Laser welding anomaly detection method and detection apparatus, electronic device and storage medium

Abstract

The present application discloses a laser welding anomaly detection method and detection apparatus, an electronic device and a storage medium. The laser welding anomaly detection method comprises: acquiring current welding data in a laser welding process; performing time sequence feature extraction on the current welding data to obtain current time sequence feature data; and performing anomaly detection on the laser welding process on the basis of the current time sequence feature data. According to the method, anomaly detection is performed on the laser welding process on the basis of the current time sequence feature data that is extracted on the basis of the current welding data; and the time sequence feature data corresponding to an anomalous laser welding process also has an anomaly , thereby avoiding impossible detection of a false welding anomaly when anomaly recognition is performed on a laser welding process on the basis of a welding seam surface image, thus improving the detection accuracy of anomaly detection for the laser welding process.

Description

Laser welding anomaly detection method, detection device, electronic equipment and storage medium

[0001] This application claims priority to the Chinese patent application filed with the China Patent Office on November 27, 2023, with application number 202311606840.8 and invention name “Laser welding anomaly detection method, detection device, electronic device and storage medium”, the entire contents of which are incorporated by reference into this application. Technical Field

[0002] The present application belongs to the field of welding technology, and in particular relates to a laser welding anomaly detection method, detection device, electronic equipment and storage medium. Background Art

[0003] In recent years, the new energy lithium battery industry has developed rapidly. The structural connection and electrical connection processes of power batteries and modules are all connected by laser welding. Laser welding mainly uses a high-energy density laser beam as a heat source. By controlling the laser pulse peak power, pulse width, defocus amount, welding speed and other parameters, it promotes the melting of the workpiece and performs rapid welding. Compared with traditional welding methods, laser welding can realize automated operation and has the advantages of small heat-affected zone of welds and fast welding speed.

[0004] Due to the influence of laser welding process, environment and other factors, welding anomalies during laser welding often lead to welding leakage, welding penetration, uneven weld formation and other phenomena in welded products. Therefore, welding anomaly detection is required after welding is completed to ensure the quality of welded products.

[0005] Currently, laser welding anomaly detection primarily relies on capturing weld surface images and identifying weld anomalies based on these images. However, when a cold weld occurs during laser welding, this anomaly recognition based on the weld surface image cannot detect the cold weld, resulting in low accuracy in laser welding anomaly detection.

[0006] Application Contents

[0007] In view of this, embodiments of the present application provide a laser welding anomaly detection method, a detection device, an electronic device, and a storage medium to overcome the above problems of the prior art. Technical Solutions

[0008] The technical solution adopted in the embodiment of this application is:

[0009] In a first aspect, an embodiment of the present application provides a method for detecting anomalies in laser welding, comprising:

[0010] Obtain current welding data of the laser welding process;

[0011] Extracting time series features from current welding data to obtain current time series feature data;

[0012] Based on the current time series feature data, abnormality detection is performed on the laser welding process.

[0013] In some optional embodiments, abnormality detection of the laser welding process is performed based on the current time series feature data, including:

[0014] According to the historical time series feature data and the current time series feature data, the corresponding actual dynamic time warping distance and the predicted dynamic time warping distance are obtained respectively;

[0015] Calculate the distance difference between the actual dynamic time warping distance and the predicted dynamic time warping distance;

[0016] When the distance difference is less than or equal to the difference threshold, it is determined that the laser welding process is normal;

[0017] When the distance difference is greater than the difference threshold, it is determined that the laser welding process is abnormal.

[0018] In some optional embodiments, the historical time series feature data includes historical plasma intensity and historical welding temperature, the current time series feature data includes current plasma intensity and current welding temperature, and the actual dynamic time warping distance includes actual plasma intensity dynamic time warping distance and actual welding temperature dynamic time warping distance;

[0019] Based on historical time series feature data and current time series feature data, the corresponding actual dynamic time warping distance is obtained, including:

[0020] According to the historical plasma intensity and the current plasma intensity, the dynamic time warping algorithm is used to calculate the corresponding actual plasma intensity dynamic time warping distance;

[0021] According to the historical welding temperature and the current welding temperature, the dynamic time warping algorithm is used to calculate the corresponding actual welding temperature dynamic time warping distance.

[0022] In some optional embodiments, the predicted dynamic time warping distance includes predicting the dynamic time warping distance of the welding temperature, and obtaining the corresponding predicted dynamic time warping distance based on the historical time series feature data and the current time series feature data includes:

[0023] Determining a predicted dynamic time warping distance function, wherein the predicted dynamic time warping distance function is used to characterize the corresponding relationship between the plasma intensity dynamic time warping distance and the welding temperature dynamic time warping distance;

[0024] The predicted welding temperature dynamic time warping distance is calculated based on the actual plasma intensity dynamic time warping distance and the predicted dynamic time warping distance function.

[0025] In some optional embodiments, determining the predicted dynamic time warping distance function includes:

[0026] Obtaining a linear correlation between the dynamic time warping distance of plasma intensity and the dynamic time warping distance of welding temperature;

[0027] The linear regression parameters between the dynamic time warping distance of historical plasma intensity and the dynamic time warping distance of historical welding temperature are calculated based on the random sampling consensus algorithm;

[0028] According to the linear correlation and linear regression parameters, the predicted dynamic time warping distance function is determined.

[0029] In some optional embodiments, abnormality detection of the laser welding process is performed based on the current time series feature data, including:

[0030] Based on the current time series feature data, a clustering algorithm is used to calculate the time series feature clustering score. The time series feature clustering score is used to characterize the distribution deviation degree of the current time series feature data;

[0031] When the time series feature cluster score is greater than or equal to the cluster score threshold, it is determined that the laser welding process is normal;

[0032] When the time series feature cluster score is less than the cluster score threshold, it is determined that the laser welding process is abnormal.

[0033] In some optional embodiments, before calculating the time series feature clustering score using a clustering algorithm based on the current time series feature data, the laser welding anomaly detection method further includes:

[0034] Perform dimensionality reduction processing on the current time series feature data to obtain dimensionality reduction features;

[0035] Based on the current time series feature data, a clustering algorithm is used to calculate the time series feature clustering score, including:

[0036] Based on the dimensionality reduction features, a clustering algorithm is used to calculate the clustering scores of time series features.

[0037] In some optional embodiments, before extracting the time series features of the current welding data to obtain the current time series feature data, the laser welding anomaly detection method further includes:

[0038] Calculate the current welding data based on the area integration algorithm to obtain the current area integration value;

[0039] Determine whether the current area integral value satisfies a preset area probability distribution;

[0040] Extract the time series features of the current welding data to obtain the current time series feature data, including:

[0041] When it is determined that the current area integral value satisfies the preset area probability distribution, time series feature extraction is performed on the current welding data to obtain current time series feature data.

[0042] In some optional embodiments, the laser welding anomaly detection method further includes:

[0043] When it is determined that the current area integral value does not satisfy the preset area probability distribution, it is determined that the laser welding process is abnormal.

