A glue pouring intelligent control method and system based on internet of things data feedback

Abstract

The application discloses a kind of based on the data feedback of Internet of Things glue filling intelligent control method and system, it is related to glue filling control technical field, including: first, obtain product information from glue filling production plan, select glue type and obtain characteristic data, after determining quality index matrix, glue filling machine is pretreated, set initial glue filling basic parameter;Again, obtain the image of the part to be glued by industrial camera and pretreatment, import product model in CAD and set initial glue filling track, based on track and parameter control glue filling machine executes glue filling, real-time image information is collected, trajectory and parameter are dynamically corrected in combination with regional environmental information.The system includes data acquisition and pretreatment, parameter calculation and model construction, trajectory planning and model construction, glue filling execution and dynamic correction module, each module contains corresponding unit collaborative work.The application has the advantages that: the precision control and dynamic self-adaptive adjustment of glue filling process are realized, and the reliability of glue filling process is improved.

Description

Technical Field

[0001] This invention relates to the field of glue dispensing control technology, specifically to a glue dispensing intelligent control method and system based on Internet of Things (IoT) data feedback. Background Technology

[0002] In modern manufacturing, potting processes are widely used in fields such as electronic packaging, new energy batteries, and aerospace. The precision of its control directly affects the sealing performance, reliability, and service life of the products.

[0003] Existing intelligent dispensing control methods and systems based on IoT data feedback often rely on preset fixed paths for dispensing complex structural components. This makes them ill-suited for precisely filling curved surfaces, narrow gaps, and other complex areas, frequently resulting in uneven glue application and air bubbles. Furthermore, they cannot adjust dispensing pressure and flow rate in real time to changes in ambient temperature and glue viscosity, leading to poor consistency and low yield rates, especially in mass production. Therefore, there is a need to provide an intelligent dispensing control method and system based on IoT data feedback to address these problems. Summary of the Invention

[0004] To address the aforementioned technical problems, this paper provides an intelligent dispensing control method and system based on IoT data feedback. This technical solution solves the problem that existing intelligent dispensing control methods and systems based on IoT data feedback mentioned in the background section often use preset fixed paths for dispensing trajectories of complex structural parts, which are difficult to adapt to the precise filling of complex areas such as curved surfaces and narrow gaps. This often results in uneven glue volume and residual air bubbles. At the same time, it is impossible to correct parameters such as dispensing pressure and flow rate in real time when the ambient temperature and glue viscosity change, which can easily lead to poor consistency and low yield in mass production.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A smart dispensing control method based on IoT data feedback includes:

[0007] Obtain information on the potting product, select the adhesive type based on the potting product information, and obtain adhesive characteristic data;

[0008] Determine the initial quality index matrix of the glue, simultaneously acquire the initial state information of the glue dispensing machine, and preprocess the glue dispensing machine;

[0009] Based on the initial quality index matrix of the adhesive, the initial basic parameters for dispensing are set for the dispensing machine after pretreatment.

[0010] Obtain image information of the area to be glued, and then preprocess the area to be glued;

[0011] Import the glue-dispensing product model into CAD and set the initial glue-dispensing trajectory information;

[0012] Based on the initial glue-dispensing trajectory information and initial glue-dispensing basic parameters, the glue-dispensing machine is controlled to perform glue-dispensing on the pre-processed parts to be glued, and real-time image information of the glue-dispensing parts is acquired.

[0013] The system acquires environmental information about the glue-filling area and, in conjunction with real-time image information of the glue-filling area, dynamically corrects the initial glue-filling trajectory information and initial glue-filling baseline parameters.

[0014] In an optional embodiment, the step of obtaining the potting product information, selecting the adhesive type based on the potting product information, and obtaining adhesive characteristic data specifically includes:

[0015] Obtain glue-filling product information from the glue-filling production plan, including application scenario information, working environment parameter information, product material information, and geometric dimension information of the glue-filling part;

[0016] Set the adhesive matching rules, and then select the adhesive type based on the potting product information;

[0017] After selecting the type of adhesive, the feature vector is read from the database and normalized to obtain the adhesive feature data G={g1, g2, g3, g4}, where g1 is the viscosity, g2 is the curing time, g3 is the curing shrinkage rate, and g4 is the thermal conductivity.

[0018] In an optional embodiment, determining the initial quality index matrix of the adhesive, simultaneously acquiring the initial state information of the dispensing machine, and preprocessing the dispensing machine specifically includes:

[0019] Based on the adhesive characteristic data, define the initial quality index matrix of the adhesive;

[0020] Obtain initial status information of the dispensing machine, including sensor calibration parameters, actuator initial parameters, and equipment operating status;

[0021] Based on the equipment's operating status, determine whether to perform pre-processing of the dispensing machine. If the equipment is operating normally, continue to acquire the initial status information of the dispensing machine. If the equipment is operating in a state of pending calibration, perform zero-point calibration of the pressure sensor and range calibration of the flow sensor on the dispensing machine.

[0022] The formula for expressing the initial quality index matrix of the adhesive is as follows: In the formula, For adhesive properties, Let be the threshold value of the process requirement corresponding to the adhesive property index, j∈m;

[0023] The pressure sensor of the dispensing machine is calibrated to zero point: In the formula, This is the pressure value after the pressure sensor of the glue dispensing machine has been calibrated. This is the original pressure value of the pressure sensor on the dispensing machine. This refers to the zero-point deviation of the pressure sensor in the dispensing machine.

[0024] Calibrate the flow sensor range of the dispensing machine: In the formula, The flow rate value after calibration of the dispensing machine's flow sensor. This is the original flow rate value from the dispensing machine's flow sensor. This refers to the calibration coefficient of the flow sensor for the glue dispensing machine.

[0025] In an optional embodiment, setting initial dispensing parameters for the pre-treated dispensing machine based on the initial glue quality index matrix specifically includes:

[0026] The glue dispensing machine after pretreatment was used to conduct glue dispensing tests, and the glue tube size information was acquired and the glue dispensing test time was recorded simultaneously.

[0027] By mapping the parameters in the initial mass index matrix of the adhesive, the initial values ​​of pressure, flow rate, and temperature are obtained.

[0028] By integrating the initial values ​​of pressure, flow rate, and temperature into a single matrix, the basic parameters for glue dispensing are obtained.

[0029] The mapping formula for the initial pressure value is: In the formula, This is the initial pressure value. This is an empirical coefficient. This refers to the length of the hose. Where is the radius of the hose. This refers to the dynamic viscosity of the adhesive. , This represents the average pressure value of the dispensing machine during the dispensing test;

[0030] The mapping formula for the initial flow value is: In the formula, The initial value of the flow rate. This represents the volume of the glued area during the glue-filling test. This is the time value for the glue-pouring test. For safety factor;

[0031] Define a viscosity mapping threshold, a temperature mapping coefficient, and a temperature adjustment coefficient. If g1 is greater than the viscosity mapping threshold, then the mapping formula for the initial temperature value is: In the formula, This is the initial temperature value. This is the baseline temperature value. This refers to the dynamic viscosity of the adhesive. This is the viscosity mapping threshold. This is the temperature mapping coefficient. This is the temperature regulation coefficient.

[0032] In an optional embodiment, the step of acquiring image information of the area to be glued and then preprocessing the area to be glued specifically includes:

[0033] A grayscale image I(x,y) is acquired using an industrial camera. The grayscale image is then denoised and edge detected to obtain the contour C={c1, c2, ..., cn} of the area to be glued, where cn is the coordinate of the nth edge detection point.

[0034] Defect repair is performed on the outline of the area to be glued, and image information of the area to be glued is obtained.

[0035] Based on grayscale images, obtain surface stain information corresponding to the image information of the area to be glued;

[0036] Based on the surface stain information corresponding to the image information of the area to be glued, the surface of the area to be glued is cleaned.

