Welding support device, welding support method, and program

Abstract

To provide a welding support device which can suppress the occurrence of a welding defect in the welding work executed by a welder.SOLUTION: A welding support device comprises: an acquisition unit which acquires welding data consisting of a plurality of parameters expressing a welding state; an estimation unit which estimates a type of a welding defect generated in a welding spot on the basis of a defect estimation model that uses the welding data as an explanatory variable and the type of the welding defect as a target variable; and an analysis unit which specifies a factor parameter that contributes to the occurrence of the welding defect on the basis of the welding data and the estimated type of the welding defect, and generates analysis data expressing the correlation between the magnitude of a value of the specified factor parameter and the magnitude of contribution to the welding defect.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present disclosure relates to a welding support device, a welding support method, and a program.

Background Art

[0002] As a technique for improving welding quality, for example, Patent Document 1 describes a technique of machine learning physical quantities related to arc welding such as the appearance of a weld bead and arc welding conditions such as welding speed and protrusion length, and adjusting arc welding conditions based on the physical quantities obtained from imaging data to perform automatic welding.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, since the technique described in Patent Document 1 is targeted at an automatic welding robot, it is difficult to apply it to welding work performed by a welder. For this reason, a method for suppressing the occurrence of welding defects by supporting the welding work performed by a welder is required.

[0005] The present disclosure has been made in view of such problems, and provides a welding support device, a welding support method, and a program capable of suppressing the occurrence of welding defects in welding work performed by a welder.

Means for Solving the Problems

[0006] According to one aspect of the present disclosure, a welding support device includes: an acquisition unit that acquires welding data consisting of a plurality of parameters representing the welding state; an estimation unit that estimates the type of welding defect occurring at a welding location based on a defect estimation model in which the welding data is an explanatory variable and the type of welding defect is an objective variable; and an analysis unit that identifies factor parameters that contribute to the occurrence of the welding defect based on the welding data and the estimated type of welding defect, and generates analysis data representing the correlation between the magnitude of the value of the identified factor parameters and the magnitude of their contribution to the welding defect.

[0007] According to one aspect of the present disclosure, a welding support method includes the steps of: acquiring welding data consisting of a plurality of parameters representing the welding state; estimating the type of welding defect occurring at a welding location based on a defect estimation model in which the welding data is an explanatory variable and the type of welding defect is an objective variable; and identifying factor parameters that contribute to the occurrence of the welding defect based on the welding data and the estimated type of welding defect, and generating analysis data that represents the correlation between the magnitude of the value of the identified factor parameters and the magnitude of their contribution to the welding defect.

[0008] According to one aspect of the present disclosure, the program causes a welding support device to perform the following steps: acquire welding data consisting of a plurality of parameters representing the welding state; estimate the type of welding defect occurring at a welding location based on a defect estimation model in which the welding data is an explanatory variable and the type of welding defect is an objective variable; and identify factor parameters that contribute to the occurrence of the welding defect based on the welding data and the estimated type of welding defect, and generate analysis data representing the correlation between the magnitude of the value of the identified factor parameters and the magnitude of their contribution to the welding defect. [Effects of the Invention]

[0009] According to the welding support device, welding support method, and program described herein, the occurrence of welding defects can be suppressed in welding work performed by a welder. [Brief explanation of the drawing]

[0010] [Figure 1] This block diagram shows the functional configuration of a welding support device according to one embodiment of the present disclosure. [Figure 2] This flowchart shows an example of a learning process in a welding support device according to one embodiment of the present disclosure. [Figure 3] This flowchart shows an example of welding defect analysis processing in a welding support device according to one embodiment of the present disclosure. [Figure 4] This figure shows an example of the estimated results of welding defects according to one embodiment of the present disclosure. [Figure 5] Figure 1 shows an example of analysis data according to one embodiment of this disclosure. [Figure 6] This is a second figure showing an example of analysis data according to one embodiment of the present disclosure. [Figure 7] This is a third figure showing an example of analysis data according to one embodiment of this disclosure. [Figure 8] This flowchart shows an example of the process for presenting support information in a welding support device according to one embodiment of the present disclosure. [Figure 9] This figure shows an example of support information according to one embodiment of this disclosure. [Modes for carrying out the invention]

[0011] Hereinafter, a welding support system 1 and a welding support device 10 according to one embodiment of this disclosure will be described with reference to Figures 1 to 9.