[0044] In some optional embodiments, determining whether the current area integral value satisfies a preset area probability distribution includes:

[0045] Get the distribution parameter range of the preset area probability distribution;

[0046] When the current area integral value is within the distribution parameter range, determining that the current area integral value satisfies a preset area probability distribution;

[0047] When the current area integral value is not within the distribution parameter range, it is determined that the current area integral value does not satisfy the preset area probability distribution.

[0048] In some optional embodiments, the current welding data is subjected to time series feature extraction to obtain current time series feature data, including:

[0049] Perform frame processing on the current welding data according to the time sequence to obtain multiple segments of framed welding data;

[0050] Perform feature extraction on each segment of framed welding data to obtain a segment of feature data;

[0051] Each segment of feature data is combined in time series to obtain the current time series feature data.

[0052] In some optional embodiments, the laser welding anomaly detection method further includes:

[0053] If the laser welding process is determined to be abnormal, the laser welding process will be marked as abnormal;

[0054] When the number of abnormal annotations is greater than or equal to the number threshold, an alarm message is generated.

[0055] In a second aspect, an embodiment of the present application provides a laser welding anomaly detection device, comprising:

[0056] An acquisition module is used to acquire current welding data of the laser welding process;

[0057] The extraction module is used to extract the time series features of the current welding data to obtain the current time series feature data;

[0058] The detection module is used to detect abnormalities in the laser welding process based on the current time series feature data.

[0059] In a third aspect, an embodiment of the present application provides an electronic device, including:

[0060] Memory;

[0061] one or more processors coupled to the memory;

[0062] One or more applications, wherein the one or more applications are stored in a memory and configured to be executed by one or more processors, and the one or more applications are configured to execute the laser welding abnormality detection method provided in the first aspect above.

[0063] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which program code is stored. The program code can be called by a processor to execute the laser welding anomaly detection method provided in the first aspect above.

[0064] In a fifth aspect, an embodiment of the present application provides a computer program product. When the computer program product is run on a computer device, the computer device executes the laser welding anomaly detection method provided in the first aspect above. Beneficial effects

[0065] The first aspect of the present invention provides a beneficial effect of detecting anomalies in the laser welding process based on the current time series feature data extracted from the current welding data. Because the time series feature data corresponding to the laser welding process anomaly is abnormal, the inability to detect a cold weld anomaly when identifying anomalies in the laser welding process based on the weld surface image can be avoided, thereby improving the accuracy of anomaly detection in the laser welding process. Furthermore, performing time series detection on the laser welding process further improves the accuracy of anomaly detection in the laser welding process.

[0066] It can be understood that the beneficial effects of the second to fifth aspects of the present application can be found in the relevant description of the first aspect of the present application and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or exemplary technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0068] FIG1 shows a schematic diagram of a scenario of a laser welding anomaly detection system provided in an embodiment of the present application.

[0069] FIG2 shows a flow chart of a laser welding anomaly detection method according to an embodiment of the present application.

[0070] FIG3 shows a scenario schematic diagram of the linear correlation between the dynamic time-warping distance of plasma intensity and the dynamic time-warping distance of welding temperature in the laser welding anomaly detection method provided in an embodiment of the present application.

[0071] FIG4 shows another flow chart of the laser welding anomaly detection method provided in an embodiment of the present application.

[0072] FIG5 shows a schematic diagram of a scenario of the laser welding anomaly detection method provided in an embodiment of the present application.

[0073] FIG6 shows a structural block diagram of a laser welding anomaly detection device provided in an embodiment of the present application.

[0074] FIG7 shows a functional block diagram of an electronic device provided in an embodiment of the present application.

[0075] FIG8 shows a computer-readable storage medium provided in an embodiment of the present application for storing or carrying program codes for implementing the laser welding anomaly detection method provided in an embodiment of the present application.

[0076] FIG9 shows a computer program product provided in an embodiment of the present application for storing or carrying program codes for implementing the laser welding anomaly detection method provided in an embodiment of the present application. DETAILED DESCRIPTION

[0077] In order to make the purpose, features, and advantages of this application more obvious and easy to understand, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the embodiments described below are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0078] It will be understood that when used in this specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0079] It should also be understood that the terms used in this specification are for the purpose of describing specific embodiments only and are not intended to limit the present application. As used in this 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.

[0080] It should be further understood that the term "and / or" used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0081] In addition, in the description of the present application, the terms "first", "second", "third", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0082] In recent years, the new energy lithium battery industry has developed rapidly. The structural connection and electrical connection processes of power batteries and modules are all connected by laser welding. Laser welding mainly uses a high-energy density laser beam as a heat source. By controlling the laser pulse peak power, pulse width, defocus amount, welding speed and other parameters, it promotes the melting of the workpiece and performs rapid welding. Compared with traditional welding methods, laser welding can realize automated operation and has the advantages of small heat-affected zone of welds and fast welding speed.

[0083] Due to the influence of laser welding process, environment and other factors, welding anomalies during laser welding often lead to welding leakage, welding penetration, uneven weld formation and other phenomena in welded products. Therefore, welding anomaly detection is required after welding is completed to ensure the quality of welded products.

[0084] Currently, laser welding anomaly detection primarily relies on capturing weld surface images and identifying weld anomalies based on these images. However, when a cold weld occurs during laser welding, this anomaly recognition based on the weld surface image cannot detect the cold weld, resulting in low accuracy in laser welding anomaly detection.

[0085] To address the above-mentioned issues, the embodiments of the present application provide a laser welding anomaly detection method, detection device, electronic device, and storage medium. By acquiring current welding data of the laser welding process and extracting time-series features from the current welding data to obtain current time-series feature data, and based on the current time-series feature data, anomaly detection of the laser welding process is performed. This enables anomaly detection of the laser welding process based on the current time-series feature data extracted from the current welding data. Because the time-series feature data corresponding to the laser welding process anomaly appears abnormal, the inability to detect a cold weld anomaly when identifying anomalies in the laser welding process based on the weld surface image can be avoided, thereby improving the accuracy of anomaly detection in the laser welding process. Furthermore, time-series detection of the laser welding process further improves the accuracy of anomaly detection in the laser welding process.

[0086] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application.

[0087] Please refer to Figure 1, which shows a schematic diagram of an application scenario of the laser welding anomaly detection system provided in an embodiment of the present application, which may include a power battery 100, a laser welding machine 200, a sensor device 300 and a processing device 400. The sensor device 300 is communicatively connected to the processing device 400 and exchanges data with the processing device 400.

[0088] The power battery 100 may include but is not limited to any one of a lithium battery, a lead-acid battery, a nickel-cadmium battery, a nickel-metal hydride battery, an iron-nickel battery, a sodium-nickel chloride battery, a silver-zinc battery, a sodium-sulfur battery, an air battery, a fuel cell or a solar battery.

[0089] The laser welding machine 200 can be used to perform laser welding on the power battery 100 . The laser welding machine 200 can include, but is not limited to, any one of a laser brazing machine, a laser melting welder, or a laser wire welding machine.

[0090] The sensor device 300 can be mounted relative to the laser welding machine 200 and can be used to collect welding data during the welding process of the laser welding machine 200 and transmit the collected welding data to the processing device 400. The welding data includes, but is not limited to, at least one of plasma intensity, welding temperature, and welding light intensity.

[0091] In one embodiment, the sensor device 300 may include but is not limited to a plasma sensor, a temperature sensor, and a light intensity sensor.