[0037] In an optional embodiment, importing the dispensing product model into CAD and setting initial dispensing trajectory information specifically includes:

[0038] Import the product model into CAD, and then extract the surface S of the glued product.

[0039] Obtain the outline C of the area to be glued corresponding to the glue-filling surface S, and each point in the glue-filling surface S needs to correspond to each point in the outline C of the area to be glued.

[0040] The surface S of the glue-filling agent is divided into a plane to obtain a simple planar region and a complex structural region.

[0041] Based on the contour C of the area to be glued, the first trajectory reference contour corresponding to the simple planar area and the second trajectory reference contour corresponding to the complex structure area are obtained respectively.

[0042] Obtain the center point corresponding to the simple planar region, and simultaneously connect the center point corresponding to the simple planar region with the first trajectory reference contour to obtain the first trajectory reference line segment;

[0043] Obtain the first trajectory reference line segment and the first trajectory reference contour where they intersect;

[0044] The first trajectory reference line segment with more than two intersection points is used as the first trajectory calibration line segment;

[0045] Obtain the coordinates of the two closest intersection points in the first trajectory calibration line segment, and simultaneously obtain the center coordinates of the two closest intersection points in the first trajectory calibration line segment;

[0046] By concatenating the center coordinates of the two closest intersection points in the first trajectory calibration line segment, the first initial trajectory D1={d1, d2, ..., dz} is obtained, where dz is the z-th center coordinate of the two closest intersection points in the first trajectory calibration line segment;

[0047] The complex structure region is unfolded to obtain a complex planar region image. The center point corresponding to the complex planar region image is obtained simultaneously. The center point corresponding to the complex planar region image is connected to the second trajectory reference contour simultaneously to obtain the second trajectory reference line segment.

[0048] Obtain the second trajectory reference line segment that intersects with the second trajectory reference contour;

[0049] The second trajectory reference line segment with more than two intersection points is used as the second trajectory calibration line segment;

[0050] Obtain the coordinates of the two closest intersection points in the second trajectory calibration line segment, and simultaneously obtain the center coordinates of the two closest intersection points in the second trajectory calibration line segment;

[0051] By concatenating the center coordinates of the two closest intersection points in the second trajectory calibration line segment, the second initial trajectory D2={d1, d2, ..., dy} is obtained, where dy is the y-th center coordinate of the two closest intersection points in the second trajectory calibration line segment;

[0052] Based on the glue-filling product model, obtain the center coordinates of the glue-filling product model;

[0053] Traverse all point coordinates in the first and second initial trajectories, take the coordinates closest to the center coordinates of the glued product model as the starting point coordinates, and splice the first and second initial trajectories to obtain the initial glued trajectory information.

[0054] In an optional embodiment, the step of acquiring environmental information of the glue-filling area and dynamically correcting the initial glue-filling trajectory information and initial glue-filling basic parameters in combination with real-time image information of the glue-filling area specifically includes:

[0055] Obtain environmental information of the glue-filling area, and obtain the ambient temperature and humidity from the environmental information of the glue-filling area, and simultaneously determine the temperature correction coefficient and trajectory correction coefficient;

[0056] Based on real-time image information of the glue application site, the real-time trajectory of glue application is obtained;

[0057] Based on the real-time and initial glue-dispensing trajectory information, the trajectory offset is determined;

[0058] Based on the trajectory correction coefficient, trajectory offset, and initial glue-filling trajectory information, a new glue-filling trajectory is determined;

[0059] Real-time bubble rate is obtained based on real-time image information of the glue application site;

[0060] Pressure, flow, and temperature corrections are applied to the initial dispensing trajectory information and initial dispensing base parameters.

[0061] The formula for calculating the temperature correction factor is: In the formula, The temperature correction factor at time t is... The quantified value of the ambient temperature at time t is obtained from the environmental information of the glue-filling area.

[0062] The formula for calculating the trajectory offset is: In the formula, Let t be the trajectory offset of the real-time glue-pouring trajectory at time t. Let be the vector value of the point coordinates corresponding to the real-time trajectory of the glue application at time t. Let be the vector value of the point coordinates corresponding to the initial glue-pouring trajectory at time t;

[0063] The formula for calculating the pressure correction is: In the formula, This represents the pressure correction at time t+1. This is the proportionality coefficient. The integral coefficients over the time interval from 0 to T. The differential coefficients are the time interval from 0 to T. For pressure deviation;

[0064] The formula for calculating the flow correction is: In the formula, The corrected flow rate. The flow rate before correction. Bubble rate;

[0065] The formula for calculating the temperature correction is: In the formula, The corrected glue temperature. This is the corrected viscosity of the adhesive. The temperature of the glue before correction. The viscosity of the glue before correction. The temperature correction factor at time t is... The quantified value of the ambient temperature at time t is obtained from the environmental information of the glue-filling area.

[0066] Furthermore, a smart dispensing control system based on IoT data feedback is proposed to implement the control method described above, including:

[0067] The data acquisition and preprocessing module is used to obtain glue product information from the glue production plan, including application scenarios, working environment parameters, product materials and the geometric dimensions of the glue-filling part; to select the appropriate glue type based on preset glue matching rules; to read feature vectors from the database, normalize them to generate glue feature data; to acquire grayscale images of the glue-filling part through an industrial camera, perform noise reduction and edge detection to obtain the outline of the glue-filling area and repair outline defects; to analyze surface stain information in the grayscale image to generate surface cleaning instructions; and to simultaneously collect environmental information such as ambient temperature and humidity of the glue-filling part.

[0068] The parameter calculation and model building module is used to define an initial quality index matrix based on glue characteristic data, to preprocess the dispensing machine, including zero-point calibration of the pressure sensor and range calibration of the flow sensor, to conduct dispensing tests, record the hose size and test time, map the initial pressure value, initial flow rate and initial temperature, and integrate the pressure, flow rate and temperature parameters to generate a basic dispensing parameter matrix.

[0069] The trajectory planning and model building module is used to import the glue-filling product model into CAD, extract the contours of the glue-filling surface and the area to be glued, divide the surface into simple planar areas and complex structural areas, generate a first initial trajectory by connecting the center point and the trajectory reference contour for the simple planar areas, and generate a second initial trajectory by unfolding the complex structural areas in the same way. The module is used to splice the first initial trajectory and the second initial trajectory with the center coordinates of the glue-filling product model as a reference to determine the starting point of the initial glue-filling trajectory and generate complete initial glue-filling trajectory information.

[0070] The dispensing execution and dynamic correction module is used to control the dispensing machine to perform dispensing based on the initial dispensing trajectory and basic parameters, to collect image information of the dispensing location in real time, to obtain the real-time dispensing trajectory and bubble rate, to calculate the temperature correction coefficient, to correct the glue temperature and viscosity according to the ambient temperature, to calculate the trajectory offset, to generate a new trajectory in combination with the trajectory correction coefficient, to correct the flow rate according to the bubble rate, and to adjust the dispensing pressure through the pressure correction formula.

[0071] In an optional embodiment, the parameter calculation and model building module includes:

[0072] The glue quality index calculation unit is used to define an initial quality index matrix based on glue characteristic data, and is used to preprocess the glue dispensing machine, including zero-point calibration of the pressure sensor and range calibration of the flow sensor.

[0073] The dispensing parameter setting unit is used to perform dispensing tests, record the tube size and test time, map the initial pressure value, initial flow rate and initial temperature value, and integrate the pressure, flow rate and temperature parameters to generate a dispensing basic parameter matrix.

