A spot welding quality monitoring device and method based on dynamic machine vision

Abstract

The present invention relates to the field of resistance spot welding application technology, and in particular to a spot welding quality monitoring device and method based on dynamic machine vision. The device comprises a workpiece to be welded, a spot welding electrode, an electrode rod, an industrial square power supply and a dynamic machine vision monitoring system, wherein the spot welding electrode is formed at the end of the electrode rod, and the workpiece to be welded is arranged between two spot welding electrodes; the industrial square power supply is arranged outside one end of the workpiece to be welded, and the dynamic machine vision monitoring system is arranged outside the other end of the workpiece to be welded; the method comprises acquiring data associated with a resistance spot welding process, detecting parameters associated with spatter from the acquired data, analyzing the detected parameters, photographing spatter during spot welding with a high-speed industrial camera, determining the amount of spatter generated, spatter morphology and spatter direction by image processing, and establishing a database to monitor spatter generation events during the resistance spot welding process.

Description

Technical Field

[0001] The invention relates to the technical field of resistance spot welding applications, and in particular to a spot welding quality monitoring device and method based on dynamic machine vision. Background Art

[0002] Resistance spot welding is a method of pressing the workpiece to be welded between two electrodes and passing an electric current through it, using the resistance heat generated by the current flowing through the contact surface of the workpiece and the adjacent area to heat it to a molten state, so as to form a metal bond. Resistance spot welding has been widely used in the fields of automobiles, rail transportation, aerospace, low-voltage electrical appliances, home appliances, batteries, etc. due to its low cost, high production efficiency and easy automation. Especially in the automotive industry, more than 90% of the body assembly work is completed by resistance spot welding. However, spatter is inevitable in the process of resistance spot welding. The root cause of spatter events is that the plastic ring is not fully formed or broken, so that it loses its containment effect on the molten pool metal. In various industries, workpieces welded with spatter events are considered to be poor quality characteristics. In the large-scale, fast-beat body assembly production process, it is usually accompanied by interference conditions such as electrode wear, assembly gap, and workpiece surface contamination. These uncertain factors increase the instability of the welding process, resulting in an increased probability of incomplete or broken plastic ring formation, which in turn causes welding quality problems such as spatter and reduces the reliability of the weld. In order to ensure the quality of welds, domestic and foreign automobile companies generally adopt the strategy of "manual sampling after welding + offline industrial adjustment". The quality of welds is evaluated by manual quality inspection methods such as ultrasonic flaw detection or dissection experiments, and then offline feedback adjustment of welding process parameters is performed based on the statistical pass rate. This type of quality inspection method is inefficient and has poor timeliness of quality feedback. It cannot meet the needs of fast-paced production and cannot guarantee the inspection of 100% welds. Enterprises can only ensure quality by increasing the number of welds, which is bound to increase costs and reduce efficiency. Therefore, the present invention proposes a spot welding quality monitoring method and device based on dynamic machine vision. By monitoring the spatter events during spot welding and analyzing and comparing them with the data in the database, the cause of the spatter event is determined, thereby determining the quality of the weld and adjusting the process parameters.

[0003] Based on the above reasons, the present invention provides a spot welding quality monitoring device and method based on dynamic machine vision. Summary of the invention

[0004] The purpose of the present invention is to improve the defects of the prior art and provide a spot welding quality monitoring device and method based on dynamic machine vision, which can judge the quality of resistance spot welding welds by monitoring welding spatter.

[0005] In order to achieve the purpose of the present invention, the technical solution adopted by the present invention is:

[0006] The invention discloses a spot welding quality monitoring device based on dynamic machine vision, comprising a workpiece to be welded, a spot welding electrode, an electrode rod, an industrial square power supply and a dynamic machine vision monitoring system, wherein the spot welding electrode is formed at the end of the electrode rod, the workpiece to be welded is arranged between two spot welding electrodes, and the angle between the electrode rod and the outer wall of the workpiece to be welded is 90°; the industrial square power supply is arranged on the outside of one end of the workpiece to be welded, and the dynamic machine vision monitoring system is arranged on the outside of the other end of the workpiece to be welded.

[0007] The industrial square power supply is connected to the light source trigger simulator through a light source fixing frame, the light source fixing frame is an inverted "L" shaped structure, the light source trigger simulator is electrically connected to the industrial square power supply, and a controller switch and a brightness adjustment button are provided thereon.

