CN117259976A - Laser welding penetration on-line detection method - Google Patents

Abstract

The invention relates to an online detection method for laser welding penetration. The invention relates to welding on-line detection and intelligent control, which selects a thermally excited mesoscopic signal of a penetration characteristic region in a keyhole as a detection object to obtain a mesoscopic detection signal; preprocessing the mesoscopic detection signal to obtain computer identifiable type data; dividing the data into a training set, a testing set and a verification set after calibrating according to the preprocessed data and the corresponding actual penetration state; establishing a neural network model, training the penetration state identification model through data in a training set until a result is converged, and adjusting model super-parameters through a verification set to obtain a penetration optimal identification model; and checking the reliability of the model through the test set, and carrying out laser welding penetration on-line detection according to the optimal model.

Description

An online detection method for laser welding penetration

Technical field

The invention relates to the technical field of welding online detection and intelligent control, and is an online detection method for laser welding penetration.

Background technique

Laser and laser composite energy field welding are one of the mainstream technologies in the field of intelligent manufacturing. However, during the laser welding process, the keyhole will eject a large amount of smoke, spatter, welding plume and other substances, which will greatly interfere with the energy input of the laser. Therefore, laser and laser composite energy field welding often have deep penetration in engineering applications. Unstable quality control has seriously restricted the development of this technology. Therefore, reliable online penetration detection and quality closed-loop control are one of the important issues in the field of laser welding intelligent manufacturing. However, the characteristic area most directly related to the weld penetration state is the metal vapor recoil pressure area at the bottom of the laser keyhole. The diameter of its core area is only 0.05~0.4mm, which belongs to the mesoscopic scale category, and is completely sprayed in large quantities. Covered by materials and surrounding high radiation signals, it has strong concealment. Moreover, the extraction position and detection resolution of the penetration mesoscopic signal also have a great impact on the detection results. In addition, the mesoscopic signal changes in the keyhole are also extremely complex, with no obvious regularity, and the radiation changes violently and is highly disruptive. However, existing detection methods, due to limitations of macroscopic sampling methods, cannot effectively separate mesoscopic scale signals with melt depth characteristics from other interference signals such as keyhole inner wall radiation signals, resulting in a too low proportion of effective signals in the detection data and melting. It is more difficult to extract deep information. At the same time, because existing data analysis methods are difficult to effectively process complex and violently fluctuating mesoscopic signals, it is difficult to extract welding penetration features and the reliability of online detection is difficult to guarantee.

These factors have had a great impact on existing detection technology. The main technical problems are as follows: 1. Existing detection methods are limited by macro sampling methods and cannot effectively obtain mesoscopic feature information. 2. Existing signal recognition methods It is greatly interfered by the strong fluctuation of the signal in the keyhole, and the detection reliability is difficult to guarantee.

Contents of the invention

In order to overcome the shortcomings of the existing technology, the present invention uses online detection of laser welding penetration based on machine identification of keyhole mesoscopic features, and proposes a method of using array sensing means to collect thermally excited state mesoscopic detection signals of the characteristic area inside the keyhole. , and establish a neural network model through machine learning, and finally call the trained recognition model to analyze signal characteristics online and extract welding penetration quality information through artificial intelligence online detection method.

It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that these entities or operations are mutually exclusive. any such actual relationship or sequence exists between them. Furthermore, the terms "comprises," "comprises," or any other variation thereof are intended to cover a non-exclusive inclusion such that a process, method, article, or apparatus that includes a list of elements includes not only those elements, but also those not expressly listed other elements, or elements inherent to the process, method, article or equipment.

The invention provides an online detection method for laser welding penetration. The invention provides the following technical solutions:

A penetration detection device, which includes: a housing, a substrate, a sensor, a narrow-band filter and a three-dimensional fine-tuning mechanism;

The connection between the base plate and the laser welding joint plays a role in fixing the entire device. There are mounting holes on the base plate that can respectively fix the device shell, the three-dimensional fine-tuning mechanism and the narrow-band filter carrier. The three-dimensional fine-tuning mechanism is connected to the sensor to adjust the three-dimensional position of the sensor sensing chip. , the narrow-band filter is located on the optical path between the optical focusing lens group and the sensing surface of the array sensor, and the filtered optical real image can be projected on the sensing surface of the array sensor.

An online detection method for laser welding penetration, the method is based on a penetration detection device, and the method includes the following steps:

Step 1: Select the thermally excited state mesoscopic signal in the penetration depth characteristic area in the keyhole as the detection object to obtain the mesoscopic detection signal;

Step 2: Preprocess the mesoscopic detection signal to obtain computer-recognizable type data;

Step 3: Based on the preprocessed data, perform calibration and divide it into training set, test set and verification set;

Step 4: Establish a neural network model, train the weight parameters of the model through the training set data, and then adjust the model hyperparameters through the verification set until the results converge;

Step 5: Check the reliability of the model through the test set;

Step 6: Call the trained recognition model to analyze signal features online and extract welding stability information.