[0074] In an optional embodiment, the trajectory planning and model building module includes:

[0075] The CAD model import and trajectory generation unit is used to import the glue-filling product model into CAD, extract the contours of the glue-filling surface and the area to be glued, divide the surface into simple planar areas and complex structural areas, generate a first initial trajectory for the simple planar areas by connecting the center point with the trajectory reference contour, and generate a second initial trajectory for the complex structural areas after unfolding them in the same way. The first and second initial trajectories are spliced ​​together with the center coordinates of the glue-filling product model as a reference to determine the starting point of the initial glue-filling trajectory and generate complete initial glue-filling trajectory information.

[0076] Compared with the prior art, the beneficial effects of the present invention are:

[0077] This solution proposes an intelligent dispensing control method based on IoT data feedback. Through intelligent planning of dispensing trajectory driven by CAD model, it realizes the precise generation of dispensing path for complex structural parts. By using the dispensing surface area division and trajectory reference contour connection algorithm, a first initial trajectory and a second initial trajectory are generated for simple planar areas and complex structural areas, respectively. After splicing, a complete trajectory is formed, which improves the fit of the trajectory. It is suitable for complex areas such as curved surfaces and narrow gaps, and avoids the problem of uneven glue amount caused by fixed paths.

[0078] This solution proposes an intelligent control method for dispensing based on IoT data feedback. Through the dynamic correction mechanism of IoT data feedback, it realizes the adaptive adjustment of the dispensing process. Based on the temperature correction coefficient formula and pressure correction formula, combined with the real-time bubble rate correction flow rate, when the ambient temperature fluctuates or the bubble rate exceeds the threshold, the system automatically adjusts the parameters to improve the dispensing yield and reduce dynamic response delay. Attached Figure Description

[0079] Figure 1 This is a flowchart of an intelligent dispensing control method based on Internet of Things data feedback proposed in this invention;

[0080] Figure 2 This is a flowchart illustrating the process of obtaining adhesive feature data in this invention.

[0081] Figure 3 This is a flowchart illustrating the process of obtaining the basic parameters for glue application in this invention.

[0082] Figure 4This is a system framework diagram of an intelligent glue dispensing control system based on Internet of Things data feedback proposed in this invention. Detailed Implementation

[0083] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0084] Reference Figure 1 - Figure 4 As shown, a smart dispensing control method based on IoT data feedback includes:

[0085] Obtain information on the potting product, select the adhesive type based on the potting product information, and obtain adhesive characteristic data;

[0086] Determine the initial quality index matrix of the glue, simultaneously acquire the initial state information of the glue dispensing machine, and preprocess the glue dispensing machine;

[0087] Based on the initial quality index matrix of the adhesive, the initial basic parameters for dispensing are set for the dispensing machine after pretreatment.

[0088] Obtain image information of the area to be glued, and then preprocess the area to be glued;

[0089] Import the glue-dispensing product model into CAD and set the initial glue-dispensing trajectory information;

[0090] Based on the initial glue-dispensing trajectory information and initial glue-dispensing basic parameters, the glue-dispensing machine is controlled to perform glue-dispensing on the pre-processed parts to be glued, and real-time image information of the glue-dispensing parts is acquired.

[0091] The system acquires environmental information about the glue-filling area and, in conjunction with real-time image information of the glue-filling area, dynamically corrects the initial glue-filling trajectory information and initial glue-filling baseline parameters.

[0092] Furthermore, obtain the potting product information, select the adhesive type based on the potting product information, and obtain adhesive characteristic data, specifically including:

[0093] Obtain glue-filling product information from the glue-filling production plan, including application scenario information, working environment parameter information, product material information, and geometric dimension information of the glue-filling part;

[0094] Set the adhesive matching rules, and then select the adhesive type based on the potting product information;

[0095] After selecting the type of adhesive, the feature vector is read from the database and normalized to obtain the adhesive feature data G={g1, g2, g3, g4}, where g1 is the viscosity, g2 is the curing time, g3 is the curing shrinkage rate, and g4 is the thermal conductivity.

[0096] Specifically, this step is implemented as follows: First, information such as the application scenario, working environment parameters (e.g., temperature and humidity tolerance range), product materials (e.g., metal, plastic), and geometric dimensions of the glued area (e.g., gap width, surface curvature) of the glued product are extracted from the glue production plan. Then, based on preset glue matching rules, such as prioritizing weather-resistant silicone for high-temperature environments and low-shrinkage epoxy resin for electronic packaging, suitable glue types are matched from the database. After selecting the glue, the feature vectors of that type of glue, such as viscosity, curing time, curing shrinkage rate, and thermal conductivity, are read. These vectors are then normalized (mapping each parameter to the [0,1] interval) to eliminate dimensional influence, ultimately generating standardized glue feature data G={g1,g2,g3,g4}.

[0097] The advantages are as follows: By linking and matching multi-dimensional product information with preset rules, accurate selection of adhesive types is achieved, avoiding selection biases caused by human experience, such as mistakenly selecting highly hygroscopic adhesives in high-humidity environments. Simultaneously, the normalization of feature vectors makes adhesive parameters of different dimensions comparable, providing a unified data foundation for subsequent quality index matrix construction and dispensing parameter calculation, thus improving the scientific rigor and consistency of dispensing process design.

[0098] Furthermore, the initial quality index matrix of the adhesive is determined, the initial state information of the dispensing machine is acquired simultaneously, and the dispensing machine is preprocessed, specifically including:

[0099] Based on the adhesive characteristic data, define the initial quality index matrix of the adhesive;

[0100] Obtain initial status information of the dispensing machine, including sensor calibration parameters, actuator initial parameters, and equipment operating status;

[0101] Based on the equipment's operating status, determine whether to perform pre-processing of the dispensing machine. If the equipment is operating normally, continue to acquire the initial status information of the dispensing machine. If the equipment is operating in a state of pending calibration, perform zero-point calibration of the pressure sensor and range calibration of the flow sensor on the dispensing machine.

[0102] The formula for expressing the initial quality index matrix of the glue is: In the formula, These are the properties of the adhesive. Let be the threshold value of the process requirement corresponding to the adhesive property index, j∈m;

[0103] Zero-point calibration of the pressure sensor on the dispensing machine: In the formula, This is the pressure value after the pressure sensor of the glue dispensing machine has been calibrated. This is the original pressure value of the pressure sensor on the dispensing machine. This refers to the zero-point deviation of the pressure sensor in the dispensing machine.

[0104] Calibrate the flow sensor range of the dispensing machine: In the formula, The flow rate value after calibration of the dispensing machine's flow sensor. This is the original flow rate value from the dispensing machine's flow sensor. This refers to the calibration coefficient of the flow sensor for the glue dispensing machine.

[0105] Specifically, firstly, based on the acquired adhesive characteristic data (such as viscosity, curing time, etc.), an initial quality index matrix Q of the adhesive is defined. The first row of the matrix corresponds to the adhesive characteristic indicators (such as g1, g2), and the second row corresponds to the process requirement thresholds for each indicator. This matrix form achieves a quantitative mapping between adhesive characteristics and process standards. Subsequently, the system automatically collects the initial status information of the dispensing machine, including the calibration parameters of the pressure / flow sensors (such as zero-point deviation, calibration coefficient), the initial parameters of the actuators (such as motor speed, cylinder pressure), and the equipment operating status (normal / to be calibrated). If the equipment operating status is displayed as "to be calibrated," the sensor preprocessing process is triggered: zero-point calibration of the pressure sensor is performed (by subtracting the zero-point deviation from the original pressure value), and range calibration of the flow sensor is performed (by multiplying the original flow value by the calibration coefficient) to ensure the accuracy of the sensor data; if the status is normal, the initial parameters are read directly.