[0008] A protective glass is provided between the dynamic machine vision monitoring system and the end surface of the workpiece to be welded.

[0009] The present invention discloses a spot welding quality monitoring method based on dynamic machine vision, comprising the following steps:

[0010] Step 910, obtaining data associated with the resistance spot welding process;

[0011] Step 920, analyzing and detecting parameters associated with splashing from the acquired data;

[0012] Step 930, analyzing the detected parameters;

[0013] Step 940, using a high-speed industrial camera to photograph the spatter during spot welding;

[0014] Step 950, using image processing to determine the amount of splashes generated, the splash morphology and the splash direction;

[0015] Step 957, establishing a database to monitor spatter generation events during the resistance spot welding process.

[0016] In step 920, detecting parameters associated with spatter includes detecting welding current, detecting electrode pressure, detecting welding time, detecting welding plate state, detecting electrode wear degree, detecting electrode rod verticality, and detecting welding spot position.

[0017] In step 930, the parameters analyzed and detected include determining a welding current greater than or equal to a value that will produce spatter, determining an electrode pressure less than or equal to a value that will produce spatter, determining a welding time period that will produce spatter, determining a relationship between the degree of wear of the spot welding electrode and spatter, determining a relationship between the angle between the electrode rod and the workpiece being welded and welding spatter, and determining a relationship between the degree to which the weld point is at the edge of the workpiece being welded and welding spatter.

[0018] In step 940, the spatter during spot welding is photographed using a high-speed industrial camera, including setting the parameters associated with spatter described in step 920 to respectively form pictures and videos of spatter generated during the spot welding process.

[0019] The method of photographing the spatter during spot welding includes setting a welding current greater than a threshold, an electrode pressure less than a threshold, a welding time greater than a threshold, a surface condition and a matching condition of a plate, an electrode wear degree, a verticality between an electrode rod and a workpiece, and a weld spot position as single variables in the resistance spot welding process, and photographing and saving pictures and videos of the spot welding spatter process caused by each variable.

[0020] In step 950, using image processing to determine the amount of splashes, splash morphology and splash direction generated includes using a differential evolution algorithm to calculate the amount of splashes generated by the splash event, and using image processing to analyze the splash morphology and splash direction generated by the splash event.

[0021] In step 957, establishing a database includes matching the splash pictures and videos processed by the image in step 930 with the variables of the splash events, saving them uniformly, and establishing a callable database.

[0022] The beneficial effects of the present invention are:

[0023] The spot welding quality monitoring device based on dynamic machine vision proposed by the present invention can judge the quality of resistance spot welding spots by monitoring welding spatter conditions.

[0024] The spot welding quality monitoring device based on dynamic machine vision proposed in the present invention can monitor the resistance spot welding process in real time. The cause of the spatter event can be known by image processing and comparison of data in the database, and the welding process can be improved by adjusting the welding parameters, which is highly efficient.

[0025] The spot welding quality monitoring device based on dynamic machine vision proposed by the present invention is very simple to build and requires low hardware cost.

[0026] The database in the spot welding quality monitoring device based on dynamic machine vision proposed in the present invention is composed of 8 sub-databases, and each sub-database corresponds to a cause of a resistance spot welding spatter event. The 8 sub-databases contain all possible causes of resistance spot welding spatter events. Therefore, the established database is rich in content, wide in coverage, and highly persuasive. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a structural schematic diagram of a spot welding quality monitoring device based on dynamic machine vision of the present invention.

[0028] Figure 2is a flow chart illustrating a method for photographing spatter during spot welding using a high-speed industrial camera according to aspects of the present invention.

[0029] Figure 3 The present invention is a flowchart showing a method for determining the distance between the generated spatter and the welding point, the spatter morphology and the spatter direction by using image processing and establishing a database.

[0030] Figure 4 This is a block diagram of a sub-database established with "welding current" as the variable.

[0031] Figure 5 FIG. 1 is a simplified diagram of an embodiment for applying the sheet matching state as a variable in a spot welding process according to an aspect of the present invention.

[0032] Figure 6 FIG. 1 is a simplified diagram of an embodiment for applying the degree of electrode cap wear as a variable in a spot welding process according to an aspect of the present invention.

[0033] Figure 7 FIG. 1 is a simplified diagram of an embodiment for using electrode rod verticality as a variable in a spot welding process according to aspects of the present invention.

