Tightening device health prediction method and apparatus, electronic device, and storage medium

Abstract

This application provides a method, apparatus, electronic device, computer-readable storage medium, and computer program product for predicting the health of tightening equipment. The method includes: obtaining a batch of tightening samples of a target tightening equipment from a preset sample trend library; obtaining target sample drift characteristics of the target tightening equipment based on the batch tightening samples and a preset standard template library; obtaining target equipment drift characteristics of the target tightening equipment based on a first normal sample of the batch tightening samples and the preset standard template library; when the number of abnormal samples in the batch tightening samples exceeds a preset threshold, performing a health prediction based on the target sample drift characteristics and the target equipment drift characteristics to obtain a health prediction result. Abnormal samples refer to workpiece samples with abnormal process inspection results. The health prediction result includes at least one of the following: abnormal component type, abnormal confidence level, and abnormal level. This application can predict potential abnormalities in tightening equipment in advance, improving the accuracy of equipment health prediction results.

Description

Technical Field

[0001] This application relates to the field of intelligent assembly technology, specifically to a method, apparatus, electronic device, computer-readable storage medium, and computer program product for predicting the health of tightening equipment. Background Technology

[0002] In the field of intelligent manufacturing assembly, tightening is a key process to ensure product assembly accuracy and operational safety. The health status of tightening equipment (tightening guns, bit bits, speed reduction mechanisms, sensors, etc.) directly determines the stability of tightening quality.

[0003] In related technologies, the monitoring of tightening quality often relies on the final value judgment of a single tightening result, or only on simple qualification tests based on real-time data; for the maintenance of tightening equipment, regular maintenance or repair after failure is usually adopted, making it difficult to predict potential abnormalities of the equipment in advance. Summary of the Invention

[0004] This application provides a method for predicting the health of tightening equipment, an electronic device, and a computer-readable storage medium, which can predict potential abnormalities in tightening equipment in advance and improve the accuracy of equipment health prediction results.

[0005] In a first aspect, this application provides a method for predicting the health of a tightening device, the method comprising: A batch of tightening samples of the target tightening equipment are obtained from a preset sample trend library, wherein the preset sample trend library includes multiple workpiece samples, and the workpiece samples are marked with process inspection results; Based on the batch tightening samples and the preset standard template library, the target sample drift characteristics of the target tightening device are obtained; Based on the first normal sample of the batch tightening samples and the preset standard template library, the target equipment drift characteristics of the target tightening equipment are obtained, wherein the first normal sample refers to the workpiece sample whose process inspection result is normal. When the number of abnormal samples in the batch tightening sample exceeds a preset quantity threshold, or when the proportion of abnormal samples in the batch tightening sample exceeds a preset proportion threshold, a health prediction is performed based on the target sample drift characteristics and the target equipment drift characteristics to obtain the health prediction result of the target tightening equipment. The abnormal sample refers to a workpiece sample with abnormal process inspection results. The health prediction result is used to indicate whether at least one of the bit, gun body, deceleration mechanism, and sensor of the target tightening equipment is abnormal. The health prediction result includes at least one of the abnormal component type, abnormal confidence level, and abnormal level.

[0006] Secondly, this application also provides a tightening equipment health prediction device, comprising: The first acquisition unit is used to acquire batch tightening samples of the target tightening equipment from a preset sample trend library, wherein the preset sample trend library includes multiple workpiece samples, and the workpiece samples are labeled with process inspection results; The second acquisition unit is used to acquire the target sample drift characteristics of the target tightening device based on the batch tightening samples and the preset standard template library; The third acquisition unit is used to acquire the target equipment drift characteristics of the target tightening device based on the first normal sample of the batch tightening sample and the preset standard template library, wherein the first normal sample refers to the workpiece sample whose process inspection result is normal. The prediction unit is used to perform a health prediction based on the target sample drift characteristics and the target equipment drift characteristics when the number of abnormal samples in the batch tightening samples is greater than a preset quantity threshold or when the proportion of abnormal samples in the batch tightening samples is greater than a preset proportion threshold, to obtain a health prediction result for the target tightening equipment. The abnormal sample refers to a workpiece sample whose process inspection result is abnormal. The health prediction result is used to indicate whether at least one of the bit, gun body, deceleration mechanism, and sensor of the target tightening equipment is abnormal. The health prediction result includes at least one of abnormal component type, abnormal confidence level, and abnormal level.

[0007] Thirdly, this application also provides an electronic device, which includes a processor and a memory, wherein the memory stores a computer program, and the processor executes any of the tightening device health prediction methods provided in this application when it calls the computer program in the memory.

[0008] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, the computer program being loaded by a processor to execute the tightening equipment health prediction method.

[0009] Fifthly, this application also provides a computer program product, including a computer program or instructions, which, when executed by a processor, implement any of the tightening equipment health prediction methods provided in the embodiments of this application.

[0010] In this application, firstly, by combining target sample drift characteristics and target equipment drift characteristics for health prediction, potential anomalies in tightening equipment can be predicted in advance, improving the accuracy of health prediction results. Specifically, by utilizing target equipment drift characteristics, health prediction can be performed based on the individual differences of the target tightening equipment, reducing misjudgments caused by the inherent drift due to the individual differences of the target tightening equipment itself. By utilizing target sample drift characteristics, drift characteristics caused by anomalies in the target tightening equipment can be identified for health prediction. Thus, combining target sample drift characteristics and target equipment drift characteristics can accurately locate the root cause of the deviation, clarifying whether the increase in abnormal samples is due to sample fluctuations or abnormal equipment components, further improving the accuracy of health prediction results. Secondly, when the number of abnormal samples in a batch of tightening samples exceeds a preset threshold, or when the proportion of abnormal samples in a batch of tightening samples exceeds a preset proportion threshold, health prediction is triggered. This avoids triggering unnecessary prediction processes due to individual abnormal samples (caused by random errors), reduces invalid calculations, and improves the accuracy of equipment health prediction. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a schematic block diagram of the structure of an electronic device provided in an embodiment of this application; Figure 2 This is a schematic flowchart of a tightening equipment health prediction method provided in an embodiment of this application; Figure 3 This is a schematic flowchart of an embodiment of step 202 provided in this application; Figure 4 This is a schematic flowchart of an embodiment of step 2022A provided in this application; Figure 5 This is a schematic flowchart of an embodiment of step 203 provided in this application; Figure 6 This is a schematic diagram of an embodiment of the tightening equipment health prediction device provided in this application. Detailed Implementation

[0013] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0014] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content and operations / steps, nor does it necessarily have to be performed in the order described. For example, some operations / steps can be broken down, combined, or partially merged, so the actual execution order may change depending on the actual situation.

[0015] In the description of the embodiments of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0016] To enable any person skilled in the art to implement and use this application, the following description is provided. In this description, details are set forth for purposes of explanation. It should be understood that those skilled in the art will recognize that this application can be implemented without using these specific details. In other instances, well-known processes will not be described in detail to avoid obscuring the description of the embodiments of this application with unnecessary detail. Therefore, this application is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in the embodiments of this application.

[0017] This application provides a method for predicting the health of tightening equipment, an electronic device, and a computer-readable storage medium. The electronic device may be a server, a mobile phone, a computer, etc.

[0018] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0019] Figure 1 This is a schematic block diagram of the structure of an electronic device provided in an embodiment of this application.

[0020] like Figure 1As shown, the electronic device 100 includes a processor 101 and a memory 102, which are connected via a bus 103, such as a PCIe (Peripheral Component Interconnect Express) bus. The electronic device 100 can establish a communication connection with the tightening device via wired or wireless communication to receive tightening operation data (such as tightening process data) from the tightening device.

[0021] Specifically, processor 101 provides computing and control capabilities to support the operation of the entire electronic device 100. Processor 101 can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0022] Specifically, the memory 102 can be a Flash chip, a read-only memory (ROM) disk, an optical disk, a USB flash drive, or a portable hard drive, etc.

[0023] Those skilled in the art will understand that Figure 1 The structures shown are merely block diagrams of some structures related to the embodiments of this application and do not constitute a limitation on the electronic devices to which the embodiments of this application are applied. Specific electronic devices may include more or fewer components than those shown in the figures, or combine certain components, or have different component arrangements.

[0024] The processor 101 is configured to run a computer program stored in the memory 102, and to implement any of the tightening equipment health prediction methods provided in this application embodiment when executing the computer program. For example, the processor 101 is configured to run a computer program stored in the memory 102, and to implement the following steps when executing the computer program: From a preset sample trend library, a batch of tightening samples of the target tightening equipment are obtained, wherein the preset sample trend library includes multiple workpiece samples, and the workpiece samples are labeled with process inspection results; based on the batch tightening samples and a preset standard template library, the target sample drift characteristics of the target tightening equipment are obtained; based on the first normal sample of the batch tightening samples and the preset standard template library, the target equipment drift characteristics of the target tightening equipment are obtained, wherein the first normal sample refers to a workpiece sample with normal process inspection results; when the number of abnormal samples in the batch tightening samples is greater than a preset quantity threshold, or when the proportion of abnormal samples in the batch tightening samples is greater than a preset proportion threshold, a health prediction is performed based on the target sample drift characteristics and the target equipment drift characteristics to obtain a health prediction result of the target tightening equipment, wherein the abnormal sample refers to a workpiece sample with abnormal process inspection results, and the health prediction result is used to indicate whether at least one of the bit, gun body, deceleration mechanism, and sensor of the target tightening equipment is abnormal, and the health prediction result includes at least one of abnormal component type, abnormal confidence level, and abnormal level.

[0025] It should be noted that those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the electronic device described above can be referred to the corresponding process in the following embodiment of the tightening device health prediction method, and will not be repeated here.

[0026] The following will be based on Figure 1 Taking the electronic device shown as the execution subject of the tightening equipment health prediction method as an example, the tightening equipment health prediction method provided in this application embodiment will be described in detail. For the sake of simplification and ease of description, the execution subject will be omitted in the subsequent method embodiments.

[0027] Please see Figure 2 , Figure 2 This is a flowchart illustrating a method for predicting the health of a tightening device according to an embodiment of this application. The method includes steps 201-204, wherein: 201. Obtain batch tightening samples of the target tightening equipment from the preset sample trend library.

[0028] The target tightening device is the tightening device for which health prediction is to be performed. The tightening device includes a drive board (which drives the tightening components / sensors according to control commands), tightening components (which may include screwdriver bits, gun body, reduction mechanism, etc., used to realize the tightening action), and sensors (used to detect the actual result parameters of the tightening process).

[0029] The preset sample trend library includes multiple workpiece samples, and the workpiece samples are labeled with process inspection results.

[0030] The batch tightening sample includes multiple workpiece samples. The batch tightening sample is a collection of all workpiece samples obtained from the preset sample trend library that have been tightened by the target tightening equipment. Each workpiece sample contains process inspection results, process context information and tightening process data. For example, if the target tightening equipment (gun number G001) tightens 100 workpieces in one production shift, the workpiece samples corresponding to these 100 workpieces together constitute the batch tightening sample of the tightening equipment.

[0031] For example, the process inspection results of a workpiece sample can be determined based on the tightening-related data of the workpiece sample. The tightening-related data includes process context information and tightening process data. Each workpiece on the production line that undergoes a tightening process using tightening equipment can be considered a workpiece sample. The process inspection results for this tightened workpiece can be determined by referring to steps A1-A4 below. Based on the tightening-related data of each workpiece sample, the process inspection results for each workpiece sample on the production line are obtained. After the process verification of the workpiece sample is completed, the workpiece sample is added to a preset sample trend library. In step 201, all workpiece samples that have undergone a tightening process using the target tightening equipment can be filtered from the preset sample trend library according to the equipment number of the target tightening equipment, thereby obtaining a batch of tightened samples.