Preferably, the thermally excited state mesoscopic signal in the penetration depth characteristic area in the keyhole is selected as the detection object, and the clear thermally excited state signal real image at the bottom of the keyhole is projected onto the sensing surface of the array sensor using the principles of optical focusing imaging and spectral transmission to obtain the result. The mesoscopic detection signal is used to obtain the sample data set for analysis, and is stored in association with the actual weld penetration. The recognition model is trained through machine learning methods, and finally the trained model is called to identify the mesoscopic signal characteristics online and the current weld penetration is obtained. Deep status information.

Preferably, the thermal excited state signal at the bottom of the keyhole is a near-infrared signal generated by the rapid melting and evaporation of the metal at the bottom of the keyhole through intense energy input after the laser beam enters the base material, and is accompanied by high-density energy excitation. .

Preferably, the mesoscopic detection signal collection method is:

S1. Through a high-power optical focusing lens group with at least a shooting working distance of 0.6-1.5m and a shooting depth of 10mm, extract a clear real image of the characteristic area at the bottom of the keyhole from the coaxial optical path of the laser welding joint. A sufficiently large depth of field can Capture clear real images of characteristic areas in a fluctuating state without changing the focal length;

S2. In the near-infrared spectrum, the welding arc, plume, laser beam and other large amounts of welding radiation signals above the keyhole are effectively shielded through narrow-band filtering, so that the thermally excited state signals in the characteristic area can be effectively separated. , thereby greatly reducing the proportion of invalid signals in the detection signal;

S3. Project the real image of the thermally excited state signal onto the sensing surface of an array sensor with a large sensing coverage area and high detection accuracy. The sensing area of the array sensor should be ≥ the characteristic area to be measured, and the resolution accuracy is ≤ 10 μm, thus obtaining the to-be-determined Mesoscopic signals at different locations in the characteristic area are measured.

Preferably, the acquisition of the penetration state recognition model is specifically:

The data within a period of time are superimposed and then averaged to obtain the trend characteristic data within this time period. When analyzing the next frame of data, the first frame of data must be removed from the superimposed data and the current After one frame of data is added to it, it is superimposed and averaged while keeping the total number of frames analyzed unchanged. This method obtains the change data set of the quasi-steady-state signal;

After the processed data set is calibrated with the actual welding penetration characteristics, it is divided into a training set, a test set and a verification set;

The weight parameters of the model are trained through the training set data until the results converge, and then the model hyperparameters are adjusted through the validation set. Finally, the reliability of the model is tested through the test set to obtain the optimal penetration model.

Preferably, the single channel/or multiple channels/or all mesoscopic detection signal data collected online are first preprocessed and converted into computer-recognizable type data through the signal preprocessing, and then the trained recognition model is called for calculation. , obtain the current penetration value of laser welding, provide online diagnosis results, or provide a basis for regulating key process parameters for the welding closed-loop control system, such as welding laser power, welding speed, and defocus amount.

A laser welding penetration online detection system, the system includes:

A data acquisition module. The data acquisition module selects the thermally excited state mesoscopic signal of the penetration characteristic area in the keyhole as the detection object to obtain the mesoscopic detection signal;

A preprocessing module, which preprocesses the mesoscopic detection signal to obtain computer-recognizable type data;

A calibration module, which performs calibration based on the preprocessed data and then divides it into a training set, a test set and a verification set;

A model building module. The model building module establishes a neural network model, builds a neural network model, trains the weight parameters of the model through the training set data until the results converge, and then adjusts the model hyperparameters through the verification set, and tests the reliability of the model through the test set. sex;

Online detection module, the online detection module calls the trained model to analyze the penetration status in real time and extract welding stability information.

A computer-readable storage medium has a computer program stored thereon, and the program is executed by a processor to implement an online detection method of laser welding penetration.

A computer device includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, it implements an online laser welding penetration detection method.

The invention has the following beneficial effects:

The present invention uses the thermal excited state signal in the characteristic area at the bottom of the keyhole as the detection signal. First, the generation position of the signal has a great correlation with the penetration depth of laser/laser composite energy field welding, so it can be used as a direct detection signal. It can avoid being affected by interference factors such as environmental humidity, temperature, and gas flow fields when using indirect detection signals. At the same time, the signal's near-infrared spectrum band enhancement characteristics also support the effective shielding of other harmful signals such as welding arcs and plumes in this spectrum band, increasing the proportion of effective signals in the detection signal and reducing the difficulty of signal analysis.