[0012] (Overall configuration of the welding support system) Figure 1 is a block diagram showing the functional configuration of a welding support device according to one embodiment of the present disclosure. As shown in Figure 1, the welding support system 1 according to this embodiment comprises a welding support device 10, a welding device 20, a data logger 30, and a display device 40.

[0013] The welding apparatus 20 includes devices such as switches and levers for adjusting various parameters that affect the welding state, and an operation panel provided with instruments. The welder performs the welding operation while adjusting each parameter on the operation panel of the welding apparatus. These parameters are, for example, welding current, welding voltage, electrode protrusion length, and the like. The range of values for each parameter is defined in advance as welding conditions, and the welder adjusts each parameter to satisfy these welding conditions and performs the welding operation. Further, the welding state is, for example, the degree of abnormality estimated based on welding data composed of a plurality of parameters measured by sensors (not shown) provided in each part such as the welding apparatus 20 and the vicinity of the welding part, and the type of welding defect.

[0014] The data logger 30 acquires the above-described welding data.

[0015] Based on the welding data acquired by the data logger 30, the welding support apparatus 10 estimates the type of welding defect and the cause of the welding defect that occur at the welding part. Further, based on the estimation result, the welding support apparatus 10 sets (changes) the recommended range of values of the parameter (factor parameter) that is the cause of the welding defect among the welding conditions defined in advance. Furthermore, during the welding operation, the welding support apparatus 10 displays the recommended range of values of each parameter defined by the welding conditions on the display device 40 to support the welder's welding operation.

[0016] (Functional Configuration of Welding Support Apparatus) The welding support apparatus 10 includes a processor 11, a memory 12, a storage 13, and a communication interface 14.

[0017] By operating according to a predetermined program, the processor 11 exhibits functions as an acquisition unit 110, an estimation unit 111, an analysis unit 112, a setting unit 113, an output unit 114, and a learning unit 115.

[0018] The acquisition unit 110 acquires welding data from the data logger 30.

[0019] The estimation unit 111 estimates the type of welding defect occurring at a weld based on a defect estimation model M, which uses welding data as an explanatory variable and the type of welding defect as the dependent variable. Examples of welding defect types include planar defects, volumetric defects, porosity, and slag inclusions.

[0020] The analysis unit 112 identifies factor parameters that contribute to the occurrence of welding defects based on the welding data and the type of welding defect estimated by the estimation unit 111, and generates analysis data that shows the correlation between the magnitude of the value of the identified factor parameters and the magnitude of their contribution to the welding defect.

[0021] The setting unit 113 sets recommended ranges for the values ​​of factor parameters in welding operations based on the analysis data, so that the contribution does not exceed a predetermined tolerance range. Alternatively, the user (such as a welder or a technician who determines welding conditions) may identify the factor parameters and specify recommended ranges based on the analysis data. In this case, the setting unit 113 sets recommended ranges for the factor parameter values ​​based on the information provided by the user. The storage unit 13 stores welding condition data T in which the recommended ranges for the values ​​of each parameter are predetermined. The setting unit 113 overwrites the welding condition data T with the set recommended ranges for the factor parameter values ​​and records them.

[0022] The output unit 114 outputs and displays the analysis data generated by the analysis unit 112 on the display device 40. In addition, during welding operations, the output unit 114 outputs and displays support information, including recommended ranges for the values ​​of each parameter, on the display device 40 based on the welding condition data T recorded in the storage 13.

[0023] The learning unit 115 performs supervised learning based on welding data collected from the start to the end of welding at a welding site, and the location and type of welding defects that occurred at the welding site, to construct a defect estimation model M.

[0024] Memory 12 has a memory area necessary for the operation of the processor 11.

[0025] Storage 13 is a so-called auxiliary storage device, such as an HDD (Hard Disk Drive) or SSD (Solid State Drive). Storage 13 stores welding data collected from the data logger 30 by the acquisition unit 110, a defect estimation model M constructed by the learning unit 115, and welding condition data T recorded by the setting unit 113.

[0026] The communication interface 14 is an interface for sending and receiving various types of information (signals) with external devices (such as the data logger 30 and the display device 40).

[0027] (Regarding the training process of the defect estimation model) Figure 2 is a flowchart showing an example of a learning process in a welding support device according to one embodiment of the present disclosure. Here, referring to Figure 2, we will explain the process flow by which the welding support device 10 learns the defect estimation model M.