[0032] A1. Obtain the first process context information and the first tightening process data of the workpiece being tightened.

[0033] The first process context information includes identification information, configuration information and batch information, and the first tightening process data includes the torque curve, angle curve, speed curve and stage markers for the current tightening process of the workpiece.

[0034] A2. Based on the first process context information and the first tightening process data, feature construction is performed to obtain the first tightening feature set and the first temporal feature of the workpiece being tightened.

[0035] A3. Based on the first tightening feature set and the first temporal feature, obtain the first verification result of the rule engine, the second verification result of the similarity model, the third verification result of the feature classification model, the fourth verification result of the temporal depth model, and the fifth verification result of the unsupervised anomaly detection model.

[0036] The first verification result is determined through the following steps a1~a5: a1. Using the rule engine, determine the final value verification result based on the pre-configured target final value window and the final torque and angle values ​​of the workpiece being tightened.

[0037] The target final value window is a final value window that matches the process scenario parameters of the workpiece being tightened. The target final value window includes at least one of the torque final value window and the angle final value window.

[0038] For example, the rule engine can call a pre-configured target final value window to extract the final torque and final angle values ​​of the workpiece being tightened. It then compares the final torque value with the final torque value window. If the final torque value is within the window's range, the final torque value is deemed acceptable (e.g., a final torque value of 5 N·m corresponds to a final torque value window of 3-7 N·m, indicating acceptable torque). If the final torque value is outside the window's range, the final torque value is deemed unacceptable. The final angle value of the workpiece being tightened is compared with the final angle value window. If the final angle value is within the window, the final angle value is considered acceptable. If it is outside the window, the final angle value is considered unacceptable. Based on the comparison results of the final torque and final angle values, the final value is determined to meet the process requirements. Finally, the final value verification result is output. For example, if both the final torque and final angle values ​​are acceptable, the final value is considered to meet the process requirements, and the final value verification result is considered acceptable.

[0039] a2. Using the rule engine, determine the stage verification result based on the pre-configured target duration range for each stage and the duration value of each process stage.

[0040] The target duration interval for each stage is the target duration interval corresponding to each process stage, and the target duration interval for each stage is the duration interval that matches the process scenario parameters.

[0041] The process scenario parameters may include, but are not limited to, one or more of the following: tightening gun number, workstation number, production line number, tightening procedure number, process version, product model, and production shift.

[0042] For example, the rule engine calls the pre-configured target time ranges corresponding to each process stage, extracts the time value of each process stage one by one, and compares the time value of each process stage with the target time range of the corresponding stage. If the time value of a certain process stage of tightening the workpiece is within the target time range of the corresponding stage, the time value of that process stage is determined to be qualified (for example, the time value of the tooth finding stage is 2s, the corresponding tooth finding stage time range is 1-3s, so the tooth finding stage time value is determined to be qualified); if the time value of a certain process stage of tightening the workpiece is outside the target time range of the corresponding stage, the time value of that process stage is determined to be unqualified; based on the comparison results of the time values ​​of each process stage, it is determined whether the time consumption of each stage meets the process requirements, and finally the stage verification result is output. For example, only when the time values ​​of all process stages of tightening the workpiece are qualified are the time consumption of each stage considered to meet the process requirements, and the stage verification result is qualified.

[0043] a3. Using the rule engine, determine the timeout verification result based on the pre-configured target total duration range and the total process duration value.

[0044] The target total duration interval is the total duration interval that matches the process scenario parameters.

[0045] For example, the rule engine can call a pre-configured target total time range, extract the total process time value of the current tightening of the workpiece, and compare the total process time value of the current tightening of the workpiece with the target total time range. If the total process time value of the current tightening of the workpiece is within the target total time range, the total process time value is determined to be qualified (for example, if the total process time value is 8 seconds, corresponding to the target total time range of 5-10 seconds, the total process time value is determined to be qualified); if the total process time value of the current tightening of the workpiece is outside the target total time range, the total process time value is determined to be unqualified. Based on the comparison result of the total process time value, it is determined whether the overall tightening time is compliant, and finally the timeout verification result is output. For example, if the total process time value of the current tightening of the workpiece is qualified, the overall tightening time is considered compliant, and the timeout verification result is qualified.

[0046] a4. Using the rule engine, determine the stage switching verification result based on the pre-configured target stage switching rules and the stage switching features.

[0047] The target stage switching rule is a stage switching rule that matches the process scenario parameters.

[0048] For example, the rule engine can call a pre-configured target stage switching rule (the target stage switching rule is used to limit the specific stage switching physical quantity, stage switching duration, and stage switching sequence), extract the stage switching features of the workpiece being tightened, and compare each stage switching feature of the workpiece being tightened (such as stage switching physical quantity, stage switching duration, and stage switching sequence) with the corresponding requirements of the target stage switching rule. If a certain stage switching feature of the workpiece being tightened matches the target stage switching rule, then the stage switching parameter is deemed qualified (for example, if the stage switching duration is 0.1s, and the corresponding target switching duration threshold is ≤0.2s, then the stage switching duration is deemed qualified). If a certain stage switching feature of the workpiece being tightened does not match the target stage switching rule, then the stage switching feature is deemed unqualified. Based on the comparison results of each stage switching feature, the engine determines whether the stage switching is standardized and finally outputs the stage switching verification result. For example, if all stage switching features of the workpiece being tightened are qualified, then the stage switching is considered standardized, and the stage switching verification result is qualified.

[0049] a5. Based on the final value verification result, the stage verification result, the timeout verification result, and the stage switching verification result, determine the first verification result.

[0050] The first verification result includes a first risk classification label. If at least one of the final value verification result, the stage verification result, the timeout verification result, and the stage switching verification result is unqualified, then the first risk classification label is risk.

[0051] The first risk classification label is a classification label determined by the rule engine to indicate whether there is a risk in the tightening of the workpiece.

[0052] For example, the rule engine summarizes the final value verification results, stage verification results, timeout verification results, and stage switching verification results, and performs logical judgment on each verification result. If any one of the verification results is unqualified, the first risk classification label is risk; if all the verification results, such as the final value verification result, stage verification result, timeout verification result, and stage switching verification result, are qualified, the first risk classification label is normal.

[0053] For example, the second verification result is determined through the following steps b1 to b5: b1. Using the similarity model, the torque reference curve is matched with the current torque curve according to the pre-configured torque reference curve, and the first curve similarity and the first deviation stage are output.

[0054] The process involves several steps: First, a standardized torque curve is generated using torque curves from multiple standard samples in a pre-defined standard template library. This curve is generated through feature extraction, curve normalization, and mean fitting. The standard samples are tightened workpieces with normal process inspection results. This standardized torque curve is configured as the torque reference curve for the similarity model. Second, a standardized torque-angle coupling curve is generated using torque-angle coupling curves from multiple standard samples in the pre-defined standard template library. This curve is configured as the torque-angle coupling reference curve for the similarity model. Third, a standardized torque-angle coupling curve is generated using torque-angle coupling curves from multiple standard samples in the pre-defined standard template library. This standardized torque-angle coupling curve is configured as the torque-angle coupling reference curve for the similarity model. The scene parameters for each standard sample used to generate the standardized torque curve are the same as the scene parameters for the workpiece being tightened. These scene parameters include at least one of the following: tightening gun number, workstation number, tightening program number, process version, product model, and production shift. The scenario parameters for each standard sample used to generate the standardized angle curve are the same as the scenario parameters for the workpiece being tightened in this instance. Similarly, the scenario parameters for each standard sample used to generate the standardized torque-angle coupling curve are the same as the scenario parameters for the workpiece being tightened in this instance.

[0055] The first curve similarity refers to the similarity between the current torque curve and the torque reference curve.

[0056] The first deviation stage refers to the process stage with deviation determined based on the current torque curve and the torque reference curve. If there is a certain deviation (such as the deviation being greater than a preset deviation threshold) between the curve segment corresponding to a certain process stage of the current torque curve (such as curve segment 1 corresponding to the tooth-finding stage) and the curve segment corresponding to the same process stage of the torque reference curve (such as curve segment 2 corresponding to the tooth-finding stage), then that process stage is regarded as the first deviation stage.

[0057] The torque reference curve is a standardized torque curve that matches the process parameters of the workpiece being tightened. It can be constructed based on standard samples from a pre-set standard template library. For example, the similarity model can call the pre-configured torque reference curve, extract the torque curve of the workpiece being tightened, perform feature alignment and similarity calculation (such as cosine similarity and Euclidean distance calculation) on the two curves, and finally output the first curve similarity (for example, the similarity calculation result is 0.92, the closer to 1, the more similar the two curves are).

[0058] The specific network structure of the similarity model can be set according to the actual business scenario requirements. For example, a DTW / Soft-DTW network structure can be used. The similarity model can use algorithms such as dynamic time warping, piecewise Euclidean distance, correlation coefficient, or cosine similarity to match curves, thereby locating the deviation stage based on the stage marker.

[0059] b2. Using the similarity model, match the current angle curve with the pre-configured angle reference curve to output the second curve similarity and the second deviation stage.

[0060] The second curve similarity refers to the similarity between the current angle curve and the angle reference curve.

[0061] The second deviation stage refers to the process stage with deviation determined based on the current angle curve and the angle reference curve. If there is a certain deviation (such as the deviation being greater than the preset deviation threshold) between the curve segment corresponding to a certain process stage of the current angle curve (such as curve segment 3 corresponding to the bonding stage) and the curve segment corresponding to the same process stage of the angle reference curve (such as curve segment 4 corresponding to the bonding stage), then that process stage is regarded as the second deviation stage.

[0062] The angle reference curve is a standardized angle curve that matches the process parameters of the tightening workpiece in this scenario. It can be constructed based on standard samples from a preset standard template library. For example, the similarity model can call the pre-configured angle reference curve, extract the current angle curve of the tightening workpiece, align the two curves, calculate the similarity, and finally output the second curve similarity.

[0063] b3. Using the similarity model, match the current torque angle coupling curve with the pre-configured torque angle coupling reference curve, and output the third curve similarity and the third deviation stage.

[0064] The third curve similarity refers to the similarity between the current torque-angle coupling curve and the torque-angle coupling reference curve.

[0065] The second deviation stage refers to the process stage with deviation determined based on the current torque-angle coupling curve and the torque-angle coupling reference curve. If there is a certain deviation (e.g., the deviation is greater than a preset deviation threshold) between the curve segment corresponding to a certain process stage of the first torque-angle coupling curve (such as curve segment 5 corresponding to the final tightening stage) and the curve segment corresponding to the same process stage of the torque-angle coupling reference curve (such as curve segment 6 corresponding to the final tightening stage), then this process stage is designated as the third deviation stage.

[0066] The torque-angle coupling reference curve is a standardized torque-angle coupling curve that matches the process parameters of the tightening workpiece in this case. It can be constructed based on standard samples from a preset standard template library. For example, the similarity model can call the pre-configured torque-angle coupling reference curve, extract the torque-angle coupling curve of the workpiece in this case, align the features of the two curves, calculate the similarity, and finally output the similarity of the third curve.