The present invention proposes a method of projecting a real image of a thermally excited state signal onto an array sensor chip to obtain a mesoscopic signal, which can achieve full coverage identification of the characteristic area at the bottom of the keyhole. Since the laser keyhole is always in a fluctuating state during the laser welding process, the position of the penetration characteristic area at the mesoscopic scale will also swing accordingly. Therefore, it is very necessary to fully cover the characteristic area at the bottom of the keyhole. This is a feature of the present invention. That is, it can adaptively track and identify the penetration characteristic area, and can accurately locate the detection signal in the key penetration area. At the same time, the high-resolution characteristics of the array sensor can be used to perform high-resolution identification of the target mesoscopic area, effectively shielding most of the interference. signal to improve detection reliability.

The present invention uses machine learning methods to analyze mesoscopic detection signals, and through a large amount of data analysis to avoid signal cases and capture signal regular characteristics, it can accurately identify the trend characteristics of mesoscopic characteristic signals at the bottom of the keyhole, and effectively avoid welding signals. Data analysis problems caused by complexity, high volatility, and many interference signals.

The present invention adopts a signal preprocessing method that superimposes data within a period of time and then performs homogenization processing, so that the trend characteristic data within this period of time can be accurately obtained, thereby effectively ignoring the impact of rapid fluctuations in keyhole depth on penetration detection. interference, and can make effective judgments on the operating trend of welding penetration.

Description of the drawings

In order to more clearly explain the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings that need to be used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description The drawings illustrate some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without exerting any creative effort.

Figure 1 is a schematic diagram of the assembly of the detection device; 1. Housing of the penetration detection device; 2. Substrate; 3. Array or image sensor; 4. Narrowband filter; 5. Three-dimensional fine-tuning mechanism

Figure 2 is a flow chart of the method of the present invention;

Figure 3 is a flow chart of the artificial intelligence detection method.

Detailed ways

The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. The indicated orientation or positional relationship is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description. It does not indicate or imply that the device or element referred to must have a specific orientation or a specific orientation. construction and operation, and therefore should not be construed as limitations of the invention. Furthermore, the terms “first”, “second” and “third” are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.

In the description of the present invention, it should be noted that, unless otherwise clearly stated and limited, the terms "installation", "connection" and "connection" should be understood in a broad sense. For example, it can be a fixed connection or a detachable connection. Connection, or integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be an internal connection between two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood on a case-by-case basis.

Specific embodiment one:

As shown in Figures 1 to 3, the specific optimization technical solution adopted by the present invention to solve the above technical problems is: the present invention relates to an online detection method of laser welding penetration.

A penetration detection device, which includes: a housing, a substrate, a sensor, a narrow-band filter and a three-dimensional fine-tuning mechanism;

The connection between the base plate and the laser welding joint plays a role in fixing the entire device. There are mounting holes on the base plate that can respectively fix the device shell, the three-dimensional fine-tuning mechanism and the narrow-band filter carrier. The three-dimensional fine-tuning mechanism is connected to the sensor to adjust the three-dimensional position of the sensor sensing chip. , the narrow-band filter is located on the optical path between the optical focusing lens group and the sensing surface of the array sensor, and the filtered optical real image can be projected on the sensing surface of the array sensor.

An online detection method for laser welding penetration, the method is based on a penetration detection device, and the method includes the following steps:

Step 1: Select the thermally excited state mesoscopic signal in the penetration depth characteristic area in the keyhole as the detection object to obtain the mesoscopic detection signal;

Step 2: Preprocess the mesoscopic detection signal to obtain computer-recognizable type data;

Step 3: Based on the preprocessed data, perform calibration and divide it into training set, test set and verification set;

Step 4: Establish a neural network model, train the weight parameters of the model through the training set data until the results converge, and then adjust the model hyperparameters through the validation set;

Step 5: Check the reliability of the model through the test set;

Step 6: Call the trained recognition model to analyze signal features online and extract welding stability information.

Specific embodiment two:

The only difference between Embodiment 2 of this application and Embodiment 1 is:

When selecting the thermally excited state mesoscopic signal in the penetration characteristic area within the keyhole as the detection object, the specific steps are:

An online detection method of laser welding penetration based on machine recognition of keyhole mesoscopic features is characterized by: proposing a method of using array sensing means to collect thermally excited state mesoscopic detection signals of the keyhole internal characteristic area, and establishing a method through machine learning. Neural network model, and finally the artificial intelligence online detection method that calls the trained recognition model to analyze signal characteristics online and extract welding penetration quality information. Specific steps are as follows:

First, the thermally excited state mesoscopic signal in the characteristic area of penetration depth in the keyhole is selected as the detection object. Secondly, the clear thermally excited state signal real image at the bottom of the keyhole is projected onto the sensing surface of the array sensor using the principle of optical focusing imaging and spectral transmission. The mesoscopic detection signal is obtained, and then the model established by the machine learning method is used to identify the characteristics of the mesoscopic signal and obtain the current penetration status information.

