Tightening state identification method and system

Abstract

The embodiment of the invention provides a tightening state identification method and system, and belongs to the technical field of tightening state identification. The identification method comprises the following steps: acquiring tightening program data and tightening process data of target equipment; inputting the tightening program data into a tightening program self-recognition model, and recognizing the tightening program of the target equipment; and under the condition that the tightening program of the target equipment is not abnormal, identifying the tightening process of the target equipment based on the tightening process data and the tightening process monitoring model, or under the condition that the tightening program of the target equipment is abnormal, identifying that the tightening program of the target equipment is abnormal. According to the tightening state identification method and system including tightening program rechecking and tightening process monitoring, the technical problems that the tightening result of a multi-variety flexible mixed production line is abnormal, tightening parameters are mistakenly selected and the like are solved, the tightening state and tightening parameter calling can be completely judged in real time, and automatic checking and confirmation of the tightening result are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of tightening state recognition, and in particular to a tightening state recognition method and system. Background Art

[0002] As market demand increases, flexible production of multiple products is becoming increasingly important in the engine manufacturing process. Configuring different equipment, procedures, and parameters on the same production line to achieve the production of multiple product models on the same production line and at the same workstation can greatly reduce the potential costs of adding new production lines and improve the market competitiveness of the entire factory. Among them, the engine manufacturing process with the largest proportion and the most adjustments is the bolt tightening process.

[0003] For example, the smart manufacturing and flexible workshops in the prior art usually use Mes (Manufacturing Execution System) configuration and socket selector picking directions to control the calling of correct tightening parameters of engine products, and confirm the correctness of on-site operations based on the number of tightening times. Although this solution can solve the flexible tightening process requirements to a certain extent, it needs to rely on the training of operators to implement the correct tightening parameters. This means that when manual operation errors occur, such as errors in Mes configuration information and incorrect picking of the socket position, the operation abnormality cannot be automatically obtained. Specifically, if the assembly status is confirmed based on the number of error-proofing times, there will be interference from the wrong tightening program. Mes cannot distinguish the source of qualified tightening, so even if the operator uses sockets of different specifications, as long as the total number of tightening times is the same, the qualified status will still be fed back.

[0004] In addition, even if the operator does not make an operational error, there is still a situation where the abnormal state of the tightening process cannot be accurately determined. For example, if the sleeve is taken correctly, if the output head of the tightening shaft is disengaged from the bolt and then reengaged, although the final tightening angle or torque is qualified, the tightening parameters do not reach the target, resulting in low clamping force, which will also cause a fault.

[0005] These situations will cause batch tightening process errors, or occasional tightening process abnormalities, resulting in batch market problems or occasional market failures. To solve such problems, manual confirmation of the tightening status is required, but the actual factory staffing cannot meet this requirement at all. Summary of the invention

[0006] The purpose of the embodiments of the present invention is to provide a method and system for identifying a tightening status, which can address the technical problems of abnormal tightening results and incorrect selection of tightening parameters in multi-variety and flexible mixed production lines, as well as the lack of error prevention in the tightening process of the above-mentioned engine and the insufficient monitoring capability of the tightening process, thereby identifying abnormal tightening procedures and abnormal tightening processes of the engine and realizing automatic inspection and confirmation of the tightening results.

[0007] In order to achieve the above-mentioned purpose, an embodiment of the present invention provides a method for identifying a tightening state, the identification method comprising: acquiring tightening program data and tightening process data of a target device; inputting the tightening program data into a tightening program self-identification model to identify the tightening program of the target device; and in the case where there is no abnormality in the tightening program of the target device, identifying the tightening process of the target device based on the tightening process data and the tightening process monitoring model, or in the case where there is an abnormality in the tightening program of the target device, identifying the abnormality of the tightening program of the target device.

[0008] Optionally, the tightening procedure self-identification model is constructed in the following manner: based on the tightening procedure data, system configuration data and design standard data of the feature sample device, a tightening procedure distribution map of the feature sample device is established; the tightening procedure distribution map is deviation-optimized using the tightening procedure data of multiple training sample devices to output an updated tightening procedure distribution map; and the updated tightening procedure distribution map is determined as the tightening procedure self-identification model.

