CN-121833671-B - Ship liquid cargo stainless steel structure welding quality data processing method and system

CN121833671BCN 121833671 BCN121833671 BCN 121833671BCN-121833671-B

Abstract

The invention provides a method and a system for processing welding quality data of a ship liquid cargo stainless steel structure, and relates to the technical field of welding quality control; the method comprises the steps of carrying out matching association on a multi-source data set and three-dimensional space coordinates of a welding seam to construct a welding quality database, carrying out intelligent recognition on nondestructive testing data in the welding quality database based on a preset stainless steel welding defect feature database to obtain a defect feature data set, and constructing an initial virtual reference plane based on the welding quality database through space coordinate data of a longitudinal and transverse bone intersection structure of a cargo tank and the position of a boundary corner point of a bulkhead. The invention ensures the service safety of corrosion resistance and crack resistance of the cargo tank welding structure.

Inventors

LIU BIN
JIANG KEWEI
LIU YINGFENG
WU HAITAO

Assignees

黄海造船有限公司

Dates

Publication Date: 20260508
Application Date: 20260313

Claims (10)

1. The method for processing the welding quality data of the ship liquid cargo stainless steel structure is characterized by comprising the following steps: Acquiring welding process parameters, nondestructive testing data and stainless steel material characteristic data, preprocessing the acquired data to obtain a multi-source data set, and carrying out matching association on the multi-source data set and the three-dimensional space coordinates of the welding seam to construct a welding quality database; Based on a preset stainless steel welding defect feature library, carrying out intelligent recognition on nondestructive testing data in a welding quality database to obtain a defect feature data set; According to the space included angle between the main direction and the normal vector of the initial virtual reference plane, rotating the initial virtual reference plane through rotation transformation around any space axis to obtain a self-adaptive reference plane with the normal vector perpendicular to the main direction; The method comprises the steps of calculating orthogonal projection points from each defect point in defect characteristic data set to a self-adaptive reference plane, taking a point set formed by all projection points as a calibrated defect space position, mapping the calibrated defect space position and corresponding defect attribute to a cargo tank three-dimensional model, and obtaining a digital twin characteristic layer with defect space distribution; and carrying out statistical analysis on the digital twin characteristic layers of the defect space distribution of the batches to obtain an analysis result, and automatically generating a welding quality assessment report according to the analysis result.
2. The method for processing the welding quality data of the ship liquid cargo stainless steel structure according to claim 1, wherein the method for processing the welding quality data is characterized by collecting welding process parameters, nondestructive testing data and stainless steel material characteristic data, preprocessing the collected data to obtain a multi-source data set, matching and correlating the multi-source data set with three-dimensional space coordinates of a welding seam, and constructing a welding quality database, and comprises the following steps: Respectively acquiring welding current, arc voltage, welding speed, argon flow and interlayer temperature data to form a welding process parameter original set, respectively acquiring detection waveforms, images and numerical data from phased array ultrasonic, ray and vortex detection equipment to form a nondestructive detection original set; Respectively cleaning the welding process parameter original set, the nondestructive testing original set and the material characteristic original set, removing noise data with abnormal jump values and obvious deviation threshold values in the acquisition process, and unifying the timestamp format and the measurement unit of each data source to obtain a standardized welding process parameter set, a standardized nondestructive testing data set and a standardized material characteristic data set; And carrying out data fusion on the standardized welding process parameter set, the nondestructive testing data set and the material characteristic data set, and carrying out one-to-one matching association with the three-dimensional space coordinates of the welding seam extracted from the digital design drawing by taking the welding seam number as a main key to form a welding quality database containing welding parameters, detection results, material characteristics and space coordinates.
