Self-recognition gluing system

Abstract

The invention discloses a self-recognition gluing system which comprises a positioning mold and a gluing device, the gluing device is integrated with a glue gun, a glue cleaning scraper, a multi-axis movement assembly and a glue overflow detection module, a two-dimensional contour and three-dimensional point cloud model is constructed through fusion of a linear array visual sensor and a laser distance measuring sensor, preset gluing path deviation is compared in real time, and the glue overflow detection module is used for detecting glue overflow. The system comprises a driving glue cleaning scraper, a bearing platform, a glue gun, an AI self-learning module and a motion error compensation module, wherein the driving glue cleaning scraper is used for realizing self-adaptive glue cleaning through a flexible hinge, a force control sensor and a shape memory alloy mechanism; a variable-density gluing path can be generated, the glue overflow risk can be predicted, process parameters can be optimized, through multi-sensor fusion detection, intelligent control and self-adaptive mechanism design, the problems that a traditional gluing system is insufficient in positioning precision, incomplete in detection and poor in glue cleaning flexibility are solved, and the automation and intelligence level of sealing ring gluing is remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic gluing equipment, in particular to a self-identification gluing system. BACKGROUND

[0002] The traditional sealing ring gluing system has the following technical bottlenecks: first, the positioning device usually adopts fixed clamping force, which cannot adapt to sealing rings of different hardness materials, and is prone to cause workpiece deformation or positioning failure, especially for materials with large difference in elastic modulus such as silicone and fluororubber, the fixed clamping force will cause the cross-section of the sealing ring to deform or the positioning to deviate; second, the glue overflow detection relies on a single visual sensor, which can only obtain two-dimensional plane information, and it is difficult to accurately identify the three-dimensional space glue shape deviation, resulting in uneven glue thickness or boundary glue overflow problem difficult to effectively control; third, the glue removal process is mechanically operated, and lacks self-adaptive protection for glue adhesion shape and mold surface, and the traditional rigid scraper is easy to damage the mold surface and difficult to completely remove the sticky glue; fourth, the gluing path planning and process parameter adjustment rely on manual experience, and the intelligent degree is low, which is difficult to meet the high-precision and flexible production demand, especially in the multi-variety and small-batch production scene, the process switching efficiency is low, and in the prior art, the collaborative optimization scheme for multi-dimensional dynamic positioning control, three-dimensional glue overflow detection and intelligent glue removal is still blank, and there is lack of data interaction and collaborative control mechanism between the functional modules, and the whole process closed-loop management and control of the gluing quality cannot be realized. SUMMARY

[0003] The purpose of the present application is to provide a self-identification gluing system and its positioning mold, gluing device, glue overflow detection module, glue removal scraper, bearing platform and glue gun, which has the advantages of being able to dynamically adjust the clamping force of the sealing ring to prevent deformation, accurately detect the three-dimensional space glue shape, and adaptively remove the glue overflow.

[0004] In order to solve the above technical problems, the present application solves the problems by the following technical scheme: a self-identification gluing system, comprising a positioning mold and a gluing device; the positioning mold is provided with a bearing platform for driving it to rotate, and the surface is provided with a profiling positioning groove matched with the profile of the sealing ring, and the inner wall of the positioning groove is distributed with elastic limiting protrusions; the gluing device comprises a glue gun, a glue removal scraper, a multi-axis motion assembly for driving the two to move, and a glue overflow detection module, characterized in that: The elastic limiting protrusions are connected with pressure sensors to form a clamping force closed-loop control system, which can dynamically adjust the clamping force according to the hardness of the sealing ring material; The glue overflow detection module detects the three-dimensional shape and boundary profile of the glue in real time through multi-sensor fusion technology, and compares the three-dimensional space deviation with the preset gluing path; The glue removal scraper performs adaptive glue removal operation based on the deviation comparison result.

[0005] By adopting the technical scheme, the clamping force closed-loop control system formed by the elastic limiting protrusion and the pressure sensor realizes dynamic self-adaptive adjustment of the clamping force of the sealing ring of different hardness materials, avoids workpiece deformation or positioning failure caused by fixed clamping force, the multi-sensor fusion glue overflow detection module can accurately capture the three-dimensional space deviation of the glue form, and the detection accuracy is significantly improved, the glue scraping knife performs adaptive removal based on the deviation data, solves the damage of the mold surface caused by mechanical glue removal, significantly improves the glue removal qualification rate, and improves the flexible control ability of the glue application system.

[0006] The application further provides that the glue overflow detection module comprises a line array vision sensor and a laser ranging sensor. The line array vision sensor constructs a two-dimensional edge contour model of the glue application area, and the laser ranging sensor generates a three-dimensional point cloud model of the glue three-dimensional form and boundary contour. The glue overflow detection module is internally provided with a space geometry matching processing unit for deviation analysis of real-time detection data and a three-dimensional digital model of a preset glue application path.

[0007] By adopting the technical scheme, the fusion application of the line array vision sensor and the laser ranging sensor constructs a composite detection system of "two-dimensional edge contour + three-dimensional point cloud model" of the glue application area, breaks through the detection dimension limitation of the traditional single sensor, realizes high-precision detection of the glue form, the internally provided space geometry matching processing unit can calculate the three-dimensional space deviation in real time, provides accurate position compensation data for the glue scraping execution mechanism, effectively improves the space positioning accuracy, and provides data support for high-precision glue scraping.

[0008] The application further provides that the glue scraping knife is connected to a multi-axis motion assembly through a flexible hinge structure and is integrated with a force control sensor. The force control sensor realizes contact force closed-loop control, so that the scraper automatically adjusts the scraping trajectory according to the glue adhesion form. The glue scraping knife is covered with a super-hydrophobic coating, and the positioning mold is provided with a residual glue pre-detection module, the residual glue pre-detection module is in communication connection with the glue overflow detection module and is used for secondary verification of glue removal, and the residual glue pre-detection module comprises an infrared thermal imaging unit and a structured light scanner.