[0106] The advantages are as follows: By constructing an initial adhesive quality index matrix, abstract adhesive characteristics are transformed into structured data, providing a standardized benchmark for subsequent mapping calculations of dispensing parameters and avoiding parameter setting deviations caused by differences in adhesive characteristics. Furthermore, the state-driven preprocessing mechanism of the dispensing machine achieves automated closed-loop control of sensor calibration: when the equipment is in a calibration-ready state, the system automatically performs calibration of the pressure / flow sensors, eliminating measurement errors caused by sensor drift. This mechanism ensures the reliability of the equipment's initial state without manual intervention, laying the hardware foundation for precise control of the dispensing process, and is particularly suitable for ensuring equipment stability in long-term continuous production scenarios.

[0107] Furthermore, based on the initial quality index matrix of the adhesive, initial dispensing parameters are set for the pre-treated dispensing machine, specifically including:

[0108] The glue dispensing machine after pretreatment was used to conduct glue dispensing tests, and the glue tube size information was acquired and the glue dispensing test time was recorded simultaneously.

[0109] By mapping the parameters in the initial mass index matrix of the adhesive, the initial values ​​of pressure, flow rate, and temperature are obtained.

[0110] By integrating the initial values ​​of pressure, flow rate, and temperature into a single matrix, the basic parameters for glue dispensing are obtained.

[0111] The formula for mapping the initial pressure value is: In the formula, This is the initial pressure value. This is an empirical coefficient. This refers to the length of the hose. Where is the radius of the hose. This refers to the dynamic viscosity of the adhesive. , This represents the average pressure value of the dispensing machine during the dispensing test;

[0112] The mapping formula for the initial flow value is: In the formula, The initial value of the flow rate. This represents the volume of the glued area during the glue-filling test. This is the time value for the glue-pouring test. For safety factor;

[0113] Set the viscosity mapping threshold, temperature mapping coefficient, and temperature adjustment coefficient. If g1 is greater than the viscosity mapping threshold, the mapping formula for the initial temperature value is: In the formula, This is the initial temperature value. This is the baseline temperature value. This refers to the dynamic viscosity of the adhesive. This is the viscosity mapping threshold. This is the temperature mapping coefficient. This is the temperature regulation coefficient.

[0114] Specifically, this step begins by using a pre-treated dispensing machine to perform a dispensing test on the sample workpiece, simultaneously collecting geometric dimensional information such as the inner diameter and length of the dispensing tube, and recording the test time from dispensing initiation to completion. Subsequently, based on parameters in the initial adhesive quality index matrix (such as viscosity g1, curing time g2, etc.), initial process parameters, including initial pressure, initial flow rate, and initial temperature, are calculated using a mapping algorithm. Finally, these three types of parameters are integrated into basic dispensing parameters, which are in matrix form, consisting of one row and three columns.

[0115] The advantages are: integration of physical model and experience: the pressure formula combines fluid dynamics principles (Hagen-Poiseuille equation) and empirical coefficients, which improves the matching accuracy between the initial pressure value and the hose geometry and glue viscosity, and avoids insufficient glue or overflow caused by coarse parameter settings.

[0116] Dynamic temperature adaptation: Based on the viscosity threshold temperature correction mechanism, the heating temperature can be automatically increased when dispensing high viscosity adhesives (e.g., when g1>5000cP, T0 can be increased by 10-15℃) to ensure the fluidity of the adhesive and reduce incomplete dispensing caused by insufficient temperature.

[0117] Closed-loop optimization of test data: Real-time acquisition of dispensing test time and volume provides a data foundation for subsequent parameter iteration. For example, in continuous production, the β value can be automatically adjusted based on historical test data to improve the dynamic adaptability of flow parameters.

[0118] Understandably, hose dimensions—the inner diameter (r) and length (L) of the hose—directly affect the flow resistance of the adhesive. For example, a 50% decrease in r will increase the pressure P0 by 16 times (because r is a 4th power in the formula), making it a core physical quantity for pressure parameter calculation. The dispensing test time (t) and volume (V) are used to calibrate the base flow rate Q0. For example, when V=100mm³ and t=10s, Q0=10mm³ / s. Combined with a safety factor β=0.1, this avoids insufficient adhesive due to flow fluctuations during actual dispensing. Viscosity mapping threshold (η) thr ): As a trigger condition for temperature regulation, when the viscosity of the glue exceeds this value (such as 3000 cP), the temperature compensation mechanism is activated to prevent high viscosity glue from causing dispensing defects due to poor flowability; Empirical coefficient (α): Taking into account factors that are difficult to quantify, such as equipment wear and environmental interference, it is obtained by fitting historical dispensing data (usually with a value of 0.8-1.2) and is used to correct the deviation between theoretical calculations and actual working conditions.

[0119] Furthermore, image information of the area to be glued is acquired, and then preprocessing is performed on the area, specifically including:

[0120] A grayscale image I(x,y) is acquired using an industrial camera. The grayscale image is then denoised and edge detected to obtain the contour C={c1, c2, ..., cn} of the area to be glued, where cn is the coordinate of the nth edge detection point.

[0121] Defect repair is performed on the outline of the area to be glued, and image information of the area to be glued is obtained.

[0122] Based on grayscale images, obtain surface stain information corresponding to the image information of the area to be glued;

[0123] Based on the surface stain information corresponding to the image information of the area to be glued, the surface of the area to be glued is cleaned.

[0124] Furthermore, import the product model into CAD and set the initial dispensing trajectory information, specifically including:

[0125] Import the product model into CAD, and then extract the surface S of the glued product.

[0126] Obtain the outline C of the area to be glued corresponding to the glue-filling surface S, and each point in the glue-filling surface S needs to correspond to each point in the outline C of the area to be glued.

[0127] The surface S of the glue-filling agent is divided into a plane to obtain a simple planar region and a complex structural region.

[0128] Based on the contour C of the area to be glued, the first trajectory reference contour corresponding to the simple planar area and the second trajectory reference contour corresponding to the complex structure area are obtained respectively.

[0129] Obtain the center point corresponding to the simple planar region, and simultaneously connect the center point corresponding to the simple planar region with the first trajectory reference contour to obtain the first trajectory reference line segment;

[0130] Obtain the first trajectory reference line segment and the first trajectory reference contour where they intersect;

[0131] The first trajectory reference line segment with more than two intersection points is used as the first trajectory calibration line segment;

[0132] Obtain the coordinates of the two closest intersection points in the first trajectory calibration line segment, and simultaneously obtain the center coordinates of the two closest intersection points in the first trajectory calibration line segment;

[0133] By concatenating the center coordinates of the two closest intersection points in the first trajectory calibration line segment, the first initial trajectory D1={d1, d2, ..., dz} is obtained, where dz is the z-th center coordinate of the two closest intersection points in the first trajectory calibration line segment;

[0134] The complex structure region is unfolded to obtain a complex planar region image. The center point corresponding to the complex planar region image is obtained simultaneously. The center point corresponding to the complex planar region image is connected to the second trajectory reference contour simultaneously to obtain the second trajectory reference line segment.

[0135] Obtain the second trajectory reference line segment that intersects with the second trajectory reference contour;

[0136] The second trajectory reference line segment with more than two intersection points is used as the second trajectory calibration line segment;

[0137] Obtain the coordinates of the two closest intersection points in the second trajectory calibration line segment, and simultaneously obtain the center coordinates of the two closest intersection points in the second trajectory calibration line segment;

[0138] By concatenating the center coordinates of the two closest intersection points in the second trajectory calibration line segment, the second initial trajectory D2={d1, d2, ..., dy} is obtained, where dy is the y-th center coordinate of the two closest intersection points in the second trajectory calibration line segment;

[0139] Based on the glue-filling product model, obtain the center coordinates of the glue-filling product model;

[0140] Traverse all point coordinates in the first and second initial trajectories, take the coordinates closest to the center coordinates of the glued product model as the starting point coordinates, and splice the first and second initial trajectories to obtain the initial glued trajectory information.

