An on-line real-time monitoring device for the quality of resistance spot welding

Abstract

The present invention relates to the field of online spot welding quality monitoring and discloses a device for online real-time monitoring of resistance spot welding quality. The device comprises a host computer system and a slave computer system. The slave computer system collects welding current, welding voltage, and welding gas pressure signals during the operation of the spot welder, as well as the corresponding welding time, processes the collected data, and transmits it to the host computer system. The host computer system uses an improved BP neural network model in a spot welding quality assessment system, taking as input a feature vector constructed from the data received by the host computer system and outputting the nugget diameter and tensile shear strength. The present invention can achieve high-precision detection and assessment of post-weld quality and is suitable for welding lines with large-scale production characteristics.

Description

Technical Field

[0001] The invention relates to the field of online spot welding quality monitoring, and in particular to an online real-time monitoring device for resistance spot welding quality. Background Art

[0002] Resistance spot welding is a point-connection process widely used in the manufacture of thin-plate structures. Its principle is to apply a certain amount of pressure to press the weldment between upper and lower electrodes, and then use the resistance heat generated by the current passing through the weldment to melt the local metal and form a weld spot, thereby connecting the weldment. Due to its advantages of simple operation, low cost, high production efficiency, material conservation, and ease of automation, resistance spot welding has been widely used in the manufacturing of automobiles, ships, rail transportation, aerospace, precision electronics, household appliances, and other fields. However, due to the influence of various factors, defects such as shrinkage cavities and lack of fusion are prone to occur in the weld nugget area during the resistance spot welding process. The quality of resistance spot welding will directly affect the service life of the weldment. Therefore, it is necessary to monitor the resistance spot welding process to achieve the purpose of stabilizing the welding process parameters and ensuring the welding quality.

[0003] Currently, resistance spot welding quality monitoring primarily relies on post-weld manual spot checks and online non-destructive full inspection. However, manual spot checks only test the performance of a subset of welds, making it difficult to guarantee the quality of all welds. Furthermore, weld quality assessment is often influenced by human factors, resulting in time-consuming and labor-intensive processes, low efficiency, and detection lag. Furthermore, in the strong magnetic field created by high welding currents, the time-varying nature of online non-destructive full inspections generates significant induced noise in the signal, resulting in low weld quality inspection accuracy and an inability to meet the demands of fast-paced production. Therefore, a solution is urgently needed to address these issues. Summary of the Invention

[0004] The present invention provides an online real-time monitoring device for resistance spot welding quality, which solves the problems of low efficiency and detection hysteresis of manual sampling, low accuracy of online non-destructive full inspection, and inability to meet the needs of fast-paced production.

[0005] In order to solve the above technical problems, the present invention provides an online real-time monitoring device for resistance spot welding quality, comprising a host computer system and a slave computer system; wherein,

[0006] The lower computer system includes an adjustable current sampling module, a voltage and pressure sampling module, and a single chip computer system; wherein,

[0007] The adjustable current sampling module is used to detect the welding current signal of the resistance spot welding machine;

[0008] The voltage and air pressure sampling module is used to detect the welding voltage signal and welding air pressure signal of the resistance spot welding machine;

[0009] The single-chip computer system is used to process the welding current signal detected by the adjustable current sampling module, and the welding voltage signal and welding gas pressure signal detected by the voltage and gas pressure sampling module to obtain welding point parameter information; the welding point parameter information includes at least the effective value of current, the effective value of voltage, the effective value of gas pressure, and welding time;

[0010] The host computer system includes a spot welding quality assessment system; wherein, the spot welding quality assessment system is used to optimize the BP neural network model through a genetic algorithm, construct a feature vector of the weld parameter information, input it into the optimized BP neural network model, and output weld quality information; the weld quality information includes the weld core diameter and shear strength.

[0011] Furthermore, the adjustable current sampling module includes an adjustable current sampling circuit and a zero-crossing detection circuit; wherein,

[0012] The adjustable current sampling circuit is used to perform voltage conversion, amplification, integral reduction and regulation on the AC current of the resistance spot welding machine to obtain a DC current;

[0013] The zero-crossing detection circuit is used to perform waveform conversion and zero-crossing detection on the DC current to obtain the welding current signal.