[0092] The plasma sensor is communicatively connected to the processing device 400 and exchanges data with the processing device 400. The plasma sensor can be used to collect the plasma intensity during the welding process of the laser welding machine 200 and send the collected plasma intensity to the processing device 400.

[0093] The temperature sensor is communicatively connected to the processing device 400 and exchanges data with the processing device 400. The temperature sensor can be used to collect the welding temperature during the welding process of the laser welding machine 200 and send the collected welding temperature to the processing device 400.

[0094] The light intensity sensor is communicatively connected to the processing device 400 and exchanges data with the processing device 400. The light intensity sensor can be used to collect the welding light intensity during the welding process of the laser welding machine 200 and send the collected welding light intensity to the processing device 400.

[0095] In some embodiments, the sensor device 300 may be an integrated sensor that integrates plasma intensity, welding temperature, and welding light intensity collection functions. The sensor 300 can be used to collect the plasma intensity, welding temperature, and welding light intensity during the welding process of the laser welding machine 200 and transmit the collected plasma intensity, welding temperature, and welding light intensity to the processing device 400.

[0096] The processing device 400 can be used to receive the welding data sent by the sensor device 300 and perform abnormality detection on the laser welding process based on the welding data. The processing device 400 can include but is not limited to any one of a server or a terminal device.

[0097] The server may include but is not limited to an independent physical server, a server cluster or distributed system consisting of multiple physical servers, a cloud server, etc.

[0098] Terminal devices may include but are not limited to mobile terminal devices (for example, mobile phones, personal digital assistants (PDAs), tablet personal computers (Tablet PCs), laptop computers, smart watches, smart bracelets, etc.) and fixed terminal devices (for example, desktop computers, smart panels, all-in-one computers, etc.).

[0099] Please refer to Figure 2, which shows a flow chart of a laser welding anomaly detection method provided by one embodiment of the present application. In a specific embodiment, the laser welding anomaly detection method can be applied to processing device 400 in the laser welding anomaly detection system shown in Figure 1. The following will use processing device 400 as an example to explain the process shown in Figure 2 in detail. The laser welding anomaly detection method can include the following steps S110 to S130.

[0100] Step 110: Acquire current welding data of the laser welding process.

[0101] In an embodiment of the present application, during the laser welding process, if the user needs to perform abnormal detection on the laser welding process, a detection instruction can be sent to the processing device, and the processing device receives and responds to the detection instruction to obtain the current welding data of the laser welding process.

[0102] The current welding data may include but is not limited to at least any one of current plasma intensity, current welding temperature, and current welding light intensity.

[0103] Specifically, during the laser welding process, if the user needs to perform abnormal detection on the laser welding process, a detection instruction can be sent to the processing device. The processing device receives and responds to the detection instruction and sends an acquisition instruction to the sensor device. The sensor device receives and responds to the acquisition instruction, collects the current welding data of the laser welding process, and sends the collected current welding data to the processing device. The processing device receives the current welding data returned by the sensor device.

[0104] In some embodiments, the processing device may be provided with an input panel. When the user needs to perform abnormality detection on the laser welding process, the user may input a detection instruction on the input panel of the processing device, and the processing device receives the detection instruction through the input panel.

[0105] In some embodiments, the processing device may be provided with a voice recognition module. When the user needs to perform abnormality detection on the laser welding process, voice information may be sent within the voice collection range of the voice recognition module. The voice recognition module collects the voice information sent by the user, and performs voice recognition on the collected voice information. According to the recognition result of the voice recognition, it is determined that the recognition result contains keywords for indicating abnormality detection of the laser welding process. For example, the keyword is "laser welding abnormality detection", and for another example, the keywords are "laser welding" and "abnormality detection", etc., then it is determined that a detection instruction for abnormality detection of the laser welding process has been received.

[0106] As an example, the voice message sent by the user is: perform abnormality detection on the laser welding process, and the recognition result of the voice recognition contains the keywords "laser welding" and "abnormality detection", then it is determined that the detection instruction for performing abnormality detection on the laser welding process has been received.

[0107] In some embodiments, the laser welding anomaly detection system may further include a client, which is connected to the processing device via a network and exchanges data with the processing device via the network.

[0108] When the user needs to perform abnormality detection on the laser welding process, a detection instruction can be sent to the client. The client receives and responds to the detection instruction and forwards the detection instruction to the processing device through the network. The processing device receives the detection instruction forwarded by the client.

[0109] Among them, the client can include but is not limited to any one of a mobile client (for example, a mobile phone client, a PDA client, a Tablet PC client, a laptop client, a smart watch client, a smart bracelet client or a wearable client, etc.) or a fixed client (for example, a desktop computer client, a smart panel client, etc.).

[0110] The network may include, but is not limited to, a ZigBee network, a Bluetooth (BT) network, a Wireless Fidelity (Wi-Fi) network, a Thread network, a Long Range Radio (LoRa) network, a Low-Power Wide-Area Network (LPWAN), an infrared network, a Narrow Band Internet of Things (NB-IoT), a Controller Area Network (CAN), a Digital Living Network Alliance (DLNA) network, a Wide Area Network (WAN), a Local Area Network (LAN), a Metropolitan Area Network (MAN), or a Wireless Personal Area Network (WPAN).

[0111] Step 120: extracting the time series features of the current welding data to obtain the current time series feature data.

[0112] In an embodiment of the present application, the processing device can extract the time series features of the current welding data to obtain the current time series feature data.

[0113] Among them, the time series characteristics may include but are not limited to maximum value characteristics, minimum value characteristics, mean characteristics, range characteristics, standard deviation characteristics, kurtosis characteristics, skewness characteristics, peak factor characteristics, waveform factor characteristics, pulse factor characteristics and margin factor characteristics, etc.

[0114] The current timing characteristic data may include but is not limited to the current welding maximum characteristic data, the current welding minimum characteristic data, the current welding mean characteristic data, the current welding range characteristic data, the current welding standard deviation characteristic data, the current welding kurtosis characteristic data, the current welding skewness characteristic data, the current welding peak factor characteristic data, the current welding waveform factor characteristic data, the current welding pulse factor characteristic data and the current welding margin factor characteristic data, etc.

[0115] Specifically, the processing equipment can frame the current welding data according to the time sequence to obtain multiple segments of framed welding data, and extract features from each segment of framed welding data to obtain a segment of feature data, and combine each segment of feature data according to the time sequence to obtain the current time sequence feature data, which is conducive to detecting the time sequence changes of welding data during the laser welding process and improving the detection accuracy of abnormality detection of laser welding.

[0116] In some embodiments, the processing device can calculate the current welding data based on the area integration algorithm to obtain the current area integration value, and determine whether the current area integration value satisfies the preset area probability distribution, and when it is determined that the current area integration value satisfies the preset area probability distribution, perform time series feature extraction on the current welding data to obtain current time series feature data, thereby realizing time series feature extraction on the current welding data when it is determined that the laser welding is highly normal, and avoiding time series feature extraction on welding data with a high degree of laser welding abnormality, which is beneficial to reducing the amount of computational complexity for abnormality detection of laser welding and improving the detection efficiency of abnormality detection of laser welding.