[0009] Optionally, the tightening program data includes one or more of the following parameters: equipment ID, program number, work station, tightening quantity and tightening status.

[0010] Optionally, the system configuration data includes the device type, tightening times and tightening procedures configured for the characteristic sample device in the system configuration database; the design standard data includes the device type, tightening parameters and tightening times of the characteristic sample device in the design stage.

[0011] Optionally, the tightening process data includes torque data, angle data and timing data of the target device, and the tightening process monitoring model includes one or more of the following models: a parameter review model, a process monitoring model, a yield monitoring model and a statistical process control model.

[0012] Optionally, in the case where the tightening process monitoring model is a parameter review model, the identifying the tightening process of the target device based on the tightening process data and the tightening process monitoring model includes: drawing a torque angle curve based on the torque data and angle data of the target device; obtaining a full tightening process curve starting from a threshold torque from the torque angle curve; and monitoring the full tightening process curve through the parameter review model, and outputting a signal indicating that the tightening process of the target device is abnormal when a curve drop occurs in the full tightening process curve.

[0013] Optionally, in the case where the tightening process monitoring model is a process monitoring model, the identifying the tightening process of the target device based on the tightening process data and the tightening process monitoring model includes: drawing a torque timing curve based on the torque data and timing data of the target device; inputting the torque timing curve of the target device into the process monitoring model, wherein the process monitoring model includes a characteristic torque timing curve; and monitoring the torque timing curve of the target device by means of a sliding window full-quantity identification through the process monitoring model, and outputting a signal that the tightening process of the target device is abnormal when the similarity between the torque timing curve of the target device and the characteristic torque timing curve is less than a first set threshold.

[0014] Optionally, in the case where the tightening process monitoring model is a yield monitoring model, the identifying the tightening process of the target device based on the tightening process data and the tightening process monitoring model includes: drawing a torque angle curve based on the torque data and angle data of the target device; obtaining a curve range with a slope of 0 from the torque angle curve; and monitoring the curve range through the yield monitoring model, and outputting a signal that the tightening process of the target device is abnormal when the curve range is greater than a second set threshold.

[0015] Optionally, in the case where the tightening process monitoring model is a statistical process control model, the identifying the tightening process of the target device based on the tightening process data and the tightening process monitoring model includes: drawing a torque timing curve and an angle timing curve based on the torque data, angle data and timing data of the tightening process data; dynamically determining process performance indexes of multiple continuous data in the torque timing curve and the angle timing curve; and monitoring the process performance indexes of the multiple continuous data through the statistical process control model, and outputting a signal indicating that the tightening process of the target device is abnormal when the fluctuation between the multiple continuous data and the set target process data is greater than a third set threshold.

[0016] Optionally, the tightening program self-identification model is a tightening program self-identification model deployed in the cloud; and / or the tightening process monitoring model is a tightening process monitoring model deployed in the cloud.

[0017] On the other hand, the present invention also provides a tightening status identification system, the identification system comprising: a data acquisition module, used to acquire tightening program data and tightening process data of a target device; a first identification module, used to input the tightening program data into a tightening program self-identification model to identify the tightening program of the target device; and a second identification module, used to identify the tightening process of the target device based on the tightening process data and the tightening process monitoring model when there is no abnormality in the tightening program of the target device, or to identify the abnormality of the tightening program of the target device when there is an abnormality in the tightening program of the target device.

[0018] Through the above technical solution, the present invention has the following advantages and outstanding effects: by proposing a tightening state identification method and system including tightening program re-inspection and tightening process monitoring, the tightening state and tightening parameter call can be determined in real time and in full, ensuring fast identification of tightening anomalies and automatically realizing secondary monitoring of the tightening state. At the same time, the present invention can achieve a tightening state equivalent to 100% manual confirmation.