[0034] Figure 8 FIG. 1 is a simplified diagram of an embodiment for applying the weld spot position as a variable in a spot welding process according to an aspect of the present invention.

[0035] Fig. 9 A flow chart of an embodiment of a method for spot welding quality monitoring based on dynamic machine vision according to an aspect of the present invention.

[0036] Fig.10 The invention is used for spot welding quality based on dynamic machine vision in actual industrial production. DETAILED DESCRIPTION

[0037] The present invention is further described below in conjunction with the accompanying drawings and embodiments:

[0038] See also Figure 1-10 .

[0039] The present invention discloses a spot welding quality monitoring device based on dynamic machine vision, such as Figure 1As shown, the workpiece 100 to be welded is placed between two identical spot welding electrodes 200, and the verticality between the electrode rod 300 and the workpiece 100 to be welded is 90°, which ensures the best spot welding angle and the weld core position 400 is in the center of the workpiece, so as to ensure that the pressure applied to the workpiece 100 to be welded during the spot welding process is uniform. The industrial square light source 500 is parallel to the right side of the workpiece 100 to be welded. The industrial square light source 500 can provide high brightness and high uniformity irradiation light and is fixed by the light source fixing frame 600 to maintain stability. The light source trigger simulator 700 is connected to the industrial square light source 500, and the light source trigger simulator 700 can control the opening and closing of the industrial square light source 500 through the controller switch 701. The brightness adjustment knob 702 can be rotated to adjust the brightness of the industrial square light source 500 according to actual needs. Clockwise rotation increases the brightness, and counterclockwise rotation reduces the brightness. The protective glass 800 and the dynamic machine vision monitoring system 900 are located opposite to the industrial square light source 500. The protective glass 800 is placed between the dynamic machine vision monitoring system 900 and the welded workpiece 100. It can protect the dynamic machine vision monitoring system 900 from being damaged by spatter during spot welding.

[0040] Generally speaking, factors that can cause spatter during spot welding include excessive welding current, low electrode pressure, long welding time, electrode wear, surface cleanliness of the workpiece, i.e., the plate, the matching degree of the plate, the location of the weld, and the verticality between the electrode rod and the workpiece. Figure 2The above factors that produce spatter events in the spot welding process are respectively applied as single variables to the process of shooting spatter events, that is, three groups of currents greater than the threshold current are set, and the remaining variables are selected to select the best values ​​and applied to the spot welding process, and the spatter pictures and videos 941 of the spot welding process under the three groups of currents greater than the threshold current are collected; three groups of electrode pressures less than the threshold voltage are set, and the remaining variables are selected to select the best values ​​and applied to the spot welding process, and the spatter pictures and videos 942 of the spot welding process under the three groups of electrode pressures less than the threshold pressure are collected; three groups of welding times greater than the threshold time are set, and the remaining variables are selected to select the best values ​​and applied to the spot welding process, and the spatter pictures and videos 943 of the spot welding process under the three groups of welding times greater than the threshold are collected; the spot welding electrode 200, the slightly worn electrode 201, and the severely worn electrode 202 are set to select the best values ​​of the remaining variables and applied to the spot welding process, and the remaining variables are selected to select the best values, and the spatter pictures and videos 944 of the spot welding process under the three electrodes are collected; the welded workpiece 100 is set to be completely matched The three plate matching modes of matching, i.e., the gap between the workpiece 100 and the welded workpiece is small 101, and the gap between the workpiece 100 and the welded workpiece is large 102, are respectively applied to the spot welding process, and the other variables select the optimal values, and the spatter pictures and videos 945 of the spot welding process under the three plate matching modes are collected; the molten core position 400, the welding point position 401 where about two-thirds of the molten core is formed, the welding point position 402 where about half of the molten core is formed, and the welding point position 403 where about one-third of the molten core is formed are set to be applied to the spot welding process respectively, and the other variables select the optimal values ​​to collect the spatter pictures and videos 946 of the spot welding process under different welding point positions; the verticality between the electrode rod 300 and the workpiece is set to 90°, the verticality between the electrode rod 300 and the workpiece is large but less than 90° 301, and the verticality between the electrode rod 300 and the workpiece is small 302, which are respectively applied to the spot welding process, and the other variables select the optimal values ​​to collect the spatter pictures and videos 947 of the spot welding process under different verticalities between the electrode rod and the workpiece;Set the plate surface state to be good (0 oil stains and oxides), plate surface state to be general (less oil stains and oxides), and plate surface state to be poor (more oil stains and oxides) respectively for the spot welding process, select the best values ​​for the remaining variables, collect the splash pictures and videos of the spot welding process under different plate surface states 948, screen all the above pictures and videos, that is, by comparing the pictures and videos of the splash events under the same variable, determine the pictures with the most serious splash under the variable to save, and archive them together with the corresponding variables 949. For example, assuming that the threshold current is 7KA, 8KA, 9KA, and 10KA splash currents are set in the spot welding process and the splash pictures and videos are collected. The splash video under the 8KA current is slowed down to intercept the two pictures with the most serious splash, and the pictures intercepted from the video are compared with the pictures directly collected by the high-speed industrial camera, accurately find the picture with the most serious splash under the 8KA current, and archive the picture together with 8KA, named "Event A1", and the splash event picture generated by 9KA and it can be named "Event A2", and so on. ;

