CN-116070910-B - In-service pipeline girth weld risk ordering method, device, equipment and medium

CN116070910BCN 116070910 BCN116070910 BCN 116070910BCN-116070910-B

Abstract

The invention relates to a method, a device, equipment and a medium for ordering risk of in-service pipeline girth welds, which comprise the steps of obtaining a plurality of risk factors of a weld leg to be risk ordered, obtaining weight corresponding to each risk factor and corresponding horizontal total value of the risk factors under specified attribute characteristics for each risk factor, determining comprehensive risk values of the risk factors according to the weight and the horizontal total value of the risk factors for each risk factor, and ordering pipeline girth weld risks according to the comprehensive risk values of the risk factors. The method can be used for sequencing the weld joint risks of in-service gas pipelines, can guide the current domestic girth weld joint investigation, effectively reduces the excavation amount and improves the excavation precision.

Inventors

YU DONGLIANG
LI HONGTAO
YANG CHUAN
YAO DENGZUN
DAI LIANSHUANG
FENG QINGSHAN
WANG AILING
XUAN HENG
SU XIN
WANG YE

Assignees

国家石油天然气管网集团有限公司
国家管网集团西南管道有限责任公司

Dates

Publication Date: 20260505
Application Date: 20230113

Claims (8)

1. The in-service pipeline girth weld risk ordering method is characterized by comprising the following steps of: acquiring a plurality of risk factors of a welded junction to be risk sequenced; For each risk factor, acquiring a weight corresponding to the risk factor and a horizontal total value corresponding to the risk factor under a specified attribute characteristic, wherein the weight represents the risk degree of the risk factor compared with each risk factor, and the horizontal total value represents the risk degree of the specified attribute characteristic of the risk factor compared with each attribute characteristic of the risk factor; for each risk factor, determining a comprehensive risk value of the risk factor according to the weight and the horizontal total value of the risk factor; according to the comprehensive risk values of the risk factors, the ordering of the pipeline girth weld risks is realized; For each risk factor, the obtaining the weight corresponding to the risk factor and the horizontal sum value corresponding to the risk factor under the specified attribute feature includes: For each risk factor, determining the weight of the risk factor through a pre-trained decision tree, wherein the decision tree comprises a plurality of nodes, for each node, each node represents one risk factor and the attribute characteristics of the risk factor, each node corresponds to the weight of the risk factor corresponding to the node, and for a connecting line between every two nodes, the connecting line represents the relationship between the two nodes; For each risk factor, determining a horizontal total value corresponding to the risk factor under a designated attribute feature through a first corresponding relation between each predetermined risk factor and the horizontal total value corresponding to each risk factor under different attribute features, wherein each attribute feature comprises the designated attribute feature; The pre-trained decision tree is built by: S1, constructing a plurality of preprocessed risk factors into a training sample set and a testing sample set; s2, dividing nodes according to the training sample set, and determining a data set D corresponding to each node according to a preset sample constraint condition corresponding to each node of each level, wherein the sample constraint condition can be set based on attribute characteristics of each risk factor; S3, the sample constraint conditions further comprise conditions determined based on the coefficient of the foundation of each sample, after the sample constraint conditions in S2 are met, processing is carried out on each node based on the coefficient of the foundation, if the coefficient of the foundation is smaller than a threshold value, a decision tree subtree is returned, and the node stops recursion; S4, for each node, calculating the coefficient of each attribute feature of each risk factor existing in the node to the data set D based on the attribute feature of the risk factor corresponding to the node, selecting a risk factor A and a corresponding attribute feature a with the minimum coefficient of the coefficient of each attribute feature of each attribute factor to the data set D, taking the risk factor A and the corresponding attribute feature a with the minimum coefficient of the attribute and the coefficient of the attribute as optimal features and optimal feature values respectively, dividing the data set corresponding to the node into two parts D1 and D2 according to the optimal features and the optimal feature values, and simultaneously establishing left and right nodes of the node as two leaf nodes of the node, wherein a connecting line is respectively arranged between the node and the two corresponding leaf nodes; determining a coefficient of a foundation corresponding to the node according to each risk factor corresponding to the node and attribute characteristics corresponding to each risk factor, and assuming that the node corresponds to K risk factors, wherein the probability of the K risk factors is that Namely, the probability that the number of samples corresponding to the Kth risk factor accounts for the total number of samples under the node, and the coefficient of the probability distribution, namely, the weight expression, is: wherein Gini (p) represents the weight corresponding to the node; S5, obtaining a plurality of nodes through the node dividing mode, wherein each node corresponds to a data set, the data set comprises at least one risk factor meeting the constraint condition of a sample and attribute characteristics of each risk factor, each node also corresponds to a weight, and a decision tree is built through the plurality of nodes, the weights corresponding to each node and the connecting lines among the nodes.