[0117] The preset area probability distribution can be used to characterize the area distribution of welding data with a high degree of normality in laser welding.

[0118] The processing device may obtain a distribution parameter range of a preset area probability distribution, and determine whether the current area integral value satisfies the preset area probability distribution based on the current area integral value and the distribution parameter range.

[0119] When the current area integral value is within the distribution parameter range, it is determined that the current area integral value satisfies the preset area probability distribution; when the current area integral value is not within the distribution parameter range, it is determined that the current area integral value does not satisfy the preset area probability distribution.

[0120] In some embodiments, the processing device can preprocess the current welding data to obtain preprocessed data, and calculate the preprocessed data based on the area integration algorithm to obtain the current area integral value, and determine whether the current area integral value satisfies the preset area probability distribution, and when it is determined that the current area integral value satisfies the preset area probability distribution, perform time series feature extraction on the current welding data to obtain current time series feature data, thereby eliminating abnormal data from the current welding data and ensuring that the data for time series feature extraction are all normal data, which is beneficial to improving the detection accuracy of abnormality detection of laser welding.

[0121] The preprocessing may include but is not limited to at least any one of low-pass filtering processing and effective segment extraction processing.

[0122] Low-pass filtering is a process based on a low-pass filtering algorithm to reduce the volatility of signal data and eliminate abnormal data in the current welding data.

[0123] The effective segment extraction process is a process of segmenting and extracting the effective segments of the current welding data through peak-finding algorithms, threshold-exceeding monitoring, or welding start and end time control, and eliminating the fluctuation of welding start and end.

[0124] Step 130: Based on the current time series feature data, perform abnormality detection on the laser welding process.

[0125] In an embodiment of the present application, the processing device can perform anomaly detection on the laser welding process based on the current time series feature data extracted from the current welding data. Because the time series feature data corresponding to the laser welding process anomaly appears abnormal, the inability to detect a cold weld anomaly when identifying anomalies in the laser welding process based on the weld surface image can be avoided, thereby improving the accuracy of anomaly detection in the laser welding process. Furthermore, performing time series detection on the laser welding process further improves the accuracy of anomaly detection in the laser welding process.

[0126] In some embodiments, the processing device can obtain the corresponding actual dynamic time warping distance and the predicted dynamic time warping distance based on the historical time series feature data and the current time series feature data, and calculate the distance difference between the actual dynamic time warping distance and the predicted dynamic time warping distance, and determine whether the laser welding process is normal based on the distance difference and the difference threshold, thereby realizing abnormality detection of the laser welding process according to the dynamic time warping algorithm, reducing the error of abnormality detection caused by the phase difference of the current welding data, and further improving the detection accuracy of abnormality detection of the laser welding process.

[0127] When the distance difference is less than or equal to the difference threshold, it is determined that the laser welding process is normal; when the distance difference is greater than the difference threshold, it is determined that the laser welding process is abnormal.

[0128] Among them, the difference threshold can be used to characterize the maximum distance difference between the actual dynamic time regularization distance and the predicted dynamic time regularization distance when the laser welding process is normal. The difference threshold may include but is not limited to the distance difference pre-set by the user, and the distance difference automatically generated by the processing equipment based on multiple abnormality detection processes of laser welding.

[0129] In some embodiments, historical timing feature data may include historical plasma intensity and historical welding temperature, current timing feature data may include current plasma intensity and current welding temperature, and actual dynamic time-warping distance may include actual plasma intensity dynamic time-warping distance and actual welding temperature dynamic time-warping distance.

[0130] The processing equipment can use a dynamic time warping algorithm to calculate the corresponding actual plasma intensity dynamic time warping distance based on the historical plasma intensity and the current plasma intensity, and use a dynamic time warping algorithm to calculate the corresponding actual welding temperature dynamic time warping distance based on the historical welding temperature and the current welding temperature.

[0131] In some embodiments, the predicted dynamic time warping distance may include a predicted welding temperature dynamic time warping distance, and the processing device may determine a predicted dynamic time warping distance function and calculate the predicted welding temperature dynamic time warping distance based on the actual plasma intensity dynamic time warping distance and the predicted dynamic time warping distance function.

[0132] The predicted dynamic time warping distance function is used to characterize the corresponding relationship between the plasma intensity dynamic time warping distance and the welding temperature dynamic time warping distance.

[0133] The processing equipment can obtain the linear correlation between the dynamic time warping distance of plasma intensity and the dynamic time warping distance of welding temperature, and calculate the linear regression parameters between the historical dynamic time warping distance of plasma intensity and the historical dynamic time warping distance of welding temperature based on the random sampling consensus algorithm (RANdom SAmple Consensus, RANSAC), and determine the predicted dynamic time warping distance function based on the linear correlation and the linear regression parameters.

[0134] In one application scenario, the plasma intensity dynamic time warping distance is P_DTW_dis, and the welding temperature dynamic time warping distance is T_DTW_dis.

[0135] By analyzing a large amount of welding data, it is found that the linear correlation between P_DTW_dis and T_DTW_dis is: T_DTW_dis = k*P_DTW_dis+bias, where k is the correlation coefficient and bias is the correlation deviation.

[0136] The linear regression parameters k′ and bias′ of the historical plasma intensity dynamic time warping distance P′_DTW_dis and the historical T′_DTW_dis are calculated by the RANSAC algorithm, and T_DTW_dis=k′*P_DTW_dis+bias′ is determined as the predicted dynamic time warping distance function.

[0137] The actual plasma intensity dynamic time warping distance P″_DTW_dis is substituted into the predicted dynamic time warping distance function T_DTW_dis=k′*P_DTW_dis+bias′ to calculate the corresponding predicted welding temperature dynamic time warping distance, and the distance difference abs(T″_DTW_dis-) between the actual welding temperature dynamic time warping distance T″_DTW_dis and the predicted welding temperature dynamic time warping distance is calculated. Based on the distance difference abs(T″_DTW_dis-) and the difference threshold abs′, it is determined whether the laser welding process is normal.

[0138] When abs(T″_DTW_dis-) is less than or equal to abs′, it is determined that the laser welding process is normal; when abs(T″_DTW_dis-) is greater than abs′, it is determined that the laser welding process is normal.

[0139] As an example, as shown in FIG3 , the linear correlation between P_DTW_dis and T_DTW_dis is shown by the dotted line in FIG3 , and the current welding data of the laser welding process includes normal welding data (normal) used to characterize a normal welding process, and abnormal welding data (abnormal) used to characterize an abnormal welding process.

[0140] P″_DTW_dis and T″_DTW_dis corresponding to normal satisfy the linear relationship, while P″_DTW_dis and T″_DTW_dis corresponding to normal do not satisfy the linear relationship.

[0141] In some embodiments, the processing device can use a clustering algorithm to calculate the time series feature clustering score based on the current time series feature data, and determine whether the laser welding process is normal based on the time series feature clustering score and the clustering score threshold, thereby realizing abnormality detection of the laser welding process based on the clustering algorithm. Since the clustering algorithm converges quickly, the detection efficiency of abnormality detection of the laser welding process can be improved.