The thermally excited state signal at the bottom of the keyhole is a near-infrared signal generated by the rapid melting and evaporation of the metal at the bottom of the keyhole through intense energy input after the laser beam enters the base material, and is accompanied by high-density energy excitation. The hole is formed under the recoil pressure of metal evaporation, and the thermal excited state signal happens to be generated in the main recoil pressure area where the keyhole is formed, so the strength of the signal can effectively reflect the pressure situation in the recoil pressure area, and its trend The sexual changes can have good consistency with the keyhole depth, which can reflect the changes in the actual penetration depth of the weld.

The mesoscopic detection signal collection method is to first extract clear images of the characteristic area at the bottom of the keyhole from the coaxial optical path of the laser welding joint through a high-power optical focusing lens group with at least a shooting working distance of 0.6-1.5m and a shooting depth of 10mm. Real image, a large enough depth of field can capture a clear real image of the characteristic area in a fluctuating state without changing the focal length. Secondly, under the near-infrared spectrum, the welding plume above the keyhole and the welding plume above the keyhole are filtered through narrow-band filtering. The laser beam and other large amounts of welding radiation signals are effectively shielded, so that the thermally excited state signals in the characteristic area can be effectively separated, thereby greatly reducing the proportion of invalid signals in the detection signal. Then, the real image of the thermally excited state signals is projected onto a transmitter On the sensing surface of an array sensor with large sensing coverage area and high detection accuracy, the sensing area of the array sensor should be ≥ the characteristic area to be measured, and the resolution accuracy should be ≤ 10 μm, so that mesoscopic signals at different positions in the characteristic area to be measured can be obtained. This method can not only obtain thermal excitation signals at all locations in the area to be measured, and accurately analyze the location of the key characteristic signals of the penetration depth, but can also further improve the effective signals in the detection data by directly extracting mesoscopic signals at key locations. Proportion, reduce the amount of data for signal analysis, and provide data guarantee for the next step of detection signal analysis.

The establishment method of the penetration state identification model is proposed based on the fluctuation characteristics of the penetration signal. Since the penetration signal has been in a rapid up and down state as the keyhole depth changes during the welding process, and in each small time period The distribution trend of signal fluctuations is directly related to the welding penetration, so the penetration signal is a quasi-steady signal with trend characteristics. It is different from the need to accurately identify each instantaneous feature when identifying transient signals. When analyzing quasi-steady-state signals, it is necessary to ignore the transient features and make effective judgments on the operating trend of the signal. Therefore, when training the penetration state recognition model, first of all, when preprocessing the penetration signal, the data within a period of time need to be superimposed and then averaged, so as to obtain the trend characteristic data within this period of time. Then when analyzing the next frame of data, the first frame of data must be removed from the superimposed data and the current frame of data must be added to it. Then, the superposition and averaging processing must be performed while keeping the total number of frames analyzed unchanged. , this method obtains the change data set of quasi-steady-state signals. Secondly, after calibrating the processed data set with the actual welding penetration characteristics, it is divided into a training set, a test set and a verification set. Thirdly, use a computer to build a neural network model for a regression task, train the penetration state recognition model through the data in the training set until the results converge, adjust the model hyperparameters through the verification set, test the reliability of the model through the test set, and call The trained recognition model analyzes signal features online and extracts welding stability information.

The real-time detection method of penetration characteristics is to first convert the single channel/or multiple channels/or all mesoscopic detection signal data collected online into computer-recognizable type data through the signal preprocessing, and then call the trained A good recognition model is used to perform calculations to obtain the current penetration value of laser welding, and provide online diagnosis results or provide a basis for regulating key process parameters for the welding closed-loop control system, such as welding laser power, welding speed, defocus, etc.

The method for establishing the penetration state recognition model is to convert the analysis sample data collected by the array sensor or image sensor into computer-recognizable type data after certain data processing, and then compare the analysis sample with the actual welding The penetration characteristics are calibrated and divided into a training set, a test set and a verification set. Then a computer is used to build a neural network model, and the penetration recognition model is trained using the data in the training set until the results converge. The model hyperparameters are adjusted through the verification set. The test set tests the reliability of the model, calls the trained recognition model to analyze signal features online, and extracts welding stability information.

The present invention uses the thermal excited state signal in the characteristic area at the bottom of the keyhole as the detection signal. First, the generation position of the signal has a great correlation with the penetration depth of laser/laser composite energy field welding, so it can be used as a direct detection signal. It can avoid being affected by interference factors such as environmental humidity, temperature, and gas flow fields when using indirect detection signals. At the same time, the signal's near-infrared spectrum band enhancement characteristics also support the effective shielding of other harmful signals such as welding arcs and plumes in this spectrum band, increasing the proportion of effective signals in the detection signal and reducing the difficulty of signal analysis.