3. The method for processing the welding quality data of the ship cargo tank stainless steel structure according to claim 2 is characterized by intelligently identifying nondestructive testing data in a welding quality database based on a preset stainless steel welding defect feature library to obtain a defect feature data set, constructing an initial virtual reference plane based on the welding quality database through space coordinate data of a cargo tank cross structure and bulkhead boundary corner positions, and comprising the following steps: extracting phased array ultrasonic detection waveform data, radial detection image data and eddy current detection numerical data associated with a weld joint number from a welding quality database to form a nondestructive detection data set to be identified; Inputting a nondestructive testing data set to be identified into a preset stainless steel welding defect feature library for feature matching, pre-storing waveform features, image gray features and numerical threshold ranges of hot cracks, intergranular corrosion trends and unfused defect types in the defect feature library, and identifying the defect types, sizes and space coordinates contained in the nondestructive testing data through traversal comparison to generate a defect feature data set; and extracting predefined cross-bone nodes and bulkhead boundary corner coordinates in the structural design data of the cargo tank from a welding quality database, taking the cross-bone nodes and bulkhead boundary corner coordinates as space base characteristic points, and generating an initial virtual reference plane according to the three-dimensional coordinate fitting of the space base characteristic points.
4. The method for processing welding quality data of ship liquid cargo stainless steel structures according to claim 3, wherein the method for processing welding quality data of ship liquid cargo stainless steel structures is characterized by solving eigenvectors of a space coordinate covariance matrix by calculating space coordinate covariance matrices of all defect points in defect characteristic data sets, taking eigenvectors corresponding to maximum eigenvalues as main directions of defect distribution, rotating an initial virtual reference plane by rotation transformation around a space arbitrary axis according to space included angles between the main directions and normal vectors of the initial virtual reference plane, and obtaining an adaptive reference plane with normal vectors perpendicular to the main directions, and comprises the following steps: Extracting three-dimensional space coordinates of all defect points from the defect characteristic data set to form a defect point coordinate matrix, and calculating a covariance matrix of the defect point coordinate matrix to obtain a space coordinate covariance matrix; Performing eigenvalue decomposition on the space coordinate covariance matrix, solving to obtain three eigenvalues and corresponding eigenvectors, and selecting the eigenvector corresponding to the eigenvalue with the largest numerical value as a main direction of defect distribution, wherein the main direction of defect distribution represents the maximum discrete extending direction of defect points in a three-dimensional space; extracting a method vector from the initial virtual reference plane, calculating a space included angle between the main direction of defect distribution and the normal vector of the initial virtual reference plane, and determining the angle required to rotate the initial virtual reference plane and the direction of a rotating shaft according to the space included angle; and (3) taking the rotation axis as a benchmark, performing rotation transformation on the initial virtual reference plane around any spatial axis according to the spatial included angle, enabling the normal vector of the rotated plane and the main direction of defect distribution to be mutually perpendicular, and taking the plane subjected to rotation transformation as a self-adaptive reference plane.
5. The method for processing welding quality data of ship liquid cargo stainless steel structures according to claim 4, wherein calculating orthogonal projection points from each defect point in defect feature data set to a self-adaptive reference plane, using a point set formed by all projection points as a calibrated defect space position, mapping the calibrated defect space position and corresponding defect attribute to a liquid cargo tank three-dimensional model to obtain a digital twin feature layer with defect space distribution, and comprising: extracting three-dimensional space coordinates of each defect point from the defect characteristic data set, respectively calculating vertical distances from each defect point to the self-adaptive reference plane, solving the vertical foot coordinates of each defect point on the self-adaptive reference plane based on a space equation of the self-adaptive reference plane, taking the vertical foot coordinates as space positions after the corresponding defect points are calibrated, and constructing a calibration defect point set according to the space positions after the calibration of all the defect points; Extracting the defect type, the defect size and the original space coordinates corresponding to each defect point from the defect characteristic data set as defect attribute information, and carrying out one-to-one association on the defect attribute information and the calibrated space positions in the calibrated defect point set to generate a calibrated defect characteristic data set containing the calibrated positions and the attribute information; And (3) introducing the calibrated defect characteristic data set into a three-dimensional model of the cargo tank, positioning each defect point and attribute information to the corresponding coordinate position of the three-dimensional model of the cargo tank according to the calibrated space position, and forming a digital twin characteristic layer of defect space distribution accurately corresponding to the spatial position of the structural entity in the three-dimensional model of the cargo tank.
