Intelligent mistake-proof assembly system and method combining vision and torque detection

Abstract

The application belongs to the technical field of high-end equipment intelligent manufacturing, and particularly discloses an intelligent mistake-proof assembly system and method combined with vision and torque detection, which deeply combines process torque data and microscopic visual information, can effectively identify faults such as false torque and floating lock that cannot be found by traditional methods, realizes online real-time determination of an assembly state, reduces the dependence on personnel experience, greatly improves connection reliability, builds a strong mistake-proof closed loop, takes the quality determination result as a key control condition of a process flow procedure, forms a rigid quality gate of detection-decision-execution-tracing, effectively prevents errors caused by human negligence from flowing in the system layer, realizes online real-time full detection in a true sense, embeds detection depth into execution actions, determines the assembly quality of each screw at the moment when the screw is tightened, eliminates defects from flowing into the next process, and can effectively solve the industry common problem of floating lock.

Description

Technical Field

[0001] This invention belongs to the field of intelligent manufacturing technology for high-end equipment, and specifically discloses an intelligent error-proofing assembly system and method that integrates vision and torque detection. Background Technology

[0002] In high-reliability precision assembly fields such as aerospace fastener assembly, stringent requirements are placed on component size and assembly space, assembly positions are heterogeneous, and fastening requirements are diverse. The assembly quality of screw connections directly affects the safety and performance of the overall structure. Currently, this process mainly relies on two separate technical methods:

[0003] (1) Torque control method: Use electric or pneumatic wrenches with constant torque or torque-angle control to ensure that the applied tightening torque meets the process requirements and prevent over-tightening or under-tightening. However, this method cannot identify defects such as missing screws, incorrect screws, and floating locks. Missing screws refer to missing screws, incorrect screws refer to screws with incorrect specifications and / or models, and floating locks refer to the screws reaching the set torque value and stopping rotating before forming an effective axial clamping force due to foreign objects in the threaded hole, thread damage, or insufficient pressure of the washer. In this case, the torque parameters appear to be qualified, but the actual connection pair does not establish an effective preload, and it is prone to loosening under vibration, temperature difference, or impact load, posing a serious safety hazard.

[0004] (2) Manual visual inspection or offline inspection method: The operator performs visual inspection after assembly or performs offline sampling inspection using tools such as go / no-go gauges. This method is inefficient, relies on personnel experience, is prone to fatigue, and cannot achieve online inspection throughout the entire process, making it difficult to detect potential quality problems in a timely manner.

[0005] While existing technologies employ independent machine vision systems for assembly inspection, they typically operate from a fixed global perspective, lacking sufficient accuracy in identifying minute features of screw heads and the presence of washers. Furthermore, they are disconnected from torque data during the tightening process, making it difficult to comprehensively assess complex defects such as floating locks. Simultaneously, the inspection results are not sufficiently linked to the standard operating procedures (SOPs) of the execution end, failing to form a real-time closed loop of "inspection-decision-guidance."

[0006] Therefore, there is an urgent need for an integrated solution that can deeply integrate process torque data and microscopic visual information to achieve online real-time determination of assembly status and intelligently drive the next process. Summary of the Invention

[0007] To address the problems in the background art, this invention discloses an intelligent error-proofing assembly system and method that integrates vision and torque detection. It deeply integrates process torque data and microscopic visual information to achieve online real-time determination of assembly status and intelligently drive the next process, realizing real-time assembly status detection and process guidance for high-end equipment, forming a closed-loop quality control.

[0008] To achieve the above-mentioned objectives, the present invention adopts the following technical solution:

[0009] An intelligent error-proofing assembly system integrating vision and torque detection includes an industrial control computer and an intelligent torque gun with integrated vision. The industrial control computer runs a central control system and a SOP (Standard Operating Procedure) process management system, which are connected to the intelligent torque gun with integrated vision and the SOP process management system.

[0010] The integrated vision intelligent torque gun includes a torque gun with a vision module integrated on the side or rear of the torque gun sleeve for real-time capture of images of the screw head at the front end of the torque gun sleeve and its surrounding local area. An information acquisition module is set inside the torque gun for real-time acquisition and output of the torque value, rotation angle value and timestamp of the torque gun.