[0142] When the time series feature cluster score is greater than or equal to the cluster score threshold, the laser welding process is determined to be normal; when the time series feature cluster score is less than the cluster score threshold, the laser welding process is determined to be abnormal.

[0143] Among them, the time series feature clustering score can be used to characterize the degree of distribution deviation of the current time series feature data, the clustering score threshold can be used to characterize the minimum clustering score with a smaller degree of distribution deviation of the time series feature data, and the clustering score threshold can include but is not limited to the clustering score pre-set by the user, and the clustering score automatically generated by the processing equipment according to the process of multiple abnormality detection of laser welding, etc.

[0144] Clustering algorithms may include, but are not limited to, a Local Outlier Factor (LOF) algorithm, a K-Nearest Neighbor (KNN) algorithm, and a Markov Chain (MC) algorithm.

[0145] In some embodiments, the processing device can perform dimensionality reduction processing on the current time series feature data to obtain dimensionality reduction features, and based on the dimensionality reduction features, use a clustering algorithm to calculate the time series feature clustering score, thereby realizing the calculation of the time series feature clustering score based on the dimensionality reduction features and clustering algorithm obtained by dimensionality reduction processing of the current time series feature data, which can reduce the data calculation amount of the clustering algorithm and improve the calculation efficiency of the clustering algorithm.

[0146] In some embodiments, the processing device determines that the laser welding process is abnormal when it determines that the current area integral value does not meet the preset area probability distribution, thereby quickly judging the laser welding abnormality based on the abnormal area distribution of the current welding data, thereby improving the detection efficiency of abnormal laser welding.

[0147] The solution provided by this application obtains current welding data of the laser welding process, extracts time-series features from the current welding data, obtains current time-series feature data, and performs anomaly detection on the laser welding process based on the current time-series feature data. This enables anomaly detection of the laser welding process based on the current time-series feature data extracted from the current welding data. Because the time-series feature data corresponding to anomalies in the laser welding process appear abnormal, this avoids the failure to detect cold weld anomalies when performing anomaly identification on the laser welding process based on weld surface images, thereby improving the accuracy of anomaly detection in the laser welding process. Furthermore, performing time-series detection on the laser welding process further improves the accuracy of anomaly detection in the laser welding process.

[0148] Please refer to Figure 4, which shows a flow chart of a laser welding anomaly detection method provided by another embodiment of the present application. In a specific embodiment, the laser welding anomaly detection method can be applied to processing device 400 in the laser welding anomaly detection system shown in Figure 1. The process shown in Figure 4 will be described in detail below using processing device 400 as an example. The laser welding anomaly detection method can include the following steps S210 to S250.

[0149] Step 210: Acquire current welding data of the laser welding process.

[0150] Step 220: extracting the time series features of the current welding data to obtain the current time series feature data.

[0151] Step 230: Based on the current time series feature data, perform abnormality detection on the laser welding process.

[0152] In this embodiment, step 210, step 220 and step S230 may refer to the contents of the corresponding steps in the aforementioned embodiment, and will not be repeated here.

[0153] Step 240: When it is determined that the laser welding process is abnormal, an abnormality mark is made for the laser welding process.

[0154] In this embodiment, when the processing device determines that the laser welding process is abnormal, it can mark the laser welding process as abnormal.

[0155] In some embodiments, when the processing device determines that the laser welding process is abnormal, it can mark the power battery as abnormal.

[0156] In some embodiments, the laser welding anomaly detection system may further include a marking device, which is communicatively connected to the processing device and exchanges data with the processing device. The marking device may be used to mark anomalies of the power battery.

[0157] When the processing device determines that the laser welding process is abnormal, it can send a marking instruction to the marking device. The marking device receives and responds to the marking instruction and marks the power battery as abnormal.

[0158] In some embodiments, when the processing device determines that the laser welding process is abnormal, it can mark the laser welding process as abnormal and count the abnormal marks.

[0159] Step 250: When the count number of abnormal annotations is greater than or equal to the quantity threshold, generate an alarm message.

[0160] In this embodiment, when the count number of abnormal annotations is greater than or equal to the quantity threshold, an alarm message can be generated so that the user can handle the abnormal welding process according to the alarm message to avoid abnormal power battery outflow, thereby improving the user's detection experience of abnormal detection of the laser welding process.

[0161] The quantity threshold may include, but is not limited to, a quantity value preset by a user, and a quantity value automatically generated by a processing device based on multiple anomaly detections of laser welding. The warning information may include, but is not limited to, at least one of a text warning message, an audible warning message, and a light warning message.

[0162] In some embodiments, the laser welding anomaly detection system may further include an intelligent failure analysis (FA) system, wherein the intelligent FA is communicatively connected to the processing device and performs data exchange with the processing device.

[0163] When the count number of abnormal annotations is greater than or equal to the quantity threshold, the current welding parameters can be obtained, and based on the current welding parameters and the set welding parameter range, the cause of the welding abnormality and the processing strategy can be determined, and an alarm message carrying the cause of the welding abnormality and the processing strategy can be generated, so that the user can quickly process the abnormal welding process according to the welding cause and the processing strategy, further improving the user's detection experience of abnormal detection of the laser welding process, and facilitating the user's processing efficiency of the abnormal welding process.

[0164] The current welding parameters may include but are not limited to air knife air flow rate, laser power, defocusing amount, and welding speed.

[0165] In an application scenario, as shown in FIG5 , the laser welding anomaly detection system may further include an intelligent FA system, the current welding data may include plasma intensity, welding temperature, and welding light intensity, and the laser welding anomaly detection method may include steps S301 to S317.

[0166] Step S301: obtaining plasma intensity, welding temperature, and welding light intensity.

[0167] Step S302: performing low-pass filtering on the plasma intensity, welding temperature, and welding light intensity respectively to obtain filtered plasma intensity, filtered welding temperature, and filtered welding light intensity.

[0168] Step S303: extracting effective segments of the filtered plasma intensity, the filtered welding temperature, and the filtered welding light intensity respectively to obtain effective filtered plasma intensity, effective filtered welding temperature, and effective filtered welding light intensity.

[0169] Step S304: Calculate the effective filtered plasma intensity based on the area integration algorithm to obtain the plasma intensity area integral value.

[0170] Step S305: determining whether the plasma intensity area integral value satisfies a preset plasma intensity area probability distribution.

[0171] When it is determined that the plasma intensity area integral value satisfies the preset plasma intensity area probability distribution, step S306 is executed.

[0172] If it is determined that the plasma intensity area integral value does not satisfy the preset plasma intensity area probability distribution, step S307 is executed.

[0173] Step S306: extracting the time series characteristics of the effective filtered plasma intensity to obtain the plasma intensity time series characteristics.

[0174] Step S307: Determine if the laser welding process is abnormal and mark the abnormality.

[0175] Step S308: Calculate the effective filtered welding temperature based on the area integration algorithm to obtain the welding temperature area integral value.

[0176] Step S309: determining whether the welding temperature area integral value satisfies a preset welding temperature area probability distribution.

[0177] When it is determined that the welding temperature area integral value satisfies the preset welding temperature area probability distribution, step S310 is executed.

[0178] If it is determined that the welding temperature area integral value does not satisfy the preset welding temperature area probability distribution, step S307 is executed.