[0019] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific implementations, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings:

[0021] Figure 1 is a schematic diagram of a tightening state identification system architecture provided according to an embodiment of the present invention;

[0022] Figure 2 is a flow chart of a method for identifying a tightening state provided according to an embodiment of the present invention;

[0023] Figure 3 is a screenshot of a Mes system configuration interface provided according to an embodiment of the present invention;

[0024] Figure 4a-4b is a screenshot of a sleeve selector program configuration function interface provided according to an embodiment of the present invention;

[0025] Figure 5 is a schematic diagram of a method for constructing a self-identification model of a tightening procedure provided according to an embodiment of the present invention;

[0026] Figure 6 is a schematic diagram of a tightening program distribution process provided according to an embodiment of the present invention;

[0027] Figure 7 is a schematic diagram of a tightening procedure principle provided according to an embodiment of the present invention;

[0028] Figure 8a-8b is a schematic diagram of an embodiment of an abnormal tightening state provided according to an embodiment of the present invention;

[0029] Figure 8c-8d is a schematic diagram of another embodiment of an abnormal tightening state provided by an embodiment of the present invention;

[0030] Fig. 9 is a functional schematic diagram of a parameter review module provided according to an embodiment of the present invention;

[0031] Fig.10 is a functional schematic diagram of a process monitoring module provided according to an embodiment of the present invention;

[0032] Fig.11 is a functional schematic diagram of a yield determination module provided according to an embodiment of the present invention;

[0033] Fig.12 is a schematic diagram of a tightening state identification system provided according to an embodiment of the present invention;

[0034] Fig.13 It is a schematic diagram of a design architecture of a tightening state identification system provided according to an embodiment of the present invention. DETAILED DESCRIPTION

[0035] The specific implementation of the embodiment of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the embodiment of the present invention, and is not used to limit the embodiment of the present invention.

[0036] In order to better explain the present invention, it is first necessary to introduce the overall concept and system architecture of the present invention. Figure 1As shown, the tightening status identification system of the present invention can adopt a cloud deployment mode, and the data source can come from the tightening data management system ToolsNet. After the edge data of the main line tightening station is aggregated into the ToolsNet database, the data is further parsed into the Hadoop smart manufacturing platform, and the Hadoop smart manufacturing platform is used to monitor the new data in real time, and then obtain the program number, tightening process data, error information, tightening status, tightening time and engine ID and other information. Then, based on the timestamp, the self-built model status is fully verified, and it is automatically run in the built tightening process monitoring model and the tightening program re-inspection model to confirm whether there is an abnormality. After all the judgment results are finally summarized, the PLC is issued through the engine health comprehensive judgment system to notify the on-site personnel, and the relevant information is emailed to the key management and control personnel to achieve closed-loop control of the tightening status.

[0037] Specifically, the present invention provides a method for identifying a tightening state, such as Figure 2 As shown, the identification method may include steps S110 - S130 .

[0038] Step S110, acquiring tightening program data and tightening process data of the target device.

[0039] In one embodiment, the full tightening data of the target device of the line-side tightening station can be uploaded to the ToolsNet database through an open protocol. Then, the full tightening data of the target device can be extracted from the ToolsNet database.

[0040] Generally speaking, tightening data anomalies can be divided into tightening program data anomalies and tightening process data anomalies. In the actual production process, if the tightening program data is abnormal, it is usually a manual selection error, that is, the operator selects the wrong tightening program due to operational errors. For details, please refer to the following Figure 4a and 4b In one embodiment, the tightening program data may include parameters such as the device ID of the target device, the program number, the workstation, the tightening quantity, and the tightening status.

[0041] If the tightening process data is abnormal, it can be divided into the following two situations: one is manual selection error, that is, the operator selects the wrong tightening program data. In this case, the tightening process data output based on the wrong tightening program data is also wrong. The other is that the manually selected tightening program is normal, that is, there is no manual operation error. But the tightening process data output based on the correct tightening program data is wrong. The reason for this situation may be that the tightening shaft output head is disengaged from the bolt and then engaged again, or the washer is missing. For details, please refer to the following. Figures 8a-8dTwo examples are shown. In one embodiment, the tightening process data may include parameters such as torque data, angle data, and timing data of the target device.

[0042] Therefore, it is necessary to first identify whether the tightening program data is abnormal (the following step S120 ). If the tightening program data is not abnormal, it is necessary to further identify whether the tightening process data is abnormal (the following step S130 ).

[0043] Step S120, inputting the tightening program data into a tightening program self-identification model to identify the tightening program of the target device. The tightening program self-identification model may be a tightening program self-identification model deployed in the cloud.