[0011] The central control system includes a real-time image processing and status recognition module, a torque-angle timing analysis module, and an information fusion decision-making and error prevention control module. The real-time image processing and status recognition module receives real-time video streams captured by the vision module and uses a lightweight deep learning model to analyze each frame of the image in real time, dynamically identifying screws, washers, and the contact status between screws and workpiece surfaces or washers. The torque-angle timing analysis module synchronously receives torque-angle-time curves uploaded by the information acquisition module and identifies abnormal patterns by analyzing curve characteristics. The information fusion decision-making and error prevention control module fuses the final visual status and torque curve within milliseconds, completing the integrated acquisition of image and torque data during the assembly process. It analyzes the assembly status of the screws based on the integrated data and makes decisions based on the analysis results. If the assembly is qualified, the SOP process management system is instructed to automatically jump to the next step; otherwise, the alarm device is triggered and the SOP process management system is instructed to lock.

[0012] The SOP process management system includes a digital SOP management module and an operation process control module. The digital SOP management module displays the standard process flow, standard operation of the specified process, and screw specifications for assembly through the SOP process guidance panel according to the instructions of the central control system, providing operators with visual process guidance. The operation process control module automatically jumps to the next step or locks the current process according to the instructions of the central control system.

[0013] Furthermore, the intelligent error-proof assembly system that integrates vision and torque detection, and the SOP process management system also include a data management and traceability module, which records the entire process torque curve, key node images before and after tightening, judgment results, time, and operator information for each screw position, forming a unique and traceable assembly data package.

[0014] Furthermore, the intelligent error-proofing assembly system integrating vision and torque detection also includes an assembly table, with workpiece assembly points set in the middle of the assembly table, conveyor belts set at the inlet and outlet ends of the workpiece assembly points, an assembly stand on the inner side of the assembly table, and cameras, lighting lights and an intelligent torque gun with integrated vision set at intervals on the assembly stand above the workpiece assembly points.

[0015] Furthermore, the intelligent error-proofing assembly system that integrates vision and torque detection is also equipped with a real-time vision detection display panel on the assembly stand. The real-time vision detection display panel is used to simultaneously display the images captured in real time by the vision module and the analysis results of the real-time image processing and status recognition module.

[0016] Furthermore, the intelligent error-proof assembly system that integrates vision and torque detection is also equipped with a torque control panel and a torque display panel on the assembly stand. The torque control panel is used to preset the threshold parameters of the torque gun assembly torque, and the torque display panel is used to display the real-time torque data of the torque gun collected by the information acquisition module in real time.

[0017] Furthermore, the intelligent error-proofing assembly system that integrates vision and torque detection is also equipped with material boxes and tool placement trays on the assembly stand.

[0018] A smart error-proofing assembly method integrating vision and torque detection is disclosed. This method utilizes the aforementioned smart error-proofing assembly system integrating vision and torque detection to achieve smart error-proofing assembly, specifically including the following steps:

[0019] S1. The real-time image processing and status recognition module receives a set of images of the screw head at the front end of the torque gun sleeve and its surrounding local area captured in real time by the vision module. It uses a lightweight deep learning model to analyze each frame of the image in real time and identify the features of the assembly-related components and operations.

[0020] S2. The torque-angle-time sequence analysis module synchronously receives the torque-angle-time curve of the torque gun that is collected and uploaded in real time by the information acquisition module. By analyzing the characteristics of the torque-angle-time curve, it confirms whether the torque curve meets the standard process window or is abnormal, and identifies the abnormal mode and issues an abnormal reminder.

[0021] S3, the information fusion decision-making and error prevention control module receives the visual recognition results from the real-time image processing and state recognition module and the torque curve analysis results from the torque-rotation time series analysis module. It then combines the visual recognition results and the torque curve analysis results to make a judgment and issue corresponding control commands:

[0022] When the visual recognition result is that the screws and washers are correct and the torque curve meets the standard process window, the step is deemed qualified. The information fusion decision and error prevention control module sends a completion signal to the SOP process management system, the operation process control module automatically unlocks, and the SOP digital management module pushes graphic and animation guidance for the next assembly operation.