[0041] Process each image obtained in 949, and calculate the distance between the spatter and the solder joint in the image. Blur the background color of the image (here we assume that white is the foreground color and black is the background color), and convert the value of the pixel in the foreground into the distance from the point to the nearest background point 952.

[0042] (1)

[0043] |X| is the Euclidean distance from the point (x2+y2) to the origin. That is, a two-dimensional rectangular coordinate system is established with the welding point as the origin, and the position of each spatter point (x i +y i)(i=1.2.3......), respectively calculate the distances between the farthest spatter point and the nearest spatter point and the weld, and organize and archive the distances between the spatter and the weld of spatter events caused by the same variable 953. For example, when the welding current is 8KA, the distance between the spatter and the weld is "1", when the welding current is 9KA, the distance between the spatter and the weld is "2", and when the welding current is 10KA, the distance between the spatter and the weld is "3", and the distance between the farthest spatter point and the weld when the welding current is 8KA is smaller than that when the welding current is 9KA, and the distance between the farthest spatter point and the weld when the welding current is 10KA is smaller than that when the welding current is 10KA. The distance between the spatter point and the weld point, the distance between the nearest spatter point and the weld point when the welding current is 8KA is smaller than the distance between the nearest spatter point and the weld point when the welding current is 9KA, which is smaller than the distance between the nearest spatter point and the weld point when the welding current is 10KA. Then the distance between the spatter and the weld point caused by the welding current can be sorted as "8KA<9KA<10KA", and the distance "1" and the welding current 8KA are archived together, named "Event A1.1", the distance "2" and the welding current 9KA are archived together, named "Event A2.1", the distance "3" and the welding current 10KA are archived together, named "Event A3.1", and so on. Each image obtained in 949 is processed, and the image processing software is used to perform digital image processing on the image to obtain the morphology and contour of the spatter, and the direction of the spatter can also be obtained more accurately after image processing. The data and images obtained after processing and their corresponding parameters are archived together 954. For example, the spatter event image when the welding current is 8KA is processed by the image processing software, and the spatter morphology is named "T1", and the spatter direction is "D1". "T1" and "D1" are archived together and named "Event A1.2". The processed data is sorted, and the spatter events caused by the same variable are used as a sub-database 956, and all sub-databases are merged to obtain a total database 957. For example, the spatter event caused by the welding current is a sub-database and is named "I", and "I" includes "Event A1" generated when the welding current is 8KA, and "Event A1" includes "Event A1.1" and "Event A1.2"; "I" also includes "Event A2" generated when the welding current is 9KA, and "Event A2" includes "Event A2.1" and "Event A2.2".