[0141] Specifically, firstly, a grayscale image I(x,y) of the area to be glued is acquired using an industrial camera. Gaussian filtering or median filtering algorithms are then used to denoise the image. Next, edge detection is performed using Canny or Sobel operators to extract the contour C={c1,c2,…,cn} of the area to be glued. For breaks or burrs in the contour, Bézier curve interpolation or morphological operations are used to repair defects, obtaining complete contour information. Simultaneously, surface stains (such as oil and dust) are identified based on the grayscale value distribution, and surface cleaning is performed using a robotic arm or air gun.

[0142] In CAD model processing, after importing the 3D model of the glue-filling product, the glue-filling surface S is extracted, and its point coordinate correspondence with the image contour C is matched. Surface S is divided into simple planar regions (such as rectangles and circles) and complex structural regions (such as curved surfaces and narrow slits), and trajectories are generated for each: Simple regions: The center point of the region is calculated, and line segments are generated by connecting it to the first trajectory reference contour. Line segments with ≥2 intersection points with the contour are selected as calibration line segments, and the center coordinates of the nearest intersection point are concatenated to form the first initial trajectory D1; Complex regions: After unfolding into a planar image, the second initial trajectory D2 is generated in the same way. Finally, using the model center coordinates as the reference, the trajectory point closest to the center is selected as the starting point, and D1 and D2 are spliced ​​together to form a complete initial trajectory.

[0143] The advantages are: Accuracy of image preprocessing: Denoising, edge detection, and defect repair increase the accuracy of contour extraction and improve the accuracy of surface stain recognition, ensuring the cleanliness of the glue-filling area and avoiding adhesion failure caused by impurities; Adaptability of trajectory planning: It handles simple / complex region segmentation, and the trajectory generated after complex structure unfolding increases its fit, solving the glue-filling blind spot problem of traditional fixed trajectories on curved surfaces and narrow gaps; Automation and intelligence: The entire process from image acquisition to trajectory generation is automated, eliminating the need for manual path drawing. Furthermore, the trajectory stitching mechanism based on the model center coordinates ensures the continuity of multi-region trajectories, improving glue-filling efficiency.

[0144] Understandably, the grayscale image I(x,y) reflects the surface reflectivity of pixels (x,y) and is used for edge detection and stain recognition. For example, the grayscale value of a stained area is usually lower than that of a clean surface. The contour point coordinates C={c1,…,cn} constitute the geometric boundary of the area to be glued and are the basis for trajectory planning. For example, cn=(xn,yn) defines the shape of the area. The glued surface S is the three-dimensional surface in the CAD model that needs to be glued. The correspondence between the points and the image contour C ensures the spatial matching between the virtual model and the actual workpiece. The trajectory reference contour is the boundary line of a simple / complex area used to generate trajectory segments. For example, the first trajectory reference contour defines the boundary of the glued path for a simple plane. The center coordinates are the center point of the area or the center point of the model. They serve as the reference point for trajectory generation and ensure the symmetry of the trajectory layout. For example, the center point of a complex area after expansion is used to evenly distribute the glued path.

[0145] Furthermore, environmental information of the glue-filling area is acquired, and combined with real-time image information of the glue-filling area, the initial glue-filling trajectory information and initial glue-filling basic parameters are dynamically corrected, specifically including:

[0146] Obtain environmental information of the glue-filling area, and obtain the ambient temperature and humidity from the environmental information of the glue-filling area, and simultaneously determine the temperature correction coefficient and trajectory correction coefficient;

[0147] Based on real-time image information of the glue application site, the real-time trajectory of glue application is obtained;

[0148] Based on the real-time and initial glue-dispensing trajectory information, the trajectory offset is determined;

[0149] Based on the trajectory correction coefficient, trajectory offset, and initial glue-filling trajectory information, a new glue-filling trajectory is determined;

[0150] Real-time bubble rate is obtained based on real-time image information of the glue application site;

[0151] Pressure, flow, and temperature corrections are applied to the initial dispensing trajectory information and initial dispensing base parameters.

[0152] The formula for calculating the temperature correction factor is: In the formula, The temperature correction factor at time t is... The quantified value of the ambient temperature at time t is obtained from the environmental information of the glue-filling area.

[0153] The formula for calculating trajectory offset is: In the formula, Let t be the trajectory offset of the real-time glue-pouring trajectory at time t. Let be the vector value of the point coordinates corresponding to the real-time trajectory of the glue application at time t. Let be the vector value of the point coordinates corresponding to the initial glue-pouring trajectory at time t;

[0154] The formula for calculating pressure correction is: In the formula, This represents the pressure correction at time t+1. The proportional gain is one of the core parameters in PID control. Its function is to "quickly respond" to pressure deviations. When the deviation is large, it relies on K... p Rapid adjustments bring the pressure close to the set value quickly, similar to an "emergency corrective force against current deviations." It is the integral coefficient for the time interval from 0 to T. It is responsible for handling the "cumulative deviation" when there is a continuous small deviation in pressure (such as the pressure slowly deviating due to changes in glue viscosity). It will be done through points ( By gradually correcting and eliminating long-term static errors, the pressure eventually stabilizes at the set value. The differential coefficients are defined for the time interval from 0 to T. It focuses on the "trend of change in deviation" by calculating the rate of change of the deviation. This allows for anticipating pressure fluctuations in advance. For example, if a rapid increase in pressure is detected, it can be "braked" (pressure reduced) in advance to suppress overshoot and allow for smoother pressure adjustment. For pressure deviation, e in the formula p =P set -P real That is, "pressure setpoint P" set "and actual measured pressure P" real The difference between "" and "". Positive deviation (e p A deviation >0 indicates that the actual pressure is lower than the set value, and the pressure needs to be increased; a negative deviation indicates the opposite.

[0155] The formula for calculating flow correction is: In the formula, The corrected flow rate. The flow rate before correction. Bubble rate;

[0156] The formula for calculating temperature correction is: In the formula, The corrected adhesive temperature was adjusted to match the process requirements, indirectly ensuring viscosity stability. To correct the viscosity of the adhesive, the temperature is adjusted to make the viscosity suitable for dispensing requirements (such as maintaining stable viscosity to ensure good adhesive flow and filling properties). This refers to the glue temperature before correction, i.e., the glue viscosity at the current temperature before adjustment, i.e., the glue temperature setting before adjustment at the current time of glue application. The viscosity of the glue before correction. is the temperature correction factor at time t, which relates to the relationship between temperature and viscosity changes. When the temperature changes, kT is adjusted to ensure the viscosity changes meet process requirements (e.g., maintaining stable viscosity for smooth dispensing). The ambient temperature at time t is the quantified value obtained from the environmental information of the dispensing area, representing the actual temperature around the dispensing equipment. Because ambient temperature affects the adhesive temperature and thus its viscosity, the heating / cooling strategy for the adhesive needs to be adjusted accordingly.

[0157] Specifically, firstly, environmental information at the dispensing site is collected in real time using IoT sensors (such as temperature and humidity sensors) deployed at the dispensing station, from which ambient temperature Te(t) and humidity data are extracted. Based on the temperature data, the following formula is used...

[0158] Calculate the temperature correction factor (25 is the standard temperature quantization value), and simultaneously obtain the trajectory correction factor by fitting historical trajectory deviation data. Continuously acquire real-time images of the dispensing area using an industrial camera, extract the point coordinate vectors of the real-time dispensing trajectory F(t) using an image recognition algorithm, and compare them with the initial trajectory Dpre(t).

[0159] The trajectory offset is calculated. Then, based on the trajectory correction coefficient and the offset, a new dispensing trajectory is generated using an interpolation algorithm. Simultaneously, the bubble distribution is analyzed from the real-time image, and the real-time bubble rate Br is calculated. For the dispensing parameters, corrections are made according to the following logic: Pressure correction: Based on the pressure deviation ep, the formula is used... Calculate the correction amount, integrating proportional, integral, and derivative control logic; flow correction: according to... Dynamically reduce flow rate based on bubble rate; temperature correction: via and Adjust the glue temperature and viscosity simultaneously.