[0067] In this way, the similarity model can be used to locate the deviation stages of the current torque curve, current angle curve, and current torque-angle coupling curve, thereby uncovering the process stages where anomalies truly exist and improving the accuracy of identifying abnormal stages in the current workpiece tightening.

[0068] b4. Based on the first curve similarity, the second curve similarity, and the third curve similarity, determine the first risk score and / or the second risk classification label.

[0069] The first risk score is the risk score of the workpiece being tightened in this instance, determined based on the second verification result. For example, the first risk score can be determined based on the first curve similarity, the second curve similarity, and the third curve similarity. The first curve similarity, the second curve similarity, and the third curve similarity are all negatively correlated with the first risk score; that is, the greater the first curve similarity, the smaller the first risk score; the greater the second curve similarity, the smaller the first risk score; and the greater the third curve similarity, the smaller the first risk score.

[0070] The second risk classification label is determined based on a similarity model to indicate whether there is a risk in the tightening of the workpiece. In some embodiments, classification can be based on a first risk score of the workpiece being tightened. If the first risk score is greater than a preset risk score threshold, the second risk classification label is "risk"; if the first risk score is less than or equal to the preset risk score threshold, the second risk classification label is "normal". In some embodiments, classification can be based on a first curve similarity, a second curve similarity, and a third curve similarity. If the first curve similarity, the second curve similarity, or the third curve similarity is less than a preset similarity threshold, the second risk classification label is "risk"; if the first curve similarity, the second curve similarity, and the third curve similarity are all greater than or equal to the preset similarity threshold, the second risk classification label is "normal".

[0071] b5. Based on the first deviation stage, the second deviation stage, and the third deviation stage, determine the first abnormal probability of each process stage of the current workpiece tightening.

[0072] For example, based on the first deviation stage, the second deviation stage, and the third deviation stage, the number of times each process stage is judged as a deviation stage can be counted; and the ratio of the number of times each process stage is judged as a deviation stage to the number of curve types used for deviation stage judgment can be used as the anomaly probability of that process stage. For example, assuming there are a total of 3 process stages, including the tooth finding stage, the bonding stage, and the final tightening stage, and the number of curve types used for deviation stage judgment is 3 (i.e., torque curve, angle curve, torque-angle coupling curve), the first deviation stage is the tooth finding stage, the second deviation stage is the tooth finding stage and the bonding stage, and the third deviation stage is the tooth finding stage, then the number of times the tooth finding stage, the bonding stage, and the final tightening stage are judged as deviation stages are 3, 1, and 0, respectively. Therefore, the anomaly probabilities of the tooth finding stage, the bonding stage, and the final tightening stage are 3 / 3, 1 / 3, and 0 / 3, respectively.

[0073] For example, the third verification result is determined through the following steps c1~c2: c1. Based on the first tightening process features and the first context process features, a combined encoding is performed to obtain the first multi-dimensional feature vector of the workpiece being tightened.

[0074] c2. Using the feature classification model, classify based on the first multi-dimensional feature vector and output the third verification result.

[0075] The third verification result includes a second risk score, a third risk classification label, and the probability and / or quality level of each candidate anomaly type. Candidate anomaly types include at least one of the following: under-tightening, over-tightening, stripped threads, floating threads, missed tightening, fit abnormalities, and tool abnormalities.

[0076] In some embodiments, the third verification result includes a second risk score of the workpiece being tightened, which can be obtained by prediction based on the first multi-dimensional feature vector using a feature classification model.

[0077] In some embodiments, the third verification result also includes a third risk classification label, which is classified based on the second risk score of the workpiece being tightened. If the second risk score is greater than a preset risk score threshold, the third risk classification label is risky; if the second risk score is less than or equal to the preset risk score threshold, the third risk classification label is normal.

[0078] The third risk classification label is a classification label determined by a similarity model to indicate whether there is a risk in the workpiece being tightened.

[0079] In some embodiments, the third verification result may further include the anomaly probability of the candidate anomaly type of the workpiece being tightened. This anomaly probability is obtained by predicting the candidate anomaly type of the workpiece being tightened based on the first multi-dimensional feature vector using a feature classification model. Thus, anomaly type identification can be achieved through the feature classification model, thereby uncovering the attribution of the anomaly in the workpiece being tightened.

[0080] In some embodiments, the third verification result may also include a quality level, such as a quality level divided into four levels: A, B, C, and D. The quality level of the workpiece being tightened can be obtained by classifying it based on the first multi-dimensional feature vector using a feature classification model, thereby classifying the quality level of the workpiece being tightened.

[0081] The specific network structure of the feature classification model can be set according to the actual business scenario requirements. For example, XGBoost, LightGBM, CatBoost and other network structures can be used.

[0082] For example, the fourth verification result is determined by the following step d1: d1. Based on the first temporal features, anomaly detection is performed using the temporal depth model to obtain the fourth verification result.

[0083] In some embodiments, the fourth verification result includes the second anomaly probability of each process stage. Anomaly prediction is performed based on the first temporal features using a time-series depth model, and the second anomaly probability of each process stage is output. Thus, the time-series depth model enables anomaly identification in the time-series process, improving the comprehensiveness of anomaly identification for this workpiece tightening operation.

[0084] In some embodiments, the fourth verification result includes a third risk score, which is generated by using a time-series depth model to predict anomalies based on the first time-series features and outputting the third risk score for the workpiece being tightened.

[0085] In some embodiments, the fourth verification result includes a fourth risk classification label. For example, it can be classified based on the third risk score of the workpiece being tightened. If the third risk score is greater than a preset risk score threshold, the fourth risk classification label is risky; if the third risk score is less than or equal to the preset risk score threshold, the fourth risk classification label is normal.

[0086] Among them, the fourth risk classification label is a classification label for whether there is a risk in the tightening of the workpiece, which is determined based on the time-series depth model.

[0087] The specific network structure of the temporal deep model can be set according to the actual business scenario requirements. For example, network structures such as ID-CNN, TCN, and CNN-LSTM can be used.

[0088] For example, the fifth verification result is determined through the following steps e1~e2: e1. Based on the first tightening process features and the first context process features, a combined encoding is performed to obtain the first multi-dimensional feature vector of the workpiece being tightened.

[0089] e2. Using the unsupervised anomaly detection model, anomaly detection is performed based on the first multi-dimensional feature vector and the first temporal feature, and the fifth verification result is output.

[0090] In some embodiments, the fifth verification result includes a fourth risk score, which indicates the risk score of the workpiece being tightened. The unsupervised anomaly detection model performs anomaly detection based on the first multi-dimensional feature vector and the first temporal feature, outputting the fourth risk score of the workpiece being tightened.

[0091] In some embodiments, the fifth verification result also includes a fifth risk classification label. For example, it can be classified based on the fourth risk score of the workpiece being tightened. If the fourth risk score is greater than a preset risk score threshold, the fifth risk classification label is risky; if the fourth risk score is less than or equal to the preset risk score threshold, the fifth risk classification label is normal.

[0092] The fifth risk classification label is a classification label for whether there is a risk in the tightening of the workpiece, determined by an unsupervised anomaly detection model.

[0093] The specific network structure of an unsupervised anomaly detection model can be set according to the actual business scenario requirements. For example, network structures such as Isolation Forest and AutoEncoder can be used. In this way, the unsupervised anomaly detection model can discover novel anomalies not covered by the training labels, reducing the problem of limited anomaly coverage by the training labels of similarity models, feature classification models, and temporal deep models. This allows for the discovery of unpredictable unknown anomalies, thereby improving the anomaly recognition rate of the final process inspection results.

[0094] Therefore, the decision is made by integrating the first verification result of the rule engine, the second verification result of the similarity model, the third verification result of the feature classification model, the fourth verification result of the time-series deep model, and the fifth verification result of the unsupervised anomaly detection model. This allows for compliance vetoes using the rule engine, the discovery of truly abnormal process stages using the similarity model, the discovery of the anomaly attribution for the tightened workpiece using the feature classification model, the identification of anomalies in the time-series process using the time-series deep model, and the discovery of unknown anomalies using the unsupervised anomaly detection model. This approach can improve the comprehensiveness and accuracy of anomaly identification for the tightened workpiece.

[0095] A4. Based on the first verification result, the second verification result, the third verification result, the fourth verification result, and the fifth verification result, a fusion decision is made to obtain the process inspection result of the workpiece being tightened.

[0096] The integrated decision-making process includes making a compliance decision on the workpiece being tightened based on the first verification result, and determining, based on the second verification result, the third verification result, the fourth verification result, and the fifth verification result, whether the workpiece being tightened has at least one of the following: risk level, abnormal stage, abnormal attribution, and process optimization suggestion information.

[0097] In some embodiments, step A4 may specifically include: A41. Based on the first verification result, conduct a compliance test on the workpiece being tightened.

[0098] In some embodiments, compliance checks are performed using the first verification result. For example, compliance checks are performed using the first risk classification label contained in the first verification result. If the first risk classification label is normal, the workpiece being tightened is determined to be a compliant product; if the first risk classification label is risky, the workpiece being tightened is determined to be a non-compliant product. In this way, compliance is vetoed using the rule engine.

[0099] In some embodiments, the first verification result and other verification results (such as at least one of the second, third, fourth, and fifth verification results) can be combined to perform compliance testing on the tightened workpiece. For example, the first verification result may include a first risk classification label, the second verification result may include a second risk classification label, the third verification result may include a third risk classification label, the fourth verification result may include a fourth risk classification label, and the fifth verification result may include a fifth risk classification label. If all of the first, second, third, fourth, and fifth risk classification labels are normal, the tightened workpiece is determined to be a compliant product. If any one of the first, second, third, fourth, and fifth risk classification labels is risky, the tightened workpiece is determined to be a non-compliant product.

[0100] A42. If the workpiece being tightened is a compliant product, then based on the second verification result, the third verification result, the fourth verification result, and the fifth verification result, determine whether there is any risk to the workpiece being tightened.

[0101] For example, the second verification result may include a first risk score, the third verification result may include a second risk score, the fourth verification result may include a third risk score, and the fifth verification result may include a fourth risk score. A weighted sum of these four risk scores can be used as the target risk score for the workpiece being tightened. If the target risk score is greater than a preset risk score threshold, the workpiece is considered to have a risk; if it is less than or equal to the threshold, it is considered to have no risk. Similarly, the second verification result may include a second risk classification label, a third risk classification label, a fourth risk classification label, and a fifth risk classification label. The risk label percentage (the ratio of the number of risk classification labels to the total number of risk classification labels) can be calculated based on these labels. If the percentage is greater than a preset threshold (e.g., 1 / 2), the workpiece is considered to have a risk; if it is less than or equal to the threshold, it is considered to have no risk.

[0102] 202. Based on the batch tightening samples and the preset standard template library, obtain the target sample drift characteristics of the target tightening device.

[0103] The preset standard template library includes multiple standard samples. The process inspection results of the standard samples are determined based on their own tightening-related data (including process context information and tightening process data), and are used as a benchmark for comparing the drift characteristics of the target sample and the drift characteristics of the target equipment.

[0104] There are several ways to implement step 202, including, for example: (1) In some embodiments, the target sample drift feature is a first sample drift feature, a second sample drift feature, and an overall sample drift feature. In this case, such as Figure 3 As shown, Figure 3 This is a schematic flowchart of an embodiment of step 202 provided in this application. Step 202 may specifically include the following steps 2021A to 2024A: 2021A. According to the preset division strategy, the batch tightening samples are divided to obtain the first division sample data and the second division sample data.