[0014] Furthermore, the adjustable current sampling circuit includes a flexible Rogowski coil, an inverting amplifier circuit, a high-resistance feedback integration circuit, a filter circuit, an adjustable signal circuit, a rectifier circuit and a voltage follower; wherein,

[0015] The flexible Rogowski coil is used to convert the AC current of the resistance spot welding machine into a voltage to obtain an AC voltage signal;

[0016] The inverting amplifier circuit is used to amplify the AC voltage signal;

[0017] The high-resistance feedback integration circuit is used to integrate and restore the amplified AC voltage signal to obtain a restored AC current;

[0018] The filtering circuit is used to filter the restored AC current;

[0019] The adjustable signal circuit is used to adjust the magnitude of the filtered alternating current;

[0020] The rectifier circuit is used to convert the regulated alternating current into a direct current;

[0021] The voltage follower is used to buffer and isolate the direct current.

[0022] Furthermore, the zero-crossing detection circuit includes a comparator and an optocoupler isolation circuit; wherein,

[0023] The comparator is used to convert the DC current into a square wave signal according to a reference level;

[0024] The optical coupler isolation circuit is used to perform zero-crossing detection on the square wave signal.

[0025] Furthermore, the single chip microcomputer system includes an ADC sampling module;

[0026] The ADC sampling module is used to collect the welding current signal and calculate the welding current signal through the point-by-point integration method to obtain the effective value of the current; collect the welding voltage signal and calculate the welding voltage signal through the point-by-point integration method to obtain the effective value of the voltage; collect the welding gas pressure signal and calculate the average value of the welding gas pressure signal to obtain the effective value of the gas pressure.

[0027] Furthermore, the lower computer system further includes a first data display module, an alarm module and a first wireless communication module; wherein,

[0028] The first data display module is used to display the soldering point parameter information when the soldering point parameter information meets the upper and lower error limit requirements;

[0029] The alarm module is used to generate an alarm signal when the soldering point parameter information does not meet the upper and lower error limit requirements, so as to turn on the LED light and sound a buzzer to warn;

[0030] The first wireless communication module is used to send the welding point parameter information to the host computer system.

[0031] Furthermore, the host computer system also includes a second wireless communication module and a spot welding online real-time monitoring system; wherein, the spot welding online real-time monitoring system is connected to the first wireless communication module through the second wireless communication module; the spot welding online real-time monitoring system is used to receive the welding point parameter information in real time.

[0032] Furthermore, the spot welding quality assessment system includes a sample information input module, a genetic algorithm optimization module and a weight threshold update module; wherein,

[0033] The sample information input module is used to input the solder joint parameter information sample and its corresponding solder joint quality information label into the BP neural network model to determine the initial weight and initial threshold of the BP neural network model;

[0034] The genetic algorithm optimization module is used to calculate the fitness value of the solder joint parameter information sample according to the genetic algorithm, and when the first constraint condition is met, optimize the initial weight and initial threshold according to the fitness value to obtain the optimal weight and optimal threshold;

[0035] The weight threshold updating module is used to calculate the error between the optimal weight and the optimal threshold, and cyclically update the optimal weight and the optimal threshold according to the calculation result until the second constraint condition is met, and output the updated optimal weight and optimal threshold and their corresponding optimized BP neural network model.

[0036] Furthermore, the genetic algorithm optimization module includes an initial population establishment unit, a fitness value calculation unit and a cycle condition judgment unit; wherein,

[0037] The initial population establishing unit is configured to encode the solder joint parameter information samples and randomly select a number of individuals from the encoded solder joint parameter information samples as the initial population;

[0038] The fitness value calculation unit is used to perform selection, crossover, and mutation operations on the initial population to obtain a new population, and calculate the fitness value of the new population;

[0039] The loop condition judgment unit is configured to optimize the initial weight and initial threshold using the new population and its fitness value to obtain an optimal weight and optimal threshold when the fitness value of the new population satisfies the first constraint condition; and

[0040] When the fitness value of the new population does not satisfy the first constraint condition, selection, crossover, and mutation operations are performed on the new population until the fitness value of the new population obtained by the operations satisfies the first constraint condition.

[0041] Furthermore, the host computer system further includes a spot welding process database, which is used to store the welding point parameter information and welding point parameter information samples, and update the welding point parameter information samples through the welding point parameter information.