6. The method for processing welding quality data of a cargo tank stainless steel structure according to claim 5, wherein the step of introducing the calibration defect feature data set into the cargo tank three-dimensional model, positioning each defect point and attribute information to a corresponding coordinate position of the cargo tank three-dimensional model according to the calibrated spatial position, and forming a digital twin feature layer of defect spatial distribution accurately corresponding to the spatial position of the structural entity in the cargo tank three-dimensional model, comprises: Analyzing a bottom data structure of the three-dimensional model of the cargo tank, and acquiring geometric topology information of all bulkheads, longitudinal and transverse bones and weld joint members contained in the three-dimensional model of the cargo tank and a spatial index under a model coordinate system; the searched target welding seam component is used as a defect attachment carrier, corresponding defect attribute information is written into an expansion data storage area of the target welding seam component in an attribute field mode, and a defect identification node is generated at a coordinate point corresponding to the calibrated space position on the target welding seam component; according to the spatial distribution of each defect identification node in the cargo tank three-dimensional model, clustering and correlating the defect identification nodes belonging to the same weld joint to form a defect cluster with spatial semantics, and superposing the defect cluster to a corresponding structural area of the cargo tank three-dimensional model in the form of an independent layer to construct a digital twin characteristic layer for obtaining the spatial distribution of the defects.
7. The method for processing welding quality data of a ship liquid cargo stainless steel structure according to claim 6, wherein the statistical analysis is performed on the defect space distribution digital twin feature layers of the plurality of batches to obtain an analysis result, and the welding quality assessment report is automatically generated according to the analysis result, and the method comprises the steps of: extracting a defect space distribution digital twin characteristic layer corresponding to each batch from a welding quality database, wherein the defect space distribution digital twin characteristic layer comprises a calibrated space position, a defect type, a defect size and welding process parameters related to each defect point; the method comprises the steps of performing data fusion on the extracted defect space distribution digital twin feature layers of a plurality of batches to construct a multi-batch defect data set comprising a time dimension, a space dimension and an attribute dimension; carrying out statistical analysis on the multi-batch defect data set, calculating the overall density of defects of each batch, the duty ratio distribution of various defects and the statistical characteristics of defect sizes, carrying out correlation regression analysis on the spatial distribution position of the defects and welding process parameters, and identifying the correlation rule between abnormal fluctuation of the welding parameters and defect generation; According to a correlation rule, combining a preset intergranular corrosion tendency threshold value and a preset thermal cracking sensitivity threshold value, carrying out trend prediction on the welding quality of the current batch and the subsequent batch, and judging whether a batch quality risk exists; and carrying out structural organization on defect distribution rules, parameter association analysis results and quality risk prediction conclusions obtained by statistical analysis, automatically filling data according to a preset report template, and generating a welding quality assessment report.
8. A ship liquid cargo stainless steel structure welding quality data processing system implementing the method according to any one of claims 1 to 7, comprising: The acquisition module is used for acquiring welding process parameters, nondestructive testing data and stainless steel material characteristic data, preprocessing the acquired data to obtain a multi-source data set, and carrying out matching association on the multi-source data set and the three-dimensional space coordinates of the welding seam to construct a welding quality database; The device comprises a recognition module, a welding quality database, an initial virtual reference plane, a welding quality database and a bulkhead boundary corner point position detection module, wherein the recognition module is used for intelligently recognizing nondestructive testing data in the welding quality database based on a preset stainless steel welding defect feature database to obtain a defect feature data set; The computing module is used for solving the feature vector of the space coordinate covariance matrix by computing the space coordinate covariance matrix of all defect points in the defect feature data set, taking the feature vector corresponding to the maximum feature value as the main direction of defect distribution; The calibration module is used for calculating the orthogonal projection point from each defect point in the defect characteristic data set to the self-adaptive reference plane, and taking a point set formed by all projection points as a calibrated defect space position; the analysis module is used for carrying out statistical analysis on the digital twin feature layers of the defect space distribution of the batches to obtain an analysis result, and automatically generating a welding quality assessment report according to the analysis result.
9. A computing device, comprising: One or more processors; storage means for storing one or more programs which when executed by the one or more processors cause the one or more processors to implement the method of any of claims 1 to 7.
10. A computer readable storage medium, characterized in that the computer readable storage medium has stored therein a program which, when executed by a processor, implements the method according to any of claims 1 to 7.