[0160] The advantages are as follows: This step utilizes real-time feedback from IoT data to construct a dynamic and adaptive glue-filling parameter correction system; strong environmental adaptability: the temperature correction mechanism can control glue viscosity fluctuations within a set range when ambient temperature fluctuates, avoiding abnormal glue flow caused by temperature changes; high trajectory accuracy: real-time calculation and correction of trajectory offset improves the glue-filling trajectory fit of complex structural parts, especially reducing trajectory deviation when filling curved surfaces; significant defect suppression effect: flow correction based on bubble rate can reduce the glue bubble rate, and combined with PID correction of pressure, it can eliminate the problem of uneven glue volume caused by pressure fluctuations; high degree of intelligence: the entire process requires no manual intervention, reducing the response delay from environmental data acquisition to parameter correction, and realizing real-time closed-loop control of the glue-filling process.

[0161] Understandably, the quantified value of ambient temperature T e(t): Real-time ambient temperature of the glue-dispensing area (unit: °C), used to calculate the temperature correction factor. 25 °C is the standard reference temperature. The larger the temperature deviation, the smaller the correction factor.

[0162] Real-time trajectory vector F(t): The actual trajectory coordinates (three-dimensional vector) of the dispensing head at time t, reflecting the deviation between the current dispensing path and the theoretical path;

[0163] Pressure deviation e p The difference between the actual dispensing pressure and the preset pressure is the core basis for pressure correction. Three coefficients achieve proportional-integral-derivative control; bubble rate B r : Percentage of air bubbles in the dispensing area. For every 10% increase in the air bubble rate (set by experienced personnel), the flow rate automatically decreases by 1% to compensate for excess adhesive caused by air bubbles; Temperature correction factor k T (t): A dimensionless coefficient reflecting the degree of influence of ambient temperature on the properties of the adhesive, for example, T e (t) = 30℃ when k T (t)=0.9 indicates that the viscosity of the glue needs to be reduced by 10% to maintain its fluidity.

[0164] Furthermore, based on the initial glue-dispensing trajectory information and initial glue-dispensing basic parameters, the glue-dispensing machine is controlled to perform glue-dispensing processing on the pre-processed area to be glued, and real-time image information of the glue-dispensing area is acquired, specifically including:

[0165] Importing basic parameters: The basic parameter matrix of dispensing (including the initial pressure value P0, the initial flow rate value Q0, and the initial temperature value T0) is transmitted to the dispensing machine control system. The system automatically parses the parameters and verifies their rationality (such as whether the pressure exceeds the equipment's range).

[0166] Trajectory information mapping: Convert the initial glue dispensing trajectory information (such as the splicing path of simple planar area trajectory D1 and complex structure area trajectory D2) into glue dispensing machine motion control commands to determine the spatial motion path and coordinate point sequence of the glue dispensing head;

[0167] Equipment status verification: The dispensing machine executes the initialization process to check the sealing of the dispensing hose connection, the heating device (such as preheating to the set value at temperature T0), and the start / stop status of the pressure pump to ensure that the equipment is in an executable state;

[0168] Dry run test: Perform a dry run without glue according to the initial trajectory to verify whether the trajectory path matches the part to be glued, and avoid mechanical collisions or glue position deviations caused by incorrect trajectory coordinates;

[0169] Industrial camera deployment: Fix an industrial camera above or to the side of the dispensing station, and set the acquisition frequency (e.g., 10 frames / second), resolution (e.g., 1280×720 pixels) and grayscale / color mode to ensure that the area to be dispensed is completely within the field of view;

[0170] Real-time image acquisition: When the glue dispensing machine starts, the camera simultaneously begins acquiring grayscale images I(x,y); each frame of image is preprocessed (denoising, brightness normalization) and transmitted to the system memory in real time; the contour features (such as glue line edges, bubble distribution) of the real-time glue dispensing trajectory F(t) are extracted using an image recognition algorithm and compared with the initial trajectory D. pre (t) Perform pixel-level comparison.

[0171] Furthermore, a smart dispensing control system based on IoT data feedback is proposed to implement the control method described above, including:

[0172] The data acquisition and preprocessing module is used to obtain glue product information from the glue production plan, including application scenarios, working environment parameters, product materials and the geometric dimensions of the glue-filling area. Based on preset glue matching rules, it selects the appropriate glue type and reads feature vectors (viscosity, curing time, shrinkage rate, thermal conductivity) from the database. After normalization, it generates glue feature data. It is used to acquire grayscale images of the area to be glued through an industrial camera, perform noise reduction and edge detection to obtain the outline of the area to be glued, and repair outline defects. It is used to analyze surface stain information in the grayscale image, generate surface cleaning instructions, and simultaneously collect environmental information such as ambient temperature and humidity of the glue-filling area.

[0173] The parameter calculation and model building module is used to define an initial quality index matrix based on glue characteristic data, to preprocess the dispensing machine, including zero-point calibration of the pressure sensor and range calibration of the flow sensor, to conduct dispensing tests, record the tube size and test time, and map the initial pressure, initial flow and initial temperature values ​​(dynamically adjusted according to the viscosity threshold). The pressure, flow and temperature parameters are integrated to generate the basic dispensing parameter matrix.

[0174] The trajectory planning and model building module is used to import the glue-filling product model into CAD, extract the contours of the glue-filling surface and the area to be glued, divide the surface into simple planar areas and complex structural areas, generate a first initial trajectory by connecting the center point of the simple planar area with the trajectory reference contour, and generate a second initial trajectory by unfolding the complex structural area in the same way. The module is used to stitch together the first and second initial trajectories with the center coordinates of the glue-filling product model as a reference to determine the starting point of the initial glue-filling trajectory and generate complete initial glue-filling trajectory information.

[0175] The dispensing execution and dynamic correction module is used to control the dispensing machine to perform dispensing based on the initial dispensing trajectory and basic parameters. It collects image information of the dispensing location in real time, obtains the real-time dispensing trajectory and bubble rate, calculates the temperature correction coefficient, corrects the glue temperature and viscosity according to the ambient temperature, calculates the trajectory offset, generates a new trajectory in combination with the trajectory correction coefficient, corrects the flow rate according to the bubble rate, and adjusts the dispensing pressure through the pressure correction formula.

[0176] The system monitoring and feedback module is used to monitor parameter deviations (such as pressure, flow rate, and temperature) and trajectory deviations during the glue dispensing process. When the threshold is exceeded, an early warning is triggered, and abnormal data is recorded synchronously for subsequent analysis. It is used to integrate historical glue dispensing data, environmental parameters, and correction records to optimize glue matching rules, sensor calibration parameters, and dynamic correction coefficients, forming a closed-loop control process of "data acquisition-analysis-correction-optimization".

[0177] Furthermore, the data acquisition and preprocessing module includes:

[0178] The product information acquisition unit is used to obtain product information for dispensing from the dispensing production plan, including application scenarios, working environment parameters, product materials and geometric dimensions of the dispensing part. Based on preset adhesive matching rules, it selects the appropriate adhesive type and reads feature vectors (viscosity, curing time, shrinkage rate, thermal conductivity) from the database, and generates adhesive feature data after normalization.

[0179] The image and environment acquisition unit is used to acquire grayscale images of the area to be glued using an industrial camera, perform noise reduction and edge detection to obtain the outline of the area to be glued, and repair outline defects. It is also used to analyze surface stain information in the grayscale image, generate surface cleaning instructions, and simultaneously acquire environmental information such as temperature and humidity of the area to be glued.