[0023] If the visual recognition result is missing or incorrect installation, the step is judged to be unqualified. The information fusion decision and error prevention control module sends an unqualified signal to the SOP process management system. The operation process control module automatically locks and sends a disable command to the torque gun. At the same time, it triggers an audible and visual alarm and highlights the error type and image on the real-time visual inspection display panel. After replacing the correct screw and / or washer, the vision module re-captures the image. Only when the visual recognition result of the real-time image processing and status recognition module is that the screw and washer are correct will the operation process control module automatically unlock and send an enable command to the torque gun. Then, steps S2 and S3 are repeated.

[0024] If the visual recognition result is qualified, but the torque curve analysis indicates abnormal torque, then the step is judged to be unqualified. The information fusion decision and error prevention control module sends a defect reminder to the SOP process management system. The operation process control module automatically locks the current process and highlights the error type and image on the torque display panel. When the torque curve meets the standard process window, the step is judged to be qualified. The information fusion decision and error prevention control module sends a completion signal to the SOP process management system. The operation process control module automatically unlocks, and the SOP digital management module pushes graphic and animation guidance for the next assembly operation.

[0025] S4. The data management and traceability module records the torque curve, key node images before and after tightening, judgment results, time and operator information for each screw position throughout the entire process, forming a unique and traceable assembly data package.

[0026] Furthermore, in the intelligent error-proof assembly method that integrates vision and torque detection, the identification features in step S1 include the presence and quantity of screws, the screw head groove type, markings, and color characteristics, the application of thread-locking adhesive and the control of the settling time of thread accelerator, the presence and type of washers, the contact state between the screw and the workpiece surface, and monitoring the disappearance of the gap between the screw head and the workpiece or washer during the tightening process.

[0027] Compared with the prior art, the beneficial effects of the present invention are:

[0028] The intelligent error-proofing assembly system disclosed in this invention integrates vision and torque detection. It deeply integrates process torque data and microscopic visual information to achieve online real-time determination of assembly status and intelligently drive the next process. It improves intelligence and traceability, replaces manual visual inspection, reduces reliance on human experience, and provides rich data assets for quality analysis, process optimization, and problem tracing through automatic guidance and accurate alarms of the SOP system. It reduces the cognitive load on operators, improves work standardization and efficiency, and has a user-friendly human-machine interaction.

[0029] This invention discloses an intelligent error-proof assembly method that integrates vision and torque detection. It innovatively combines microscopic visual fitting status monitoring with macroscopic torque curve analysis, which can effectively identify false torque faults that traditional methods cannot detect, greatly improving connection reliability and constructing a strong error-proof closed loop. It uses the quality judgment result as a key control condition for the process sequence, forming a rigid quality gate of detection-decision-execution-traceability. It effectively prevents errors caused by human negligence from flowing to the next process at the system level, realizing true online real-time full inspection. It embeds the detection depth into the execution action, and the assembly quality of each screw is judged the moment it is tightened, preventing defects from flowing into the next process. It can effectively solve the common problem of floating locks in the industry. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of the overall structure of the intelligent error-proofing assembly system that integrates vision and torque detection according to the present invention.

[0031] In the diagram: 1-Assembly stand; 2-SOP process guidance panel; 3-Real-time visual inspection display panel; 4-Material box; 5-Conveyor belt; 6-Camera; 7-Lighting; 8-Alarm device; 9-Torque control panel; 10-Tool placement tray; 11-Intelligent torque gun with integrated vision; 12-Torque display panel; 13-Workpiece assembly point; 14-Assembly table; 15-Industrial computer;

[0032] Figure 2 This is a flowchart illustrating the workflow and decision-making logic of the intelligent error-proofing assembly system that integrates vision and torque detection according to the present invention.

[0033] Figure 3 This is a schematic diagram of the visual inspection system in this invention;

[0034] Figure 4 This is a schematic diagram of the SOP process management system in this invention;

[0035] Figure 5 This is a schematic diagram showing the normal and abnormal curves of the endoscopic intelligent torque gun in this invention. Detailed Implementation

[0036] To better understand the present invention, the following embodiments further illustrate the content of the invention, but the scope of protection of the present invention is not limited to the following embodiments. Numerous specific details are set forth in the following description to provide a more thorough understanding of the invention. However, it will be apparent to those skilled in the art that the present invention can be practiced without one or more of these details.