[0044] After the overall database is completed, practical production can be carried out. Fig.10This is a flow chart of an implementation example of a spot welding quality monitoring method based on dynamic machine vision in actual industrial production. As the resistance spot welding process begins 1, the spot welding process is photographed 2 to determine whether there is spatter production 3. If there is no spatter event in the process, the weld quality can be considered good 3.1. If a spatter event occurs, the collected pictures and videos need to be processed to obtain the farthest spatter point of the spatter event, the distance between the nearest spatter point and the weld, the spatter morphology and the spatter direction 3.2. The sub-data in the database are retrieved, and the data of this spot welding process is compared with the sub-data for reference. If the comparison result has a high degree of consistency, that is, the spatter contour is the same, the spatter direction is the same, and the distance between the spatter and the molten core is the same, then it can be Get the sub-database 4 to which this spatter event belongs, find the variable 5 corresponding to this sub-database, and then improve the spot welding process 6 by adjusting this variable. For example, if the spatter event data obtained in actual production is consistent with the data in the sub-database "I", that is, the distances between the farthest spatter point, the nearest spatter point and the weld point are between "Event A1.1" and "Event A2.1", the spatter morphology is the same as "Event A" and the spatter direction is the same as "Event A", then it can be confirmed that this spatter event is caused by excessive welding current, and a weld with good performance can be obtained by adjusting the process parameters, that is, reducing the welding current.

[0045] The described embodiments are only a part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

Claims

1. A spot welding quality monitoring method based on dynamic machine vision, characterized in that: The spot welding quality monitoring device comprises a workpiece to be welded, a spot welding electrode, an electrode rod, an industrial square power supply and a dynamic machine vision monitoring system, wherein the spot welding electrode is formed at the end of the electrode rod, the workpiece to be welded is arranged between the two spot welding electrodes, and the angle between the electrode rod and the outer wall of the workpiece to be welded is 90°; the industrial square power supply is arranged outside one end of the workpiece to be welded, and the dynamic machine vision monitoring system is arranged outside the other end of the workpiece to be welded; the industrial square power supply is connected to a light source trigger simulator through a light source fixing frame, and the light source fixing frame is an inverted "L" shaped structure, and the light source trigger simulator is electrically connected to the industrial square power supply, and a controller switch and a brightness adjustment button are arranged on the industrial square power supply; a protective glass is arranged between the dynamic machine vision monitoring system and the end face of the workpiece to be welded; The spot welding quality monitoring method of the spot welding quality monitoring device comprises the following steps: Step 910, obtaining data associated with the resistance spot welding process; Step 920, analyzing and detecting parameters associated with splashing from the acquired data; Step 930, analyzing the detected parameters; Step 940, using a high-speed industrial camera to photograph the spatter during spot welding; Step 950, using image processing to determine the amount of splashes generated, the splash morphology and the splash direction; Step 957, establishing a database to monitor spatter generation events during the resistance spot welding process; In step 920, detecting parameters associated with spatter includes detecting welding current, detecting electrode pressure, detecting welding time, detecting welding plate state, detecting electrode wear degree, detecting electrode rod verticality, and detecting welding spot position.

2. A spot welding quality monitoring method based on dynamic machine vision according to claim 1, characterized in that: In step 930, the parameters analyzed and detected include determining a welding current greater than or equal to a value that will produce spatter, determining an electrode pressure less than or equal to a value that will produce spatter, determining a welding time period that will produce spatter, determining a relationship between the degree of wear of the spot welding electrode and spatter, determining a relationship between the angle between the electrode rod and the workpiece being welded and welding spatter, and determining a relationship between the degree to which the weld point is at the edge of the workpiece being welded and welding spatter.

3. The spot welding quality monitoring method based on dynamic machine vision according to claim 1, characterized in that: In step 940, the spatter during spot welding is photographed using a high-speed industrial camera, including setting the parameters associated with spatter described in step 920 to respectively form pictures and videos of spatter generated during the spot welding process.

4. The spot welding quality monitoring method based on dynamic machine vision according to claim 1 is characterized in that: The method of photographing the spatter during spot welding includes setting a welding current greater than a threshold, an electrode pressure less than a threshold, a welding time greater than a threshold, a surface condition and a matching condition of a plate, an electrode wear degree, a verticality between an electrode rod and a workpiece, and a weld spot position as single variables in the resistance spot welding process, and photographing and saving pictures and videos of the spot welding spatter process caused by each variable.

5. The spot welding quality monitoring method based on dynamic machine vision according to claim 1 is characterized in that: In step 950, using image processing to determine the amount of splashes, splash morphology and splash direction generated includes using a differential evolution algorithm to calculate the amount of splashes generated by the splash event, and using image processing to analyze the splash morphology and splash direction generated by the splash event.

6. The spot welding quality monitoring method based on dynamic machine vision according to claim 1, characterized in that: In step 957, establishing a database includes matching the splash pictures and videos processed by the image in step 930 with the variables of the splash events, saving them uniformly, and establishing a callable database.

Images