2. The method of claim 1, wherein the first correspondence is determined by: for each node, determining a horizontal score corresponding to the risk factor corresponding to the node under the specified attribute characteristic according to the attribute characteristic of the risk factor corresponding to the node, wherein the horizontal score represents the proportion of the number of the risk factors with the same specified attribute characteristic in the training sample to the total number of samples in the training sample; for each node, determining a horizontal total value corresponding to the risk factors corresponding to the node under the appointed attribute characteristics according to the horizontal value corresponding to the node; and establishing the first corresponding relation according to the risk factors corresponding to the nodes and the horizontal sum values corresponding to the nodes.
3. The method of claim 1, wherein for each of the nodes, determining the coefficient of the foundation corresponding to the node based on the respective risk factor corresponding to the node and the attribute feature corresponding to each of the risk factors comprises: For each node, determining a coefficient of a foundation corresponding to the node according to each risk factor corresponding to the node and attribute characteristics corresponding to each risk factor by a first formula, wherein the first formula is as follows: wherein D represents risk factors of the same type corresponding to the nodes, gini (D) represents a coefficient of keni, k represents the number of attribute features possessed by the risk factors, Representing the number of risk factors that are present, Representing the number of kth attribute features.
4. The method according to claim 2, wherein for each node, the determining, according to the attribute characteristics of the risk factors corresponding to the node, the horizontal score corresponding to the risk factors corresponding to the node under the specified attribute characteristics includes: for each node, determining a horizontal score corresponding to the risk factor corresponding to the node under the appointed attribute according to the attribute characteristics of the risk factor corresponding to the node by a second formula, wherein the second formula is as follows: Wherein H ik represents a horizontal score corresponding to a kth attribute feature of the ith risk factor, T ik represents the number of risk factors having the kth attribute feature in the training sample, S ik represents the total number of samples in the training sample, and the kth attribute feature is the specified attribute feature.
5. The method of claim 4, wherein for each of the nodes, the determining a horizontal aggregate value corresponding to the node from the horizontal scores corresponding to the nodes comprises: for each node, determining a horizontal total value corresponding to the node according to a third formula according to the horizontal value corresponding to the node, wherein the third formula is as follows: Wherein f ik represents a horizontal combination value corresponding to the kth attribute feature of the ith risk factor, and n i represents the number of attribute features of the ith risk factor.
6. An in-service pipeline girth weld risk ordering apparatus, characterized in that an in-service pipeline girth weld risk ordering method according to claim 1 is adopted, the apparatus comprises: The risk factor acquisition module is used for acquiring a plurality of risk factors of the welded junction to be risk ordered; The weight and level total value determining module is used for obtaining, for each risk factor, a weight corresponding to the risk factor and a level total value corresponding to the risk factor under a specified attribute characteristic, wherein the weight represents the risk degree of the risk factor compared with each risk factor, and the level total value represents the risk degree of the specified attribute characteristic of the risk factor compared with each attribute characteristic of the risk factor; The comprehensive risk value determining module is used for determining the comprehensive risk value of each risk factor according to the weight and the horizontal total value of the risk factor; and the risk ranking module is used for ranking the risks of the pipeline girth welds according to the comprehensive risk values of the risk factors.
7. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the method of any one of claims 1-5 when the computer program is executed.
8. A computer readable storage medium, characterized in that the computer readable storage medium has stored thereon a computer program which, when executed by a processor, implements the method of any of claims 1-5.