[0042] Compared with the prior art, the embodiments of the present invention have the following advantages:

[0043] The present invention provides an online real-time monitoring device for resistance spot welding quality. The device solves the problems of low current measurement accuracy and low stability through an adjustable current sampling module based on a flexible Rogowski coil. The adjustable circuit can realize three-speed adjustment measurement of high, medium and low according to different welding currents of a resistance spot welding machine, ensures stable parameter measurement, and can meet the accuracy and reliability requirements for welding spot current measurement. At the same time, it has strong anti-interference ability and low production cost. A spot welding quality evaluation method based on an improved genetic algorithm to optimize a BP neural network can solve the problem of easily falling into a local minimum during use, compensate for the defects of a traditional BP neural network, and improve the efficiency of welding spot quality evaluation. Compared with the traditional BP algorithm, the improved BP algorithm has a reduced number of training steps during use, saves training time, realizes high-precision and rapid detection and evaluation of post-weld quality, and is suitable for welding lines with large-scale production characteristics. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solution of the present invention, the following is a brief introduction to the drawings required for use in the implementation. Obviously, the drawings described below are only some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0045] Figure 1 This is a structural diagram of an online real-time monitoring device for resistance spot welding quality provided by one embodiment of the present invention;

[0046] Figure 2 is a structural diagram of an adjustable current sampling module 11 provided in one embodiment of the present invention;

[0047] Figure 3 is a structural diagram of an adjustable current sampling circuit 111 provided in one embodiment of the present invention;

[0048] Figure 4 is a structural diagram of a zero-crossing detection circuit 112 provided in one embodiment of the present invention;

[0049] Figure 5 is a structural diagram of a single-chip computer system 12 provided in one embodiment of the present invention;

[0050] Figure 6 This is a specific structural diagram of a lower computer system 1 provided in one embodiment of the present invention;

[0051] Figure 7 This is a specific workflow diagram of the lower computer system 1 provided in one embodiment of the present invention;

[0052] Figure 8 This is a specific structural diagram of the host computer system 2 provided in one embodiment of the present invention;

[0053] Figure 9 is a structural diagram of a spot welding quality assessment system 21 provided in one embodiment of the present invention;

[0054] Figure 10 is a structural diagram of a genetic algorithm optimization module 212 provided in one embodiment of the present invention;

[0055] Figure 11 This is a diagram of the formation process of the optimized BP neural network model provided by a certain embodiment of the present invention. DETAILED DESCRIPTION

[0056] The following is a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings and embodiments. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts are within the scope of protection of the present invention.

[0057] It should be understood that the step numbers used herein are only for convenience of description and are not intended to limit the order in which the steps are to be executed.

[0058] It should be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, the singular forms "a", "an" and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0059] The terms “include” and “comprising” indicate the presence of described features, integers, steps, operations, elements and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0060] The term "and / or" refers to and includes any and all possible combinations of one or more of the associated listed items.

[0061] In one embodiment, if Figure 1 As shown, the present invention provides an online real-time monitoring device for resistance spot welding quality, comprising a host computer system 2 and a slave computer system 1; wherein,

[0062] The lower computer system 1 includes an adjustable current sampling module 11, a voltage and pressure sampling module 13, and a single chip computer system 12; wherein,

[0063] The adjustable current sampling module 11 is used to detect the welding current signal of the resistance spot welding machine;

[0064] The voltage and air pressure sampling module 13 is used to detect the welding voltage signal and welding air pressure signal of the resistance spot welding machine;

[0065] The single-chip computer system 12 is used to process the welding current signal detected by the adjustable current sampling module 11 and the welding voltage signal and welding gas pressure signal detected by the voltage and gas pressure sampling module 13 to obtain welding point parameter information; the welding point parameter information includes at least the effective value of current, the effective value of voltage, the effective value of gas pressure, and welding time;

[0066] The host computer system 2 includes a spot welding quality assessment system 21; wherein, the spot welding quality assessment system 21 is used to optimize the BP neural network model through a genetic algorithm, construct a feature vector of the weld parameter information, input it into the optimized BP neural network model, and output weld quality information; the weld quality information includes the weld core diameter and shear strength.