[0179] Step S310: extracting the timing characteristics of the effective filtered welding temperature to obtain the welding temperature timing characteristics.

[0180] Step S311: Calculate the effective filtered welding light intensity based on the area integration algorithm to obtain the welding light intensity area integral value.

[0181] Step S312: Determine whether the welding light intensity area integral value satisfies a preset welding light intensity area probability distribution.

[0182] When it is determined that the welding light intensity area integral value satisfies the preset welding light intensity area probability distribution, step S313 is executed.

[0183] If it is determined that the welding light intensity area integral value does not satisfy the preset welding light intensity area probability distribution, step S307 is executed.

[0184] Step S313: extracting the timing characteristics of the effective filtered welding light intensity to obtain the timing characteristics of the welding light intensity.

[0185] Step S314: judging whether the laser welding process is normal according to the plasma intensity timing characteristics, the welding temperature timing characteristics and the welding light intensity timing characteristics.

[0186] If the laser welding process is judged to be normal, step S315 is executed;

[0187] If it is determined that the laser welding process is abnormal, step S307 is executed.

[0188] Step S315: End abnormality detection.

[0189] Step S316: Determine whether the counted number of abnormal annotations exceeds a threshold.

[0190] If the number of abnormal annotations exceeds the threshold, step S317 is executed.

[0191] If it is determined that the counted number of abnormal annotations does not exceed the number threshold, step S315 is executed.

[0192] Step S317: Trigger the intelligent FA system to perform FA.

[0193] The solution provided in this embodiment obtains current welding data of the laser welding process and extracts time series features from the current welding data to obtain current time series feature data. Based on the current time series feature data, the laser welding process is detected for anomalies. When the laser welding process is determined to be abnormal, the laser welding process is annotated for anomalies. When the number of anomaly annotated counts is greater than or equal to a threshold, an alarm message is generated. This achieves anomaly detection of the laser welding process based on the current time series feature data extracted from the current welding data. Since the time series feature data corresponding to the laser welding process anomaly is abnormal, the inability to detect a cold weld anomaly when performing anomaly identification on the laser welding process based on the weld surface image can be avoided, thereby improving the accuracy of anomaly detection in the laser welding process. Furthermore, performing time series detection on the laser welding process further improves the accuracy of anomaly detection in the laser welding process.

[0194] Furthermore, when the count number of abnormal markings in laser welding is greater than or equal to the quantity threshold, an alarm message is generated so that the user can handle the abnormal welding process according to the alarm message to avoid abnormal power battery outflow, thereby improving the user's detection experience of abnormal detection of the laser welding process.

[0195] Please refer to Figure 6, which shows a laser welding anomaly detection device 500 provided by an embodiment of the present application. The laser welding anomaly detection device 500 can be applied to the processing device 400 in the laser welding anomaly detection system shown in Figure 1. The laser welding anomaly detection device 500 shown in Figure 6 will be described in detail below using the processing device 400 as an example. The laser welding anomaly detection device 500 may include an acquisition module 510, an extraction module 520 and a detection module 530.

[0196] The acquisition module 510 can be used to obtain the current welding data of the laser welding process; the extraction module 520 can be used to extract the timing features of the current welding data to obtain the current timing feature data; the detection module 530 can be used to perform abnormality detection on the laser welding process based on the current timing feature data.

[0197] In some implementations, the detection module 530 may include a first acquiring unit, a first calculating unit, a first determining unit, and a second determining unit.

[0198] The first acquisition unit can be used to obtain the corresponding actual dynamic time warping distance and the predicted dynamic time warping distance based on the historical time series feature data and the current time series feature data; the first calculation unit can be used to calculate the distance difference between the actual dynamic time warping distance and the predicted dynamic time warping distance; the first determination unit can be used to determine that the laser welding process is normal when the distance difference is less than or equal to the difference threshold; the second determination unit can be used to determine that the laser welding process is abnormal when the distance difference is greater than the difference threshold.

[0199] In some embodiments, historical timing feature data may include historical plasma intensity and historical welding temperature, current timing feature data may include current plasma intensity and current welding temperature, actual dynamic time warping distance may include actual plasma intensity dynamic time warping distance and actual welding temperature dynamic time warping distance; the first acquisition unit may include a first calculation subunit and a second calculation subunit.

[0200] The first calculation subunit can be used to calculate the corresponding actual plasma intensity dynamic time warping distance based on the historical plasma intensity and the current plasma intensity using a dynamic time warping algorithm; the second calculation subunit can be used to calculate the corresponding actual welding temperature dynamic time warping distance based on the historical welding temperature and the current welding temperature using a dynamic time warping algorithm.

[0201] In some embodiments, predicting the dynamic time warping distance may include predicting the welding temperature dynamic time warping distance, and the first acquisition unit may further include a determination subunit and a third calculation subunit.

[0202] The determination subunit can be used to determine the predicted dynamic time warping distance function, and the predicted dynamic time warping distance function can be used to characterize the corresponding relationship between the plasma intensity dynamic time warping distance and the welding temperature dynamic time warping distance; the third calculation subunit can be used to calculate the predicted welding temperature dynamic time warping distance based on the actual plasma intensity dynamic time warping distance and the predicted dynamic time warping distance function.

[0203] In some embodiments, determining the sub-unit may include acquiring the sub-unit, calculating the sub-unit, and determining the sub-unit.

[0204] The acquisition sub-unit can be used to obtain the linear correlation between the dynamic time warping distance of plasma intensity and the dynamic time warping distance of welding temperature; the calculation sub-unit can be used to calculate the linear regression parameters between the historical plasma intensity dynamic time warping distance and the historical welding temperature dynamic time warping distance based on the random sampling consensus algorithm; the determination sub-unit can be used to determine the predicted dynamic time warping distance function based on the linear correlation and the linear regression parameters.

[0205] In some implementations, the detection module 530 may further include a second calculation unit, a third determination unit, and a fourth determination unit.

[0206] The second calculation unit can be used to calculate the time series feature cluster score based on the current time series feature data using a clustering algorithm. The time series feature cluster score can be used to characterize the degree of distribution deviation of the current time series feature data; the third determination unit can be used to determine that the laser welding process is normal when the time series feature cluster score is greater than or equal to the cluster score threshold; the fourth determination unit can be used to determine that the laser welding process is abnormal when the time series feature cluster score is less than the cluster score threshold.

[0207] In some embodiments, the laser welding abnormality detection device 500 may further include a processing module.

[0208] The processing module can be used for performing dimensionality reduction processing on the current time series feature data to obtain dimensionality reduction features before the second calculation unit calculates the time series feature clustering score based on the current time series feature data using a clustering algorithm.

[0209] In some embodiments, the second computing unit may include a fourth computing subunit.

[0210] The fourth calculation subunit can be used to calculate the time series feature clustering score based on the dimensionality reduction feature using a clustering algorithm.

[0211] In some embodiments, the laser welding abnormality detection device 500 may further include a calculation module and a first determination module.

[0212] The calculation module can be used to extract the time series features of the current welding data by the extraction module 520. Before obtaining the current time series feature data, the current welding data is calculated based on the area integration algorithm to obtain the current area integration value; the first determination module can be used to determine whether the current area integration value meets the preset area probability distribution.