The present invention proposes a method of projecting a real image of a thermally excited state signal onto an array sensor chip to obtain a mesoscopic signal, which can achieve full coverage identification of the characteristic area at the bottom of the keyhole. Since the laser keyhole is always in a fluctuating state during the laser welding process, the position of the penetration characteristic area at the mesoscopic scale will also swing accordingly. Therefore, it is very necessary to fully cover the characteristic area at the bottom of the keyhole. This is a feature of the present invention. That is, it can adaptively track and identify the penetration characteristic area, and can accurately locate the detection signal in the key penetration area. At the same time, the high-resolution characteristics of the array sensor can be used to perform high-resolution identification of the target mesoscopic area, effectively shielding most of the interference. signal to improve detection reliability.

The present invention uses machine learning methods to analyze mesoscopic detection signals, and through a large amount of data analysis to avoid signal cases and capture signal regular characteristics, it can accurately identify the trend characteristics of mesoscopic characteristic signals at the bottom of the keyhole, and effectively avoid welding signals. Data analysis problems caused by complexity, high volatility, and many interference signals.

The present invention adopts a signal preprocessing method that superimposes data within a period of time and then performs homogenization processing, so that the trend characteristic data within this period of time can be accurately obtained, thereby effectively ignoring the impact of rapid fluctuations in keyhole depth on penetration detection. interference, and can make effective judgments on the operating trend of welding penetration.

Specific embodiment three:

The only difference between Embodiment 3 and Embodiment 2 of this application is:

The invention provides an online detection system for laser welding penetration. The system includes:

A data acquisition module. The data acquisition module selects the thermally excited state mesoscopic signal of the penetration characteristic area in the keyhole as the detection object to obtain the mesoscopic detection signal;

A preprocessing module, which preprocesses the mesoscopic detection signal to obtain computer-recognizable type data;

A calibration module, which performs calibration based on the preprocessed data and then divides it into a training set, a test set and a verification set;

A model building module. The model building module establishes a neural network model, builds a neural network model, trains the weight parameters of the model through the training set data until the results converge, and then adjusts the model hyperparameters through the verification set, and tests the reliability of the model through the test set. sex;

Online detection module, the online detection module calls the trained model to analyze the penetration status in real time and extract welding stability information.

Specific embodiment four:

The only difference between Embodiment 4 and Embodiment 3 of this application is:

The invention provides a computer-readable storage medium on which a computer program is stored, and the program is executed by a processor to implement an online detection method of laser welding penetration.

The specific steps are:

First, the thermally excited state mesoscopic signal in the characteristic area of penetration depth in the keyhole is selected as the detection object. Secondly, the clear thermally excited state signal real image at the bottom of the keyhole is projected onto the sensing surface of the array sensor using the principle of optical focusing imaging and spectral transmission. The mesoscopic detection signal is obtained, and then the model established by the machine learning method is used to identify the characteristics of the mesoscopic signal and obtain the current penetration status information.

The thermally excited state signal at the bottom of the keyhole is a near-infrared signal generated by the rapid melting and evaporation of the metal at the bottom of the keyhole through intense energy input after the laser beam enters the base material, and is accompanied by high-density energy excitation. The hole is formed under the recoil pressure of metal evaporation, and the thermal excited state signal happens to be generated in the main recoil pressure area where the keyhole is formed, so the strength of the signal can effectively reflect the pressure situation in the recoil pressure area, and its trend The sexual changes can have good consistency with the keyhole depth, which can reflect the changes in the actual penetration depth of the weld.

The mesoscopic detection signal collection method is to first extract clear images of the characteristic area at the bottom of the keyhole from the coaxial optical path of the laser welding joint through a high-power optical focusing lens group with at least a shooting working distance of 0.6-1.5m and a shooting depth of 10mm. Real image, a large enough depth of field can capture a clear real image of the characteristic area in a fluctuating state without changing the focal length. Secondly, under the near-infrared spectrum, the welding plume above the keyhole and the welding plume above the keyhole are filtered through narrow-band filtering. The laser beam and other large amounts of welding radiation signals are effectively shielded, so that the thermally excited state signals in the characteristic area can be effectively separated, thereby greatly reducing the proportion of invalid signals in the detection signal. Then, the real image of the thermally excited state signals is projected onto a transmitter On the sensing surface of an array sensor with large sensing coverage area and high detection accuracy, the sensing area of the array sensor should be ≥ the characteristic area to be measured, and the resolution accuracy should be ≤ 10 μm, so that mesoscopic signals at different positions in the characteristic area to be measured can be obtained. This method can not only obtain thermal excitation signals at all locations in the area to be measured, and accurately analyze the location of the key characteristic signals of the penetration depth, but can also further improve the effective signals in the detection data by directly extracting mesoscopic signals at key locations. Proportion, reduce the amount of data for signal analysis, and provide data guarantee for the next step of detection signal analysis.