[0105] In this configuration, the tightening time of the second segmented sample data is later than that of the first segmented sample data. The first segmented sample data reflects the early sample conditions, while the second segmented sample data reflects the sample change trend over a long period. This allows for the capture of target sample drift characteristics at different time periods for health prediction, thereby identifying whether the target tightening equipment is aging or experiencing other equipment malfunctions over time.

[0106] The first sample drift feature and the second sample drift feature are used to capture the drift trend change of the target device.

[0107] The preset partitioning strategy refers to a pre-defined strategy for splitting batch tightening samples according to the time dimension, such as a proportional partitioning strategy, a sliding window partitioning strategy, or an equal-time-window partitioning strategy. Taking the proportional partitioning strategy as an example, it refers to a pre-defined proportional rule for splitting batch tightening samples according to the time dimension, used to distinguish sample data from different time periods. For example, a preset ratio of 7:3 divides 70% of the early tightening samples in the batch into the first partition and 30% into the second partition; or a preset ratio of 1:1 divides the batch tightening samples into two parts equally according to the tightening time. The first sample drift feature and the second sample drift feature extracted based on the first and second partition sample data can then be used to capture the drift trend changes of the target device.

[0108] The first segment of sample data refers to the portion of sample data separated from the batch tightening samples according to a preset segmentation strategy, which consists of samples with relatively earlier tightening times. For example, if there are 100 batch tightening samples, divided in a 7:3 ratio, the sample data with the first 70 tightening times constitutes the first segment of sample data.

[0109] The second segment of sample data refers to the portion of sample data separated from the batch tightening samples according to a preset segmentation strategy, where the tightening time is later than that of the first segment of sample data. For example, in the aforementioned 100 batch tightening samples, the sample data with the last 30 tightening times constitutes the second segment of sample data.

[0110] For example, firstly, a preset proportional division strategy is retrieved; then, the tightening timestamp of each workpiece sample in the batch tightening samples is extracted, and all workpiece samples are sorted from earliest to latest according to the tightening time; finally, according to the preset division strategy, the sorted samples are split into two parts, with the part with the earlier tightening time being the first division sample data and the part with the later tightening time being the second division sample data, so that the time dimension of the two parts of the samples is clearly distinguished, providing a data basis for subsequently capturing the drift characteristics of target samples in different time periods. For example, the preset division strategy is 6:4 (60% of early samples and 40% of later samples); 200 batch tightening samples of the target tightening equipment (gun number G002) are obtained from the preset sample trend library, and each workpiece sample contains a tightening timestamp; the 200 samples are sorted from early to late according to the tightening time, and the first 120 (60%) samples are taken as the first division sample data (tightening time range: April 10, 2026, 9:00-10:30), and the last 80 (40%) samples are taken as the second division sample data (tightening time range: April 10, 2026, 10:31-12:00); through this division, the tightening samples of the target tightening equipment in different time periods can be clearly distinguished, which is convenient for comparative analysis of sample drift differences.

[0111] 2022A. Based on the first segmented sample data and the preset standard template library, obtain the first sample drift feature.

[0112] The first sample drift feature refers to the offset feature of the first segmented sample data relative to the standard template, obtained by comparing the first segmented sample data with the preset standard template library. It is used to reflect the sample offset in the early working stage of the target tightening equipment. For example, the offset of the center of the torque-time curve of the first segmented sample data relative to the center of the torque-time curve of the standard template is the first curve offset in the first sample drift feature.

[0113] For example, such as Figure 4 As shown, Figure 4 This is a schematic flowchart of an embodiment of step 2022A provided in this application. Step 2022A may specifically include the following steps B1 to B3: B1. Obtain the first central feature of the first segmented sample data.

[0114] The first central feature refers to the central feature determined based on the first segmented sample data. It is the core representation of the first segmented sample data in dimensions such as curve, process stage, stiffness, and ripple, including but not limited to at least one of the first curve center, first stage center, first stiffness center, and first ripple center. In this embodiment, the curve corresponding to the curve center is at least one of the torque-time curve, angle-time curve, speed-time curve, and torque-angle curve. For example, the curve corresponding to the first curve center is at least one of the torque-time curve, angle-time curve, speed-time curve, and torque-angle curve.

[0115] The tightening process data includes torque-time curves, angle-time curves, and speed-time curves. The first curve center refers to the central representation determined based on the first segmented sample data, reflecting the core characteristics of the tightening process curves of all workpiece samples within the first segmented sample data. By retrieving the tightening process data (torque-time curves, angle-time curves, and speed-time curves) of all workpiece samples in the first segmented sample data, statistical analysis (such as calculating the average or median) is performed on curves of the same type (e.g., the torque-time curves of all workpiece samples). The average or median curve of this type is then fitted, and this average or median curve is the first curve center for that type. Integrating the average or median curves of all types forms the first curve center. (For example, taking the torque-time curve as an example, the average curve of the torque-time curves of all workpiece samples in the first segmented sample data is the first curve center.) The first-stage center refers to the central representation determined based on the first segmented sample data, reflecting the core characteristics of each tightening process stage of all workpiece samples in the first segmented sample data. By retrieving the tightening stage data of all workpiece samples in the first segmented sample data, the key tightening process stages of each workpiece sample (such as the tooth finding stage, the fitting stage, and the final tightening stage) are first identified, and the key parameters of each process stage (including stage start time, stage end time, stage start torque, stage end torque, stage duration, etc.) are extracted; statistical analysis (such as calculating the average value and median value) is performed on the same type of key parameters of the same process stage to obtain the average value or median value of each key parameter of the process stage; the average value or median value of the key parameters of all tightening key process stages is integrated to form the first-stage center.

[0116] The first stiffness center refers to the central representation determined based on the first segmented sample data, reflecting the core characteristics of the torque-angle coupling stiffness of all workpiece samples in the first segmented sample data during the tightening process. It is a key parameter for quantifying the stiffness characteristics during the tightening process and is used to capture stiffness changes during sample drift. By retrieving the tightening process data (torque data, angle data) of all workpiece samples in the first segmented sample data, the torque-angle coupling stiffness during the tightening process is calculated for each workpiece sample (the calculation method is the ratio of the torque change to the corresponding angle change during the tightening process); statistical analysis is performed on the torque-angle coupling stiffness values ​​of all workpiece samples (such as calculating the average or median), and the obtained average or median is the first stiffness center.

[0117] The first ripple center refers to the central representation determined based on the first segmented sample data, reflecting the core characteristics of the torque curve ripple during the tightening process of all workpiece samples in the first segmented sample data. It focuses on the subtle fluctuations of the torque curve and can be used to capture sample drift corresponding to hidden anomalies that are difficult to detect by traditional methods. By retrieving the tightening process data (torque-time curve) of all workpiece samples in the first segmented sample data, the ripple amplitude of the torque-time curve of each workpiece sample is extracted (calculated as the difference between the peak and valley values ​​in the torque curve); statistical analysis is performed on the torque curve ripple amplitude of all workpiece samples (such as calculating the average or median), and the obtained average or median is the first ripple center.

[0118] For example, the tightening process data (torque-time curve, angle-time curve, and speed-time curve) and stage data of all workpiece samples in the first segmentation sample data are retrieved; statistical analysis (such as calculating the average and median) is performed on the data of the same dimension to obtain the central characterization of that dimension; for the curve dimension, the average curve of the corresponding curves of all workpiece samples is calculated as the curve center; for the process stage dimension, the time nodes of the tooth finding stage, fitting stage, and final tightening stage of all workpiece samples are extracted, and the average duration and average starting / ending torque of each process stage are calculated as the stage center; for the stiffness dimension, the average value of the torque-angle coupling stiffness of all workpiece samples is calculated as the stiffness center; for the ripple dimension, the average value of the ripple amplitude of the torque curve of all workpiece samples is calculated as the ripple center; and the above one or more central characterizations are integrated to obtain the first central feature. For example, the first set of sample data consists of 120 early samples, each containing torque-time curves, angle-time curves, speed-time curves, and data for each stage. The torque-time curves of the 120 samples are processed to calculate the average torque at each time point, and an average torque-time curve is fitted to obtain the torque-time curve center in the first curve center. The start and end times of the bonding stage of the 120 samples are extracted, and the average start time (e.g., 2.5 seconds), average end time (e.g., 4.8 seconds), and average starting torque (e.g., 5 N·m) of the bonding stage are calculated as the first stage center. The torque-angle coupling stiffness (torque change / angle change) of the 120 samples is calculated, and the average value (e.g., 8 N·m / deg) is taken as the first stiffness center. The ripple amplitude (difference between peak and trough values) of the torque curves of the 120 samples is calculated, and the average value (e.g., 0.3 N·m) is taken as the first ripple center. The above four center characteristics are integrated as the first center feature.

[0119] B2. Obtain the template center features of the preset standard template library.

[0120] The template center feature refers to the central feature determined based on a preset standard template library. It is the core representation of the standard sample in dimensions such as curve, stage, stiffness, and ripple, including but not limited to at least one of the template curve center, template stage center, template stiffness center, and template ripple center. For example, the average curve of the angle-time curves of all standard samples in the preset standard template library is the template curve center.

[0121] Among them, the template center feature is a center feature determined based on a preset standard template library, including but not limited to at least one of the template curve center, template stage center, template stiffness center, and template ripple center.

[0122] Step B2 is implemented in a similar manner to step B1. For details, please refer to the relevant explanations above, which will not be repeated here.

[0123] B3. Based on the first central feature and the template central feature, determine the first sample drift feature.

[0124] The first center feature includes at least one of the following: first curve center, first stage center, first stiffness center, and first ripple center; the template center feature includes at least one of the following: template curve center, template stage center, template stiffness center, and template ripple center; and the first sample drift feature includes at least one of the following: first curve offset, first stage offset, first stiffness offset, and first ripple offset. Taking "the first sample drift feature includes the first curve offset, first stage offset, first stiffness offset, and first ripple offset" as an example, step B3 may specifically include: determining the first curve offset based on the first curve center and the template curve center; determining the first stage offset based on the first stage center and the template stage center; determining the first stiffness offset based on the first stiffness center and the template stiffness center; and determining the first ripple offset based on the first ripple center and the template ripple center. Thus, by obtaining the first central features of the first segmented sample data, such as the first curve center, the first stage center, the first stiffness center, and the first ripple center, and using them to determine the first sample drift features, it is possible to perform feature processing on batch tightening samples, construct the long-term change trend of tightening equipment, and enable the health prediction results to comprehensively consider the long-term change trend and explore long-term evolution trends such as aging, thereby improving the accuracy of health prediction.

[0125] For example, the first center feature is matched with the template center feature in terms of dimension correspondence (e.g., the first curve center corresponds to the template curve center, and the first stiffness center corresponds to the template stiffness center); a preset deviation calculation algorithm (e.g., the difference method or the ratio method) is used to calculate the deviation value of the corresponding dimension respectively; the deviation values ​​of each dimension are standardized to remove the influence of the dimension; the processed deviation values ​​of each dimension are integrated to form the first sample drift feature. In this way, the first sample drift feature can quantitatively reflect the degree of deviation of the first partitioned sample data from the standard template. The larger the deviation value, the more obvious the drift of the early sample.