[0009] By adopting the technical scheme, the integration of the flexible hinge structure and the force control sensor enables the glue scraping knife to realize accurate contact force closed-loop control when contacting the glue, avoids mold scratching caused by rigid contact, the super-hydrophobic coating significantly reduces the adhesion between the glue and the scraper surface, improves the glue removal efficiency, and the residual glue pre-detection module can identify trace residual glue through infrared thermal imaging and structured light scanning, solves the problem of missed detection caused by traditional manual detection or single sensor, and provides a pretreatment basis for subsequent precise glue removal.

[0010] The application is further provided with the residual glue pre-detection module which identifies the trace residual glue by the temperature field distribution difference relative to the mold base; The glue cleaning scraper is provided with a height self-adaptive mechanism driven by a shape memory alloy, and the working angle is adjusted in real time according to the residual glue three-dimensional topography obtained by the structured light scanner. The height self-adaptive mechanism is integrated with a vibration frequency adjusting unit for peeling off the cured glue marks and protecting the mold surface.

[0011] By adopting the above technical scheme, the height self-adaptive mechanism driven by the shape memory alloy can quickly complete the dynamic adjustment of the scraper working angle and adapt to the residual glue topography of different heights; the vibration frequency adjusting unit can effectively peel off the cured glue marks and reduce the mold surface wear; combined with the three-dimensional topography data of the structured light scanner, the complex curved surface residual glue can be completely removed in all attitudes, and the success rate of removing the cured glue marks is improved.

[0012] The application is further provided with the bearing platform which is redundantly driven by double servo motors and realizes full closed loop angle control through a resolver. The bottom of the positioning mold is provided with a quick-change base which is integrated with a mechanical self-locking structure, an electromagnetic positioning unit and a chip type identity recognition module and is used for automatically matching the glue applying parameters of different sealing ring types.

[0013] By adopting the above technical scheme, the double servo motor redundancy driving and the resolver full closed loop control significantly improve the angle positioning accuracy of the bearing platform and meet the high-precision rotary glue applying requirements; the chip type identity recognition module integrated with the quick-change base can quickly complete the mold type identification and automatically match the glue applying parameters, greatly improves the efficiency compared with the traditional manual parameter adjustment and significantly improves the change production efficiency of various sealing rings.

[0014] The application is further provided with the glue gun which is integrated with an ultrasonic viscosity sensor and a precision valve needle driving system. The glue gun is built-in with a glue applying path planning module which can generate a glue line density variable glue applying path in combination with the sealing ring cross section curvature and realize the glue deposition control.

[0015] By adopting the above technical scheme, the ultrasonic viscosity sensor can monitor the glue viscosity change in real time, and the variable density glue applying path generated in combination with the sealing ring cross section curvature can significantly improve the glue deposition uniformity of the complex curved surface; the precision valve needle driving system cooperates with the path planning module to realize the high-precision glue output control, solves the glue accumulation or shortage problem in the curvature change area of the traditional fixed path glue applying and improves the glue applying consistency.

[0016] The application is further provided with the AI self-learning module which constructs a multi-modal neural network model and is used for fusing the visual detection data and the process parameters. The AI self-learning module can predict the glue overflow probability under different working conditions and generate a risk heat map, and realize interactive optimization of process parameters through a man-machine collaborative interface.

[0017] By adopting the above technical scheme, the AI self-learning module fuses visual detection data and process parameters through a multi-modal neural network, establishes a high-precision overflow risk prediction model, and can early warn of overflow risk; the generated risk heat map supports man-machine collaborative parameter optimization, significantly shortens process debugging time, and continuously improves system adaptability through long-term learning, and realizes intelligent upgrading.

[0018] The application further provides that the servo drive system of the multi-axis motion assembly is integrated with a motion error compensation module for real-time correction of the positioning accuracy of the end of the mechanical arm. The motion error compensation module comprises a sensor detection unit and a control unit, the sensor detection unit is used to acquire joint angle deviation, angular velocity and acceleration parameters, and the control unit generates a position compensation amount according to the parameters.

[0019] By adopting the above technical scheme, the motion error compensation module acquires joint angle deviation, angular velocity and other parameters in real time, generates a position compensation amount through a control algorithm, significantly improves the positioning accuracy of the end of the mechanical arm, effectively compensates for the error influence of mechanical transmission clearance, thermal deformation and the like, ensures the trajectory accuracy of the glue gun and the glue removing scraper in high-speed motion, and meets the stringent requirements of high-precision glue coating processes on motion control.

[0020] The application further provides that the elastic limiting protrusions on the inner wall of the profiling positioning groove are distributed in an array, and each elastic limiting protrusion is independently connected with a pressure sensor. The bearing platform is provided with an adaptive leveling unit for adjusting the levelness of the bearing platform according to the placement posture of the sealing ring.

[0021] By adopting the above technical scheme, the arrayed independent pressure sensors realize distributed clamping force monitoring of the sealing ring, and the adaptive leveling unit dynamically adjusts the levelness of the bearing platform, thereby solving the positioning inclination problem of the sealing ring caused by placement posture deviation, effectively improving the flatness accuracy of workpiece positioning, and improving the position consistency of subsequent glue coating.

[0022] The application further provides that the glue coating device further comprises a glue preheating module for pre-treating the glue to maintain the stability of the rheological properties of the glue, so that the temperature of the glue is maintained within a preset range. The glue gun is provided with a posture adjusting unit capable of adjusting the glue discharging direction of the glue gun according to the spatial angle of the glue coating position.