[0028] First, the acquisition unit 110 acquires the welding data collected by the data logger 30 during the period from the start to the end of welding at the welding site (step S10).

[0029] Furthermore, once the welding work is complete, the user performs non-destructive testing (radioactive testing, ultrasonic testing, etc.) on the welded area to identify the location and type of welding defects. The acquisition unit 110 acquires the location and type of welding defects detected by the user from the welded area (step S11).

[0030] Next, the learning unit 115 performs supervised learning based on the welding data acquired by the acquisition unit 110 and the location and type of welding defects to construct a defect estimation model M (step S12). The defect estimation model M is a model that takes welding data as explanatory variables and outputs the location and type of welding defects as dependent variables. For supervised learning, for example, a random forest is used. The defect estimation model M constructed by the learning unit 115 is recorded in the storage 13.

[0031] The welding support device 10 may collect this learning data each time it performs a welding operation and retrain (update) the defect estimation model M.

[0032] (Regarding the analysis process for welding defects) Figure 3 is a flowchart showing an example of welding defect analysis processing in a welding support device according to one embodiment of the present disclosure. Here, with reference to Figure 3, we will explain the flow of the welding defect analysis process of the welding support device 10.

[0033] First, the acquisition unit 110 acquires the welding data collected by the data logger 30 during the period from the start to the end of welding at the welding site (step S20).

[0034] Next, the estimation unit 111 estimates the location and type of welding defects occurring at the weld site based on the defect estimation model M recorded in the storage 13 and the acquired welding data (step S21).

[0035] Figure 4 shows an example of the estimated results of welding defects according to one embodiment of the present disclosure. Figure 4 shows an example of the estimation results for the location and type of welding defects by the estimation unit 111. The estimation unit 111 may output the estimation results for the location and type of welding defects as a heat map, as shown in Figure 4. Welding defects 1, 2, 3, 4, ... each represent a different type of welding defect (e.g., planar defects, volumetric defects, porosity, slag inclusions, etc.). The vertical axis of the heat map represents the welding layers and passes, and the horizontal axis represents the position in the welding direction. For example, when welding an annular structure such as a pipe while rotating it, the position in the welding direction may be represented by the angle of the structure.

[0036] Next, the analysis unit 112 analyzes the parameters that influenced the estimation results, i.e., the parameters that cause welding defects, based on the welding data and the estimation results of the estimation unit 111, for each type of welding defect.

[0037] In a model trained using supervised learning, it is not possible to know which explanatory variables (parameters) and what values ​​are estimated to be the factors causing welding defects. As a result, while the user can recognize the location and type of welding defects occurring using the defect estimation model M, it becomes difficult to determine how to change the welding conditions (values ​​of each parameter) to suppress the occurrence of welding defects.

[0038] Therefore, the analysis unit 112 in this embodiment uses SHAP (Shapley additive explanations) technology to generate analysis data that visualizes the explanatory variables and their physical quantities that are the causes of welding defects.

[0039] Specifically, first, the analysis unit 112 identifies the factor parameters that contributed to the occurrence of the welding defects estimated by the estimation unit 111 (step S22). For example, suppose that the occurrence of four types of welding defects 1 to 4 is estimated, as shown in Figure 4. In this case, the analysis unit 112 identifies the factor parameters for each of these four types of welding defects 1 to 4.

[0040] Figure 5 is a first figure showing an example of analysis data according to one embodiment of the present disclosure. Here, with reference to Figure 5, an example of how the analysis unit 112 identifies the factor parameters for welding defect 1 will be described. As shown in Figure 5, the analysis unit 112 uses SHAP to calculate SHAP values ​​(contributions) that represent the degree to which each parameter contributes to the occurrence of welding defect 1, and generates analysis data D1 sorted in descending order of contribution. The analysis unit 112 also identifies a predetermined number of parameters (for example, 3) with large SHAP values ​​as factor parameters that cause welding defect 1. In the example in Figure 5, the analysis unit 112 identifies three parameters 1, 5, and 2 with large SHAP values ​​as factor parameters for welding defect 1. The analysis unit 112 similarly identifies the factor parameters for welding defects 2 to 4.