[0067] Specifically, the present invention addresses the technical issues of low measurement accuracy, low monitoring efficiency, and inability to quickly assess weld quality in existing resistance spot welding monitoring devices. The present invention proposes an online, real-time monitoring device for resistance spot welding quality, comprising a lower computer system 1 and a host computer system 2. Lower computer system 1 collects welding current, welding voltage, and welding gas pressure signals during the operation of the spot welder, along with their corresponding welding time, processes the collected data, and transmits it to host computer system 2. Host computer system 2 uses an improved BP neural network model in a spot welding quality assessment system 21, taking as input a feature vector constructed from the data received by host computer system 2, and outputting nugget diameter and tensile shear strength. The present invention optimizes a BP neural network spot welding quality assessment method based on an improved genetic algorithm, thereby resolving issues such as the tendency to fall into local minima during use, addressing shortcomings of traditional BP neural networks and improving the efficiency of weld quality assessment. Furthermore, compared to traditional BP algorithms, the improved BP algorithm requires fewer training steps and saves training time, enabling high-precision, rapid detection and assessment of post-weld quality. The improved BP algorithm is suitable for use in welding lines with mass production characteristics.

[0068] In one embodiment, the structure of the adjustable current sampling module 11 is as follows: Figure 2 As shown, it includes an adjustable current sampling circuit 111 and a zero-crossing detection circuit 112; wherein,

[0069] The adjustable current sampling circuit 111 is used to perform voltage conversion, amplification, integral reduction and regulation on the AC current of the resistance spot welding machine to obtain a DC current;

[0070] The zero-crossing detection circuit 112 is used to perform waveform conversion and zero-crossing detection on the DC current to obtain the welding current signal.

[0071] Specifically, to address the shortcomings of existing resistance spot welding monitoring devices such as low welding current measurement accuracy and stability, the adjustable current sampling circuit 111 in the present invention uses a flexible Rogowski coil 1111 as a current sensor to detect the welding current signal of the resistance spot welding machine.

[0072] In one embodiment, the structure of the adjustable current sampling circuit 111 is as follows: Figure 3 As shown, it includes a flexible Rogowski coil 1111, an inverting amplifier circuit 1112, a high-resistance feedback integration circuit 1113, a filter circuit 1114, an adjustable signal circuit 1115, a rectifier circuit 1116 and a voltage follower 1117; wherein,

[0073] The flexible Rogowski coil 1111 is used to convert the AC current of the resistance spot welding machine into a voltage to obtain an AC voltage signal;

[0074] The inverting amplifier circuit 1112 is used to amplify the AC voltage signal;

[0075] The high-resistance feedback integration circuit 1113 is used to integrate and restore the amplified AC voltage signal to obtain a restored AC current;

[0076] The filter circuit 1114 is used to filter the restored AC current;

[0077] The adjustable signal circuit 1115 is used to adjust the magnitude of the filtered AC current;

[0078] The rectifier circuit 1116 is used to convert the regulated AC current into a DC current;

[0079] The voltage follower 1117 is used to buffer and isolate the DC current.

[0080] Specifically, the AC voltage signal output by the flexible Rogowski coil 1111 is essentially the differential signal of the measured AC current; the inverting amplifier circuit 1112 amplifies the AC voltage signal to facilitate subsequent circuit processing; by improving the traditional integration circuit, a high-resistance feedback integration circuit 1113 is added between the input and output terminals of the operational amplifier to truly restore the input current, effectively solve the integration drift problem of the traditional integration circuit, reduce the influence of the input offset voltage, and attenuate the low-frequency quantity; the filter circuit 1114 is used to remove interference signals; the adjustable signal circuit 1 115 is used to adjust the size of the voltage signal. According to the different welding currents of the resistance spot welder, the measurement is divided into three levels: high, medium and low. When the welding current of the resistance spot welder is large, the signal is reduced by the adjustable signal circuit 1115. When the welding current of the resistance spot welder is small, the signal is amplified by the adjustable signal circuit 1115, which can significantly improve the welding current measurement accuracy; the rectifier circuit 1116 converts AC power into DC power for the single-chip microcomputer system 12 to collect signals; the voltage follower 1117 is connected to the external port of the single-chip microcomputer system 12, mainly playing a buffering and isolating role.