[0037] In the description of this invention, unless otherwise explicitly specified and limited, the terms "connected," "linked," and "fixed" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0038] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" of the second feature includes the first feature directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature includes the first feature directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.

[0039] In the description of this embodiment, the terms "upper," "lower," "left," and "right," etc., refer to the orientation or positional relationship shown in the accompanying drawings. They are used only for ease of description and simplification of operation, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the present invention. In addition, the terms "first" and "second" are used only for distinction in description and have no special meaning.

[0040] Combined with appendix Figure 1-5 This invention details an intelligent error-proofing assembly system integrating vision and torque detection, comprising an industrial control computer 15 and an intelligent torque gun 11 with integrated vision. The industrial control computer 15 runs a central control system and a SOP process management system, which are connected to the intelligent torque gun 11 with integrated vision and the SOP process management system.

[0041] The integrated vision intelligent torque gun 11 is an endoscopic intelligent torque gun. The vision module integrates a miniature high-resolution endoscope camera and a matching ring LED illumination source on the side or rear of the torque gun sleeve. The optical axis of the camera is parallel to or at a fixed small angle to the central axis of the sleeve, ensuring that its field of view can clearly capture images of the screw head and washer at the front end of the sleeve, as well as the local area around the workpiece surface. The information acquisition module is a built-in sensor of the torque gun. This sensor collects and outputs torque value, angle value, and timestamp in real time. It should be noted that the specific mechanical structure of the endoscopic intelligent torque gun can adopt existing mature solutions and is existing technology, so it will not be described in detail here.

[0042] The central control system includes a real-time image processing and status recognition module, a torque-angle timing analysis module, and an information fusion decision and error prevention control module. The real-time image processing and status recognition module receives real-time video streams captured by the vision module and uses a lightweight deep learning model to perform real-time analysis on each frame of the image, dynamically identifying the following features: (1) the presence and quantity of screws to prevent omissions; (2) the characteristics of the screw head groove type, markings, and colors, used to compare with preset standard parts to prevent mis-installation; (3) the omission of thread adhesive coating and the control of the static time of thread accelerator; (4) the presence and type of gaskets to prevent omissions or misuse of gaskets; (5) the contact state between the screw and the workpiece surface, and during the tightening process, monitoring the disappearance of the gap between the screw head and the workpiece or gasket.

[0043] The torque-angle timing analysis module synchronously receives the torque-angle-time curve uploaded by the information acquisition module. By analyzing the curve characteristics, it identifies abnormal patterns, such as premature torque saturation indicating float lock and abnormal torque fluctuation. The information fusion decision and error prevention control module fuses the final visual state with the torque curve in real time, completing the integrated acquisition of image and torque data during the assembly process. It analyzes the assembly status of the screws based on the integrated data and makes decisions based on the analysis results. If the assembly is qualified, it instructs the SOP process management system to automatically jump to the next step; otherwise, it triggers the alarm device 8 and instructs the SOP process management system to lock.

[0044] The SOP process management system includes a digital SOP management module and an operation process control module. The digital SOP management module is responsible for the digital storage, updating, and retrieval of assembly process documents. Based on the instructions from the central control system, the digital SOP management module displays the standard process flow, the standard operation of the specified process, and the screw specifications for assembly through the SOP process guidance panel 2, providing operators with visual process guidance. The operation process control module only allows the SOP process management system to automatically jump to the next process after receiving the assembly qualification instruction from the central control system; otherwise, it locks the current process. The operation process control module realizes real-time guidance of assembly operation steps and supervision of operation specifications.

[0045] As an optional design, the intelligent error-proof assembly system that integrates vision and torque detection is preferred. The SOP process management system also includes a data management and traceability module, which records the torque curve of each screw position throughout the entire process, images of key nodes before and after tightening, judgment results, time and operator information, forming a unique and traceable assembly data package.