[0023] By adopting the technical scheme, the glue preheating module controls the glue temperature in a preset range, ensures stable glue viscosity, and significantly reduces glue coating flow fluctuation; the glue gun posture adjusting unit can realize full-angle glue discharging direction adjustment, meets the glue coating demand of complex structures such as deep cavity and reverse buckle, avoids glue accumulation or wire drawing defects caused by fixed glue discharging angle, and improves the glue coating yield of special sealing rings.

[0024] The self-identification glue coating system provided by the application breaks through the precision bottleneck and adaptability limitation of the traditional glue coating system through deep cooperation of mechanical structure innovation, sensor fusion detection and intelligent algorithm driving, and significantly improves the automation level, production quality stability and complex working condition adaptation capability of the sealing ring glue coating process, thereby providing an intelligent solution for high-end sealing component manufacturing. The positioning link: through distributed clamping force control of the elastic limiting protrusion and the pressure sensor array, combined with adaptive leveling of the bearing platform and double servo motor redundant driving technology, dynamic and accurate positioning of sealing rings with different materials and postures is realized, clamping deformation and positioning deviation are avoided, and a foundation for high-precision glue coating is laid. The detection link: a multi-modal detection scheme of line array vision sensor and laser ranging sensor fusion is adopted, a two-dimensional contour and three-dimensional point cloud composite model of glue form is constructed, the dimensional limitation of traditional two-dimensional detection is broken through, and spatial deviation in the glue coating process is captured in real time, thereby providing high-precision data support for the subsequent execution mechanism. The execution link: the glue cleaning scraper integrates a flexible hinge, a force control sensor and a shape memory alloy driving mechanism, contact force closed-loop control and working angle adaptive adjustment are realized based on detection data, combined with super-hydrophobic coating and vibration peeling technology, the problem of incomplete removal of solidified glue marks and mold damage is solved, and full-posture flexible removal of residual glue on complex curved surfaces is realized; the glue gun is adapted to the change of the cross-sectional curvature of the sealing ring and the complex spatial structure through the ultrasonic viscosity sensor and the variable density path planning technology, and cooperates with the posture adjusting unit to realize uniformity control of glue deposition. Intelligence and automation: the AI self-learning module fuses multi-source data to construct an overflow glue risk prediction model, supports man-machine collaborative parameter optimization, significantly improves process debugging efficiency and system adaptability, and realizes automatic identification and high-precision motion control of multiple varieties of molds through the chip type identification module and motion error compensation technology integrated in the quick-change base, thereby greatly shortening the changeover time and meeting the flexible production demand. Overall, through mechanical structure innovation, sensor fusion detection and intelligent algorithm driven deep cooperation, the system breaks through the precision bottleneck and adaptability limitation of the traditional glue coating system, significantly improves the automation level, production quality stability and complex working condition adaptation capability of the sealing ring glue coating process, and provides an intelligent solution for high-end sealing component manufacturing. BRIEF DESCRIPTION OF DRAWINGS

[0025] Figure 1 is a schematic view of a positioning mold and a glue applying device; Figure 2 is Figure 1 is an enlarged schematic view at A in FIG. 1.

[0026] The parts referred to by the reference numerals in the above drawings are as follows: 1, positioning mold; 2, glue applying device; 11, bearing platform; 12, profiling positioning groove; 21, glue gun; 22, glue removing scraper; 23, multi-axis motion assembly. DETAILED DESCRIPTION

[0027] The application will be further described in detail below with reference to the drawings and examples.

[0028] Example 1 In the prior art, the sealing ring glue applying system generally adopts a fixed clamping force positioning mode, which is difficult to adapt to workpieces of different hardness materials, is prone to cause deformation or positioning deviation of the sealing ring, and mainly relies on a two-dimensional vision sensor for glue overflow detection, which cannot obtain actual form data of the glue in three-dimensional space, resulting in a blind area in glue removing operation. The glue removing mechanism usually adopts a rigid scraper for mechanical operation, lacks dynamic response capability to the glue sticking state, and is prone to damage the mold surface. The glue applying process parameters are adjusted depending on human experience, and precise control is difficult to achieve.

[0029] The researchers found that the difference in sealing ring material causes the traditional fixed clamping force to easily cause damage to the workpiece, and a dynamic clamping control mechanism needs to be established. For the problem of three-dimensional glue overflow detection, a spatial form recognition capability needs to be built by integrating multiple sensing technologies. The glue removing process needs to adjust the operation parameters in real time in combination with the glue state to avoid excessive scraping. By introducing closed-loop control, multi-modal sensing and intelligent algorithms, an adaptive glue applying solution is formed.

[0030] The present application proposes a system including a positioning mold and a glue applying device. The positioning mold is provided with a rotatable bearing platform, the surface of which is provided with a profiling positioning groove matching the profile of the sealing ring. The inner wall of the positioning groove is distributed with elastic limiting protrusions. The glue applying device includes a glue gun, a glue removing scraper, a multi-axis motion assembly and a glue overflow detection module. The elastic limiting protrusions are connected with a pressure sensor to form a clamping force closed-loop control system. The glue overflow detection module detects the three-dimensional form and boundary profile of the glue in real time through multi-sensor fusion technology and compares the three-dimensional deviation with the preset path. The glue removing scraper performs adaptive removal operation based on the deviation comparison result.

[0031] The elastic limiting protrusion connected pressure sensor refers to integrating the contact pressure detection unit in the positioning structure, which can be specifically implemented by a piezoresistive sensor array, and constructing a clamping force distribution model by real-time collection of pressure data of each contact point. The multi-sensor fusion technology refers to combining detection devices of different principles, which can be specifically implemented by cooperative work of a linear array vision sensor and a laser ranging sensor to respectively acquire two-dimensional contour features and three-dimensional spatial coordinate data. The three-dimensional spatial deviation comparison refers to establishing matching analysis of the actual colloid form and the digital model, which can be specifically implemented by a point cloud registration algorithm to calculate the spatial offset. The adaptive glue removal operation refers to dynamically adjusting the scraping parameters according to the colloid state, which can be specifically implemented by a path re-planning algorithm to generate a compensation motion trajectory.