[0213] In some implementations, the extraction module 520 may include a first extraction unit.

[0214] The first extraction unit may be configured to extract time series features from the current welding data to obtain current time series feature data when it is determined that the current area integral value satisfies a preset area probability distribution.

[0215] In some embodiments, the laser welding abnormality detection device 500 may further include a second determination module.

[0216] The second determination module may be configured to determine that the laser welding process is abnormal when it is determined that the current area integral value does not satisfy a preset area probability distribution.

[0217] In some implementations, the first determining module may include a second acquiring unit, a fifth determining unit, and a sixth determining unit.

[0218] The second acquisition unit can be used to obtain the distribution parameter range of the preset area probability distribution; the fifth determination unit can be used to determine that the current area integral value satisfies the preset area probability distribution when the current area integral value is within the distribution parameter range; the sixth determination unit can be used to determine that the current area integral value does not satisfy the preset area probability distribution when the current area integral value is not within the distribution parameter range.

[0219] In some implementations, the extraction module 520 may further include a processing unit, a second extraction unit, and a combination unit.

[0220] The processing unit can be used to frame the current welding data according to the time sequence to obtain multiple segments of framed welding data; the second extraction unit can be used to extract features from each segment of framed welding data to obtain a segment of feature data; the combination unit can be used to combine each segment of feature data according to the time sequence to obtain the current time sequence feature data.

[0221] In some embodiments, the laser welding anomaly detection device 500 may further include a marking module and a generation module.

[0222] The marking module can be used to mark the laser welding process as abnormal when it is determined that the laser welding process is abnormal; the generation module can be used to generate an alarm message when the count number of abnormal markings is greater than or equal to the quantity threshold.

[0223] The solution provided in this embodiment obtains current welding data of the laser welding process, extracts time-series features from the current welding data, obtains current time-series feature data, and performs anomaly detection on the laser welding process based on the current time-series feature data. This enables anomaly detection of the laser welding process based on the current time-series feature data extracted from the current welding data. Because the time-series feature data corresponding to the laser welding process anomaly appears abnormal, the inability to detect a cold weld anomaly when performing anomaly identification on the laser welding process based on the weld surface image can be avoided, thereby improving the accuracy of anomaly detection in the laser welding process. Furthermore, performing time-series detection on the laser welding process further improves the accuracy of anomaly detection in the laser welding process.

[0224] It should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to in detail. For the device embodiments, since they are basically similar to the method embodiments, the description is relatively simple. For the relevant parts, refer to the partial description of the method embodiments. Any processing method described in the method embodiment can be implemented by the corresponding processing module in the device embodiment, and will not be repeated in detail in the device embodiment.

[0225] In addition, the functional modules in the various embodiments of the present application may be integrated into a processing module, or each module may exist physically separately, or two or more modules may be integrated into a single module. The above-mentioned integrated modules may be implemented in the form of hardware or software functional modules.

[0226] Please refer to Figure 7, which shows a functional block diagram of an electronic device 600 provided by an embodiment of the present application. The electronic device 600 may include one or more of the following components: a memory 610, a processor 620, and one or more applications, wherein the one or more applications may be stored in the memory 610 and configured to be executed by the one or more processors 620, and the one or more applications are configured to execute the method described in the aforementioned method embodiment.

[0227] The memory 610 may include a random access memory (RAM) or a read-only memory (ROM). The memory 610 may be used to store instructions, programs, codes, code sets or instruction sets. The memory 610 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as obtaining current welding data, extracting timing features, obtaining current timing feature data, abnormality detection, obtaining actual dynamic time warping distance, obtaining predicted dynamic time warping distance, calculating distance difference, determining that the laser welding process is normal, determining that the laser welding process is abnormal, calculating actual plasma intensity dynamic time warping distance, calculating actual welding temperature dynamic time warping distance, determining predicted dynamic time warping distance function, calculating predicted welding temperature dynamic time warping distance function). separation, obtaining linear correlation, calculating linear regression parameters, calculating time series feature clustering scores, dimensionality reduction processing of current time series feature data, obtaining dimensionality reduction features, area integral algorithm calculating current welding data, obtaining current area integral value, determining whether the current area integral value satisfies a preset area probability distribution, determining whether the current area integral value satisfies a preset area probability distribution, determining that the current area integral value does not satisfy a preset area probability distribution, obtaining a distribution parameter range, frame processing, obtaining multi-segment framed welding data, obtaining a segment of feature data, combining each segment of feature data, anomaly marking, and generating alarm information, etc.), instructions for implementing the following various method embodiments, etc. The storage data area can also store data created by the electronic device 600 during use (such as current welding data, current timing feature data, historical timing feature data, actual dynamic time warping distance, predicted dynamic time warping distance, distance difference, difference threshold, historical plasma intensity, historical welding temperature, current plasma intensity, current welding temperature, actual plasma intensity dynamic time warping distance, actual welding temperature dynamic time warping distance, dynamic time warping algorithm, predicted welding temperature dynamic time warping distance, predicted dynamic time warping distance function, plasma intensity dynamic time warping distance, welding temperature dynamic time warping distance, linear correlation, random sampling consensus algorithm, linear regression parameters, clustering algorithm, timing feature clustering score, distribution deviation degree, clustering score threshold, dimensionality reduction feature, area integral algorithm, current area integral value, preset area probability distribution, distribution parameter range, timing, multi-segment framed welding data and one segment of feature data), etc.

[0228] The processor 620 may include one or more processing cores. The processor 620 utilizes various interfaces and circuits to connect various components within the electronic device 600. It executes instructions, programs, code sets, or instruction sets stored in the memory 610, and accesses data stored in the memory 610 to perform various functions and process data within the electronic device 600. Optionally, the processor 620 may be implemented using at least one of the following hardware forms: a digital signal processing (DSP), a field-programmable gate array (FPGA), or a programmable logic array (PLA). The processor 620 may integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. The CPU primarily processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing display content; and the modem handles wireless communications. It is understood that the modem may not be integrated into the processor 620 and may be implemented separately via a communication chip.

[0229] Please refer to Figure 8, which shows a block diagram of a computer-readable storage medium provided in an embodiment of the present application. The computer-readable storage medium 700 stores program code 710, which can be called by a processor to execute the method described in the above method embodiment.

[0230] The computer-readable storage medium 700 can be an electronic memory such as a flash memory, an EEPROM (Electrically Erasable Programmable Read-Only Memory), an EPROM, a hard disk, or a ROM. Alternatively, the computer-readable storage medium 700 includes a non-transitory computer-readable storage medium. The computer-readable storage medium 700 has storage space for program code 710 for executing any of the method steps described above. These program codes can be read from or written to one or more computer program products. The program code 710 can be compressed, for example, in a suitable form.

[0231] Please refer to Figure 9, which shows a block diagram of a computer program product 800 provided in an embodiment of the present application. Computer program product 800 includes a computer program / instructions 810, which is stored in a computer-readable storage medium of a computer device. When computer program product 800 is executed on a computer device, the computer device's processor reads computer program / instructions 810 from the computer-readable storage medium and executes computer program / instructions 810, causing the computer device to perform the method described in the above method embodiment.