[0081] In one embodiment, the structure of the zero-crossing detection circuit 112 is as follows: Figure 4 As shown, it includes a comparator 1121 and an optocoupler isolation circuit 1122; wherein,

[0082] The comparator 1121 is configured to convert the DC current into a square wave signal according to a reference level;

[0083] The optical coupler isolation circuit 1122 is used to perform zero-crossing detection on the square wave signal.

[0084] Specifically, the zero detection circuit part is composed of a comparator 1121 and a zero-crossing detection circuit based on optocoupler isolation - an optocoupler isolation circuit 1122. The DC current signal utilizes the characteristics of the comparator 1121, and the threshold of the DC current can be converted into a square wave signal by adjusting the reference level. The zero-crossing point is then detected by the zero-crossing detection circuit 112 based on optocoupler isolation, and finally connected to the external port of the single-chip computer system 12, and implemented through the external interrupt collection method of the lower computer system 1.

[0085] In one embodiment, the structure of the single chip computer system 12 is as follows: Figure 5 As shown, it includes an ADC sampling module 121;

[0086] The ADC sampling module 121 is used to collect the welding current signal and calculate the welding current signal through the point-by-point integration method to obtain the effective value of the current; collect the welding voltage signal and calculate the welding voltage signal through the point-by-point integration method to obtain the effective value of the voltage; collect the welding gas pressure signal and calculate the average value of the welding gas pressure signal to obtain the effective value of the gas pressure.

[0087] Specifically, the single-chip microcomputer system 12 is an STM32 single-chip microcomputer, and the calculation formulas for the effective value of current, the effective value of voltage, and the effective value of air pressure are respectively:

[0088]

[0089]

[0090]

[0091] Where, I is the effective value of current; U is the effective value of voltage; P is the effective value of air pressure; I k is the current value corresponding to the welding current signal at time k; V k is the voltage value corresponding to the welding voltage signal at time k; P k is the air pressure value corresponding to the welding air pressure signal at time k; n is the number of sampling times in the entire welding cycle.

[0092] In one embodiment, the structure of the slave system 1 is as follows: Figure 6 As shown, it also includes a first data display module 14, an alarm module 15 and a first wireless communication module 17; wherein,

[0093] The first data display module 14 is configured to display the soldering point parameter information when the soldering point parameter information meets the upper and lower error limit requirements;

[0094] The alarm module 15 is used to generate an alarm signal when the soldering point parameter information does not meet the upper and lower error limits, so as to turn on the LED light and sound a buzzer.

[0095] The first wireless communication module 17 is used to send the welding point parameter information to the host computer system 2; wherein the wireless communication method is not specifically limited here.

[0096] Specifically, the specific working process of the lower computer system 1 is as follows: Figure 7As shown, the lower computer system 1 first performs system-level and user-level initialization, then displays the initial interface and mounts the SD card, waiting for the ADC sampling module 121 to complete sampling; if the ADC sampling module 121 completes sampling, it further determines whether the collected welding point parameters (welding effective current, voltage, gas pressure and welding time) meet the upper and lower error limit requirements. If so, the data is displayed on the display screen of the first data module 14, and the device time is updated. If not, the alarm module 15 generates an alarm signal to alarm for abnormal welding point parameters (including welding current, voltage, gas pressure and welding time that are too small or too large), so that the LED light is on and the buzzer is on.

[0097] Among them, the lower computer system 1 also includes a power supply module 18 that provides different voltages to the current, voltage and air pressure sampling modules and other electrical devices according to different needs; and a button module 19 for realizing reset functions, such as device reset, buzzer reset, and LED light reset.

[0098] In one embodiment, the structure of the host computer system 2 is as follows: Figure 8 As shown, it also includes a second wireless communication module 25 and a spot welding online real-time monitoring system 23; wherein, the spot welding online real-time monitoring system 23 is connected to the first wireless communication module 17 through the second wireless communication module 25 and exchanges information; the spot welding online real-time monitoring system 23 is used to receive the welding point parameter information in real time.