[0180] Furthermore, the parameter calculation and model building module includes:

[0181] The glue quality index calculation unit is used to define an initial quality index matrix based on glue characteristic data. It is used for preprocessing of the glue dispensing machine, including zero-point calibration of the pressure sensor and range calibration of the flow sensor.

[0182] The dispensing parameter setting unit is used to perform dispensing tests, record the tube size and test time, map the initial pressure, initial flow rate and initial temperature values ​​(dynamically adjusted according to the viscosity threshold), and integrate the pressure, flow rate and temperature parameters to generate the dispensing basic parameter matrix.

[0183] Furthermore, the trajectory planning and model building module includes:

[0184] The CAD model import and trajectory generation unit is used to import the glue-filling product model into CAD, extract the contours of the glue-filling surface and the area to be glued, divide the surface into simple planar areas and complex structural areas, generate a first initial trajectory for the simple planar areas by connecting the center point with the trajectory reference contour, and generate a second initial trajectory for the complex structural areas after unfolding them in the same way. The first and second initial trajectories are spliced ​​together with the center coordinates of the glue-filling product model as a reference to determine the starting point of the initial glue-filling trajectory and generate complete initial glue-filling trajectory information.

[0185] Furthermore, the dispensing execution and dynamic correction module includes:

[0186] The dispensing execution control unit is used to control the dispensing machine to perform dispensing based on the initial dispensing trajectory and basic parameters, and to collect image information of the dispensing area in real time to obtain the real-time dispensing trajectory and bubble rate.

[0187] The parameter dynamic correction unit is used to calculate the temperature correction coefficient, correct the glue temperature and viscosity according to the ambient temperature, calculate the trajectory offset, generate a new trajectory in combination with the trajectory correction coefficient, correct the flow rate according to the bubble rate, and adjust the dispensing pressure through the pressure correction formula.

[0188] Furthermore, the system monitoring and feedback module includes:

[0189] The real-time monitoring and early warning unit is used to monitor parameter deviations (such as pressure, flow rate, and temperature) and trajectory deviations during the dispensing process. When the threshold is exceeded, an early warning is triggered, and abnormal data is recorded simultaneously for subsequent analysis.

[0190] The data closed-loop optimization unit integrates historical glue dispensing data, environmental parameters, and correction records to optimize glue matching rules, sensor calibration parameters, and dynamic correction coefficients, forming a closed-loop control process of "data acquisition-analysis-correction-optimization".

[0191] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A glue pouring intelligent control method based on Internet of Things data feedback, characterized in that, The method comprises the following steps: Obtain glue product information, select glue type based on glue product information, and obtain glue characteristic data; Determine the initial quality index matrix of the glue, simultaneously obtain the initial state information of the glue filling machine, and pretreat the glue filling machine; Based on the initial quality index matrix of the glue, set the initial glue filling basic parameters for the pretreated glue filling machine; Obtain the image information of the part to be filled with glue, and then pretreat the part to be filled with glue; Import the glue filling product model into CAD, set the initial glue filling track information by extracting the glue filling surface, dividing simple plane area and complex structure area, generating track reference line segment, screening calibration line segment, and solving the mode of concatenating center coordinates; Based on the initial glue filling track information and the initial glue filling basic parameters, control the glue filling machine to fill the pretreated part to be filled with glue, and obtain the real-time image information of the glue filling part; Obtain the regional environment information of the glue filling part, and dynamically correct the initial glue filling track information and the initial glue filling basic parameters in combination with the real-time image information of the glue filling part. The glue initial quality index matrix is a two-dimensional matrix constructed based on glue characteristic data, and the matrix expression is: In the formula, is the jth glue characteristic index in the glue characteristic data, is the jth glue characteristic index in the glue characteristic data, is the process requirement threshold of the jth glue characteristic index, j ∈ m. The glue characteristic data is G={g1, g2, g3, g4}, wherein g1 is viscosity, g2 is curing time, g3 is curing shrinkage rate, and g4 is thermal conductivity. 2.The glue filling intelligent control method based on Internet of Things data feedback according to claim 1, characterized in that, The method for obtaining glue product information, selecting glue type based on glue product information, and obtaining glue characteristic data specifically comprises: Obtain glue product information from glue filling production plan, including application scene information, working environment parameter information, product material information, and glue filling part geometric dimension information; Set glue matching rules, and then select glue type based on glue product information; After selecting the glue type, read the characteristic vector from the database, and normalize the characteristic vector to obtain the glue characteristic data. 3.The glue filling intelligent control method based on Internet of Things data feedback according to claim 1, characterized in that, The method for determining the initial quality index matrix of the glue, simultaneously obtaining the initial state information of the glue filling machine, and pretreating the glue filling machine specifically comprises: Based on the glue characteristic data, define the initial quality index matrix of the glue; Obtain the initial state information of the glue filling machine, including sensor calibration parameters, initial parameters of the actuator, and equipment running state; Based on the equipment running state, determine whether to pretreat the glue filling machine. If the equipment running state is normal, continue to obtain the initial state information of the glue filling machine. If the equipment running state is to be calibrated, calibrate the pressure sensor zero point and calibrate the flow sensor range of the glue filling machine; The pressure sensor zero point calibration of the glue filling machine is carried out: In the formula, is the pressure value of the glue filling machine pressure sensor after calibration, is the original pressure value of the glue filling machine pressure sensor, is the zero point deviation of the glue filling machine pressure sensor; Calibration of flow sensor range for glue filling machine: In the formula, is the flow value of the glue filling machine after calibration of flow sensor, is the original flow value of the glue filling machine flow sensor, is the calibration coefficient of the glue filling machine flow sensor.

4. The glue filling intelligent control method based on Internet of Things data feedback according to claim 1, characterized in that, The method for setting the initial glue filling basic parameters for the pretreated glue filling machine based on the initial quality index matrix of the glue specifically comprises: Use the pretreated glue filling machine to conduct glue filling test, simultaneously obtain the glue pipe size information, and record the glue filling test time; Map the parameters in the initial quality index matrix of the glue to obtain the initial values of pressure, flow, and temperature; Integrate the initial values of pressure, flow, and temperature into a matrix to obtain the glue filling basic parameters; The mapping formula of the pressure initial value is: In the formula, is the pressure initial value, is an empirical coefficient, is the hose length, is the hose radius, is the dynamic viscosity of the glue, g1 is the viscosity, is the average pressure value of the glue filling machine in the glue filling test; The mapping formula of the flow initial value is: In the formula, is the flow initial value, is the volume value of the glue-filling part of the glue-filling test, is the time value of the glue-filling test, is the safety factor; A viscosity mapping threshold, a temperature mapping coefficient and a temperature adjustment coefficient are set, and if g1 is greater than the viscosity mapping threshold, the mapping formula of the temperature initial value is: In the formula, is the temperature initial value, is the temperature base value, is the dynamic viscosity of the glue, is the viscosity mapping threshold, is the temperature mapping coefficient, is the temperature adjustment coefficient.

5. The glue filling intelligent control method based on Internet of Things data feedback according to claim 1, characterized in that, The method for obtaining the image information of the part to be filled with glue, and then pretreating the part to be filled with glue specifically comprises: Obtain the gray-scale image I(x,y) through an industrial camera, denoise and edge detect the gray-scale image to obtain the contour C={c1, c2, …, cn} of the part to be filled with glue, wherein cn is the point coordinates of the nth edge detection. The profile of the area to be filled with glue is repaired for defects to obtain image information of the part to be filled with glue; Based on the gray image, surface stain information corresponding to the image information of the part to be filled with glue is obtained; Based on the surface stain information corresponding to the image information of the part to be filled with glue, the surface of the part to be filled with glue is cleaned.

