine learning model using the out-of-pipe corrosion failure sample data to obtain a trained machine learning model comprises: dividing the sample data of the corrosion failure outside the pipeline into a training set and a testing set; Inputting pipeline external corrosion failure image data in a training set into a pre-constructed external corrosion morphology recognition model based on a neural network, and optimizing super parameters of the model according to the external corrosion morphology type output by the model and an external corrosion morphology type label in the training set to enable the accuracy of the model to reach a preset accuracy requirement and obtain an optimized external corrosion morphology recognition model, wherein the super parameters comprise at least one of training times, batch training size, a loss function, an optimization function, learning rate, momentum and weight attenuation; The method comprises the steps of using a test set to verify a model, inputting pipeline external corrosion failure image data in a training set into an optimized external corrosion morphology recognition model based on a neural network, judging whether the accuracy of the model meets the preset accuracy requirement according to the external corrosion morphology type output by the model and an external corrosion morphology type label in the training set, if not, returning to continue to execute the process of training the model by using the training set, and if so, obtaining the trained external corrosion morphology recognition model.
8. A computer vision-based steel pipe external corrosion type determining device, comprising: the data acquisition module is used for acquiring the external corrosion failure image data and the external corrosion working condition data of the pipeline to be identified; The first identification module is used for identifying that the external corrosion morphology type of the pipeline to be identified is local corrosion or comprehensive corrosion by using a trained machine learning model according to the external corrosion failure image data of the pipeline to be identified; The system comprises a first identification module, a second identification module and a control module, wherein the first identification module is used for determining the corrosion type of a pipeline to be identified based on a multi-mode identification model constructed in advance according to a first set of specified parameters in the external corrosion working condition data of the pipeline to be identified if the pipeline to be identified is locally corroded, and the first set of specified parameters comprise at least one of alternating current density, cathodic protection current density, self-corrosion potential, cathodic protection setting condition, outage potential, lap joint condition and heat preservation setting condition; And the third identification module is used for determining the corrosion type of the pipeline to be identified based on a pre-constructed multi-mode identification model according to a second set of specified parameters in the pipeline external corrosion working condition data to be identified if the pipeline to be identified is subjected to comprehensive corrosion, wherein the second set of specified parameters comprises at least one of corrosion failure position, heat preservation layer setting condition and cathode protection setting condition.
9. The apparatus as recited in claim 8, further comprising: The model training module is used for acquiring pipeline external corrosion failure sample data, wherein the sample data comprises pipeline external corrosion failure image data and external corrosion morphology type labels, training the constructed machine learning model by using the pipeline external corrosion failure sample data to obtain a trained machine learning model, and the machine learning model is an external corrosion morphology identification model based on a neural network.
10. A computer storage medium having stored therein computer executable instructions which when executed by a processor implement the computer vision based method for determining the type of out-of-steel-pipe corrosion of any one of claims 1 to 7.
11. An identification device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the computer vision-based method for determining the type of out-of-steel-pipe corrosion of any one of claims 1 to 7 when executing the program.

Description

Method and device for determining type of corrosion outside steel pipe based on computer vision Technical Field The invention relates to the technical field of pipeline failure analysis, in particular to a method and a device for determining the type of corrosion outside a steel pipe based on computer vision. Background The oil gas pipeline is an important component part of oil gas development and production, and ensuring safe and stable operation of the oil gas pipeline is very important for guaranteeing national oil gas energy safety. Most of the oil and gas pipelines are laid underground due to management and economic considerations. The presence of water, various corrosive ions, bacteria, stray currents, etc. in the soil may cause corrosion perforations outside the buried pipeline, thereby affecting the integrity of the oil and gas pipeline. In order to control the external corrosion of the buried pipeline, an external corrosion-resistant layer and a cathode protection measure are often adopted, and an interference protection measure is also adopted when stray current is interfered. However, damage to the outer anticorrosive layer of the pipeline is unavoidable, cathodic protection is not applied to a part of the pipeline, the effectiveness of cathodic protection of the pipeline applied with cathodic protection is difficult to ensure sometimes, and the effect of stray current interference still leads to high risk of outer corrosion of the buried pipeline, and the type of outer corrosion failure is multiple, and the identification of the type of outer corrosion is difficult, such as natural soil corrosion, joint coating corrosion, corrosion under an insulating layer, alternating current stray current corrosion, direct current stray current corrosion, galvanic corrosion, corrosion caused by cathodic protection failure and the like. Currently, identification of the type of corrosion outside buried pipelines is mainly based on human experience and third party laboratories. Experience identification affects the efficiency and precision of failure identification, and third party detection increases workload and expense investment, reduces coverage rate, and cannot meet the requirements of oil and gas pipeline failure management and risk management. The two modes can not accurately and effectively identify the external corrosion type of the buried steel pipe, can not formulate effective targeted measures and perform effective risk management and control aiming at different external corrosion types, and can not ensure the safe and stable operation of the buried oil and gas pipeline. Disclosure of Invention In view of the above problems, the present invention has been made to provide a method and apparatus for determining an external corrosion type of a steel pipe based on computer vision, which overcome or at least partially solve the above problems, and can provide a convenient, rapid, scientific and effective method for identifying an external corrosion type for an oil and gas pipeline on site, so as to support establishment of targeted measures for external corrosion of a buried oil and gas pipeline and management of risk of external corrosion, and ensure safe and smooth operation of the buried oil and gas pipeline. The embodiment of the invention provides a method for determining the type of corrosion outside a steel pipe based on computer vision, which comprises the following steps: Acquiring outer corrosion failure image data and outer corrosion working condition data of a pipeline to be identified; According to the external corrosion failure image data of the pipeline to be identified, using a trained machine learning model to identify that the external corrosion morphology type of the pipeline to be identified is local corrosion or comprehensive corrosion; If the corrosion is local corrosion, determining the corrosion type of the pipeline to be identified based on a multi-mode identification model constructed in advance according to a first set of specified parameters in the pipeline external corrosion working condition data to be identified, wherein the first set of specified parameters comprise at least one of alternating current density, cathodic protection current density, self-corrosion potential, cathodic protection setting condition, outage potential, lap joint condition and heat preservation setting condition; And if the corrosion is comprehensive corrosion, determining the corrosion type of the pipeline to be identified based on a pre-constructed multi-mode identification model according to a second set of specified parameters in the pipeline external corrosion working condition data to be identified, wherein the second set of specified parameters comprises at least one of corrosion failure position, heat preservation layer setting condition and cathode protection setting condition. In some alternative embodiments, the determining the corrosion type of the pipe to be identified based on the pre