Description

In-service pipeline girth weld risk ordering method, device, equipment and medium Technical Field The invention relates to the technical field of oil and gas pipeline girth weld integrity management, in particular to a method, a device, equipment and a medium for in-service pipeline girth weld risk sequencing. Background In recent years, the fracture failure of the girth weld of the oil and gas long-distance pipeline has become a main accident affecting the safety of the pipeline, and in order to effectively control the girth weld risk, domestic pipeline company organizations perform girth weld quality risk investigation work, a large number of unqualified welded junctions and crack welded junctions are found, and a large number of excavated welded junction data and achievements are accumulated. For the data and results, the data statistics and analysis are carried out by adopting a manual method, so that rules are found and problems are clarified. However, due to the complexity of the girth weld defect cause and the relevance among various influencing factors, manual analysis cannot obtain a clear conclusion, and the girth weld excavation work cannot be guided well, so that an effective welding mouth investigation scheme is formed, and the welding mouth excavation investigation quantity is reduced effectively. The data mining method is an effective means for solving the problem of correlation of a large amount of data and searching data rules. Through deep excavation and mathematical statistical analysis of girth weld investigation data, a girth weld risk ordering model of the gas pipeline is established, a girth weld risk ordering technology is formed, and the excavation accuracy of in-service high-strength steel pipeline crack welded junctions is improved, so that the method is a work which needs to be developed urgently at present. Disclosure of Invention The invention aims to solve at least one technical problem by providing a method, a device, equipment and a medium for ordering the risk of circumferential welds of in-service pipelines. The technical scheme for solving the technical problems is as follows, the method for ordering the risk of the girth weld of the in-service pipeline comprises the following steps: acquiring a plurality of risk factors of a welded junction to be risk sequenced; For each risk factor, acquiring a weight corresponding to the risk factor and a horizontal total value corresponding to the risk factor under a specified attribute characteristic, wherein the weight represents the risk degree of the risk factor compared with each risk factor, and the horizontal total value represents the risk degree of the specified attribute characteristic of the risk factor compared with each attribute characteristic of the risk factor; for each risk factor, determining a comprehensive risk value of the risk factor according to the weight and the horizontal total value of the risk factor; and according to the comprehensive risk values of the risk factors, the ordering of the pipeline girth weld risks is realized. The method has the advantages that in the scheme, the multiple risk factors influencing the pipeline weld risk are considered, the risk degree of each risk factor compared with each risk factor is considered, the risk degree of the appointed attribute characteristic of each risk factor compared with each attribute characteristic of the risk factor is considered, the comprehensive risk value of each risk factor is determined from multiple aspects, and finally the ordering of pipeline girth weld risk is realized according to the comprehensive risk value of each risk factor. On the basis of the technical scheme, the invention can be improved as follows. Further, for each risk factor, the obtaining the weight corresponding to the risk factor and the horizontal combined value corresponding to the risk factor under the specified attribute feature includes: For each risk factor, determining the weight of the risk factor through a pre-trained decision tree, wherein the decision tree comprises a plurality of nodes, for each node, each node represents one risk factor and the attribute characteristics of the risk factor, each node corresponds to the weight of the risk factor corresponding to the node, and for a connecting line between every two nodes, the connecting line represents the relationship between the two nodes; And for each risk factor, determining a horizontal total value corresponding to the risk factor under the appointed attribute characteristics through a first corresponding relation between each predetermined risk factor and the horizontal total value corresponding to each risk factor under different attribute characteristics, wherein each attribute characteristic comprises the appointed attribute characteristics. The further scheme has the beneficial effect that the weight of each risk factor and the horizontal sum value of each risk factor can be more accurately determined throu