[0044] The tightening procedure self-identification model can solve the problem of interference from incorrect tightening procedures due to insufficient operating skills of employees. Specifically, Mes is usually used in production sites, which has the characteristics of greatly reducing logistics inventory costs, evenly distributing the production capacity of each production line, shortening product delivery time, and reducing the production cost of a single vehicle. A product manufacturing bill of materials (MBOM) is usually configured in Mes, which can complete product assembly on the engine manufacturing line based on the planned sequence. At the same time, special manufacturing procedures for parts and error prevention times are specified, which can realize customized manufacturing of products and the occurrence of potential problems. For example, in Figure 3 As shown in the screenshot, in the engine tightening process, Mes can issue the assembly parameter number (tightening program number) of the part by configuring the Tool Recipe, enable the triggered tightening tool by associating the Tool Number, and output the total tightening times by the Tool Click Count. In other words, in theory, the quality of on-site assembly can be effectively improved by fully configuring the three points of Tool Recipe, Tool Number and Tool Click Count.

[0045] However, due to the large number of commercial engine categories, there are thousands of engine categories in production, resulting in different tightening tools, tightening sequences, tightening parameters and tightening times for the same workstation. At this time, if all the above parameters are configured, it is necessary to maintain the Mes configuration list in real time, adjust the PLC program, and dynamically match the program settings of the tightening controller, which is difficult to implement in actual production. Therefore, the solution of using Mes to send all configuration information cannot be applied to most scenes in the manufacturing site, and is only used for special workstations and scenes where tightening and assembly requirements do not change much. For the above reasons, the prior art mainly sets the Tool Number and Tool Click Count points to achieve error-proof control, and the issuance of tightening programs is more of a human-computer interaction mode. That is, by adding a manually selectable sleeve selector configuration mode, the call of the required tightening program is realized.

[0046] Specifically, Figure 4a and 4b As can be seen from the screenshot shown, for example, tightening programs of 5Nm, 18Nm, 24Nm and 10Nm are placed in positions 1 to 4 of the socket selector respectively, and positions 1 to 4 correspond to the physical addresses of the socket selector respectively. When the operator takes the socket at the physical position, the corresponding tightening program will be called to achieve the specified number of tightening work. When the number of tightening times exceeds the Tool Click Count value, the PLC can complete the status output of the error-proofing verification and notify Mes that the tightening work at this station is completed. However, this mode of assembly status based on error-proofing times confirmation places high demands on the operating skills of employees and there is a risk of interference with the wrong tightening program. For example, you can refer to the following Figure 4b For the workstation shown, when the tightening sequence requires 5Nm×3, 18Nm×2, and 10Nm×1, the total number of tightening times is 6. However, even if the operator uses the tightening sequence of 5Nm×6, since Mes cannot distinguish the source of qualified tightening, it will still feedback the qualified status, resulting in failure to identify abnormalities and the outflow of problematic parts, so there is a risk of batch abnormalities.

[0047] In view of the above problems, in one embodiment, the self-identification model of the tightening procedure can be constructed by using techniques such as template learning, big data classification analysis and regression confirmation, so as to realize the secondary confirmation of the whole engine construction procedure. Figure 5 As shown, the construction of the design scheme for the self-identification model of the tightening procedure may include steps S121-S123.

[0048] In step S121 , a tightening program distribution map of the characteristic sample equipment may be established based on the tightening program data, system configuration data and design standard data of the characteristic sample equipment.

[0049] In one embodiment, if Figure 6 As shown, the system configuration data may include parameters such as the equipment type, tightening times, and tightening procedures configured by the characteristic sample equipment in the system configuration database of Mes. That is, parameters such as the tightening times and tightening procedures (if any) under the engine type can be obtained from the Mes database. Among them, Mes can realize reasonable arrangements of production plans according to the current status of each production line, regulate the production tasks of each workshop, and provide strong support for the effective use of various resources to ensure production. Therefore, the production data collected by Mes can be accurately connected with the enterprise's ERP and LES systems to improve data accuracy.