[0046] As an optional design, the preferred intelligent error-proofing assembly system integrating vision and torque detection also includes an assembly table 14. A workpiece assembly point 13 is located in the center of the assembly table 14. Conveyor belts 5 are installed at the inlet and outlet ends of the workpiece assembly point 13. An assembly stand 1 is installed on the inner side of the assembly table 14. Cameras 6, lighting 7, and an intelligent torque gun 11 with integrated vision are spaced apart on the assembly stand 1 above the workpiece assembly point 13. The assembly table 14 and assembly stand 1 serve as the basic load-bearing structure of the entire workstation, fixing the installation positions of all functional modules and ensuring the spatial layout stability of each component. The lighting 7 provides sufficient light for the camera 6 to capture images and for operators to work, ensuring image clarity and avoiding detection errors caused by insufficient light. The camera 6 captures real-time images of the workpiece assembly point 13 and transmits the image data to the real-time vision inspection display panel 3, providing an image basis for visual error prevention. Simultaneously, the images captured by the camera 6 can assist the industrial control computer in identifying whether the material placement and operational actions at the workpiece assembly point 13 are compliant.

[0047] As an optional design, the intelligent error-proofing assembly system that integrates vision and torque detection is preferred. A real-time vision detection display panel 3 is also provided on the assembly stand 1. The real-time vision detection display panel 3 is used to simultaneously display the images captured in real time by the vision module and the analysis results of the real-time image processing and status recognition module.

[0048] As an optional design, the intelligent error-proof assembly system that integrates vision and torque detection is preferred. On the assembly stand plate 1, a torque control panel 9 and a torque display panel 12 are also provided. The torque control panel 9 is used to preset the threshold parameter of the torque gun assembly torque, and the torque display panel 12 is used to display the real-time torque data of the torque gun collected in real time by the information acquisition module.

[0049] As an optional design, the intelligent error-proofing assembly system integrating vision and torque detection is preferred. The assembly stand 1 is also equipped with a material box 4 and a tool tray 10, which respectively store the materials and tools required for assembly, forming a close-to-work layout with the workpiece assembly point 13 to improve operational efficiency. Simultaneously, the video stream captured by the camera 6 of the material box 4 and tool tray 10 identifies whether materials are missing from the material box 4 and whether tools are properly positioned in the tool tray 10, thus preventing material / tool ​​errors. The assembly stand 1 is also equipped with an audible and visual alarm device 8, which receives trigger signals from the central control system and issues a light alarm in cases of visual detection abnormalities, excessive torque, or missing materials, reminding operators to handle the situation promptly.

[0050] A smart error-proofing assembly method integrating vision and torque detection is disclosed. This method utilizes the aforementioned smart error-proofing assembly system integrating vision and torque detection to achieve smart error-proofing assembly, specifically including the following steps:

[0051] S1. The real-time image processing and status recognition module receives a set of images of the screw head at the front end of the torque gun sleeve and its surrounding local area captured in real time by the vision module. It uses a lightweight deep learning model to analyze each frame of the image in real time and identify the features of the assembly-related components and operations.

[0052] S2. The torque-angle-time sequence analysis module synchronously receives the torque-angle-time curve of the torque gun that is collected and uploaded in real time by the information acquisition module. By analyzing the characteristics of the torque-angle-time curve, it confirms whether the torque curve meets the standard process window or is abnormal, and identifies the abnormal mode and issues an abnormal reminder.

[0053] S3, the information fusion decision-making and error prevention control module receives the visual recognition results from the real-time image processing and state recognition module and the torque curve analysis results from the torque-rotation angle timing analysis module. It then combines the visual recognition results and the torque curve analysis results to make a judgment and issue corresponding control commands:

[0054] When the visual recognition result is that the screws and washers are correct and the torque curve meets the standard process window, that is, the final torque is qualified and the torque rise slope is normal, the process step is judged to be qualified. The information fusion decision and error prevention control module sends a completion signal to the SOP process management system, the operation process control module automatically unlocks, and the SOP digital management module pushes the graphic and animation guidance for the next assembly operation.

[0055] When the visual recognition result is missing or incorrect installation, the step is judged to be unqualified. The information fusion decision and error prevention control module sends an unqualified signal to the SOP process management system. The operation process control module automatically locks and sends a disable command to the torque gun to prevent the torque gun from being used dry or accidentally tightened. At the same time, an audible and visual alarm is triggered, and the error type and image are highlighted on the visual inspection real-time display panel 3. The operation process control module will automatically unlock and send an enable command to the torque gun when the operator replaces the correct screw and / or washer. Then, steps S2 and S3 are repeated.