[0041] Next, the analysis unit 112 generates analysis data for each identified factor parameter that represents the correlation between the magnitude of the factor parameter value and the magnitude of the SHAP value (step S23).

[0042] Figure 6 is a second figure showing an example of analysis data according to one embodiment of the present disclosure. Here, referring to Figure 6, we will explain an example of generating analysis data D2 and parameter 1, which is one of the factor parameters for welding defect 1.

[0043] As shown in Figure 6, the analysis unit 112 generates analysis data D2 by plotting each of the sampled data (measured values ​​at each time point) of parameter 1 measured during welding according to the magnitude of the SHAP value. The horizontal axis represents the magnitude of the SHAP value. In addition, each plot is represented by a different color according to the magnitude of the sampled data value. In the example in Figure 6, it can be seen from the analysis data D2 that there is a tendency for the SHAP value to increase as the value of parameter 1 increases.

[0044] Figure 7 is a third figure showing an example of analysis data according to one embodiment of this disclosure. Furthermore, the analysis unit 112 generates analysis data D3, which is plotted according to the magnitude of each sampled data value and the magnitude of the SHAP value, as shown in Figure 7. The vertical axis represents the magnitude of the parameter 1 value, and the horizontal axis represents the magnitude of the SHAP value. Analysis data D3 is a scatter plot that shows the correlation between the magnitude of the parameter 1 value and the magnitude of the SHAP value. By referring to analysis data D3, it is easy to understand how large the value of parameter 1 needs to be for the SHAP value to increase. In the example in Figure 7, it can be seen from analysis data D3 that the SHAP value tends to increase when the value of parameter 1 is between 33 and 35.

[0045] Furthermore, the analysis unit 112 generates analysis data D2 and D3 similarly for other factor parameters of welding defect 1, and for the factor parameters of each of the other welding defects 2 to 4.

[0046] Next, the output unit 114 outputs and displays the analysis data D1 to D3 generated by the analysis unit 112 on the display device 40 (step S24).

[0047] The user refers to the analysis data D1 to D3 for each of the welding defects 1 to 4 and adjusts the recommended range of values ​​for the welding defect factor parameters among the parameters defined in the welding condition data T.

[0048] For example, as shown in Figure 7, the analysis data D3 for welding defect 1 indicates that when the value of parameter 1 is between 33 and 35, the SHAP value increases, making welding defect 1 more likely to occur. Therefore, the user changes the recommended range for the value of parameter 1 to make welding defect 1 less likely to occur. Specifically, the user specifies that the recommended range for the value of parameter 1 is outside the range of 33 to 35. The setting unit 113 sets the recommended range for parameter 1 according to the user's specified operation (step S25) and overwrites and records it in the welding condition data T. As a result, the welding condition data T is updated to make welding defect 1 less likely to occur.

[0049] Furthermore, in step S25, the setting unit 113 may automatically set a recommended range for parameter 1, excluding parameter 1 values ​​where the SHAP value exceeds a predetermined threshold, based on the analysis data D1 to D3 for each welding defect. The threshold is, for example, a value specified by the user.

[0050] For example, suppose the initial recommended range for the value of parameter 1 is 30-35, and the SHAP value exceeds the threshold when the value of parameter 1 is between 33 and 35. In this case, the setting unit 113 excludes the range 33-35 where the SHAP value exceeds the threshold, and sets the range 27-32, which is close to the initial setting, as the recommended range for the value of parameter 1. The setting unit 113 may also overlay the provisionally set recommended range on the analysis data D3, and formally adopt the recommended range if the user approves it. In this case, if the user changes the recommended range, the setting unit 113 will adopt the recommended range changed by the user.

[0051] Furthermore, from the analysis data D1 to D3 for each welding defect, the same parameter may be identified as a contributing parameter for multiple types of welding defects. For example, suppose parameter 1 is one of the contributing parameters for both welding defect 1 and welding defect 2. Also, suppose it can be seen that welding defect 1 is more likely to occur when the value of parameter 1 is between 33 and 35, and welding defect 2 is more likely to occur when the value of parameter 1 is between 35 and 37. In this case, the user (or the setting unit 113) sets a recommended range for the value of parameter 1 so as to avoid both the range in which welding defect 1 is likely to occur (33 to 35) and the range in which welding defect 2 is likely to occur (35 to 37). By referring to the analysis data D1 to D3 for each welding defect in this way, the welding conditions can be appropriately adjusted so that multiple types of welding defects can be suppressed.