[0050] In one embodiment, the design standard data may include parameters such as device type, tightening parameters, and tightening times of the characteristic sample device in the design stage. That is, parameters such as tightening parameters and tightening times of a specific engine model may be obtained from the design standard.

[0051] Step S122: Optimize the deviation of the tightening program distribution map by using the tightening program data of a plurality of training sample devices to output an updated tightening program distribution map.

[0052] Specifically, refer to Figure 6 , the tightening program distribution map established in step S121 can be used as Figure 6 The first-build learning model shown. Then, the tightening program data of multiple training sample devices (including the tightening program, tightening parameters and tightening quantity of the engine at each station) are classified and analyzed, and the tightening program data of the multiple training sample devices are deviated from the first-build learning model for optimization. That is, the first-build learning model is repeatedly iterated using the deviated tightening program data until the output accuracy reaches the set value. For example, the accuracy reaches 100% or 95%. Then, the version to be solidified for mass production can be monitored and confirmed manually to obtain a solidified tightening program distribution map, thereby outputting an updated tightening program distribution map.

[0053] Step S123: determining the updated tightening program distribution map as a tightening program self-identification model.

[0054] Specifically, based on the updated tightening procedure distribution map, a real-time monitoring system for engine health can be established to identify the tightening procedure of the target equipment, thereby fully monitoring the construction information. If process deviations are still involved in the updated tightening procedure distribution map, manual confirmation of the mass production curing monitoring results is required to avoid actual production line stoppages.

[0055] Through the above solution, the efficiency of building the traceability program diagram can be greatly improved, the problem of difficult implementation of statistics for the entire process of personnel can be solved, data-driven process changes can be truly realized, and abnormal conditions in the tightening procedures of target equipment can be quickly and accurately identified.

[0056] Step S130, when the tightening program of the target device is not abnormal, the tightening process of the target device is identified based on the tightening process data and the tightening process monitoring model. The tightening process monitoring model may be a tightening process monitoring model deployed in the cloud. Alternatively, when the tightening program of the target device is abnormal, the tightening program of the target device is identified to be abnormal.

[0057] Specifically, in order to ensure that the product has reliable clamping force and achieve high-quality product output, a large number of tightening processes are involved in engine assembly, such as torque method, angle method, SG, PA and other methods.

[0058] For example, the torque method is a method of tightening control in the elastic zone using the linear relationship between torque and preload. This method uses the torque value as the target and stops the rotation after reaching the set value. However, due to the fluctuation of the friction coefficient of the tightening contact surface, as well as the influence of the size, hardness and surface coating of the parts, the clamping force dispersion of the torque method is large, so it is mainly used for non-critical torque positions with low precision requirements.

[0059] The angle method is a tightening control method that rotates the bolt relative to a certain angle after the bolt reaches the set torque during tightening. This method uses the angle value as the target. After the tightening reaches the threshold torque (initial angle calculation point), the tightening angle value is fixed and is completely unaffected by the friction coefficient and component performance, which can greatly reduce the dispersion of the clamping force. The angle method can be used for tightening key bolts, and when tightening past the yield point, it can determine whether the yield point is reliable based on the tightening curve.

[0060] The SG and PA methods are combined control methods based on slope (torque / angle) and angle. This method determines the initial position of the angle by slope change or slope zero point, which can reduce the impact of part performance on the initial state. The SG and PA methods are further optimizations of the angle method. Experimental data show that the clamping force dispersion can be increased by more than 20% compared to the angle method.

[0061] Usually, after the tightening method is selected based on the tightening process, the site will develop a matching tightening procedure based on product requirements and process conditions. Figure 7 The steps shown in FIG. 1 are as follows: the identification cap (correct matching of the position of the tightening shaft output head and the bolt), pre-tightening (fitting of the connecting parts), final tightening (final execution of the specified parameters of the product), etc. These steps together constitute the entire tightening procedure. If monitoring rules (angle, slope, time, etc.) can be added to each required step of the tightening procedure, potential program anomalies can be identified. For example, step 120 can effectively identify whether there is an anomaly in the entire tightening procedure. If the identification result shows that the entire tightening process is completely matched with the set rules, the identification result that the tightening procedure is qualified is output.