[0099] The host computer system 2 also includes a spot welding process database 22, which is used to store the weld parameter information and weld parameter information samples, and update the weld parameter information samples using the weld parameter information. Specifically, the data interaction end of the spot welding process database 22 is connected to the data interaction end of the spot welding online real-time monitoring system 23 and the data input end of the spot welding quality assessment system 21, respectively, and is used to store weld parameter information, including weldment name, weld current, voltage, effective value of air pressure, and welding time, etc., and automatically refresh the database after a weld is completed to ensure timeliness, while providing a data basis for real-time analysis by the spot welding quality assessment system 21.

[0100] Among them, the host computer system 2 also includes a second data display module 24, whose input end is connected to the display data output end of the spot welding online real-time monitoring system 23 and the spot welding quality evaluation system 21, respectively, for displaying the welding point parameters and the quality of the spot welding; when necessary, the user can query and obtain the previous welding process data through the spot welding process database 22.

[0101] In one embodiment, the structure of the spot welding quality assessment system 21 is as follows: Figure 9As shown, it includes a sample information input module 211, a genetic algorithm optimization module 212 and a weight threshold update module 213; wherein,

[0102] The sample information input module 211 is used to input the solder joint parameter information sample and its corresponding solder joint quality information label into the BP neural network model to determine the initial weight and initial threshold of the BP neural network model;

[0103] The genetic algorithm optimization module 212 is configured to calculate the fitness value of the solder joint parameter information sample according to the genetic algorithm, and optimize the initial weight and initial threshold according to the fitness value when the first constraint condition is satisfied to obtain the optimal weight and optimal threshold. The first constraint condition is that the number of optimization iterations reaches the maximum number, which is used to terminate the optimization process. If the maximum number of iterations is set too large, it will inevitably affect the convergence time, and if it is too small, it will affect the accuracy of the optimal solution. Therefore, the maximum number of iterations is preferably 200.

[0104] The weight threshold updating module 213 is used to calculate the error between the optimal weight and the optimal threshold, and to perform cyclic updates on the optimal weight and the optimal threshold according to the calculation results until the second constraint condition is satisfied, and then output the updated optimal weight and the optimal threshold and the corresponding optimized BP neural network model; wherein the second constraint condition is that the error between the optimal weight and the optimal threshold is less than the target error, and the second constraint condition is used to terminate the cyclic update process. If the target error is zero, the entire network training time will increase unnecessarily and the convergence or over-fitting may occur. Therefore, the target error is preferably 1×10 -5 .

[0105] In one embodiment, the structure of the genetic algorithm optimization module 212 is as follows: Figure 10 As shown, it includes an initial population establishment unit 2121, a fitness value calculation unit 2122 and a cycle condition judgment unit 2123; wherein,

[0106] The initial population establishing unit 2121 is configured to encode the solder joint parameter information samples and randomly select a number of individuals from the encoded solder joint parameter information samples as an initial population;

[0107] The fitness value calculation unit 2122 is used to perform selection, crossover, and mutation operations on the initial population to obtain a new population, and calculate the fitness value of the new population;

[0108] The loop condition judgment unit 2123 is configured to optimize the initial weight and initial threshold using the new population and its fitness value to obtain an optimal weight and optimal threshold when the fitness value of the new population satisfies the first constraint condition; and

[0109] When the fitness value of the new population does not satisfy the first constraint condition, selection, crossover, and mutation operations are performed on the new population until the fitness value of the new population obtained by the operations satisfies the first constraint condition.

[0110] Specifically, this application optimizes the BP neural network model through an improved genetic algorithm, and uses the optimized improved BP neural network model to evaluate the spot welding quality; compared with the traditional BP algorithm, the optimized improved BP neural network model requires fewer steps to train, and training is more time-saving. It can also achieve high-precision and rapid detection and evaluation of post-weld quality, and is suitable for welding lines with large-scale production characteristics.

[0111] Among them, the formation process of the optimized BP neural network model is shown in the figure below: Figure 11 As shown, the spot welding quality evaluation method of the present application based on the improved genetic algorithm to optimize the BP neural network specifically includes: the lower computer system 1 uses the first wireless communication module 17 to transmit the welding process parameters such as the welding spot current, voltage, effective value of air pressure and welding time to the spot welding real-time online monitoring system according to the transmission protocol of the welding process parameters; the spot welding real-time online monitoring system stores the welding process parameters in real time to the spot welding process database 22, and displays the welding process parameters. These data will be stored in the database in a structured and persistent manner to provide data support for the subsequent spot welding quality evaluation system 21; based on the spot welding process database 22, the spot welding quality evaluation system 21 establishes a mapping model between the welding process parameters and the welding spot quality parameters by optimizing the BP neural network based on the improved genetic algorithm; the characteristic vector is constructed with the welding spot effective current, voltage, air pressure and welding time as parameters as the model input, and the weld core diameter and shear strength are used as output to realize the evaluation and analysis of the welding spot quality.