Description

Ship liquid cargo stainless steel structure welding quality data processing method and system Technical Field The invention relates to the technical field of welding quality control, in particular to a method and a system for processing welding quality data of a ship liquid cargo stainless steel structure. Background The ship cargo tank adopts an austenitic stainless steel welding structure, has extremely high strict requirements on the mechanical property, intergranular corrosion resistance and thermal crack resistance of a welding joint, and the welding quality data processing of the traditional ship cargo tank stainless steel structure generally has the problems that welding process parameters, nondestructive testing data and material characteristic data are heterogeneous and dispersed and isolated, various quality data cannot be accurately matched and associated with the three-dimensional space coordinates of welding seams, the welding defects depend on manual interpretation and identification efficiency and have large errors, the defect space position does not have an adaptive calibration means of a complex structure of the adaptive cargo tank, the defect digital twin mapping relation corresponding to the solid structure is difficult to construct, and the multi-batch quality data lack systematic statistical analysis and trend prediction cannot realize the full-flow digital control of the welding quality; In the actual welding operation of the longitudinal and transverse bone intersection nodes of the cargo hold and the bulkhead boundary fillet weld of a certain ocean chemical ship 316L stainless steel liquid cargo hold, the traditional management and control mode of manually recording welding electric parameters, offline reading phased array ultrasonic and ray detection images and simply judging defects by experience is adopted, the outstanding technical defects are that multi-source quality data are not fused and normalized, the matching precision of weld space coordinates and defect data is low, the defect space position self-adaptive calibration cannot be realized through the defect distribution main direction calculation, the defects cannot be accurately mapped to a cargo hold three-dimensional model to form a digital twin characteristic layer, multiple batches of defect data cannot be subjected to associated regression analysis and quality risk prediction with welding process parameters, hot crack and inter-crystal corrosion defect misjudgment is easily caused, and batch welding quality hidden dangers cannot be identified in advance. Disclosure of Invention The invention provides a method and a system for processing welding quality data of a ship liquid cargo stainless steel structure, which ensure the service safety of corrosion resistance and crack resistance of a liquid cargo tank welding structure. In order to solve the technical problems, the technical scheme of the invention is as follows: In a first aspect, a method for processing welding quality data of a ship liquid cargo stainless steel structure, the method comprising: Acquiring welding process parameters, nondestructive testing data and stainless steel material characteristic data, preprocessing the acquired data to obtain a multi-source data set, and carrying out matching association on the multi-source data set and the three-dimensional space coordinates of the welding seam to construct a welding quality database; Based on a preset stainless steel welding defect feature library, carrying out intelligent recognition on nondestructive testing data in a welding quality database to obtain a defect feature data set; According to the space included angle between the main direction and the normal vector of the initial virtual reference plane, rotating the initial virtual reference plane through rotation transformation around any space axis to obtain a self-adaptive reference plane with the normal vector perpendicular to the main direction; The method comprises the steps of calculating orthogonal projection points from each defect point in defect characteristic data set to a self-adaptive reference plane, taking a point set formed by all projection points as a calibrated defect space position, mapping the calibrated defect space position and corresponding defect attribute to a cargo tank three-dimensional model, and obtaining a digital twin characteristic layer with defect space distribution; and carrying out statistical analysis on the digital twin characteristic layers of the defect space distribution of the batches to obtain an analysis result, and automatically generating a welding quality assessment report according to the analysis result. Further, collecting welding process parameters, nondestructive testing data and stainless steel material characteristic data, preprocessing the collected data to obtain a multi-source data set, and matching and correlating the multi-source data set with three-dimensional space coordinates of a welding