[0052] Furthermore, the user may adjust the recommended range of values ​​for each factor parameter, considering the mutual influence of multiple factor parameters of welding defect 1, based on analysis data D1-D3 and metallurgical knowledge. The user may also adjust the recommended range of values ​​for other parameters that affect the factor parameters. In this way, the welding support device 10 provides the user with analysis data D1-D3, which makes it easy to understand the contribution of each parameter to welding defects, thereby enabling the user to delve deeper into the causes of welding defects in conjunction with their own knowledge. This allows for appropriate improvement of welding conditions so that welding defects are less likely to occur.

[0053] (Regarding the processing of providing support information) Figure 8 is a flowchart showing an example of the process for presenting support information in a welding support device according to one embodiment of the present disclosure. Figure 9 shows an example of support information according to one embodiment of the present disclosure. Furthermore, when a welder performs welding work, the welding support device 10 displays the welding conditions set by the setting unit 113 as support information 41 on the display device 40. The welder performs the welding work while checking this support information 41. Here, with reference to Figures 8 to 9, the process flow in which the welding support device 10 presents the support information 41 will be explained.

[0054] First, when the welding operation is started, the output unit 114 reads welding condition data T from the storage unit 13 and outputs support information 41, including the recommended range of values ​​for each parameter, to the display device 40 for display (step S30). The welding condition data T is updated in the above-mentioned analysis process, with the recommended range of values ​​for the parameters that cause welding defects being updated.

[0055] Furthermore, the acquisition unit 110 acquires welding data from the data logger 30 in real time (step S31). The welding data includes the measured values ​​of each parameter.

[0056] As shown in Figure 9, the support information 41 displays the recommended range 412 for the values ​​of each parameter read from the welding condition data T, and the measured values ​​413 for each parameter acquired by the acquisition unit 110.

[0057] Furthermore, the output unit 114 may determine whether the measured values ​​413 of each parameter exceed the recommended range 412 based on the welding condition data T and the welding data acquired by the acquisition unit 110 (step S32).

[0058] The output unit 114 sets the parameter's judgment result 414 to "OK (no problem)" if the measured value 413 of the parameter does not exceed the recommended range 412 (step S32; NO).

[0059] On the other hand, if the measured value 413 of the parameter exceeds the recommended range 412 (step S32; YES), the output unit 114 sets the judgment result 414 of this parameter to "NG (problem)". In this case, the output unit 114 also displays an improvement suggestion display 415 that highlights this parameter with a frame, background color, text color, etc., to encourage improvement (step S33).

[0060] The welding support device 10 repeatedly executes steps S31 to S33 during welding. The welder also refers to the support information 41 displayed on the display device 40 and operates the welding device 20 to change or maintain welding conditions so that the measured values ​​413 of each parameter fall within the recommended range 412. Furthermore, if the measured values ​​413 of each parameter exceed the recommended range 412 (when the improvement suggestion display 415 is displayed), the welder promptly changes the welding conditions to bring the values ​​within the recommended range 412. Since the recommended range 412 is a range of values ​​in which welding defects are less likely to occur, the occurrence of welding defects can be suppressed by the welder performing welding work in accordance with this recommended range 412.

[0061] (Effect, Action) As described above, the welding support device 10 according to this embodiment includes an acquisition unit 110 that acquires welding data, an estimation unit 111 that estimates the type of welding defect occurring at the welding location based on a defect estimation model M, and an analysis unit 112 that identifies the factor parameters of the welding defect based on the welding data and the estimation results of the estimation unit 111, and generates analysis data that represents the correlation between the magnitude of the value of the factor parameter and the magnitude of the contribution to the welding defect (SHAP value).

[0062] In this way, the welding support device 10 can provide the user with analysis data that makes it easy to understand which parameters are likely to cause welding defects and at what values ​​for each welding defect. This allows the user to appropriately set the recommended range of parameter values ​​that can suppress welding defects based on the analysis data.

[0063] Furthermore, the welding support device 10 includes a setting unit 113 that sets a recommended range for the values ​​of the factor parameters so as not to include any factor parameter values ​​whose contribution exceeds a predetermined threshold.

[0064] In this way, the welding support device 10 can automatically set a recommended range of parameter values ​​that can suppress welding defects based on the analysis data.