The establishment method of the penetration state identification model is proposed based on the fluctuation characteristics of the penetration signal. Since the penetration signal has been in a rapid up and down state as the keyhole depth changes during the welding process, and in each small time period The distribution trend of signal fluctuations is directly related to the welding penetration, so the penetration signal is a quasi-steady signal with trend characteristics. It is different from the need to accurately identify each instantaneous feature when identifying transient signals. When analyzing quasi-steady-state signals, it is necessary to ignore the transient features and make effective judgments on the operating trend of the signal. Therefore, when training the penetration state recognition model, first of all, when preprocessing the penetration signal, the data within a period of time need to be superimposed and then averaged, so as to obtain the trend characteristic data within this period of time. Then when analyzing the next frame of data, the first frame of data must be removed from the superimposed data and the current frame of data must be added to it. Then, the superposition and averaging processing must be performed while keeping the total number of frames analyzed unchanged. , this method obtains the change data set of quasi-steady-state signals. Secondly, after calibrating the processed data set with the actual welding penetration characteristics, it is divided into a training set, a test set and a verification set. Thirdly, use a computer to build a neural network model for a regression task, train the penetration state recognition model through the data in the training set until the results converge, adjust the model hyperparameters through the verification set, test the reliability of the model through the test set, and call The trained recognition model analyzes signal features online and extracts welding stability information.

The real-time detection method of penetration characteristics is to first convert the single channel/or multiple channels/or all mesoscopic detection signal data collected online into computer-recognizable type data through the signal preprocessing, and then call the trained A good recognition model is used to perform calculations to obtain the current penetration value of laser welding, and provide online diagnosis results or provide a basis for regulating key process parameters for the welding closed-loop control system, such as welding laser power, welding speed, defocus, etc.

The method for establishing the penetration state recognition model is to convert the analysis sample data collected by the array sensor or image sensor into computer-recognizable type data after certain data processing, and then compare the analysis sample with the actual welding The penetration characteristics are calibrated and divided into a training set, a test set and a verification set. Then a computer is used to build a neural network model, and the penetration recognition model is trained using the data in the training set until the results converge. The model hyperparameters are adjusted through the verification set. The test set tests the reliability of the model, calls the trained recognition model to analyze signal features online, and extracts welding stability information.

Specific embodiment five:

The only difference between Embodiment 5 and Embodiment 4 of this application is:

The invention provides a computer device, which includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, it implements an online laser welding penetration detection method.

The specific steps are:

First, the thermally excited state mesoscopic signal in the characteristic area of penetration depth in the keyhole is selected as the detection object. Secondly, the clear thermally excited state signal real image at the bottom of the keyhole is projected onto the sensing surface of the array sensor using the principle of optical focusing imaging and spectral transmission. The mesoscopic detection signal is obtained, and then the model established by the machine learning method is used to identify the characteristics of the mesoscopic signal and obtain the current penetration status information.

The thermally excited state signal at the bottom of the keyhole is a near-infrared signal generated by the rapid melting and evaporation of the metal at the bottom of the keyhole through intense energy input after the laser beam enters the base material, and is accompanied by high-density energy excitation. The hole is formed under the recoil pressure of metal evaporation, and the thermal excited state signal happens to be generated in the main recoil pressure area where the keyhole is formed, so the strength of the signal can effectively reflect the pressure situation in the recoil pressure area, and its trend The sexual changes can have good consistency with the keyhole depth, which can reflect the changes in the actual penetration depth of the weld.

The mesoscopic detection signal collection method is to first extract clear images of the characteristic area at the bottom of the keyhole from the coaxial optical path of the laser welding joint through a high-power optical focusing lens group with at least a shooting working distance of 0.6-1.5m and a shooting depth of 10mm. Real image, a large enough depth of field can capture a clear real image of the characteristic area in a fluctuating state without changing the focal length. Secondly, under the near-infrared spectrum, the welding plume above the keyhole and the welding plume above the keyhole are filtered through narrow-band filtering. The laser beam and other large amounts of welding radiation signals are effectively shielded, so that the thermally excited state signals in the characteristic area can be effectively separated, thereby greatly reducing the proportion of invalid signals in the detection signal. Then, the real image of the thermally excited state signals is projected onto a transmitter On the sensing surface of an array sensor with large sensing coverage area and high detection accuracy, the sensing area of the array sensor should be ≥ the characteristic area to be measured, and the resolution accuracy should be ≤ 10 μm, so that mesoscopic signals at different positions in the characteristic area to be measured can be obtained. This method can not only obtain thermal excitation signals at all locations in the area to be measured, and accurately analyze the location of the key characteristic signals of the penetration depth, but can also further improve the effective signals in the detection data by directly extracting mesoscopic signals at key locations. Proportion, reduce the amount of data for signal analysis, and provide data guarantee for the next step of detection signal analysis.