[0062] However, in the actual production process, due to the complexity of the tightening procedure, there will be a large number of tightening processes that cannot be kept consistent with the tightening procedure, thus triggering abnormal alarms, and then causing the problem of low OEE in the manufacturing process. Therefore, in order to be compatible with the relationship between manufacturing and quality, the tightening procedure process monitoring will be appropriately relaxed, that is, not all tightening parameters will be monitored. Under the existing edge rule solidification conditions, the market defective PPM can be controlled within the acceptable range of the factory.

[0063] However, this method of judging the tightening process based on the edge solidification rule still has the risk of missed judgment or misjudgment. Figure 8a and Figure 8b As shown in the figure, when the angle method is started, if the tightening shaft output head is disengaged from the bolt and then re-engaged, the final tightening angle or torque will be qualified, but the tightening parameters will not reach the target. This will cause low clamping force and eventually cause market failure. Or, if Figure 8c and Figure 8d As shown in the figure, the tightening curves are similar whether the gasket is missing or not, and the torque is the same, but the tightening angle is different. If the edge curing monitoring rule does not have angle monitoring, it will cause abnormal parts to be assembled and flowed out.

[0064] In response to the above problems, in one embodiment, a tightening process monitoring model can be established. By constructing four types of custom rule modules, namely, parameter review model, process monitoring model, yield monitoring model and statistical process control (SPC) model, the tightening process can be analyzed and optimized by introducing big data rules into the tightening process.

[0065] In one embodiment, when the tightening process monitoring model is a parameter review model, Fig. 9 As shown, the function of the parameter recheck model can be: obtain tightening process data from ToolsNet, and perform secondary confirmation of the set parameters based on the three parameter combinations of torque VS angle, torque VS time, and angle VS time. The confirmation scheme can be that the entire tightening process starting from the threshold torque does not have abnormal conditions such as curve drop.

[0066] In this case, the identifying the tightening process of the target device based on the tightening process data and the tightening process monitoring model may specifically include the following steps:

[0067] 1) Based on the torque data and angle data of the target device, draw a torque angle curve.

[0068] 2) Obtain the full tightening process curve starting from the threshold torque from the torque angle curve.

[0069] 3) The full tightening process curve is monitored through the parameter recheck model, and when the full tightening process curve drops, a tightening process abnormality signal of the target device is output.

[0070] In one embodiment, when the tightening process monitoring model is a process monitoring model, such as Fig.10As shown, the function of the process monitoring model can be: obtaining tightening process data from ToolsNet, and identifying abnormal conditions of the curve based on the sliding window based on the three parameter combinations of torque VS angle, torque VS time, and angle VS time.

[0071] In this case, the identifying the tightening process of the target device based on the tightening process data and the tightening process monitoring model may specifically include the following steps:

[0072] 1) Based on the torque data and timing data of the target device, draw a torque timing curve.

[0073] 2) Inputting the torque timing curve of the target device into a process monitoring model, wherein the process monitoring model includes a characteristic torque timing curve.

[0074] 3) The torque timing curve of the target device is monitored by a process monitoring model in a sliding window full-quantity recognition manner, and when the similarity between the torque timing curve of the target device and the characteristic torque timing curve is less than a first set threshold, a signal indicating that the tightening process of the target device is abnormal is output.

[0075] In one embodiment, the characteristic torque timing curve may be a curve formed by the average value or median of the torque timing curves of multiple characteristic devices. In this case, the corresponding first set threshold may be the mean or median ±3σ (standard deviation) of the multiple torque timing curves.

[0076] In one embodiment, when the tightening process monitoring model is a yield monitoring model, Fig.11 As shown, the function of the yield monitoring model can be: obtain tightening process data from ToolsNet, identify the flat area of ​​the curve from the three parameter combinations of torque VS angle, torque VS time, and angle VS time. When there is a slope of 0 area greater than the set range, the output yield occurs, which can be considered as a failure in the tightening process.

[0077] In this case, the identifying the tightening process of the target device based on the tightening process data and the tightening process monitoring model may specifically include the following steps:

[0078] 1) Based on the torque data and angle data of the target device, draw a torque angle curve.

[0079] 2) Obtain the curve range with a slope of 0 from the torque angle curve.