[0126] For example, the center of the torque time curve in the first center feature is compared with the center of the template curve, and the torque difference between the two at each time point is calculated. The maximum difference (e.g., 0.5 N·m) is taken as the first curve offset. The first stiffness center (e.g., 8 N·m / deg) is compared with the template stiffness center (e.g., 7.8 N·m / deg), and the difference (e.g., 0.2 N·m / deg) is calculated as the first stiffness offset. The first ripple center (e.g., 0.3 N·m) is compared with the template ripple center (e.g., 0.25 N·m), and the difference (e.g., 0.05 N·m) is calculated as the first ripple offset. The first curve offset, the first stiffness offset, and the first ripple offset are integrated to obtain the first sample drift feature, which is used to reflect the drift of early samples relative to the standard template.

[0127] 2023A. Based on the second segmentation sample data and the preset standard template library, obtain the second sample drift feature.

[0128] The second sample drift feature refers to the offset feature of the second sample data relative to the standard template obtained by comparing the second segmented sample data with the preset standard template library. It is used to reflect the sample offset in the later working stage of the target tightening equipment. Its representation form is consistent with the first sample drift feature (such as curve offset, stage offset, etc.).

[0129] For example, step 2023A may specifically include the following steps C1 to C3: C1. Obtain the second central feature of the second partitioned sample data.

[0130] The second center feature refers to the center feature determined based on the second partitioned sample data. It is the core representation of the second partitioned sample data in dimensions such as curve, process stage, stiffness, and ripple, including but not limited to at least one of the second curve center, second stage center, second stiffness center, and second ripple center.

[0131] The second curve center refers to the central representation determined based on the second segmentation sample data, reflecting the core characteristics of the tightening process curves of all workpiece samples in the second segmentation sample data. The curve corresponding to the second curve center is at least one of the following: torque-time curve, angle-time curve, speed-time curve, and torque-angle curve.

[0132] The second-stage center refers to the center representation determined based on the second-division sample data, which reflects the core characteristics of each process stage of tightening of all workpiece samples in the second-division sample data.

[0133] The second stiffness center refers to the central representation determined based on the second division sample data, which reflects the core characteristics of torque-angle coupling stiffness during the tightening process of all workpiece samples in the second division sample data. It is a key parameter for quantifying stiffness characteristics during the tightening process and is used to capture stiffness changes during sample drift.

[0134] The second ripple center refers to the central representation of the core characteristics of the torque curve ripple during the tightening process of all workpiece samples in the second division sample data, which is determined based on the second division sample data. It can focus on the subtle fluctuations of the torque curve and can be used to capture the sample drift corresponding to hidden anomalies that are difficult to detect by traditional methods.

[0135] It is understandable that the second curve center, second stage center, second stiffness center, and second ripple center are all components of the second center feature. Their implementation methods are similar to those of the first curve center, first stage center, first stiffness center, and first ripple center. They are all obtained by performing statistical analysis (such as calculating the average and median) on the data corresponding to the second division sample data. This data is then compared with the template center features of the preset standard template library to determine the offset of each dimension of the second sample drift feature. This reflects the sample offset in the later working stage of the target tightening equipment and provides data support for the health prediction of the target tightening equipment.

[0136] C2. Obtain the template center features of the preset standard template library.

[0137] C3. Based on the second central feature and the template central feature, determine the second sample drift feature.

[0138] Taking "the second sample drift feature includes a second curve offset, a second stage offset, a second stiffness offset, and a second ripple offset" as an example, step C3 may specifically include: determining the second curve offset based on the center of the second curve and the center of the template curve; determining the second stage offset based on the center of the second stage and the center of the template stage; determining the second stiffness offset based on the center of the second stiffness and the center of the template stiffness; and determining the second ripple offset based on the center of the second ripple and the center of the template ripple.

[0139] The implementation of steps C1 to C3 is similar to that of steps B1 to B3. For details, please refer to the relevant explanations above, which will not be repeated here.

[0140] 2024A. Based on the batch tightening samples and the preset standard template library, obtain the overall sample drift characteristics.

[0141] Among them, the overall sample drift feature refers to the overall offset feature of the batch tightening sample relative to the standard template, which is obtained by comparing all batch tightening samples with the preset standard template library. It is used to reflect the sample offset of the target tightening equipment in the overall working stage.

[0142] The implementation of step 2024A is similar to that of steps 2021B to 2023B. For details, please refer to the relevant sections. It will not be repeated here.

[0143] (2) In some embodiments, the target sample drift feature is the overall sample drift feature. In this case, step 202 may specifically include the following steps 2021B~2023B: 2021B. Obtain the overall center features of the batch tightened samples.

[0144] The overall center feature is a center feature determined based on a batch of tightening samples, including but not limited to at least one of the following: overall curve center, overall stage center, overall stiffness center, and overall ripple center. The curve corresponding to the overall curve center is at least one of the following: torque-time curve, angle-time curve, speed-time curve, and torque-angle curve.

[0145] Among them, the overall curve center refers to the central representation that reflects the core characteristics of the tightening process curve of all workpiece samples in the batch tightening sample, determined based on the batch tightening sample.

[0146] Among them, the overall stage center refers to the central representation determined based on the batch tightening samples, which reflects the core characteristics of each process stage of tightening of all workpiece samples in the batch tightening samples.

[0147] Among them, the overall stiffness center refers to the central characterization determined based on the batch tightening samples, which reflects the core characteristics of the torque-angle coupling stiffness of all workpiece samples in the batch tightening samples during the tightening process.

[0148] Among them, the overall ripple center refers to the central representation that reflects the core characteristics of the torque curve ripple characteristics of all workpiece samples in the batch tightening process, determined based on the batch tightening samples.

[0149] For example, the overall curve center features, overall stage center features, overall stiffness center features, and overall ripple center features of a batch of tightening samples are obtained.

[0150] 2022B. Obtain the template center features of the preset standard template library.

[0151] For example, the template curve center feature, template stage center feature, template stiffness center feature, and template ripple center feature of the preset standard template library are obtained.

[0152] 2023B. Based on the overall center features and the template center features, determine the overall sample drift features.

[0153] Among them, the overall center feature includes at least one of the overall curve center, overall stage center, overall stiffness center, and overall ripple center; the template center feature includes at least one of the template curve center, template stage center, template stiffness center, and template ripple center; and the overall sample drift feature includes at least one of the overall curve offset, overall stage offset, overall stiffness offset, and overall ripple offset.

[0154] Taking "the overall sample drift characteristics include overall curve offset, overall stage offset, overall stiffness offset, and overall ripple offset" as an example, step 2023B may specifically include: determining the overall curve offset based on the overall curve center and the template curve center; determining the overall stage offset based on the overall stage center and the template stage center; determining the overall stiffness offset based on the overall stiffness center and the template stiffness center; determining the overall ripple offset based on the overall ripple center and the template ripple center; determining the overall curve offset based on the overall curve center and the template curve center; determining the overall stage offset based on the overall stage center and the template stage center; determining the overall stiffness offset based on the overall stiffness center and the template stiffness center; and determining the overall ripple offset based on the overall ripple center and the template ripple center.

[0155] The implementation of step 2023B is similar to that of step 2022A. For details, please refer to the relevant explanations above. It will not be repeated here.

[0156] (3) In some embodiments, the target sample drift feature is a first sample drift feature and a second sample drift feature. In this case, step 202 may specifically include the following steps 2021C~2023C: 2021C. According to the preset division strategy, the batch tightening samples are divided to obtain the first division sample data and the second division sample data.

[0157] 2022C. Based on the first segmented sample data and the preset standard template library, obtain the first sample drift feature.

[0158] 2023C. Based on the second segmentation sample data and the preset standard template library, obtain the second sample drift feature.

[0159] The implementation of steps 2021C to 2023C is similar to that of steps 2021A to 2023A. For details, please refer to the relevant explanations above. They will not be repeated here.

[0160] 203. Based on the first normal sample of the batch tightening samples and the preset standard template library, obtain the target device drift characteristics of the target tightening device.

[0161] The first normal sample refers to a workpiece sample in the batch tightening samples whose process inspection results are normal. Specifically, the process inspection results in steps A1 to A4 can be used to determine whether a workpiece sample is normal. If steps A1 to A4 determine that the workpiece being tightened is a compliant product and there is no risk, then the workpiece being tightened is a normal workpiece sample. The first normal sample can be used to characterize the inherent offset of the target tightening equipment. Based on the first normal sample and a preset standard template library, the target equipment drift characteristics of the target tightening equipment are obtained, thereby forming a comparison with the target sample drift characteristics (constructed based on abnormal samples).

[0162] For example, such as Figure 5 As shown, Figure 5 This is a schematic flowchart of an embodiment of step 203 provided in this application. Step 203 may specifically include the following steps 2031 to 2034: 2031. Based on the batch tightening samples, determine the first normal sample set.

[0163] The first normal sample set refers to the set of the first normal samples.

[0164] 2032. Obtain the third central feature of the first normal sample set.

[0165] The third center feature includes at least one of the following: third curve center, third stage center, third stiffness center, and third ripple center. The curve corresponding to the third curve center is at least one of the following: torque-time curve, angle-time curve, speed-time curve, and torque-angle curve.

[0166] The third center feature refers to the center feature determined based on the first normal sample set. It is the core representation of the first normal sample set in dimensions such as curve, process stage, stiffness, and ripple, including but not limited to at least one of the third curve center, third stage center, third stiffness center, and third ripple center.

[0167] The third curve center refers to the central representation of the core characteristics of the tightening process curves of all workpiece samples in the first normal sample set, determined based on the first normal sample set. By retrieving tightening process data (torque-time curve, angle-time curve, and speed-time curve) of all workpiece samples in the first normal sample set, statistical analysis (such as calculating the average or median) is performed on curves of the same type (e.g., the torque-time curve of all workpiece samples). The average or median curve of this type is then fitted, and this average or median curve is the corresponding third curve center. Integrating the average or median curves of all types forms the third curve center. (For example, taking the torque-time curve as an example, the average curve of the torque-time curves of all workpiece samples in the first normal sample set is the third curve center.) The third-stage center refers to the central representation determined based on the first normal sample set, reflecting the core characteristics of each tightening process stage of all workpiece samples in the first normal sample set. By retrieving the tightening stage data of all workpiece samples in the first normal sample set, the key tightening process stages of each workpiece sample (such as the tooth finding stage, the fitting stage, and the final tightening stage) are first identified, and the key parameters of each process stage (including stage start time, stage end time, stage start torque, stage end torque, stage duration, etc.) are extracted. Statistical analysis (such as calculating the average value and median) is performed on the same type of key parameters of the same process stage to obtain the average value or median of each key parameter of the process stage. The average value or median of the key parameters of all tightening key process stages is integrated to form the third-stage center.

[0168] The third stiffness center refers to the central representation determined based on the first normal sample set, reflecting the core characteristics of the torque-angle coupling stiffness of all workpiece samples in the first normal sample set during the tightening process. It is a key parameter for quantifying the stiffness characteristics during the tightening process and is used to capture stiffness changes during sample drift. By retrieving the tightening process data (torque data, angle data) of all workpiece samples in the first normal sample set, the torque-angle coupling stiffness during the tightening process is calculated for each workpiece sample (the calculation method is the ratio of the torque change to the corresponding angle change during the tightening process); statistical analysis is performed on the torque-angle coupling stiffness values ​​of all workpiece samples (such as calculating the average or median), and the obtained average or median is the third stiffness center.