[0032] When the sealing ring is placed on the bearing platform, the array type pressure sensor monitors the pressure distribution of each contact point in real time, and the control system dynamically adjusts the clamping force according to the preset material parameters to avoid excessive extrusion of hard materials or displacement of soft materials. During the glue coating process, the linear array vision sensor scans the glue line edge along the motion trajectory, and the laser ranging sensor synchronously collects the colloid height data to generate a three-dimensional point cloud model. The spatial geometry matching processing unit performs registration analysis on the real-time model and the preset path to calculate the volume and position of the colloid overflow area. The glue removal scraper generates a compensation path according to the deviation data, adjusts the contact angle through the flexible hinge structure, and maintains a constant scraping pressure under the feedback of the force control sensor.

[0033] The traditional system adopts a fixed clamping force, which is easy to cause damage to the workpiece. The present scheme realizes dynamic optimization of the clamping force through pressure feedback closed loop. The conventional two-dimensional detection cannot identify the colloid height change. The present scheme realizes three-dimensional form recognition through multi-sensor fusion. The mechanical glue removal device has the risk of scraping overload. The present scheme realizes adaptive operation control combined with real-time detection data.

[0034] The present application effectively solves the flexible clamping problem of sealing rings of different materials, realizes accurate glue overflow detection in three-dimensional space, and guarantees the integrity of the mold surface through intelligent glue removal mechanism. The multi-sensor data fusion improves the spatial resolution of colloid form recognition. The closed loop control mechanism ensures the dynamic optimization of process parameters. The overall system has stronger environmental adaptability and operation precision.

[0035] The glue overflow detection module includes a linear array vision sensor and a laser ranging sensor. The linear array vision sensor constructs a two-dimensional edge contour model of the glue coating area, and the laser ranging sensor generates a three-dimensional point cloud model of the glue three-dimensional form and boundary contour. The glue overflow detection module is built-in with a spatial geometry matching processing unit for deviation analysis of real-time detection data and a three-dimensional digital model of the preset glue coating path.

[0036] The line array vision sensor refers to a sensor that continuously collects image data in a single direction, which can be implemented by using a CCD or CMOS linear array device, and obtains the transverse profile data of the glue application area through line-by-line scanning. The laser ranging sensor refers to a ranging device based on the principle of triangulation or time of flight, which can be implemented by using a laser line scanner combined with a photoelectric detector, and generates three-dimensional coordinate data by measuring the reflection displacement of the laser beam on the glue surface. The spatial geometry matching processing unit refers to an operation module with three-dimensional model comparison function, which can be implemented by using an embedded GPU combined with a point cloud registration algorithm, and determines the glue overflow deviation by calculating the spatial distance difference between the real-time point cloud and the preset model.

[0037] The line array vision sensor performs high-speed scanning in the vertical direction of the glue application path to obtain the width change data of the glue in the planar projection direction, forming a continuous image of the two-dimensional edge profile. The laser ranging sensor performs synchronous multi-angle scanning to generate three-dimensional point cloud data by measuring the spatial coordinates of each point on the glue surface. The spatial geometry matching processing unit aligns the two-dimensional profile data and the three-dimensional point cloud data in coordinates, constructs a complete three-dimensional glue form model containing height information, and then compares it with the three-dimensional digital model of the preset glue application path point by point to calculate the region coordinates of the glue volume exceeding the standard value.

[0038] The traditional method only relies on a single vision sensor to obtain planar projection information, which cannot detect the height change of the glue in the vertical direction, making it difficult to identify thin layer overflow or three-dimensional accumulation defects. The present scheme can accurately quantify the shape deviation of the glue in three dimensions in space by fusing two-dimensional profile and three-dimensional point cloud data, especially for the glue lateral flow phenomenon caused by the curved sealing ring.

[0039] The present application can accurately identify the deviation of the actual accumulation form of the glue in three-dimensional space from the preset path, providing accurate coordinate positioning data for subsequent glue cleaning operations, effectively avoiding the problems of glue residue or excessive cleaning caused by detection blind area, and improving the consistency of glue application quality.

[0040] The glue cleaning scraper is connected to the multi-axis motion assembly through a flexible hinge structure and integrates a force control sensor. The force control sensor realizes closed-loop control of contact force, allowing the scraper to automatically adjust the scraping trajectory according to the glue adhesion form. The surface of the glue cleaning scraper is covered with a super-hydrophobic coating, and the positioning mold is equipped with a residual glue pre-detection module. The residual glue pre-detection module is in communication connection with the glue overflow detection module for secondary verification of glue cleaning. The residual glue pre-detection module includes an infrared thermal imaging unit and a structured light scanner.

[0041] The flexible hinge structure refers to a mechanical connection mechanism with multi-degree-of-freedom deformation capability, which can be specifically implemented by a titanium alloy sheet laminated structure, and functions to allow the glue cleaning scraper to elastically deform when contacting the glue, thereby avoiding rigid collision to cause damage to the mold surface. The force control sensor refers to a sensing device capable of measuring the contact pressure in real time, which can be specifically implemented by a piezoresistive sensor, and functions to dynamically adjust the contact force between the scraper and the mold surface through a closed-loop control algorithm, thereby ensuring stable pressure during the glue cleaning process. The super-hydrophobic coating refers to a surface treatment material with low surface energy characteristics, which can be specifically implemented by a nano-silicon dioxide composite coating, and functions to reduce the adhesion of glue on the scraper surface and prevent secondary pollution. The residual glue pre-detection module refers to a residual detection system based on optical principles, which can be specifically implemented by an infrared thermal imaging unit to detect the temperature difference signal between the glue and the mold, and a structured light scanner to generate three-dimensional surface topography data, and functions to improve the identification accuracy of trace residual glue through multi-modal data fusion.