[0065] Furthermore, the analysis unit 112 generates analysis data for each type of welding defect when it is estimated that multiple different types of welding defects may occur. The setting unit 113 sets a recommended range for the values ​​of the factor parameters so as not to include values ​​of the factor parameters whose contribution to these welding defects exceeds a threshold when the factor parameters are common to multiple types of welding defects.

[0066] In this way, the welding support device 10 can appropriately adjust the welding conditions so that multiple types of welding defects can be suppressed.

[0067] Furthermore, the welding support device 10 is further equipped with an output unit 114 that outputs support information 41 including the recommended range set by the setting unit 113.

[0068] In this way, the welding support device 10 can provide the welder during welding with a range of parameter values ​​that are less likely to cause welding defects. This allows the welder, regardless of their skill level, to easily and appropriately adjust the parameter values, thereby suppressing the occurrence of welding defects.

[0069] In the above-described embodiment, the processes of various operations of the welding support device 10 are stored in program form on a computer-readable recording medium, and the various operations are performed by a computer reading and executing this program. The computer-readable recording medium refers to magnetic disks, magneto-optical disks, CD-ROMs, DVD-ROMs, semiconductor memory, etc. Alternatively, this computer program may be distributed to a computer via a communication line, and the computer that receives the distribution may execute the program.

[0070] The above program may be intended to implement some of the functions described above. Furthermore, it may be a so-called differential file (differential program) that can implement the above functions in combination with a program already recorded in the computer system.

[0071] As described above, several embodiments relating to this disclosure have been explained, but all of these embodiments are presented as examples and are not intended to limit the scope of the invention. These embodiments can be carried out in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims and their equivalents.

[0072] <Note> The welding support apparatus, welding support method, and program described in the above-described embodiment can be understood, for example, as follows.

[0073] (1) According to a first aspect of the present disclosure, the welding support device 10 includes an acquisition unit 110 that acquires welding data consisting of a plurality of parameters representing the welding state, an estimation unit 111 that estimates the type of welding defect occurring at a welding location based on a defect estimation model M that uses the welding data as an explanatory variable and the type of welding defect as an objective variable, and an analysis unit 112 that identifies factor parameters that contribute to the occurrence of welding defects based on the welding data and the estimated type of welding defect, and generates analysis data that represents the correlation between the magnitude of the value of the identified factor parameters and the magnitude of their contribution to the welding defect.

[0074] In this way, the welding support device 10 can provide the user with analysis data that makes it easy to understand which parameters are likely to cause welding defects and at what values ​​for each welding defect. This allows the user to appropriately set the recommended range of parameter values ​​that can suppress welding defects based on the analysis data.

[0075] (2) According to a second aspect of the present disclosure, the welding support device 10 according to the first aspect further includes a setting unit 113 that sets a recommended range of values ​​for factor parameters during welding work, based on analysis data, so as not to include any factor parameter values ​​whose contribution is greater than or equal to a predetermined threshold.

[0076] In this way, the welding support device 10 can automatically set a recommended range of parameter values ​​that can suppress welding defects based on the analysis data.

[0077] (3) According to a third aspect of the present disclosure, in the welding support device 10 according to the second aspect, the analysis unit 112 generates analysis data for the first welding defect and the second welding defect when the occurrence of different types of first and second welding defects is estimated, and the setting unit 113 sets a recommended range of values ​​for the factor parameters so as not to include any factor parameter values ​​whose contribution to both the first and second welding defects exceeds a threshold when the factor parameters for the first and second welding defects are common.

[0078] In this way, the welding support device 10 can appropriately adjust the welding conditions so that multiple types of welding defects can be suppressed.

[0079] (4) According to a fourth aspect of the present disclosure, the welding support device 10 according to the second or third aspect further comprises an output unit 114 that outputs support information 41 including a recommended range set by the setting unit 113.

[0080] In this way, the welding support device 10 can provide the welder during welding with a range of parameter values ​​that are less likely to cause welding defects. This allows the welder, regardless of their skill level, to easily and appropriately adjust the parameter values, thereby suppressing the occurrence of welding defects.