[0112] Among them, the size of the weld core diameter is mainly affected by the welding current and welding time. As the welding current or welding time increases, the growth rate of the weld core diameter and the final size gradually increase, and the tensile shear strength will first increase and then decrease. When the welding gas pressure is low, the welding time has a greater impact on the weld core diameter; when the welding gas pressure reaches a certain value, the impact of the welding time on the weld core diameter will decrease; because the welding process is constant current welding, and the welding voltage is determined by the dynamic resistance of the welding process. In daily production, the tensile shear strength of the weld is usually required to be greater than 3kN, and the weld core diameter needs to reach 3 times the thickness of the weld. This application will treat welds that cannot simultaneously meet the aforementioned tensile shear strength and weld core diameter requirements as unqualified welds. This application uses an improved genetic algorithm to optimize the spot welding quality assessment method of the BP neural network, which can solve problems such as easily falling into local minima during use, make up for the defects of traditional BP neural networks, and improve the efficiency of weld quality assessment.

[0113] In summary, in the embodiments of the present application, a device for online, real-time monitoring of resistance spot welding quality is designed to address the problems of low manual sampling efficiency, detection lag, low accuracy of online non-destructive full inspection, and inability to meet the demands of fast-paced production. The device includes a host computer system and a slave computer system. The slave computer system collects the welding current signal, welding voltage signal, and welding gas pressure signal during the operation of the spot welder, as well as the corresponding welding time, and processes the collected data before transmitting it to the host computer system. The host computer system uses the improved BP neural network model in the spot welding quality assessment system as input, constructing a feature vector from the data received by the host computer system, and outputs the weld nugget diameter and tensile shear strength. The present invention can achieve high-precision detection and assessment of post-weld quality and is suitable for welding lines with large-scale production characteristics.

[0114] Each embodiment in this specification is described in a progressive manner, and the same or similar parts of each embodiment can be directly referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. It should be noted that the various technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the various technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0115] The above-described embodiments merely represent several preferred implementations of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent. It should be noted that a person skilled in the art could make several improvements and substitutions without departing from the technical principles of the present invention, and these improvements and substitutions should also be considered within the scope of protection of the present application. Therefore, the scope of protection of the present patent application shall be based on the scope of protection of the claims.