[0169] The third ripple center refers to the central representation determined based on the first normal sample set, reflecting the core characteristics of the torque curve ripple properties during the tightening process of all workpiece samples in the first normal sample set. By retrieving the tightening process data (torque-time curves) of all workpiece samples in the first normal sample set, the ripple amplitude of each workpiece sample's torque-time curve is extracted (calculated as the difference between the peak and trough values ​​in the torque curve). Statistical analysis (such as calculating the average or median) is performed on the torque curve ripple amplitudes of all workpiece samples; the resulting average or median is the third ripple center.

[0170] The specific implementation of step 2032 is similar to that of step B1. For details, please refer to the relevant sections. It will not be repeated here.

[0171] 2033. Obtain the template center features of the preset standard template library.

[0172] The implementation of step 2033 is similar to that of step B2. For details, please refer to the relevant explanations above. It will not be repeated here.

[0173] 2034. Based on the third center feature and the template center feature, determine the target equipment drift feature of the target tightening device.

[0174] The third center feature includes at least one of the third curve center, third stage center, third stiffness center, and third ripple center; the template center feature includes at least one of the template curve center, template stage center, template stiffness center, and template ripple center; and the target equipment drift feature includes at least one of the equipment curve offset, equipment stage offset, equipment stiffness offset, and equipment ripple offset. Taking "target equipment drift features include equipment curve offset, equipment stage offset, equipment stiffness offset, and equipment ripple offset" as an example, step 2034 may specifically include: determining the equipment curve offset based on the third curve center and the template curve center; determining the equipment stage offset based on the third stage center and the template stage center; determining the equipment stiffness offset based on the third stiffness center and the template stiffness center; and determining the equipment ripple offset based on the third ripple center and the template ripple center.

[0175] 204. When the number of abnormal samples in the batch tightening samples is greater than a preset quantity threshold, or when the proportion of abnormal samples in the batch tightening samples is greater than a preset proportion threshold, a health prediction is performed based on the target sample drift characteristics and the target equipment drift characteristics to obtain the health prediction result of the target tightening equipment.

[0176] The abnormal sample refers to a workpiece sample whose process inspection result is abnormal. The health prediction result is used to indicate whether there is an abnormality in at least one of the bit, gun body, deceleration mechanism and sensor of the target tightening equipment. The health prediction result includes at least one of the abnormal component type, abnormal confidence level and abnormal level.

[0177] In some embodiments, workpiece samples whose process inspection results indicate they are non-conforming products can be designated as abnormal samples, while workpiece samples whose process inspection results indicate other conditions (such as being compliant products) can be designated as normal samples. In some embodiments, workpiece samples whose process inspection results indicate they are non-conforming products or workpiece samples that pose a risk can also be designated as abnormal samples.

[0178] The preset quantity threshold is a pre-set value used to determine whether the number of abnormal samples in a batch of tightening samples has reached the critical value required for equipment health prediction. The specific value of the preset quantity threshold can be set according to the actual business scenario requirements. There is no restriction on the specific value of the preset quantity threshold here.

[0179] For example, firstly, the number of all abnormal samples in the batch tightening samples is counted, and the counted number of abnormal samples is compared with a preset threshold. If the number of abnormal samples is not greater than the preset threshold, no further health prediction steps are needed, and the target tightening equipment can be determined to be in normal working condition, or the number of abnormal samples is small and negligible, and the target tightening equipment can continue to be used for normal production. If the number of abnormal samples is greater than the preset threshold, the health prediction process is triggered, and the previously acquired target sample drift characteristics and target equipment drift characteristics are retrieved for subsequent health prediction using a preset health prediction algorithm. Alternatively, the proportion of all abnormal samples in the batch tightening samples is counted, and the counted number of abnormal samples is compared with a preset proportion threshold. If the proportion of abnormal samples is not greater than the preset proportion threshold, no further health prediction steps are needed, and the target tightening equipment can be determined to be in normal working condition, or the proportion of abnormal samples is low and negligible, and the target tightening equipment can continue to be used for normal production. If the proportion of abnormal samples is greater than the preset proportion threshold, the health prediction process is triggered, and the previously acquired target sample drift characteristics and target equipment drift characteristics are retrieved for subsequent health prediction using a preset health prediction algorithm.

[0180] Then, after obtaining the target sample drift features and the target device drift features, the target sample drift features and the target device drift features are standardized to remove the influence of dimensions and then input into the preset health prediction algorithm, and the health prediction algorithm is used to make health predictions.

[0181] In some embodiments, during health prediction, the processed target sample drift features are first compared with the target equipment drift features to obtain comparison features. These comparison features are then combined with the correlation between the health prediction algorithm and component anomalies (e.g., the health prediction algorithm is trained based on historical sample drift features, historical equipment drift features, and historical fault labels, and can establish a mapping relationship between comparison features and anomalies in bits, gun bodies, deceleration mechanisms, and sensors). The algorithm then determines the equipment component corresponding to the root cause of the anomaly and finally outputs the health prediction result. For example, the health prediction algorithm determines and outputs the abnormal component type based on the differences, rates of change, and coupling correlations between the comparison features (such as target sample drift features and target equipment drift features) in four dimensions: curve offset, stage offset, stiffness offset, and ripple offset. Based on the deviation of the comparison features (such as target sample drift features and target equipment drift features) from the standard range and the similarity to historical faults, the algorithm calculates and outputs the anomaly confidence level. Combining the anomaly confidence level with the proportion of abnormal samples, the algorithm maps the anomaly level according to a preset grading rule. Finally, the health prediction result is output, including at least one of the abnormal component type, anomaly confidence level, and anomaly level.

[0182] In some embodiments, the target sample drift features include a first sample drift feature, a second sample drift feature, and an overall sample drift feature. When performing health prediction, a first comparison feature between the first sample drift feature and the target device drift feature, a second comparison feature between the second sample drift feature and the target device drift feature, and a third comparison feature between the overall sample drift feature and the target device drift feature are obtained. Based on the first comparison feature, the second comparison feature, and the third comparison feature, the abnormal component type of the target tightening device is determined. For example, the first curve offset, first stage offset, first stiffness offset, and first ripple offset in the first sample drift features are obtained, and the differences are calculated with the equipment curve offset, equipment stage offset, equipment stiffness offset, and equipment ripple offset in the target equipment drift features, respectively, to obtain the first comparison feature; the second curve offset, second stage offset, second stiffness offset, and second ripple offset in the second sample drift features are obtained, and the differences are calculated with the equipment curve offset, equipment stage offset, equipment stiffness offset, and equipment ripple offset in the target equipment drift features, respectively, to obtain the second comparison feature; the overall curve offset, overall stage offset, overall stiffness offset, and overall ripple offset in the overall sample drift features are obtained, and the differences are calculated with the equipment curve offset, equipment stage offset, equipment stiffness offset, and equipment ripple offset in the target equipment drift features, respectively, to obtain the third comparison feature; when the first When the first and second comparison features show no significant deviations in curve offset and ripple offset dimensions, but the second comparison feature shows a significant increase in curve offset and ripple offset dimensions, and the third comparison feature shows an overall deviation, it indicates that the drift trend of the target tightening device is continuously expanding, corresponding to progressive wear of the bit. When the first and second comparison features show continuous deviations in stage offset dimension, and the third comparison feature shows an overall deviation in stage offset dimension, it indicates that the drift trend of the target tightening device is stable, corresponding to an abnormality in the gun body transmission mechanism. When the first and second comparison features show a gradual increase in stiffness offset dimension, and the third comparison feature shows an overall deviation in stiffness offset dimension, it indicates that the drift trend of the target tightening device is gradually deteriorating, corresponding to aging of the deceleration mechanism. When the first and second comparison features show irregular fluctuations in each dimension, and the third comparison feature shows no significant overall deviation, it indicates that the target tightening device has no obvious trend of drift, corresponding to abnormal sensor detection accuracy. By acquiring the first sample drift feature, the second sample drift feature, and the overall sample drift feature, and comparing them with the target device drift feature respectively, the first comparison feature, the second comparison feature, and the third comparison feature are obtained. The differences are quantified from multiple dimensions, including time series and overall. Based on the differential change pattern of the three types of comparison features, the abnormal component type is determined, thereby improving the accuracy of abnormal component type determination.

[0183] The health prediction algorithm can be obtained through supervised learning based on historical workpiece sample data, historical equipment status data, and historical fault data. The historical workpiece sample data is a set of workpiece samples used for training the health prediction algorithm, and it is not limited to workpiece samples in the preset sample trend library. Specifically, during the training of the health prediction algorithm, the training sample drift features corresponding to the historical workpiece sample data and the training equipment drift features corresponding to the normal samples in the historical workpiece sample data are used as input data. The actual fault conditions of the historical equipment (i.e., which one or more of the bit, gun body, deceleration mechanism, and sensor are abnormal) are used as labels. The algorithm is trained using algorithms such as Gradient Boosting Tree (GBDT), Random Forest, or Neural Network. During the training process, the model parameters are continuously adjusted so that the model can accurately learn the correlation between the training sample drift features, the training equipment drift features, and the abnormality of equipment components, until the model prediction accuracy reaches the preset standard, the algorithm learning is completed, and it is solidified as the preset health prediction algorithm. At the same time, the algorithm supports online updates and can continuously optimize the model parameters based on newly added workpiece sample data and equipment fault data to improve prediction accuracy. For example, the correlation between learning training sample drift characteristics, training equipment drift characteristics, and equipment component anomalies (such as bit, gun body, deceleration mechanism, and sensor anomalies) is illustrated below: <1> If the curve offset (such as torque-time curve or angle-time curve offset) exceeds the preset range and the ripple offset increases abnormally, it corresponds to bit wear (bit wear will cause unstable tightening torque and angle, resulting in curve offset and abnormal ripple). <2> Abnormal stage offset (especially the start / end time and torque offset of the final tightening stage) corresponds to a gun body malfunction (poor gun body transmission will cause the parameters of each tightening stage to deviate from the standard). <3> If the stiffness deviation exceeds the preset range, it corresponds to the aging of the reduction mechanism (the aging of the reduction mechanism will cause a decrease in the torque angle coupling stiffness, resulting in stiffness deviation). <4> All curve offsets showed no obvious abnormalities, but the data fluctuated frequently and the ripple offset was abnormal, which corresponds to abnormal sensor detection accuracy (sensor failure can cause detection data distortion and irregular fluctuations).

[0184] Furthermore, the tightening equipment health prediction method further includes: obtaining trend library center features based on a second normal sample from the preset sample trend library; determining template drift features of the preset standard template library based on the trend library center features and the template center features; and outputting template reconstruction suggestions based on the template drift features. For example, the specific implementation can be as follows: The first step is to obtain the trend library center features: First, select all workpiece samples with normal process inspection results (i.e., the second normal samples) from the preset sample trend library to form the second normal sample set; then, in the implementation method of "obtaining the first center features", retrieve the tightening process data (torque-time curve, angle-time curve, speed-time curve) and tightening stage data of all workpiece samples in the second normal sample set, perform statistical analysis on the data of the same dimension (such as calculating the average and median), and obtain at least one of the trend library curve center, trend library stage center, trend library stiffness center, and trend library ripple center respectively. After integration, the trend library center features are formed to reflect the overall core characteristics of the normal samples in the preset sample trend library.