[0080] 3) The curve range is monitored by the yield monitoring model, and when the curve range is greater than a second set threshold, a signal indicating that the tightening process of the target device is abnormal is output.

[0081] In one embodiment, when the tightening process monitoring model is a statistical process control model, the function of the process control model can be: obtain tightening process data from ToolsNet, and obtain set target process data from the combination of torque VS time and angle VS time. Then, the process performance index (PPKPerformance Index of Process) of 25-30 consecutive data can be dynamically calculated, and an exception is output after the data fluctuates greater than the set value, and an exception notification message is output.

[0082] In this case, the identifying the tightening process of the target device based on the tightening process data and the tightening process monitoring model may specifically include the following steps:

[0083] 1) Based on the torque data, angle data and timing data of the tightening process data, draw a torque timing curve and an angle timing curve.

[0084] 2) Dynamically determine the process performance index of multiple continuous data in the torque timing curve and the angle timing curve.

[0085] 3) Monitoring the process performance index of multiple continuous data through a statistical process control model, and outputting a signal indicating that the tightening process of the target device is abnormal when the fluctuation between the multiple continuous data and the set target process data is greater than a third set threshold.

[0086] In summary, through the technical solution presented in the above steps S110-S130, the present invention has the following advantages and outstanding effects: by proposing a tightening state identification method including tightening program re-inspection and tightening process monitoring, the tightening state and tightening parameter call can be determined in real time and in full, ensuring fast identification of tightening anomalies and automatically realizing secondary monitoring of the tightening state. At the same time, the present invention can achieve a tightening state equivalent to 100% manual confirmation.

[0087] The present invention also provides a tightening state identification system, such as Fig.12 As shown, the recognition system may include: a data acquisition module 210 , a first recognition module 220 , and a second recognition module 230 .

[0088] The data acquisition module 210 may be used to acquire tightening program data and tightening process data of the target device.

[0089] The first identification module 220 may be used to input the tightening program data into a tightening program self-identification model to identify the tightening program of the target device.

[0090] The second identification module 230 can be used to identify the tightening process of the target device based on the tightening process data and the tightening process monitoring model when there is no abnormality in the tightening program of the target device, or to identify the tightening program abnormality of the target device when there is an abnormality in the tightening program of the target device.

[0091] Through the above technical solution, the present invention has the following advantages and outstanding effects: by proposing a tightening state identification system including tightening program re-inspection and tightening process monitoring, it is possible to determine the tightening state and tightening parameter call in real time and in full, ensure rapid identification of tightening anomalies, and automatically realize secondary monitoring of the tightening state. At the same time, the present invention can achieve a tightening state equivalent to 100% manual confirmation.

[0092] In addition, the tightening status recognition system of the present invention can also integrate data analysis and IIOT verification results, build a tightening data analysis system for the edge-cloud system, and develop the deployment of the field system based on the Hadoop distributed storage architecture and edge-cloud collaborative solution. Fig.13 shown.

[0093] Specifically, at the edge layer, all tightening data can be pushed to the ToolsNet database based on an open protocol, and key data can be captured in real time to the data center based on SQL statements. The data center can adopt the Hadoop distributed storage architecture design, and can divide the captured data into three types of storage: cold, warm, and hot based on business needs, so as to match the real-time needs of the business. In the factory cloud platform, the tightening data analysis system can be further deployed, and the monitoring model can be automatically run to solve the pain points of tightening in the industry.

[0094] Based on the calculation results of the cloud-deployed model, this system can also design an edge-cloud architecture to push the results back to the edge and send them to the terminal PLC points to achieve abnormal line-side response, display and alarm, thereby establishing a closed-loop edge-cloud tightening monitoring system.

[0095] According to the above description, the beneficial effects of the present invention also include: open system design, personalized configuration of rule models, customized settings for application scenario monitoring, and dynamic matching of business needs.

[0096] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0097] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for identifying a tightening state, characterized in that: The identification method comprises: Obtain tightening program data and tightening process data of target equipment; Inputting the tightening program data into a tightening program self-identification model to identify the tightening program of the target device; and When the tightening procedure of the target device is normal, the tightening procedure of the target device is identified based on the tightening process data and the tightening process monitoring model, or when the tightening procedure of the target device is abnormal, the tightening procedure of the target device is identified to be abnormal.