[0056] If the visual recognition result is qualified, but the torque curve analysis indicates an abnormal torque, such as when the torque is reached at too small an angle, the torque curve analysis indicates a risk of floating lock. In this case, the step is judged to be unqualified. The information fusion decision and error prevention control module sends a defect reminder to the SOP process management system. The operation process control module automatically locks the current process and highlights the error type and image on the torque display panel 12 until the operator operates correctly. When the torque curve meets the standard process window, the step is judged to be qualified. The information fusion decision and error prevention control module sends a completion signal to the SOP process management system. The operation process control module automatically unlocks. The SOP digital management module pushes graphic and animation guidance for the next assembly operation.

[0057] S4. The data management and traceability module records the torque curve, key node images before and after tightening, judgment results, time and operator information for each screw position throughout the entire process, forming a unique and traceable assembly data package.

[0058] Figure 5 The diagram shows a comparison of torque-angle curves for normal assembly and floating lock defects. This diagram is the core basis for the intelligent detection of assembly defects achieved by the present invention through torque-angle curve feature analysis. It works in conjunction with the vision module and the central control system to achieve the linkage between SOP process guidance and error prevention mechanisms during the assembly process. The diagram presents the torque-angle curve characteristics of normal assembly and various assembly defects, as well as the defect judgment logic. The horizontal axis of the coordinate system represents the tightening angle during the assembly process, and the vertical axis represents the torque. The curve in the diagram shows the torque-angle change trend during the assembly tightening process. The gray area in the middle is the normal torque range, which is the core reference range for judging the assembly qualification. Different defect zones are divided around the normal torque range, including assembly defects such as repeated tightening, excessive tightening speed, abnormal threads / threaded holes or product surfaces, insufficient screw strength, and stripped threads. Combined with the endoscopic intelligent torque gun vision module, the assembly defects such as floating locks are accurately identified and classified in real time.

[0059] Take the screw assembly of the mounting edge of an aircraft engine casing as an example:

[0060] Preparation: The SOP process guidance panel 2 shows that the screw specifications to be assembled are: M2 (M2.5, M3, M3.5), cross-slot, plated, and stainless steel elastic washers need to be installed. The operator takes the parts from the material box 4.

[0061] Alignment and Pre-inspection: The operator places the endoscopic intelligent torque gun sleeve, which integrates an endoscope, onto the screw. The miniature camera on the gun body starts working as it approaches the screw, capturing real-time images of the screw head and its surrounding local area at the front end of the torque gun sleeve. The real-time image processing and status recognition module identifies objects in the field of view in real time. If the screw and washer are identified as the correct specifications, the green light on the gun body will light up slightly, allowing start-up; if the incorrect part is identified or only a missing screw hole is visible, the red light on the gun body will light up and the trigger will be electronically locked. At the same time, the visual detection real-time display panel 3 will magnify and display the error image.

[0062] Tightening and Synchronization Monitoring: When the operator pulls the trigger, the torque gun begins to tighten, and at the same time, the central control system executes synchronously.

[0063] The real-time image processing and status recognition module performs visual monitoring: continuously analyzes the video image stream to confirm that the shim has not been ejected during the tightening process, and determines that the screw head has completely adhered to the workpiece or shim without gaps at the end;

[0064] The torque-angle timing analysis module monitors torque and records the complete torque-angle curve.

[0065] The information fusion decision and error prevention control module performs fusion judgment: At the moment the tightening action ends, the information fusion decision and error prevention control module receives the tightening completion signal. It retrieves the final visual judgment result and torque curve analysis result of the screw position. The final visual judgment result is that the fit is good, and the torque curve analysis result is that the torque value is 0.45 Nm (within the allowable tolerance range) and the 180-degree rotation angle is within the normal range. Both are qualified, and the module sends a process qualified signal to the upper MES / SOP process management system.

[0066] The operation process control module advances the process: After the SOP process management system receives the signal, the operation process control module changes the current process status mark to green "completed". The SOP digital management module automatically jumps to the SOP process guidance panel 2 and provides graphic guidance to the operator to assemble the next screw or perform subsequent processes, such as applying protective paint.