6. The intelligent control method for glue filling based on Internet of Things data feedback according to claim 1, characterized in that, The imported glue product model in CAD, setting the initial glue filling track information, specifically includes: Importing the glue filling product model in CAD, and then extracting the glue filling surface S; Obtain the profile C of the area to be filled with glue corresponding to the glue filling surface S, and each point in the glue filling surface S needs to correspond to each point in the profile C of the area to be filled with glue; The glue filling surface S is divided into a simple plane area and a complex structure area; Based on the profile C of the area to be filled with glue, the first trajectory reference contour corresponding to the simple plane area and the second trajectory reference contour corresponding to the complex structure area are obtained respectively; Obtain the center point corresponding to the simple plane area, and simultaneously connect the center point corresponding to the simple plane area with the first trajectory reference contour to obtain the first trajectory reference line segment; Obtain the first trajectory reference line segment that has an intersection with the first trajectory reference contour; The first trajectory reference line segment with more than two intersection points is taken as the first trajectory calibration line segment; Obtain the coordinates of the two closest intersection points in the first trajectory calibration line segment, and simultaneously obtain the center coordinates of the two closest intersection points in the first trajectory calibration line segment; The center coordinates of the two closest intersection points in the first trajectory calibration line segment are sequentially concatenated to obtain the first initial trajectory D1={d1, d2, …, dz}, wherein dz is the zth center coordinate of the two closest intersection points in the first trajectory calibration line segment. The complex structure area is unfolded to obtain a complex plane area image, and simultaneously the center point corresponding to the complex plane area image is obtained, and simultaneously the center point corresponding to the complex plane area image is connected with the second trajectory reference contour to obtain the second trajectory reference line segment; Obtain the second trajectory reference line segment that has an intersection with the second trajectory reference contour; The second trajectory reference line segment with more than two intersection points is taken as the second trajectory calibration line segment; Obtain the coordinates of the two closest intersection points in the second trajectory calibration line segment, and simultaneously obtain the center coordinates of the two closest intersection points in the second trajectory calibration line segment; The center coordinates of the two closest intersection points in the second trajectory calibration line segment are sequentially concatenated to obtain the second initial trajectory D2={d1, d2, …, dy}, wherein dy is the yth center coordinate of the two closest intersection points in the second trajectory calibration line segment. Based on the glue filling product model, the center coordinates of the glue filling product model are obtained; All point coordinates in the first initial trajectory and the second initial trajectory are traversed, the closest coordinate to the center coordinates of the glue filling product model is taken as the starting point coordinate, and the first initial trajectory and the second initial trajectory are spliced to obtain the initial glue filling track information.

7. The intelligent control method for glue filling based on Internet of Things data feedback according to claim 1, characterized in that, The environment information of the glue filling part region is obtained, and the initial glue filling track information and the initial glue filling basic parameters are dynamically corrected in combination with the real-time image information of the glue filling part, specifically including: Acquire the environment information of the glue-filling part region, and acquire the environment temperature and the environment humidity from the environment information of the glue-filling part region, and synchronously determine the temperature correction coefficient and the track correction coefficient; Based on the real-time image information of the glue-filling part, acquire the real-time track of glue-filling; Based on the real-time track of glue-filling and the initial glue-filling track information, determine the track offset; Based on the track correction coefficient, the track offset and the initial glue-filling track information, determine the new track of glue-filling; Based on the real-time image information of the glue-filling part, acquire the real-time bubble rate; Perform pressure correction, flow correction and temperature correction on the initial glue-filling track information and the initial glue-filling basic parameters; The calculation formula of the temperature correction coefficient is: In the formula, is the temperature correction coefficient at time t, is an environmental temperature quantization value obtained from the environmental information of the glue filling part region at time t; The calculation formula of the track offset is: In the formula, is the track offset of the real-time track of the glue filling at time t, is the vector value of the point coordinate corresponding to the real-time track of the glue filling at time t, is the vector value of the point coordinate corresponding to the initial glue filling track at time t; The calculation formula of the pressure correction is: In the formula, is the pressure correction amount at t+1 moment, is a proportional coefficient, is an integral coefficient within 0 to T time, is a differential coefficient within 0 to T time, is a pressure deviation; The calculation formula of the flow correction is: In the formula, is the corrected flow, is the uncorrected flow, is the bubble rate; The calculation formula of the temperature correction is: In the formula, is the corrected glue temperature, is the corrected glue viscosity, is the uncorrected glue temperature, is the uncorrected glue viscosity, is the temperature correction coefficient at t moment, is the environment temperature quantization value obtained from the environment information of the glue filling part region at t moment.

8. An intelligent control system for glue injection based on Internet of Things data feedback, for implementing the control method according to any one of claims 1-7, characterized in that, Comprise: The data acquisition and pretreatment module is used for acquiring glue-filling product information from the glue-filling production plan, including application scenarios, working environment parameters, product materials and glue-filling part geometric dimensions, selecting the suitable glue type based on the preset glue matching rule, reading the feature vector from the database, generating the glue feature data after normalization processing, acquiring the gray image of the part to be glued by the industrial camera, performing denoising and edge detection to obtain the contour of the part to be glued, repairing contour defects, analyzing the surface stain information in the gray image to generate surface cleaning instructions, and synchronously collecting the environment temperature and humidity of the glue-filling part region environment information; The parameter calculation and model construction module is used for defining the initial quality index matrix based on the glue feature data, pre-processing the glue-filling machine, including pressure sensor zero point calibration and flow sensor range calibration, and performing glue-filling test to record the glue pipe size and test time, map to obtain the pressure initial value, flow initial value and temperature initial value, and integrate the pressure, flow and temperature parameters to generate the glue-filling basic parameter matrix; The trajectory planning and model construction module is used for importing the glue-filling product model in CAD, extracting the glue-filling surface and the contour of the part to be glued, dividing the surface into simple plane region and complex structure region, generating the first initial track by connecting the center point and the track reference contour for the simple plane region, generating the second initial track for the complex structure region after unfolding, splicing the first initial track and the second initial track based on the center coordinates of the glue-filling product model as the reference to determine the initial glue-filling track starting point and generate the complete initial glue-filling track information; The glue-filling execution and dynamic correction module is used for controlling the glue-filling machine to perform glue-filling based on the initial glue-filling track and the basic parameters, collecting the image information of the glue-filling part in real time, acquiring the real-time track and the bubble rate of glue-filling, calculating the temperature correction coefficient to correct the glue temperature and viscosity according to the environment temperature, calculating the track offset to generate a new track in combination with the track correction coefficient, and correcting the flow according to the bubble rate and adjusting the glue-filling pressure through the pressure correction formula.

9. The intelligent control system for glue filling based on data feedback of Internet of Things according to claim 8, characterized in that, The parameter calculation and model construction module comprises: The glue quality index calculation unit is used for defining the initial quality index matrix based on the glue feature data, and pre-processing the glue-filling machine, including pressure sensor zero point calibration and flow sensor range calibration. A glue-filling parameter setting unit is configured to perform a glue-filling test, record a glue pipe size and a test time, map to obtain a pressure initial value, a flow initial value and a temperature initial value, and integrate pressure, flow and temperature parameters to generate a glue-filling basic parameter matrix.

10. The intelligent control system for glue filling based on data feedback of Internet of Things according to claim 8, characterized in that, The trajectory planning and model construction module comprises: A CAD model import and trajectory generation unit is configured to import a glue-filling product model in CAD, extract a glue-filling surface and a region to be glued contour, divide the surface into a simple plane region and a complex structure region, generate a first initial trajectory by connecting a center point and a trajectory reference contour for the simple plane region, generate a second initial trajectory for the complex structure region after unfolding, splice the first initial trajectory and the second initial trajectory with the center coordinates of the glue-filling product model as a reference, determine an initial glue-filling trajectory starting point, and generate complete initial glue-filling trajectory information.

Images