2. The identification method according to claim 1, characterized in that: The tightening procedure self-identification model is constructed in the following way: Based on the tightening program data, system configuration data and design standard data of the characteristic sample device, a tightening program distribution map of the characteristic sample device is established; Using the tightening program data of a plurality of training sample devices to perform deviation optimization on the tightening program distribution map to output an updated tightening program distribution map; as well as The updated tightening procedure profile is determined as a tightening procedure self-identification model.

3. The identification method according to claim 1 or 2, characterized in that: The tightening program data includes one or more of the following parameters: equipment ID, program number, work station, tightening quantity, and tightening status.

4. The identification method according to claim 2, characterized in that: The system configuration data includes the device type, tightening times and tightening procedures configured in the system configuration database for the characteristic sample device; The design standard data includes the device type, tightening parameters and tightening times of the characteristic sample device in the design stage.

5. The identification method according to claim 1, characterized in that: The tightening process data includes torque data, angle data and timing data of the target device, and the tightening process monitoring model includes one or more of the following models: a parameter review model, a process monitoring model, a yield monitoring model and a statistical process control model.

6. The identification method according to claim 5, characterized in that: In the case where the tightening process monitoring model is a parameter review model, identifying the tightening process of the target device based on the tightening process data and the tightening process monitoring model includes: Based on the torque data and angle data of the target device, drawing a torque angle curve; Obtaining a full tightening process curve starting from a threshold torque from the torque angle curve; and The full tightening process curve is monitored by the parameter review model, and when a curve drop occurs in the full tightening process curve, a signal indicating that the tightening process of the target device is abnormal is output.

7. The identification method according to claim 5, characterized in that: In the case where the tightening process monitoring model is a process monitoring model, identifying the tightening process of the target device based on the tightening process data and the tightening process monitoring model includes: Drawing a torque timing curve based on the torque data and timing data of the target device; inputting a torque timing curve of the target device into the process monitoring model, wherein the process monitoring model includes a characteristic torque timing curve; and The torque timing curve of the target device is monitored by the process monitoring model in a sliding window full-quantity recognition manner, and when the similarity between the torque timing curve of the target device and the characteristic torque timing curve is less than a first set threshold, a signal of an abnormal tightening process of the target device is output.

8. The identification method according to claim 5, characterized in that: In a case where the tightening process monitoring model is a yield monitoring model, identifying the tightening process of the target device based on the tightening process data and the tightening process monitoring model includes: Based on the torque data and angle data of the target device, drawing a torque angle curve; Acquire a curve range having a slope of 0 from the torque angle curve; and The curve range is monitored by the yield monitoring model, and when the curve range is greater than a second set threshold, a signal indicating that the tightening process of the target device is abnormal is output.

9. The identification method according to claim 5, characterized in that: In the case where the tightening process monitoring model is a statistical process control model, identifying the tightening process of the target device based on the tightening process data and the tightening process monitoring model includes: Based on the torque data, angle data and timing data of the tightening process data, drawing a torque timing curve and an angle timing curve; Dynamically determining a process performance index for a plurality of continuous data in the torque timing curve and the angle timing curve; and The process performance index of the plurality of continuous data is monitored by the statistical process control model, and when the fluctuation between the plurality of continuous data and the set target process data is greater than a third set threshold, a signal that the tightening process of the target device is abnormal is output.

10. The identification method according to claim 1, characterized in that: The tightening program self-identification model is a tightening program self-identification model deployed in the cloud; and / or The tightening process monitoring model is a tightening process monitoring model deployed in the cloud.

11. A tightening state identification system, characterized in that: The identification system comprises: A data acquisition module, used to acquire tightening program data and tightening process data of a target device; A first identification module, configured to input the tightening program data into a tightening program self-identification model to identify the tightening program of the target device; and The second identification module is used to identify the tightening process of the target device based on the tightening process data and the tightening process monitoring model when there is no abnormality in the tightening program of the target device, or to identify the abnormality of the tightening program of the target device when there is an abnormality in the tightening program of the target device.

Images