[0042] The glue cleaning scraper is connected to the multi-axis motion assembly through the flexible hinge structure, and forms a contact force closed-loop control under the feedback of the force control sensor. When the scraper contacts the glue, the system adjusts the cleaning trajectory and pressure parameters in real time according to the glue adhesion form. The super-hydrophobic coating on the surface of the glue cleaning scraper can reduce the adhesion of the glue and reduce the risk of residue. The residual glue pre-detection module is started after the glue cleaning is completed, detects the temperature difference distribution between the glue and the mold through infrared thermal imaging, and obtains the three-dimensional surface topography data using a structured light scanner. The detection results are compared with the original data of the glue overflow detection module to realize secondary verification. If residual glue is found, the system automatically triggers the re-scraping program.

[0043] The traditional glue cleaning device adopts a fixed trajectory and rigid contact method, which cannot adapt to the change of glue form and is prone to cause mold scratches or residual glue. The present scheme realizes dynamic pressure regulation and trajectory optimization through the synergistic effect of the flexible hinge and the force control sensor, and significantly improves the adaptability and reliability of the glue cleaning process by combining the super-hydrophobic coating and multi-modal residual glue detection technology.

[0044] The present application solves the problem of mold damage caused by rigid contact of the traditional glue cleaning device, avoids the secondary rework caused by glue residue, and ensures the accuracy and consistency of the glue cleaning operation through closed-loop control and multi-sensor verification mechanism.

[0045] The residual glue pre-detection module identifies trace residual glue through the temperature field distribution difference relative to the mold base. The glue cleaning scraper is provided with a height adaptive mechanism driven by a shape memory alloy, which adjusts the working angle in real time according to the three-dimensional topography of residual glue obtained by the structured light scanner. The height adaptive mechanism integrates a vibration frequency adjustment unit for peeling off the cured glue marks and protecting the mold surface.

[0046] The temperature field distribution difference refers to the difference in heat conduction characteristics between the mold base and the residual glue, which can be realized by capturing the temperature gradient change using an infrared thermal imaging unit, and identifying the trace amount of residual glue by analyzing the thermal radiation difference of different materials. The height self-adaptive mechanism driven by shape memory alloy refers to a structure that deforms by utilizing the phase change characteristics of the alloy, which can be realized by using a nickel-titanium alloy wire as a driving element. According to the residual glue height data obtained by the structured light scanner, the alloy deformation is triggered, thereby adjusting the contact angle between the scraper and the mold surface. The vibration frequency adjustment unit refers to a mechanical vibration device that can output different amplitudes, which can be realized by using a piezoelectric ceramic transducer and a frequency controller to work together. By matching the vibration parameters with the curing degree of the glue mark, the damage to the mold surface caused by high-frequency vibration is avoided.

[0047] The residual glue pre-detection module scans the mold surface through an infrared thermal imaging unit, captures the temperature field distribution difference between the residual glue and the mold base, generates a residual glue distribution thermograph, and transmits the three-dimensional topographic data of the residual glue to the control system of the glue removal scraper. The shape memory alloy driving mechanism triggers deformation according to the residual glue height information, drives the scraper to rotate around the flexible hinge, and forms the best contact angle between the scraper edge and the residual glue surface. The vibration frequency adjustment unit dynamically adjusts the vibration parameters according to the curing state of the glue mark, such as using high-frequency low-amplitude vibration to achieve peeling for completely cured glue marks, and using low-frequency vibration to avoid glue splashing for semi-cured glue marks.

[0048] The traditional glue removal device cannot identify trace amount of residual glue and relies on fixed-angle mechanical scraping, which easily causes mold scratches or incomplete glue removal. The present scheme realizes accurate identification of residual glue position and morphology through the dual detection mechanism of temperature field and three-dimensional topography. Combined with the dynamic angle adjustment of shape memory alloy and the matching of vibration parameters, the mold surface integrity is effectively protected while removing residual glue.

[0049] The present application can solve the problem of repeated work caused by insufficient residual glue identification accuracy in the traditional glue removal process, avoid mechanical damage to the mold surface caused by rigid scraping action, and improve the efficiency of removing cured glue marks.

[0050] The bearing platform is redundantly driven by double servo motors, and full-closed loop angle control is realized through a resolver. The bottom of the positioning mold is provided with a quick-change base, which integrates a mechanical self-locking structure, an electromagnetic positioning unit, and a chip-type identity recognition module, and is used to automatically match the glue application parameters of different sealing rings.

[0051] Dual servo motor redundant drive refers to using two groups of servo motors to synchronously drive the bearing platform to rotate. Specifically, a cross-coupling control algorithm can be used to realize balanced distribution of the torque of the dual motors, avoid single-point failure risk, and a resolver refers to an electromechanical conversion device for high-precision angle feedback. Specifically, a brushless resolver can be used to realize full-closed-loop control and real-time compensation of transmission gap error. The quick-change base refers to a mold mounting structure with a standardized interface. Specifically, a hydraulic quick clamping device can be used in cooperation with a guide pin to realize quick disassembly and assembly of the mold. The mechanical self-locking structure refers to a physical locking mechanism for preventing displacement of the mold. Specifically, a wedge block and spring composite structure can be used to realize passive self-locking. The electromagnetic positioning unit refers to an auxiliary positioning device based on electromagnetic adsorption. Specifically, an array-type electromagnet can be used to realize micron-level positioning accuracy of the mold mounting surface. The chip-type identity recognition module refers to an electronic tag for storing mold parameters. Specifically, an RFID chip and a reader can be used to realize automatic identification of the mold identity.

[0052] The dual servo drive system of the bearing platform monitors the rotation angle in real time through the resolver to form an angle closed-loop control loop. When an abnormality of a single motor is detected, the redundant drive system can automatically switch the master-slave control mode to maintain the operation of the platform. The quick-change base realizes physical fixation of the mold through the mechanical self-locking structure. The electromagnetic positioning unit generates an adsorption force to eliminate the installation gap after being powered on. After the chip-type identity recognition module is read, the system automatically calls the corresponding sealing ring type glue coating pressure and path parameters to realize seamless switching of process parameters.