Claims

1. A device for online real-time monitoring of resistance spot welding quality, characterized in that: It includes the host computer system and the slave computer system; among them, The lower computer system includes an adjustable current sampling module, a voltage and pressure sampling module, and a single chip computer system; wherein, The adjustable current sampling module is used to detect the welding current signal of the resistance spot welding machine; The voltage and air pressure sampling module is used to detect the welding voltage signal and welding air pressure signal of the resistance spot welding machine; The single-chip computer system is used to process the welding current signal detected by the adjustable current sampling module, and the welding voltage signal and welding gas pressure signal detected by the voltage and gas pressure sampling module to obtain welding point parameter information; the welding point parameter information includes at least the effective value of current, the effective value of voltage, the effective value of gas pressure, and welding time; The host computer system includes a spot welding quality assessment system; wherein the spot welding quality assessment system is used to optimize the BP neural network model through a genetic algorithm, construct a feature vector of the weld parameter information, input it into the optimized BP neural network model, and output weld quality information; the weld quality information includes the weld nugget diameter and tensile shear strength; The adjustable current sampling module includes an adjustable current sampling circuit and a zero-crossing detection circuit; wherein, The adjustable current sampling circuit is used to perform voltage conversion, amplification, integral reduction and regulation on the AC current of the resistance spot welding machine to obtain a DC current; The zero-crossing detection circuit is used to perform waveform conversion and zero-crossing detection on the DC current to obtain the welding current signal; The adjustable current sampling circuit includes a flexible Rogowski coil, an inverting amplifier circuit, a high-resistance feedback integration circuit, a filter circuit, an adjustable signal circuit, a rectifier circuit and a voltage follower; wherein, The flexible Rogowski coil is used to convert the AC current of the resistance spot welding machine into a voltage to obtain an AC voltage signal; The inverting amplifier circuit is used to amplify the AC voltage signal; The high-resistance feedback integration circuit is used to integrate and restore the amplified AC voltage signal to obtain a restored AC current; The filtering circuit is used to filter the restored AC current; The adjustable signal circuit is used to adjust the magnitude of the filtered alternating current; The rectifier circuit is used to convert the regulated alternating current into a direct current; The voltage follower is used to buffer and isolate the DC current; The spot welding quality assessment system includes a sample information input module, a genetic algorithm optimization module and a weight threshold update module; wherein, The sample information input module is used to input the solder joint parameter information sample and its corresponding solder joint quality information label into the BP neural network model to determine the initial weight and initial threshold of the BP neural network model; The genetic algorithm optimization module is used to calculate the fitness value of the solder joint parameter information sample according to the genetic algorithm, and when the first constraint condition is met, optimize the initial weight and initial threshold according to the fitness value to obtain the optimal weight and optimal threshold; The weight threshold updating module is used to calculate the error between the optimal weight and the optimal threshold, and cyclically update the optimal weight and the optimal threshold according to the calculation result until the second constraint condition is met, and output the updated optimal weight and optimal threshold and the corresponding optimized BP neural network model; The genetic algorithm optimization module includes an initial population establishment unit, a fitness value calculation unit and a cycle condition judgment unit; wherein, The initial population establishing unit is configured to encode the solder joint parameter information samples and randomly select a number of individuals from the encoded solder joint parameter information samples as the initial population; The fitness value calculation unit is used to perform selection, crossover, and mutation operations on the initial population to obtain a new population, and calculate the fitness value of the new population; The loop condition judgment unit is configured to optimize the initial weight and initial threshold using the new population and its fitness value to obtain an optimal weight and optimal threshold when the fitness value of the new population satisfies the first constraint condition; and When the fitness value of the new population does not satisfy the first constraint condition, selection, crossover, and mutation operations are performed on the new population until the fitness value of the new population obtained by the operations satisfies the first constraint condition.

2. The device for online real-time monitoring of resistance spot welding quality according to claim 1, characterized in that: The zero-crossing detection circuit includes a comparator and an optocoupler isolation circuit; wherein, The comparator is used to convert the DC current into a square wave signal according to a reference level; The optical coupler isolation circuit is used to perform zero-crossing detection on the square wave signal.

3. The device for online real-time monitoring of resistance spot welding quality according to claim 1, characterized in that: The single chip microcomputer system includes an ADC sampling module; The ADC sampling module is used to collect the welding current signal and calculate the welding current signal through the point-by-point integration method to obtain the effective value of the current; collect the welding voltage signal and calculate the welding voltage signal through the point-by-point integration method to obtain the effective value of the voltage; collect the welding gas pressure signal and calculate the average value of the welding gas pressure signal to obtain the effective value of the gas pressure.

4. The device for online real-time monitoring of resistance spot welding quality according to claim 1, characterized in that: The lower computer system also includes a first data display module, an alarm module and a first wireless communication module; wherein, The first data display module is used to display the soldering point parameter information when the soldering point parameter information meets the upper and lower error limit requirements; The alarm module is used to generate an alarm signal when the soldering point parameter information does not meet the upper and lower error limit requirements, so as to turn on the LED light and sound a buzzer to warn; The first wireless communication module is used to send the welding point parameter information to the host computer system.

5. The device for online real-time monitoring of resistance spot welding quality according to claim 4, characterized in that: The host computer system also includes a second wireless communication module and a spot welding online real-time monitoring system; wherein, the spot welding online real-time monitoring system is connected to the first wireless communication module via the second wireless communication module; the spot welding online real-time monitoring system is used to receive the welding point parameter information in real time.

6. The device for online real-time monitoring of resistance spot welding quality according to claim 1, characterized in that: The host computer system further includes a spot welding process database, which is used to store the welding point parameter information and welding point parameter information samples, and update the welding point parameter information samples according to the welding point parameter information.

Images