[0067] Anomaly Handling: If, during the fusion judgment step, the visual judgment is qualified but the torque curve shows that it reaches 12Nm after only 30 degrees of rotation, the information fusion decision and error prevention control module sends a suspected floating lock defect reminder to the SOP process management system, triggering a yellow alarm. The operation process control module is locked, and the torque display panel 12 displays the alarm information and a comparison chart of the floating lock curve and the normal curve. Process quality management personnel need to intervene, use higher precision instruments to re-examine and handle the problem, and record the cause in the system before the manual authorization unlocking process can be performed.

[0068] This invention discloses an intelligent error-proofing assembly system and method integrating vision and torque detection, aiming to solve the problem of difficulty in real-time detection of missing, incorrect, and floating screws in precision screw assembly. It achieves the linkage between SOP process guidance and error-proofing mechanisms through torque-angle curve feature analysis, the combined action of a vision module and a central control system. During operation, the central control system simultaneously acquires microscopic visual images of the screw head and the torque-angle curve of the tightening process. Image recognition technology is used to prevent missing and incorrect screws, and information fusion decision-making and error-proofing control modules combine visual fit status and torque curve features to diagnose floating defects and make a comprehensive judgment. Only when the inspection is qualified is the SOP process management system allowed to proceed to the next step; otherwise, an alarm is immediately triggered and the process is locked. Simultaneously, traceable data packets are recorded throughout the process. This invention achieves online real-time full inspection and closed-loop control of assembly quality, significantly improving the error-proofing capability and quality consistency of high-reliability assembly.

[0069] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Any other modifications or equivalent substitutions made by those skilled in the art to the technical solutions of the present invention, as long as they do not depart from the spirit and scope of the technical solutions of the present invention, should be covered within the scope of the claims of the present invention.

Claims

1. An intelligent error-proofing assembly system integrating vision and torque detection, characterized in that, The system includes an industrial control computer and an intelligent torque gun with integrated vision. The industrial control computer runs a central control system and a SOP (Standard Operating Procedure) process management system, which are connected to the intelligent torque gun with integrated vision and the SOP process management system. The integrated vision intelligent torque gun includes a torque gun with a vision module integrated on the side or rear of the torque gun sleeve for real-time capture of images of the screw head at the front end of the torque gun sleeve and its surrounding local area. An information acquisition module is set inside the torque gun for real-time acquisition and output of the torque value, rotation angle value and timestamp of the torque gun. The central control system includes a real-time image processing and status recognition module, a torque-angle timing analysis module, and an information fusion decision-making and error prevention control module. The real-time image processing and status recognition module receives real-time video streams captured by the vision module and uses a lightweight deep learning model to analyze each frame of the image in real time, dynamically identifying screws, washers, and the contact status between screws and workpiece surfaces or washers. The torque-angle timing analysis module synchronously receives torque-angle-time curves uploaded by the information acquisition module and identifies abnormal patterns by analyzing curve characteristics. The information fusion decision-making and error prevention control module fuses the final visual state with the torque curve in real time, completing the integrated acquisition of image and torque data during the assembly process. Based on the integrated data, it analyzes the assembly status of the screws and makes decisions based on the analysis results. If the assembly is qualified, the SOP process management system is instructed to automatically jump to the next step; otherwise, the alarm device is triggered and the SOP process management system is instructed to lock. The SOP process management system includes a digital SOP management module and an operation process control module. The digital SOP management module displays the standard process flow, standard operation of the specified process, and screw specifications for assembly through the SOP process guidance panel according to the instructions of the central control system, providing operators with visual process guidance. The operation process control module only allows the SOP process management system to automatically jump to the next process after receiving the assembly qualification instruction from the central control system; otherwise, it locks the current process.

2. The intelligent error-proofing assembly system integrating vision and torque detection according to claim 1, characterized in that, The SOP process management system also includes a data management and traceability module, which records the torque curve of each screw position throughout the entire process, images of key nodes before and after tightening, judgment results, time, and operator information, forming a unique and traceable assembly data package.