[0185] The second step is to determine the template drift characteristics: retrieve the template center features from the preset standard template library (corresponding to the dimensions of the trend library center features, such as the template curve center corresponding to the trend library curve center, and the template stiffness center corresponding to the trend library stiffness center); use preset deviation calculation algorithms (such as the difference method and the ratio method) to calculate the deviation values ​​of the corresponding dimensions of the trend library center features and the template center features respectively; standardize the deviation values ​​of each dimension to remove the influence of dimensions; integrate the processed deviation values ​​of each dimension to form the template drift characteristics, which quantitatively reflect the degree of deviation of the preset standard template library from the normal samples of the current preset sample trend library.

[0186] The third step is to output template reconstruction suggestions: A preset template drift threshold (set according to the actual business scenario and equipment accuracy requirements, used to determine whether the template drift has reached the point where reconstruction is necessary) is established. The offset of each dimension of the template drift feature is compared with the preset template drift threshold. If the offset of the template drift feature does not exceed the preset threshold, it indicates that the standard template can still adapt to the current sample trend, and a suggestion of "no need to rebuild the template, it is recommended to monitor the template drift regularly" can be output. If the offset exceeds the preset threshold, it indicates that the standard template has deviated from the current sample trend and cannot be used as a reliable comparison benchmark. Specific template reconstruction suggestions need to be output, specifying the reconstruction method (such as re-collecting a second normal sample to calculate the template center feature, adjusting existing template parameters, etc.) to ensure the accuracy of subsequent feature calculations and health predictions.

[0187] Furthermore, the tightening equipment health prediction method also includes: updating the preset standard template library based on the second normal sample when the template drift feature meets the preset template reconstruction conditions. This enables dynamic adaptive updating of the preset standard template library, ensuring the timeliness and accuracy of the template benchmark and adapting to changes in process benchmarks during long-term operation of the tightening production line. First, the template drift feature is calculated based on the second normal sample in the preset sample trend library. The second normal sample is a workpiece sample in the preset sample trend library whose process inspection results are normal, which can truly represent the overall benchmark state of the current qualified tightening operation on the production line. The template drift feature is used to quantify the degree of deviation between the preset standard template library and the current overall benchmark of the normal samples, reflecting whether the standard template is no longer suitable for the current production conditions. Second, the preset template reconstruction conditions are pre-set judgment thresholds used to determine whether the template drift has reached the point where updating is necessary, avoiding frequent and meaningless updates. When the template drift characteristics meet the preset template reconstruction conditions, it indicates that the original preset standard template library has deviated from the current normal sample benchmark. Continuing to use it will lead to distortion in the calculation of target sample drift characteristics and target equipment drift characteristics, thereby affecting the accuracy of health prediction results and causing misjudgments and omissions of equipment anomalies. At this time, updating the preset standard template library based on the second normal sample (a workpiece sample with normal process inspection results, representing the current qualified production benchmark) can re-establish a standardized comparison benchmark that fits the current real production conditions. This ensures the accuracy and reliability of the entire process of feature extraction of subsequent batch tightening samples, comparison calculation of target sample drift characteristics and target equipment drift characteristics, equipment trend diagnosis, and health prediction. It provides continuous and accurate benchmark support for the long-term stable diagnosis of the health status of tightening production line equipment, avoiding misjudgments and omissions caused by template aging. The preset template reconstruction conditions can be: the offset value of any one or more of the template curve offset, template stage offset, template stiffness offset, and template ripple offset in the template drift characteristics exceeds the corresponding preset benchmark threshold; or the comprehensive offset coefficient of the template drift characteristics exceeds the preset reconstruction threshold; or the template continues to drift and cannot return to the standard benchmark range on its own within a fixed period, all of which can trigger the automatic reconstruction and update process of the preset standard template library.

[0188] Furthermore, the tightening equipment health prediction method also includes: outputting parameter adjustment suggestions based on the target sample drift characteristics and preset parameter sensitivity information. These suggestions are used to adjust the process parameters of each process stage of the target tightening equipment, and the process parameters include at least one of torque, angle, and rotational speed. The preset parameter sensitivity information is pre-set and reflects the correlation between process parameters and sample drift characteristics, indicating the adjustment direction and range corresponding to different drift anomalies. This preset parameter sensitivity information can be generated from historical parameter adjustment records, process test data, or expert rules. In practice, by combining the offset of the target sample drift characteristics with the preset parameter sensitivity information, it is determined which process parameters are abnormal and causing the drift anomaly. Targeted adjustment suggestions are then output, specifying the adjusted process parameters (torque / angle / rotational speed), adjustment direction, and specific range. This allows for rapid correction of parameter deviations, reduces the generation of abnormal samples, ensures stable equipment operation, and optimizes equipment operation without downtime maintenance.

[0189] Furthermore, the tightening equipment health prediction method also includes: generating first quality analysis information of the preset sample trend library and / or second quality analysis information of each evaluation dimension based on the preset sample trend library, wherein the evaluation dimension includes at least one of batch, workstation, gun number, and shift, and the quality analysis information includes at least one of normal rate, risk rate, and non-compliance rate.

[0190] Among them, the first quality analysis information is the quality analysis information of the preset sample trend library, which is used to reflect the overall situation of the preset sample trend library. For example, based on the process inspection results of each workpiece sample in the preset sample trend library, the overall normal rate, risk rate and non-compliance rate of the preset sample trend library can be statistically analyzed.

[0191] The second quality analysis information comprises quality analysis information for each evaluation dimension. For example, based on the process inspection results of each workpiece sample in the preset sample trend library, and according to the batch number corresponding to the workpiece sample, the normal rate, risk rate, and non-compliance rate of each batch in the preset sample trend library can be calculated. Based on the workstation number corresponding to the workpiece sample, the normal rate, risk rate, and non-compliance rate of each workstation in the preset sample trend library can be calculated. Based on the gun number corresponding to the workpiece sample, the normal rate, risk rate, and non-compliance rate of each gun number in the preset sample trend library can be calculated.

[0192] Furthermore, the tightening equipment health prediction method also includes: based on the health prediction results, outputting review suggestions for the batch tightening samples and / or maintenance suggestions for the target tightening equipment. The review suggestions guide personnel to conduct targeted reviews of the batch tightening samples and investigate the causes of abnormal samples; the maintenance suggestions clarify the maintenance direction, content, and priority of the target tightening equipment. In specific implementation, based on the health prediction results, for abnormal components, equipment maintenance suggestions are output (e.g., worn bits need replacement, aging reduction mechanisms need lubrication); simultaneously, for abnormal samples in the batch tightening samples, such as workpiece samples where process inspection results indicate risk, review suggestions are output (e.g., focusing on reviewing tightened workpieces completed by the target tightening equipment and where process inspection results indicate risk); in addition, review suggestions can be output by associating abnormal component type, abnormal confidence level, abnormal level, abnormal workstation number, and abnormal tightening gun number, such as prioritizing review of abnormal samples completed by abnormal gun numbers and with high risk levels, to improve the targeting of sample quality control and equipment maintenance and reduce production risks.

[0193] As can be seen from the above, firstly, combining target sample drift characteristics and target equipment drift characteristics for health prediction can improve the accuracy of health prediction results. Specifically, utilizing target equipment drift characteristics allows for health prediction based on individual differences in the target tightening equipment, reducing misjudgments caused by inherent drift due to these individual differences. Utilizing target sample drift characteristics allows for the discovery of drift characteristics caused by abnormalities in the target tightening equipment for health prediction. Thus, combining target sample drift characteristics and target equipment drift characteristics can accurately pinpoint the root cause of the deviation, clarifying whether the increase in abnormal samples is due to sample fluctuations or equipment component malfunctions, further improving the accuracy of health prediction results. Secondly, when the number of abnormal samples in a batch of tightening samples exceeds a preset threshold, or when the proportion of abnormal samples in a batch of tightening samples exceeds a preset proportion threshold, health prediction is triggered. This avoids unnecessary prediction processes triggered by individual abnormal samples (due to random errors), reducing invalid calculations and improving the accuracy of equipment health prediction. Thirdly, the health prediction results can indicate whether there are any abnormalities in key components of the target tightening equipment, such as the bit, gun body, deceleration mechanism, and sensors. This can directly provide a clear basis for equipment maintenance, help staff quickly locate faulty parts, and carry out timely maintenance or replacement, thereby reducing equipment downtime, lowering production costs, and reducing the number of defective workpieces caused by equipment abnormalities.

[0194] Furthermore, to better implement the tightening equipment health prediction method in the embodiments of this application, based on the tightening equipment health prediction method, the embodiments of this application also provide a tightening equipment health prediction device, such as... Figure 6The diagram shown is a structural schematic of one embodiment of the tightening equipment health prediction device provided in this application. The tightening equipment health prediction device 600 includes: The first acquisition unit 601 is used to acquire batch tightening samples of the target tightening equipment from a preset sample trend library, wherein the preset sample trend library includes multiple workpiece samples, and the workpiece samples are labeled with process inspection results. The second acquisition unit 602 is used to acquire the target sample drift characteristics of the target tightening device based on the batch tightening samples and the preset standard template library; The third acquisition unit 603 is used to acquire the target equipment drift characteristics of the target tightening device based on the first normal sample of the batch tightening sample and the preset standard template library, wherein the first normal sample refers to the workpiece sample whose process inspection result is normal. The prediction unit 604 is used to perform a health prediction based on the target sample drift characteristics and the target equipment drift characteristics when the number of abnormal samples in the batch tightening samples is greater than a preset quantity threshold or when the proportion of abnormal samples in the batch tightening samples is greater than a preset proportion threshold, to obtain a health prediction result of the target tightening equipment. The abnormal sample refers to a workpiece sample whose process inspection result is abnormal. The health prediction result is used to indicate whether at least one of the bit, gun body, deceleration mechanism, and sensor of the target tightening equipment is abnormal. The health prediction result includes at least one of abnormal component type, abnormal confidence level, and abnormal level.

[0195] In some embodiments, the target sample drift features include a first sample drift feature, a second sample drift feature, and an overall sample drift feature; the second acquisition unit 602 is used for: According to a preset division strategy, the batch tightening samples are divided to obtain first division sample data and second division sample data, wherein the tightening time of the second division sample data is later than the tightening time of the first division sample data. Based on the first segmented sample data and the preset standard template library, the first sample drift feature is obtained; Based on the second segmentation sample data and the preset standard template library, the second sample drift feature is obtained; Based on the batch tightening samples and the preset standard template library, the overall sample drift characteristics are obtained; In some embodiments, the prediction unit 604 is used for: Obtain a first comparison feature between the first sample drift feature and the target device drift feature, a second comparison feature between the second sample drift feature and the target device drift feature, and a third comparison feature between the overall sample drift feature and the target device drift feature; Based on the first comparison feature, the second comparison feature, and the third comparison feature, the abnormal component type of the target tightening device is determined.

[0196] In some embodiments, the second acquisition unit 602 is used for: Obtain the first central feature of the first segmented sample data; Obtain the template center features of the preset standard template library; Based on the first central feature and the template central feature, the first sample drift feature is determined.

[0197] In some embodiments, the first center feature includes at least one of a first curve center, a first stage center, a first stiffness center, and a first ripple center; the template center feature includes at least one of a template curve center, a template stage center, a template stiffness center, and a template ripple center; the first sample drift feature includes at least one of a first curve offset, a first stage offset, a first stiffness offset, and a first ripple offset; the curve corresponding to the curve center is at least one of a torque-time curve, an angle-time curve, a rotational speed-time curve, and a torque-angle curve; the second acquisition unit 602 is used for: The first curve offset is determined based on the center of the first curve and the center of the template curve; Based on the first stage center and the template stage center, determine the first stage offset; Based on the first stiffness center and the template stiffness center, the first stiffness offset is determined; The first ripple offset is determined based on the first ripple center and the template ripple center.