[0053] The traditional mold replacement relies on manual calibration and positioning reference, and has problems such as poor repeatability, poor parameter setting lag, etc. The present scheme realizes automatic parameter matching of the mold replacement process through the combination of the electromechanical composite positioning mechanism and the identity recognition technology, and eliminates human operation errors.

[0054] The present application effectively solves the technical defects of low positioning accuracy and poor parameter matching efficiency in the mold replacement process, realizes quick switching of the glue coating process of different specifications of sealing rings, ensures the accurate correspondence between the glue coating parameters and the mold type, and improves the flexible production capacity of the production line.

[0055] The glue gun integrates an ultrasonic viscosity sensor and a precision valve needle driving system. The glue gun has a built-in glue coating path planning module, which can generate a glue coating path with variable glue line density in combination with the cross-sectional curvature of the sealing ring, and realize glue deposition control.

[0056] The ultrasonic viscosity sensor refers to a device for monitoring the viscosity parameters of glue in real time through the change of the propagation characteristics of ultrasonic waves in the glue medium. Specifically, it can be realized by combining a piezoelectric transducer with a signal processing circuit, and is used for dynamically sensing the flow state of glue. The precision valve needle driving system refers to an actuator for controlling the opening degree of the valve needle through a micro-step motor or a piezoelectric ceramic driver. Specifically, it can be realized by using a closed-loop feedback control algorithm, and is used for adjusting the glue flow in real time according to the viscosity detection result. The glue application path planning module refers to an operation unit for generating a glue line distribution strategy based on the geometric characteristics of the sealing ring. Specifically, it can be realized by combining the curvature radius interpolation algorithm with the path optimization model, and is used for matching the glue deposition requirements of different cross-sectional shapes.

[0057] During the glue application process, the glue gun continuously obtains glue viscosity data through the ultrasonic viscosity sensor and transmits the detection results to the precision valve needle driving system. The driving system adjusts the opening degree of the valve needle in real time according to the viscosity change, ensures that the glue flow is consistent with the preset process parameters, and the glue application path planning module analyzes the three-dimensional model of the sealing ring, extracts the cross-sectional curvature distribution characteristics, generates glue line density gradient adjustment instructions based on the curvature change amplitude, so that the glue gun adopts high-density glue application mode in straight-line segments and switches to low-density glue application mode in areas with sudden curvature changes, thereby avoiding glue accumulation or glue breakage.

[0058] The traditional glue application system relies on fixed flow control and manual preset path, and cannot adapt to glue viscosity fluctuations and complex curved surface characteristics, which easily leads to uneven glue line width. The present scheme realizes adaptive control of the glue deposition process by integrating real-time viscosity detection and dynamic path planning, and eliminates the dependence on manual experience.

[0059] The present application solves the problem of inaccurate flow control caused by changes in glue viscosity in traditional glue application systems, and ensures that the glue line density and structural strength requirements of different cross-sectional areas of the sealing ring are accurately matched through curvature-responsive path planning, effectively improving the uniformity of the glue layer and the sealing performance.

[0060] It also includes an AI self-learning module that constructs a multi-modal neural network model for fusing visual detection data and process parameters. The AI self-learning module can predict the glue overflow probability under different working conditions and generate a risk heat map, and realize interactive optimization of process parameters through a man-machine cooperation interface.

[0061] The multi-modal neural network model refers to a deep learning architecture capable of processing image data and numerical data simultaneously, which can be implemented by a hybrid model of a convolutional neural network and a fully connected network in parallel, and integrates visual inspection data and process parameters through a feature-level fusion method. The risk heat map refers to a two-dimensional visual map representing the probability distribution of glue overflow with a color gradient, which can be generated by a probability density function combined with a spatial interpolation algorithm, and is used to visually display high-risk areas in the glue application path. The human-machine collaborative interface refers to an interactive terminal with bidirectional parameter adjustment function, which can be implemented by integrating parameter sliding bar and heat map superimposed display module on a touch screen to realize real-time feedback adjustment of process parameters.

[0062] The visual inspection data includes real-time monitoring results of glue three-dimensional morphology and boundary contour, and the process parameters include glue gun moving speed, glue discharge pressure and glue viscosity data. The multi-modal neural network model establishes a mapping relationship between glue overflow probability and multi-dimensional input parameters through offline training, generates a risk heat map in the online prediction stage combined with the current working conditions. The operator observes the heat map distribution through the human-machine collaborative interface and interactively adjusts the process parameters corresponding to the high-probability overflow area. The adjusted parameters are input into the neural network model as new samples to realize self-learning optimization.

[0063] The process parameter adjustment of the traditional glue application system relies on manual experience judgment and lacks data-driven decision support, resulting in low parameter optimization efficiency and difficulty in dealing with complex working condition changes. The AI self-learning module of the present scheme realizes automatic correlation analysis of process parameters and detection data, and provides accurate overflow risk prediction combined with visual heat map, effectively solving the problem of strong dependence on manual experience.

[0064] The present application realizes intelligent dynamic optimization of glue application process parameters, establishes a glue overflow prediction model by fusing multi-source data, reduces the time cost of manual debugging, improves the stability of glue application quality, and the visual presentation of the risk heat map enables the operator to quickly identify the key risk area and make targeted parameter adjustment to avoid glue overflow defects.

[0065] The servo drive system of the multi-axis motion assembly integrates a motion error compensation module for real-time correction of the positioning accuracy of the end of the mechanical arm. The motion error compensation module includes a sensor detection unit and a control unit. The sensor detection unit is used to obtain joint angle deviation, angular velocity and acceleration parameters, and the control unit generates a position compensation amount according to the parameters.