[0081] (5) According to a fifth aspect of the present disclosure, a welding support method includes the steps of: acquiring welding data consisting of a plurality of parameters representing the welding state; estimating the type of welding defect occurring at a welding site based on a defect estimation model M in which the welding data is an explanatory variable and the type of welding defect is an objective variable; and identifying factor parameters that contribute to the occurrence of welding defects based on the welding data and the estimated type of welding defect, and generating analysis data that represents the correlation between the magnitude of the value of the identified factor parameters and the magnitude of their contribution to the welding defect.

[0082] (6) According to a sixth aspect of the present disclosure, the program causes the welding support device 10 to perform the following steps: acquire welding data consisting of a plurality of parameters representing the welding state; estimate the type of welding defect occurring at the welding site based on a defect estimation model M in which the welding data is an explanatory variable and the type of welding defect is an objective variable; and identify factor parameters that contribute to the occurrence of welding defects based on the welding data and the estimated type of welding defect, and generate analysis data that represents the correlation between the magnitude of the value of the identified factor parameters and the magnitude of their contribution to the welding defect. [Explanation of Symbols]

[0083] 1. Welding support system 10 Welding support device 11 processors 110 Acquisition Department 111 Estimation Department 112 Analysis Department 113 Settings Section Output section of 114 115 Learning Department 12 memory 13 Storage 14. Communication Interface 20 Welding equipment 30 Data Loggers 40 Display device

Claims

1. An acquisition unit that acquires welding data consisting of multiple parameters representing the welding state, A learning unit learns a defect estimation model using welding data as explanatory variables and the type of welding defect as the dependent variable, based on training welding data, which is welding data collected from the start to the end of welding at previously welded locations, and the location and type of welding defects identified by non-destructive testing performed on the said welding locations. Based on the defect estimation model learned by the learning unit, an estimation unit estimates the type of welding defect that occurs at the welding site, An analysis unit identifies factor parameters that contribute to the occurrence of the welding defect based on the welding data and the estimated type of welding defect, and generates analysis data that shows the correlation between the magnitude of the value of the identified factor parameters and the magnitude of their contribution to the welding defect. A welding support device equipped with the following features.

2. The analysis unit calculates the contribution value of each parameter included in the welding data using SHAP (Shapley additive explanations), A predetermined number of parameters are identified as factors contributing to the occurrence of welding defects, starting with those with the greatest contribution. The welding support device according to claim 1.

3. The system further includes a setting unit that sets a recommended range for the values ​​of the factor parameters during welding operations, based on the analysis data, so as not to include any values ​​of the factor parameters whose contribution exceeds a predetermined threshold. The welding support device according to claim 1.

4. When the analysis unit estimates the occurrence of different types of first and second welding defects, it generates the analysis data for the first and second welding defects, respectively. The setting unit sets the recommended range of values ​​for the factor parameters such that, when the factor parameters for the first welding defect and the second welding defect are common, the value of the factor parameter does not include the value of the factor parameter whose contribution to both the first welding defect and the second welding defect is equal to or greater than the threshold. The welding support device according to claim 3.

5. The system further includes an output unit that outputs support information including the recommended range set by the setting unit. The welding support device according to claim 3 or 4.

6. A step of acquiring welding data consisting of multiple parameters representing the welding state, The steps include: training a defect estimation model using welding data as explanatory variables and the type of welding defect as the dependent variable, based on training welding data, which is welding data collected from the start to the end of welding at previously welded locations, and the location and type of welding defects identified by non-destructive testing of the said welding locations; The steps include: estimating the type of welding defect that will occur at the welding site based on the previously trained defect estimation model; The steps include: identifying factor parameters that contribute to the occurrence of the welding defect based on the welding data and the estimated type of welding defect, and generating analysis data that shows the correlation between the magnitude of the value of the identified factor parameters and the magnitude of their contribution to the welding defect; A welding support method having

7. A step of acquiring welding data consisting of multiple parameters representing the welding state, The steps include: training a defect estimation model using welding data as explanatory variables and the type of welding defect as the dependent variable, based on training welding data, which is welding data collected from the start to the end of welding at previously welded locations, and the location and type of welding defects identified by non-destructive testing of the said welding locations; The steps include: estimating the type of welding defect that will occur at the welding site based on the previously trained defect estimation model; The steps include: identifying factor parameters that contribute to the occurrence of the welding defect based on the welding data and the estimated type of welding defect, and generating analysis data that shows the correlation between the magnitude of the value of the identified factor parameters and the magnitude of their contribution to the welding defect; A program that causes a welding support device to execute a command.

Images