The establishment method of the penetration state identification model is proposed based on the fluctuation characteristics of the penetration signal. Since the penetration signal has been in a rapid up and down state as the keyhole depth changes during the welding process, and in each small time period The distribution trend of signal fluctuations is directly related to the welding penetration, so the penetration signal is a quasi-steady signal with trend characteristics. It is different from the need to accurately identify each instantaneous feature when identifying transient signals. When analyzing quasi-steady-state signals, it is necessary to ignore the transient features and make effective judgments on the operating trend of the signal. Therefore, when training the penetration state recognition model, first of all, when preprocessing the penetration signal, the data within a period of time need to be superimposed and then averaged, so as to obtain the trend characteristic data within this period of time. Then when analyzing the next frame of data, the first frame of data must be removed from the superimposed data and the current frame of data must be added to it. Then, the superposition and averaging processing must be performed while keeping the total number of frames analyzed unchanged. , this method obtains the change data set of quasi-steady-state signals. Secondly, after calibrating the processed data set with the actual welding penetration characteristics, it is divided into a training set, a test set and a verification set. Thirdly, use a computer to build a neural network model for a regression task, train the penetration state recognition model through the data in the training set until the results converge, adjust the model hyperparameters through the verification set, test the reliability of the model through the test set, and call The trained recognition model analyzes signal features online and extracts welding stability information.

The real-time detection method of penetration characteristics is to first convert the single channel/or multiple channels/or all mesoscopic detection signal data collected online into computer-recognizable type data through the signal preprocessing, and then call the trained A good recognition model is used to perform calculations to obtain the current penetration value of laser welding, and provide online diagnosis results or provide a basis for regulating key process parameters for the welding closed-loop control system, such as welding laser power, welding speed, defocus, etc.

The method for establishing the penetration state recognition model is to convert the analysis sample data collected by the array sensor or image sensor into computer-recognizable type data after certain data processing, and then compare the analysis sample with the actual welding The penetration characteristics are calibrated and divided into a training set, a test set and a verification set. Then a computer is used to build a neural network model, and the penetration recognition model is trained using the data in the training set until the results converge. The model hyperparameters are adjusted through the verification set. The test set tests the reliability of the model, calls the trained recognition model to analyze signal features online, and extracts welding stability information.

In the description of this specification, reference to the terms "one embodiment," "some embodiments," "an example," "specific examples," or "some examples" or the like means that specific features are described in connection with the embodiment or example. , structures, materials or features are included in at least one embodiment or example of the invention. In this specification, the schematic expressions of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the specific features, structures, materials or characteristics described may be combined in any suitable manner in any one or N embodiments or examples. Furthermore, those skilled in the art may combine and combine different embodiments or examples and features of different embodiments or examples described in this specification unless they are inconsistent with each other. In addition, the terms “first” and “second” are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity of indicated technical features. Therefore, features defined as "first" and "second" may explicitly or implicitly include at least one of these features. In the description of the present invention, "N" means at least two, such as two, three, etc., unless otherwise clearly and specifically limited. Any process or method descriptions in flowcharts or otherwise described herein may be understood to represent modules, segments, or portions of code that include one or more executable instructions for implementing customized logical functions or steps of the process. , and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed out of the order shown or discussed, including in a substantially simultaneous manner or in the reverse order, depending on the functionality involved, which shall It should be understood by those skilled in the art to which embodiments of the present invention belong. The logic and/or steps represented in the flowcharts or otherwise described herein, for example, may be considered a sequenced list of executable instructions for implementing the logical functions, and may be embodied in any computer-readable medium, For use by, or in combination with, instruction execution systems, devices or devices (such as computer-based systems, systems including processors or other systems that can fetch instructions from and execute instructions from the instruction execution system, device or device) or equipment. For the purposes of this specification, a "computer-readable medium" may be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of computer readable media include the following: electrical connections with one or N wires (electronic device), portable computer disk cartridge (magnetic device), random access memory (RAM), Read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium may even be paper or other suitable medium on which the program may be printed, as the paper or other medium may be optically scanned, for example, and subsequently edited, interpreted, or otherwise suitable as necessary. process to obtain the program electronically and then store it in computer memory. It should be understood that various parts of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods may be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if it is implemented in hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: discrete logic gate circuits with logic functions for implementing data signals. Logic circuits, application specific integrated circuits with suitable combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), etc.

The above is only a preferred implementation mode of a laser welding penetration online detection method. The protection scope of a laser welding penetration online detection method is not limited to the above embodiments. All technical solutions that fall under this idea belong to this invention. protection scope of the invention. It should be pointed out that for those skilled in the art, several improvements and changes can be made without departing from the principles of the present invention, and these improvements and changes should also be regarded as the protection scope of the present invention.