3. The intelligent error-proofing assembly system integrating vision and torque detection according to claim 2, characterized in that, It also includes an assembly table, with workpiece assembly points set in the middle of the assembly table. Conveyor belts are set at the inlet and outlet ends of the workpiece assembly points. An assembly stand is set on the inner side of the assembly table. Cameras, lighting and intelligent torque guns with integrated vision are set at intervals on the assembly stand above the workpiece assembly points.

4. The intelligent error-proofing assembly system integrating vision and torque detection according to claim 3, characterized in that, in The assembly stand is also equipped with a real-time visual inspection display panel, which is used to simultaneously display the images captured in real time by the vision module and the analysis results of the real-time image processing and status recognition module.

5. The intelligent error-proofing assembly system integrating vision and torque detection according to claim 3, characterized in that, in The assembly stand is also equipped with a torque control panel and a torque display panel. The torque control panel is used to preset the threshold parameters of the torque gun assembly torque, and the torque display panel is used to display the real-time torque data of the torque gun collected by the information acquisition module.

6. The intelligent error-proofing assembly system integrating vision and torque detection according to claim 3, characterized in that, Material boxes and tool trays are also provided on the assembly stand.

7. An intelligent error-proofing assembly method integrating vision and torque detection, characterized in that, The intelligent error-proofing assembly system that integrates vision and torque detection as described in any one of claims 1-6 is used to achieve intelligent error-proofing assembly, specifically including the following steps: S1. The real-time image processing and status recognition module receives a set of images of the screw head at the front end of the torque gun sleeve and its surrounding local area captured in real time by the vision module. It uses a lightweight deep learning model to analyze each frame of the image in real time and identify the features of the assembly-related components and operations. S2. The torque-angle-time sequence analysis module synchronously receives the torque-angle-time curve of the torque gun that is collected and uploaded in real time by the information acquisition module. By analyzing the characteristics of the torque-angle-time curve, it confirms whether the torque curve meets the standard process window or is abnormal, and identifies the abnormal mode and issues an abnormal reminder. S3, the information fusion decision-making and error prevention control module receives the visual recognition results from the real-time image processing and state recognition module and the torque curve analysis results from the torque-rotation angle timing analysis module. It then combines the visual recognition results and the torque curve analysis results to make a judgment and issue corresponding control commands: When the visual recognition result is that the screws and washers are correct and the torque curve meets the standard process window, the step is deemed qualified. The information fusion decision and error prevention control module sends a completion signal to the SOP process management system, the operation process control module automatically unlocks, and the SOP digital management module pushes graphic and animation guidance for the next assembly operation. If the visual recognition result is missing or incorrect installation, the step is judged to be unqualified. The information fusion decision and error prevention control module sends an unqualified signal to the SOP process management system. The operation process control module automatically locks and sends a disable command to the torque gun. At the same time, it triggers an audible and visual alarm and highlights the error type and image on the real-time visual inspection display panel. After replacing the correct screw and / or washer, the vision module re-captures the image. Only when the visual recognition result of the real-time image processing and status recognition module is that the screw and washer are correct will the operation process control module automatically unlock and send an enable command to the torque gun. Then, steps S2 and S3 are repeated. If the visual recognition result is qualified, but the torque curve analysis indicates abnormal torque, then the step is judged to be unqualified. The information fusion decision and error prevention control module sends a defect reminder to the SOP process management system. The operation process control module automatically locks the current process and highlights the error type and image on the torque display panel. When the torque curve meets the standard process window, the step is judged to be qualified. The information fusion decision and error prevention control module sends a completion signal to the SOP process management system. The operation process control module automatically unlocks, and the SOP digital management module pushes graphic and animation guidance for the next assembly operation. S4. The data management and traceability module records the torque curve, key node images before and after tightening, judgment results, time and operator information for each screw position throughout the entire process, forming a unique and traceable assembly data package.

8. The intelligent error-proofing assembly method integrating vision and torque detection according to claim 7, characterized in that, The identification features in step S1 include the presence and number of screws, the screw head groove type, markings and color characteristics, the application of threadlocker and the control of the settling time of thread accelerator, the presence and type of washer, the contact state between the screw and the workpiece surface, and the disappearance of the gap between the screw head and the workpiece or washer during the tightening process.

Images