[0066] The motion error compensation module refers to a real-time error correction device embedded in the servo driving system, which can be specifically realized by combining a multi-axis linkage control algorithm with a feedback mechanism, and is used to eliminate the cumulative error caused by joint clearance or load change in the motion process of the robot arm. The sensor detection unit refers to a detection device installed at each joint of the robot arm, which can be specifically realized by using a high-precision encoder or an inertial measurement unit, and is used to collect real-time motion state parameters. The control unit refers to an operation module with dynamic compensation function, which can be specifically realized by using a PID controller or a fuzzy control algorithm, and is used to calculate the compensation amount according to the real-time motion parameters and drive the actuator.

[0067] During the motion process of the robot arm, the sensor detection unit continuously monitors the angle deviation, angular velocity and acceleration parameters of each joint. The control unit compares the collected real-time parameters with the theoretical motion trajectory based on the preset kinematic model, calculates the deviation between the actual position of the robot arm end and the target position, and transmits the compensation signal to the actuator of each joint of the robot arm through the servo driving system to form a closed-loop control loop, dynamically corrects the spatial pose of the robot arm end effector, for example, in the glue coating path tracking process, when the abnormal fluctuation of the joint angular velocity of the robot arm is detected, the control unit can generate a reverse compensation amount in real time to ensure that the glue gun moves stably along the preset trajectory.

[0068] The traditional system relies on manual experience to adjust the motion parameters of the robot arm and cannot eliminate the dynamic error generated in the motion process in real time. The present scheme realizes the autonomous correction of the positioning accuracy of the robot arm end by integrating a closed-loop error compensation module, which can effectively suppress the trajectory deviation caused by mechanical vibration or load mutation in the multi-axis linkage high-speed motion scene.

[0069] The present application solves the technical problem that the positioning accuracy of the robot arm of the traditional glue coating system is affected by motion error, realizes high-precision positioning control of the end effector of the glue coating device under complex motion trajectory, ensures the matching accuracy of the glue line deposition position and the profiled positioning groove of the sealing ring, and avoids glue overflow or glue coating loss caused by robot arm positioning deviation.

[0070] The elastic limiting protrusions on the inner wall of the profiled positioning groove are arrayed, and each elastic limiting protrusion is independently connected to a pressure sensor; the bearing platform is provided with an adaptive leveling unit for adjusting the levelness of the bearing platform according to the placement posture of the sealing ring.

[0071] The array distribution refers to that the elastic limiting protrusions are arranged in a matrix in the inner wall of the positioning groove according to a preset interval, and can be implemented by equidistantly distributed hemispherical silica gel protrusions, and a uniform distributed clamping force is formed through multi-point contact, the independent connection pressure sensor refers to that a micro pressure sensing unit is embedded at the bottom of each protrusion, and can be implemented by integrating a piezoresistive sensor and a signal acquisition circuit, and the real-time pressure change of each contact point is monitored, and the self-adaptive leveling unit refers to a mechanism capable of dynamically adjusting the inclination angle of the platform, and can be implemented by connecting a three-axis gyroscope and an electric leveling leg, and the leveling action is triggered by detecting the placement posture of the sealing ring.

[0072] When the sealing ring is placed on the bearing platform, the array distributed elastic limiting protrusions feedback the contact pressure of each point in real time through the independent pressure sensor, if the local pressure exceeds the preset threshold, the bearing platform rotation angle is adjusted through the clamping force closed-loop control system, so that the pressure distribution tends to be balanced, at the same time, the self-adaptive leveling unit obtains the sealing ring placement posture data through the three-axis gyroscope, if the inclination angle deviation is detected, the electric leg is driven to adjust the levelness of the platform, so as to ensure that the bottom surface of the sealing ring is completely matched with the positioning groove.

[0073] In some specific embodiments, the pressure sensor can adopt a thin film pressure sensing array, and the thickness is controlled within 0.5 mm to avoid affecting the structure of the positioning groove; the electric leg of the self-adaptive leveling unit can be equipped with a ball screw transmission mechanism, and the leveling accuracy can reach ±0.1 degree.

[0074] The traditional positioning mold adopts a fixed limiting structure, and cannot sense the actual contact state of the sealing ring and the positioning groove, which is easy to cause local overpressure or positioning deviation due to the difference in material hardness, and the scheme realizes the spatial distribution optimization of the contact pressure through the synergistic action of the array pressure sensor and the dynamic leveling, and eliminates the positioning error caused by the workpiece deformation or the placement inclination.

[0075] The application effectively solves the problem of uneven distribution of clamping force caused by the difference in material hardness of the sealing ring, avoids the workpiece deformation or positioning failure; at the same time, the dynamic leveling mechanism compensates for the placement posture deviation of the sealing ring, improves the positioning accuracy and system adaptability, and provides a stable and reliable positioning reference for the subsequent gluing process.

[0076] The gluing device also includes a glue preheating module for maintaining the rheological property stability of the glue, so that the temperature of the glue is maintained within a preset range; the glue gun setting posture adjusting unit can adjust the glue discharging direction of the glue gun according to the spatial angle of the gluing position.

[0077] The glue preheating module refers to a device for maintaining the viscosity and fluidity of glue through temperature control. Specifically, it can be implemented by using a resistance heating band combined with a PID temperature control algorithm. Through closed-loop feedback adjustment of heating power, it ensures that the glue is in the best rheological state during transportation. The posture adjustment unit refers to a mechanism for controlling the spatial orientation of the glue gun. Specifically, it can be implemented by using a three-degree-of-freedom parallel mechanical arm combined with an angle encoder. Through real-time analysis of three-dimensional coordinate data of the glue application path, it dynamically adjusts the normal angle between the glue gun and the surface of the sealing ring.