Claims (10)

1. A penetration detection device, characterized in that: the device includes: a housing, a substrate, a sensor, a narrow-band filter and a three-dimensional fine-tuning mechanism; The connection between the base plate and the laser welding joint plays a role in fixing the entire device. There are mounting holes on the base plate that can respectively fix the device shell, the three-dimensional fine-tuning mechanism and the narrow-band filter carrier. The three-dimensional fine-tuning mechanism is connected to the sensor to adjust the three-dimensional position of the sensor sensing chip. , the narrow-band filter is located on the optical path between the optical focusing lens group and the sensing surface of the array sensor, and the filtered optical real image can be projected on the sensing surface of the array sensor. 2. An online detection method for laser welding penetration, the method is based on the penetration detection device as claimed in claim 1, characterized in that: the method includes the following steps: Step 1: Select the thermally excited state mesoscopic signal in the penetration depth characteristic area in the keyhole as the detection object to obtain the mesoscopic detection signal; Step 2: Preprocess the mesoscopic detection signal to obtain computer-recognizable type data; Step 3: Based on the preprocessed data, perform calibration and divide it into training set, test set and verification set; Step 4: Establish a neural network model, train the weight parameters of the model through the training set data until the results converge, and then adjust the model hyperparameters through the validation set; Step 5: Check the reliability of the model through the test set; Step 6: Call the trained recognition model to analyze signal features online and extract welding stability information. 3. The method according to claim 2, characterized by: selecting the thermally excited state mesoscopic signal of the penetration characteristic area in the keyhole as the detection object, and using the principle of optical focusing imaging and spectral transmission to excite the clear thermal energy at the bottom of the keyhole. The real image of the state signal is projected onto the sensing surface of the array sensor to obtain the mesoscopic detection signal. The sample data set for analysis is obtained and stored in association with the actual weld penetration. The recognition model is trained through machine learning methods, and finally the trained The model identifies mesoscopic signal characteristics online and obtains current penetration status information. 4. The method according to claim 3, characterized in that: the thermal excited state signal at the bottom of the keyhole is that after the laser beam enters the base material, the metal at the bottom of the keyhole is rapidly melted and evaporated through violent energy input, and accompanied by A near-infrared signal generated by high-density energy excitation. 5. The method according to claim 4, characterized in that: the mesoscopic detection signal collection method is: S1. Through a high-power optical focusing lens group with at least a shooting working distance of 0.6-1.5m and a shooting depth of 10mm, extract a clear real image of the characteristic area at the bottom of the keyhole from the coaxial optical path of the laser welding joint. A sufficiently large depth of field can Capture clear real images of characteristic areas in a fluctuating state without changing the focal length; S2. In the near-infrared spectrum, the welding arc, plume, laser beam and other large amounts of welding radiation signals above the keyhole are effectively shielded through narrow-band filtering, so that the thermally excited state signals in the characteristic area can be effectively separated. , thereby greatly reducing the proportion of invalid signals in the detection signal; S3. Project the real image of the thermally excited state signal onto the sensing surface of an array sensor with a large sensing coverage area and high detection accuracy. The sensing area of the array sensor should be ≥ the characteristic area to be measured, and the resolution accuracy is ≤ 10 μm, thus obtaining the to-be-determined Mesoscopic signals at different locations in the characteristic area are measured. 6. The method according to claim 5, characterized in that: the acquisition of the penetration state recognition model is specifically: The data within a period of time are superimposed and then averaged to obtain the trend characteristic data within this period of time. When analyzing the next frame of data, the first frame of data must be eliminated from the superimposed data and the current After one frame of data is added to it, it is superimposed and averaged while keeping the total number of frames analyzed unchanged. This method obtains the change data set of the quasi-steady-state signal; After the processed data set is calibrated with the actual welding penetration characteristics, it is divided into a training set, a test set and a verification set; The weight parameters of the model are trained through the training set data until the results converge, and then the model hyperparameters are adjusted through the validation set. Finally, the reliability of the model is tested through the test set to obtain the optimal penetration model. 7. The method according to claim 6, characterized in that: single channel/or multiple channels/or all mesoscopic detection signal data collected online are first processed through the signal preprocessing and converted into a computer-recognizable type. data, and then call the already trained recognition model to perform calculations to obtain the current penetration value of laser welding, provide online diagnosis results or provide a basis for regulating key process parameters for the welding closed-loop control system, such as welding laser power, welding speed, Defocus amount. 8. A laser welding penetration online detection system, characterized by: the system includes: A data acquisition module. The data acquisition module selects the thermally excited state mesoscopic signal of the penetration characteristic area in the keyhole as the detection object to obtain the mesoscopic detection signal; A preprocessing module, which preprocesses the mesoscopic detection signal to obtain computer-recognizable type data; A calibration module, which performs calibration based on the preprocessed data and then divides it into a training set, a test set and a verification set; A model building module. The model building module establishes a neural network model, builds a neural network model, trains the weight parameters of the model through the training set data, and then adjusts the model hyperparameters through the verification set, and tests the reliability of the model through the test set; Online detection module, the online detection module calls the trained model to analyze the penetration status in real time and extract welding stability information. 9. A computer-readable storage medium with a computer program stored thereon, characterized in that the program is executed by a processor to implement the method of claims 2-7. 10. A computer device, comprising a memory and a processor, the memory stores a computer program, characterized in that when the processor executes the computer program, the method of claims 2-7 is implemented.