[0198] In some embodiments, the tightening equipment health prediction device further includes a first suggestion unit (not shown in the figure), the first suggestion unit being used for: Based on the second normal sample of the preset sample trend library, obtain the central features of the trend library; Based on the trend library center features and the template center features of the preset standard template library, the template drift features of the preset standard template library are determined; Based on the template drift characteristics, output template reconstruction suggestions; When the template drift feature meets the preset template reconstruction conditions, the preset standard template library is updated based on the second normal sample.

[0199] In some embodiments, the tightening equipment health prediction device further includes a second recommendation unit (not shown in the figure), the second recommendation unit being used for: Based on the target sample drift characteristics and preset parameter sensitivity information, parameter adjustment suggestions are output. The parameter adjustment suggestions are used to adjust the process parameters of each process stage of the target tightening equipment. The process parameters include at least one of torque, angle, and rotational speed.

[0200] In some embodiments, the tightening equipment health prediction device further includes a generation unit (not shown in the figure), the generation unit being used for: Based on the preset sample trend library, generate first quality analysis information of the preset sample trend library and / or second quality analysis information of each evaluation dimension, wherein the evaluation dimension includes at least one of batch, workstation, gun number, and shift, and the quality analysis information includes at least one of normal rate, risk rate, and non-compliance rate.

[0201] In some embodiments, the tightening equipment health prediction device further includes a third recommendation unit (not shown in the figure), which is used for: Based on the health prediction results, output review recommendations for the batch tightening samples and / or maintenance recommendations for the target tightening equipment.

[0202] Those skilled in the art will understand that all or part of the steps in the above-described method for predicting the health of tightening equipment can be accomplished by instructions or by controlling related hardware with instructions. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0203] Therefore, embodiments of this application provide a computer-readable storage medium storing a plurality of computer programs that can be loaded by a processor to execute any of the tightening equipment health prediction methods provided in embodiments of this application. For example, the computer program can be loaded by a processor to perform the following steps: From a preset sample trend library, a batch of tightening samples of the target tightening equipment are obtained, wherein the preset sample trend library includes multiple workpiece samples, and the workpiece samples are labeled with process inspection results; based on the batch tightening samples and a preset standard template library, the target sample drift characteristics of the target tightening equipment are obtained; based on the first normal sample of the batch tightening samples and the preset standard template library, the target equipment drift characteristics of the target tightening equipment are obtained, wherein the first normal sample refers to a workpiece sample with normal process inspection results; when the number of abnormal samples in the batch tightening samples is greater than a preset quantity threshold, or when the proportion of abnormal samples in the batch tightening samples is greater than a preset proportion threshold, a health prediction is performed based on the target sample drift characteristics and the target equipment drift characteristics to obtain a health prediction result of the target tightening equipment, wherein the abnormal sample refers to a workpiece sample with abnormal process inspection results, and the health prediction result is used to indicate whether at least one of the bit, gun body, deceleration mechanism, and sensor of the target tightening equipment is abnormal, and the health prediction result includes at least one of abnormal component type, abnormal confidence level, and abnormal level.

[0204] The computer-readable storage medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0205] Furthermore, according to the tightening equipment health prediction method of this application, a computer program product or computer program is also provided, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the methods provided in the various optional implementations of the above embodiments.

[0206] In the above embodiments of the tightening equipment health prediction method, apparatus, electronic device, computer-readable storage medium, computer program product, or computer program, the descriptions of each embodiment have different focuses. Parts not described in detail in a particular embodiment can be referred to in the relevant descriptions of other embodiments. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes and beneficial effects of the computer-readable storage medium, electronic device, and their corresponding units described above can be referred to the description of the tightening equipment health prediction method in the above embodiments, and will not be repeated here.

[0207] The description of the various embodiments above tends to emphasize the differences between the various embodiments. The similarities or similarities between them can be referred to, and for the sake of brevity, they will not be repeated here.

[0208] The foregoing has provided a detailed description of a tightening equipment health prediction method, apparatus, electronic device, computer-readable storage medium, and computer program product provided in the embodiments of this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for predicting the health of tightening equipment, characterized in that, The method includes: A batch of tightening samples of the target tightening equipment are obtained from a preset sample trend library, wherein the preset sample trend library includes multiple workpiece samples, and the workpiece samples are marked with process inspection results; Based on the batch tightening samples and the preset standard template library, the target sample drift characteristics of the target tightening device are obtained; Based on the first normal sample of the batch tightening samples and the preset standard template library, the target equipment drift characteristics of the target tightening equipment are obtained, wherein the first normal sample refers to the workpiece sample whose process inspection result is normal. When the number of abnormal samples in the batch tightening sample is greater than a preset quantity threshold, or when the proportion of abnormal samples in the batch tightening sample is greater than a preset proportion threshold, a health prediction is performed based on the target sample drift characteristics and the target equipment drift characteristics to obtain the health prediction result of the target tightening equipment. The abnormal sample refers to a workpiece sample with abnormal process inspection results. The health prediction result is used to indicate whether at least one of the bit, gun body, deceleration mechanism, and sensor of the target tightening equipment is abnormal. The health prediction result includes at least one of the abnormal component type, abnormal confidence level, and abnormal level. The target sample drift features include first sample drift features, second sample drift features, and overall sample drift features; The step of obtaining the target sample drift characteristics of the target tightening device based on the batch tightening samples and the preset standard template library includes: According to a preset division strategy, the batch tightening samples are divided to obtain first division sample data and second division sample data, wherein the tightening time of the second division sample data is later than the tightening time of the first division sample data. Based on the first segmented sample data and the preset standard template library, the first sample drift feature is obtained; Based on the second segmentation sample data and the preset standard template library, the second sample drift feature is obtained; Based on the batch tightening samples and the preset standard template library, the overall sample drift characteristics are obtained; The health prediction based on the target sample drift characteristics and the target equipment drift characteristics, to obtain the health prediction result of the target tightening equipment, includes: Obtain a first comparison feature between the first sample drift feature and the target device drift feature, a second comparison feature between the second sample drift feature and the target device drift feature, and a third comparison feature between the overall sample drift feature and the target device drift feature; Based on the first comparison feature, the second comparison feature, and the third comparison feature, the abnormal component type of the target tightening device is determined.

2. The method for predicting the health of tightening equipment according to claim 1, characterized in that, The step of obtaining the first sample drift feature based on the first segmented sample data and the preset standard template library includes: Obtain the first central feature of the first segmented sample data; Obtain the template center features of the preset standard template library; Based on the first central feature and the template central feature, the first sample drift feature is determined.

3. The method for predicting the health of tightening equipment according to claim 2, characterized in that, The first center feature includes at least one of a first curve center, a first stage center, a first stiffness center, and a first ripple center; the template center feature includes at least one of a template curve center, a template stage center, a template stiffness center, and a template ripple center; the first sample drift feature includes at least one of a first curve offset, a first stage offset, a first stiffness offset, and a first ripple offset; and the curve corresponding to the curve center is at least one of a torque-time curve, an angle-time curve, a rotational speed-time curve, and a torque-angle curve. The step of determining the first sample drift feature based on the first central feature and the template central feature includes: The first curve offset is determined based on the center of the first curve and the center of the template curve; Based on the first stage center and the template stage center, determine the first stage offset; Based on the first stiffness center and the template stiffness center, the first stiffness offset is determined; The first ripple offset is determined based on the first ripple center and the template ripple center.

4. The method for predicting the health of tightening equipment according to claim 1, characterized in that, The method further includes: Based on the second normal sample of the preset sample trend library, obtain the central features of the trend library; Based on the trend library center features and the template center features of the preset standard template library, the template drift features of the preset standard template library are determined; Based on the template drift characteristics, output template reconstruction suggestions; When the template drift feature meets the preset template reconstruction conditions, the preset standard template library is updated based on the second normal sample.

5. The method for predicting the health of tightening equipment according to claim 1, characterized in that, The method further includes: Based on the target sample drift characteristics and preset parameter sensitivity information, parameter adjustment suggestions are output. The parameter adjustment suggestions are used to adjust the process parameters of each process stage of the target tightening equipment. The process parameters include at least one of torque, angle, and rotational speed.

6. The method for predicting the health of tightening equipment according to claim 1, characterized in that, The method further includes: Based on the preset sample trend library, generate first quality analysis information of the preset sample trend library and / or second quality analysis information of each evaluation dimension, wherein the evaluation dimension includes at least one of batch, workstation, gun number, and shift, and the quality analysis information includes at least one of normal rate, risk rate, and non-compliance rate.

7. The method for predicting the health of tightening equipment according to claim 1, characterized in that, The method further includes: Based on the health prediction results, output review recommendations for the batch tightening samples and / or maintenance recommendations for the target tightening equipment.

8. A health prediction device for tightening equipment, characterized in that, The tightening equipment health prediction device includes: The first acquisition unit is used to acquire batch tightening samples of the target tightening equipment from a preset sample trend library, wherein the preset sample trend library includes multiple workpiece samples, and the workpiece samples are labeled with process inspection results; The second acquisition unit is used to acquire the target sample drift characteristics of the target tightening device based on the batch tightening samples and the preset standard template library; The third acquisition unit is used to acquire the target equipment drift characteristics of the target tightening device based on the first normal sample of the batch tightening sample and the preset standard template library, wherein the first normal sample refers to the workpiece sample whose process inspection result is normal. The prediction unit is used to perform a health prediction based on the target sample drift characteristics and the target equipment drift characteristics when the number of abnormal samples in the batch tightening samples is greater than a preset quantity threshold or when the proportion of abnormal samples in the batch tightening samples is greater than a preset proportion threshold, to obtain a health prediction result for the target tightening equipment. The abnormal sample refers to a workpiece sample with an abnormal process inspection result. The health prediction result is used to indicate whether at least one of the bit, gun body, deceleration mechanism, and sensor of the target tightening equipment is abnormal. The health prediction result includes at least one of the abnormal component type, abnormal confidence level, and abnormal level. The target sample drift features include first sample drift features, second sample drift features, and overall sample drift features; The second acquisition unit is configured to: divide the batch tightening samples according to a preset division strategy to obtain first division sample data and second division sample data, wherein the tightening time of the second division sample data is later than the tightening time of the first division sample data; acquire a first sample drift feature based on the first division sample data and a preset standard template library; acquire a second sample drift feature based on the second division sample data and the preset standard template library; and acquire an overall sample drift feature based on the batch tightening samples and the preset standard template library. The prediction unit is used to: acquire a first comparison feature between the first sample drift feature and the target device drift feature, a second comparison feature between the second sample drift feature and the target device drift feature, and a third comparison feature between the overall sample drift feature and the target device drift feature; and determine the abnormal component type of the target tightening device based on the first comparison feature, the second comparison feature, and the third comparison feature.

9. An electronic device, characterized in that, It includes a processor and a memory, wherein the memory stores a computer program, and the processor executes the tightening equipment health prediction method as described in any one of claims 1 to 7 when it invokes the computer program in the memory.

10. A computer-readable storage medium, characterized in that, It stores a computer program, which is loaded by a processor to execute the tightening equipment health prediction method according to any one of claims 1 to 7.

11. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by the processor, they implement the tightening equipment health prediction method as described in any one of claims 1 to 7.

Images