[0078] The glue preheating module monitors the temperature distribution inside the glue pipe through embedded thermocouples. When the detected temperature deviates from the preset range, it triggers the heating band partition compensation function to eliminate local temperature gradients. The posture adjustment unit receives the pose data of the multi-axis motion assembly and generates servo motor driving signals through inverse kinematics calculation to ensure that the glue gun outlet axis always maintains a preset offset angle with the tangent direction of the sealing ring curve, ensuring the uniformity of glue line width.

[0079] Traditional glue application systems do not configure active glue temperature control devices, and the viscosity of glue is significantly affected by environmental temperature fluctuations, leading to glue line defects such as glue breakage or accumulation. At the same time, the fixed installation method of the glue gun cannot adapt to changes in the angle of curved surface glue application, which easily causes glue splashing or path deviation. This scheme integrates temperature control and dynamic steering functions, achieving dual improvement in the stability of glue physical properties and the spatial adaptability of glue application trajectory.

[0080] This application solves the problem of rheological property fluctuation caused by temperature change of glue, avoiding defects such as inconsistent glue line width and glue layer thickness. Through real-time matching of glue gun glue outlet direction and workpiece surface, it eliminates the glue splashing phenomenon during curved surface glue application, improving the glue application path precision of complex profile sealing rings.

Claims

1. A self-identifying gluing system comprising a positioning mold and a gluing device; the positioning mold is provided with a supporting platform for driving its rotation, the positioning mold surface is provided with a contoured positioning groove matching the contour of the sealing ring, and the inner wall of the positioning groove is provided with elastic limiting protrusions; the gluing device includes a glue gun, a glue scraper, a multi-axis motion assembly that drives the movement of the two, and a glue overflow detection module, characterized by: The elastic limiting protrusion is connected to a pressure sensor to form a clamping force closed-loop control system, which can dynamically adjust the clamping force according to the hardness of the sealing ring material; The glue overflow detection module uses multi-sensor fusion technology to detect the three-dimensional shape and boundary contour of the glue in real time, and compares the three-dimensional spatial deviation with the preset glue coating path; The glue removing scraper performs an adaptive glue removing operation based on the deviation comparison result.

2. The self-identifying gluing system according to claim 1, characterized in that: The overflow detection module includes a linear array vision sensor and a laser ranging sensor; The linear array vision sensor constructs a two-dimensional edge contour model of the glue coating area, and the laser ranging sensor generates a three-dimensional point cloud model of the three-dimensional shape of the glue and the boundary contour; The glue overflow detection module has a built-in spatial geometry matching processing unit for performing deviation analysis between real-time detection data and a three-dimensional digital model of a preset glue coating path.

3. The self-identifying gluing system according to claim 1, characterized in that: The glue cleaning scraper is connected to the multi-axis motion component through a flexible hinge structure and integrated with a force control sensor; The force control sensor realizes closed-loop control of the contact force, so that the scraper automatically adjusts the scraping trajectory according to the glue adhesion form; The surface of the glue cleaning scraper is covered with a super-hydrophobic coating, and the positioning mold is equipped with a residual glue pre-detection module. The residual glue pre-detection module is communicatively connected to the overflow glue detection module for secondary verification of glue cleaning. The residual glue pre-detection module includes an infrared thermal imaging unit and a structured light scanner.

4. The self-identifying gluing system according to claim 3, characterized in that: The residual adhesive pre-detection module identifies trace residual adhesive by the difference in temperature field distribution relative to the mold base; The adhesive removal scraper is equipped with a highly adaptive mechanism driven by a shape memory alloy, which adjusts the working angle in real time according to the three-dimensional morphology of the residual adhesive obtained by the structured light scanner; The highly adaptive mechanism is integrated with a vibration frequency adjustment unit for peeling off cured glue marks and protecting the mold surface.

5. The self-identifying gluing system according to claim 1, characterized in that: The carrying platform is redundantly driven by dual servo motors, and full closed-loop angle control is achieved through a rotary transformer; A quick-change base is provided at the bottom of the positioning mold, which integrates a mechanical self-locking structure, an electromagnetic positioning unit and a chip-type identity recognition module for automatically matching the gluing parameters of different sealing ring types.

6. The self-identifying gluing system according to claim 1, characterized in that: The glue gun is integrated with an ultrasonic viscosity sensor and a precision valve needle drive system; The glue gun has a built-in glue path planning module, which can generate a glue path with variable glue line density based on the curvature of the sealing ring cross section, thereby realizing glue deposition control.

7. The self-identifying gluing system according to claim 1, characterized in that: It also includes an AI self-learning module that builds a multimodal neural network model to fuse visual inspection data with process parameters; The AI ​​self-learning module can predict the probability of glue overflow under different working conditions and generate a risk heat map, and realize interactive optimization of process parameters through a human-machine collaborative interface.

8. The self-identifying gluing system according to claim 1, characterized in that: The servo drive system of the multi-axis motion assembly is integrated with a motion error compensation module for correcting the positioning accuracy of the end of the robotic arm in real time; The motion error compensation module includes a sensor detection unit and a control unit. The sensor detection unit is used to obtain joint angle deviation, angular velocity and acceleration parameters. The control unit generates a position compensation amount according to the parameters.

9. The self-identifying gluing system according to claim 1, characterized in that: The elastic limiting protrusions on the inner wall of the contoured positioning groove are distributed in an array, and each elastic limiting protrusion is independently connected to a pressure sensor; The carrying platform is provided with an adaptive leveling unit for adjusting the horizontality of the carrying platform according to the placement posture of the sealing ring.

10. The self-identifying gluing system according to claim 1, characterized in that: The glue coating device also includes a glue preheating module for pre-treating the glue to maintain the stability of the glue rheological properties so that the glue temperature is kept within a preset range; The glue gun is provided with a posture adjustment unit, which can adjust the glue discharge direction of the glue gun according to the spatial angle of the glue coating position.

Images