[0232] The solution provided in this embodiment obtains current welding data of the laser welding process, extracts time-series features from the current welding data, obtains current time-series feature data, and performs anomaly detection on the laser welding process based on the current time-series feature data. This enables anomaly detection of the laser welding process based on the current time-series feature data extracted from the current welding data. Because the time-series feature data corresponding to the laser welding process anomaly appears abnormal, the inability to detect a cold weld anomaly when performing anomaly identification on the laser welding process based on the weld surface image can be avoided, thereby improving the accuracy of anomaly detection in the laser welding process. Furthermore, performing time-series detection on the laser welding process further improves the accuracy of anomaly detection in the laser welding process.

[0233] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for detecting abnormality in laser welding, wherein: include: Get the current welding data of the laser welding process; Extracting time series features from the current welding data to obtain current time series feature data; Based on the current time series feature data, abnormality detection is performed on the laser welding process.

2. The laser welding abnormality detection method according to claim 1, wherein: The abnormality detection of the laser welding process based on the current time series feature data includes: According to the historical time series feature data and the current time series feature data, respectively obtain the corresponding actual dynamic time warping distance and the predicted dynamic time warping distance; Calculating a distance difference between the actual dynamic time warping distance and the predicted dynamic time warping distance; When the distance difference is less than or equal to the difference threshold, determining that the laser welding process is normal; When the distance difference is greater than a difference threshold, it is determined that the laser welding process is abnormal.

3. The laser welding abnormality detection method according to claim 2, wherein: The historical time series characteristic data include historical plasma intensity and historical welding temperature, the current time series characteristic data include current plasma intensity and current welding temperature, and the actual dynamic time warping distance includes actual plasma intensity dynamic time warping distance and actual welding temperature dynamic time warping distance; The obtaining, according to the historical time series feature data and the current time series feature data, a corresponding actual dynamic time warping distance includes: According to the historical plasma intensity and the current plasma intensity, a dynamic time warping algorithm is used to calculate the corresponding actual plasma intensity dynamic time warping distance; According to the historical welding temperature and the current welding temperature, the dynamic time warping algorithm is used to calculate the corresponding actual welding temperature dynamic time warping distance.

4. The laser welding abnormality detection method according to claim 3, wherein: The predicted dynamic time warping distance includes predicted welding temperature dynamic time warping distance, and the corresponding predicted dynamic time warping distance is obtained according to the historical time series feature data and the current time series feature data, including: Determine a predicted dynamic time warping distance function, wherein the predicted dynamic time warping distance function is used to characterize the corresponding relationship between the plasma intensity dynamic time warping distance and the welding temperature dynamic time warping distance; The predicted welding temperature dynamic time warping distance is calculated according to the actual plasma intensity dynamic time warping distance and the predicted dynamic time warping distance function.

5. The laser welding abnormality detection method according to claim 4, wherein: The determining and predicting the dynamic time warping distance function comprises: Obtaining a linear correlation between the dynamic time-warping distance of plasma intensity and the dynamic time-warping distance of welding temperature; The linear regression parameters between the dynamic time warping distance of historical plasma intensity and the dynamic time warping distance of historical welding temperature are calculated based on the random sampling consensus algorithm; According to the linear correlation and the linear regression parameters, a predicted dynamic time warping distance function is determined.

6. The laser welding abnormality detection method according to any one of claims 1 to 5, wherein: The abnormality detection of the laser welding process based on the current time series feature data includes: Based on the current time series feature data, a clustering algorithm is used to calculate the time series feature clustering score. The class score is used to characterize the distribution deviation degree of the current time series feature data; When the time series feature cluster score is greater than or equal to the cluster score threshold, determining that the laser welding process is normal; When the temporal feature cluster score is less than the cluster score threshold, it is determined that the laser welding process is abnormal.

7. The laser welding abnormality detection method according to claim 6, wherein: Before calculating the time series feature clustering score based on the current time series feature data using a clustering algorithm, the laser welding anomaly detection method further includes: Performing dimensionality reduction processing on the current time series feature data to obtain dimensionality reduction features; The step of calculating the time series feature clustering score based on the current time series feature data by using a clustering algorithm includes: Based on the dimensionality reduction features, a clustering algorithm is used to calculate the time series feature clustering scores.

8. The laser welding abnormality detection method according to any one of claims 1 to 7, wherein: Before extracting the time series features of the current welding data to obtain the current time series feature data, the laser welding anomaly detection method further includes: Calculating the current welding data based on an area integration algorithm to obtain a current area integration value; Determining whether the current area integral value satisfies a preset area probability distribution; The step of extracting the time series characteristics of the current welding data to obtain the current time series characteristic data includes: When it is determined that the current area integral value satisfies the preset area probability distribution, time series feature extraction is performed on the current welding data to obtain current time series feature data.

9. The laser welding abnormality detection method according to claim 8, wherein: The current welding data is calculated based on the area integration algorithm to obtain the current area integration value, including: Preprocessing the current welding data to obtain preprocessed data; The preprocessed data is calculated based on an area integration algorithm to obtain a current area integration value.

10. The laser welding abnormality detection method according to claim 9, wherein: The preprocessing includes at least any one of a low-pass filtering process and a valid segment extraction process.

11. The laser welding abnormality detection method according to any one of claims 8 to 10, wherein: Also includes: When it is determined that the current area integral value does not satisfy the preset area probability distribution, it is determined that the laser welding process is abnormal.

12. The laser welding abnormality detection method according to any one of claims 8 to 11, wherein: The determining whether the current area integral value satisfies a preset area probability distribution includes: Obtaining a distribution parameter range of a preset area probability distribution; When the current area integral value is within the distribution parameter range, determining that the current area integral value satisfies the preset area probability distribution; When the current area integral value is not within the distribution parameter range, it is determined that the current area integral value does not satisfy the preset area probability distribution.

13. The laser welding abnormality detection method according to any one of claims 1 to 12, wherein: The step of extracting the time series characteristics of the current welding data to obtain the current time series characteristic data includes: Performing frame processing on the current welding data according to the time sequence to obtain multiple segments of framed welding data; Extract features from each segment of framed welding data to obtain a segment of feature data; Each segment of feature data is combined according to the time sequence to obtain current time sequence feature data.

14. The laser welding abnormality detection method according to any one of claims 1 to 13, wherein: Also includes: In the case where it is determined that the laser welding process is abnormal, marking the laser welding process as abnormal; When the counted number of the abnormal annotations is greater than or equal to the quantity threshold, an alarm message is generated.

15. A laser welding anomaly detection device, wherein: include: An acquisition module, used for acquiring current welding data of the laser welding process; An extraction module, used for extracting time series features from the current welding data to obtain current time series feature data; A detection module is used to perform abnormality detection on the laser welding process based on the current time series feature data.

16. An electronic device, wherein: include: Memory; One or more processors coupled to the memory; One or more applications, wherein the one or more applications are stored in the memory and are configured to be executed by one or more processors, and the one or more applications are configured to execute the laser welding anomaly detection method as described in any one of claims 1 to 14.

17. A computer-readable storage medium, wherein: The computer-readable storage medium stores program codes, and the program codes can be called by a processor to execute the laser welding anomaly detection method according to any one of claims 1 to 14.

Images