Intelligent coating system for silicone adhesives based on multi-source data fusion

Abstract

This invention relates to the field of adhesive coating quality control technology, specifically to an intelligent coating system for silicone adhesives based on multi-source data fusion. The system includes: a data acquisition module for collecting multi-source data; a thermal impedance determination module for calculating the thermal impedance fluctuation characterization value of the adhesive layer; a state diversion module for determining the internal density state of the adhesive layer based on the thermal impedance fluctuation characterization value; a density attenuation determination module for determining the density attenuation characterization value of the adhesive layer; a density detection module for determining whether density degradation exists in the warning area based on the density attenuation characterization value when the internal density state is in a warning state; an acoustic emission determination module for determining the frequency value of acoustic emission events; and an acoustic field evaluation module for determining the adjustment strategy of the adhesive layer based on the frequency value of acoustic emission events when the internal density state is in an abnormal state, thereby improving the coating efficiency of silicone adhesives.

Description

Technical Field

[0001] This invention relates to the field of adhesive coating quality control technology, specifically to an intelligent coating system for silicone adhesives based on multi-source data fusion. Background Technology

[0002] Silicone adhesives are widely used for waterproofing and sealing building joints due to their excellent weather resistance, resistance to high and low temperatures, and elasticity. These applications include concealed works such as joints in prefabricated building exterior walls, and overlapping joints in metal cladding of storage tanks and pipelines. The sealing quality of these joints directly affects the waterproofing performance and service life of the building.

[0003] During the adhesive application process, silicone adhesives exhibit moisture-curing characteristics, with preferential surface skinning and delayed internal curing. After application, the surface layer of the adhesive rapidly forms a dense skin due to contact with moisture in the air, while trace amounts of gas generated during the deeper curing of the internal adhesive layer are trapped beneath the skin, easily forming subsurface microbubble clusters invisible to the surface. This type of defect presents a latent characteristic of "normal surface, abnormal interior," and is a significant microscopic cause of premature seal failure.

[0004] Currently, the inspection of sealant application quality mainly relies on manual visual inspection and post-application sampling. Manual visual inspection struggles to accurately determine the thickness, fullness, and internal air bubbles of the sealant layer, especially in high-altitude or poorly lit areas, resulting in a high rate of missed detections. Methods such as water spray tests and ultrasonic testing are post-application sampling methods and cannot detect internal defects in the sealant layer in real-time and comprehensively during application. Furthermore, existing inspection methods mostly rely on single sensors, only acquiring surface information and failing to effectively identify hidden defects that appear normal on the surface but are internally abnormal. Moreover, they cannot convert inspection results into real-time process parameter adjustments before the sealant layer cures. Therefore, there is an urgent need in this field for an intelligent system capable of real-time, quantitatively detecting the internal density of the sealant layer during coating and adaptively adjusting coating process parameters accordingly. Summary of the Invention

[0005] To address this issue, the present invention provides an intelligent coating system for silicone adhesives based on multi-source data fusion. This system overcomes the problem in the prior art that does not consider the influence of the adhesive layer characteristics of silicone adhesives on the internal density of the adhesive layer, and cannot judge the internal defects of the adhesive layer and adjust the coating process in a timely manner based on the difference in multi-physics field signals generated by surface skinning and internal delayed curing, thus resulting in poor coating efficiency of silicone adhesives.

[0006] To achieve the above objectives, the present invention provides an intelligent coating system for silicone adhesives based on multi-source data fusion, comprising: The data acquisition module is used to collect multi-source data during the coating process of silicone adhesive. The multi-source data includes the current coating rate, the current coating pressure, the surface temperature of the adhesive layer, the contact force signal of the adhesive layer, and the acoustic emission signal inside the adhesive layer. A thermal impedance determination module, which is connected to the data acquisition module, is used to calculate the thermal impedance fluctuation characterization value of the adhesive layer based on the surface temperature; A state diversion module, which is connected to the thermal impedance determination module, is used to determine the internal density state of the adhesive layer based on the thermal impedance fluctuation characterization value, wherein the internal density state includes a normal state, a warning state, and an abnormal state. A density attenuation determination module, which is connected to the data acquisition module, is used to determine the density attenuation characterization value inside the adhesive layer based on the contact force signal corresponding to the warning area. A density detection module, which is connected to the state diversion module and the density attenuation determination module respectively, is used to determine whether there is density degradation in the warning area based on the density attenuation characterization value when the internal density state is a warning state. An acoustic emission determination module, which is connected to the data acquisition module, determines the frequency value of acoustic emission events based on the acoustic emission signal; The acoustic field evaluation module is connected to the acoustic emission determination module, the data acquisition module, and the state diversion module, respectively. It is used to determine the adjustment strategy of the adhesive layer based on the frequency value of the acoustic emission event when the internal density state is abnormal. The adjustment strategy is to increase the coating pressure or output a shutdown command.

[0007] Preferably, the state diversion module determines that the internal density state of the adhesive layer is normal in response to the thermal impedance fluctuation characterization value being less than the first preset impedance fluctuation threshold. The state diversion module, in response to the thermal impedance fluctuation characterization value being greater than or equal to the second preset impedance fluctuation threshold, determines that the internal density state of the adhesive layer is abnormal and triggers the sound field evaluation module to determine the adjustment strategy of the adhesive layer based on the frequency value of the acoustic emission event; wherein, the first preset impedance fluctuation threshold is less than the second preset impedance fluctuation threshold.

[0008] Preferably, the state diversion module responds to the thermal impedance fluctuation characterization value being greater than or equal to the first preset impedance fluctuation threshold and less than the second preset impedance fluctuation threshold by determining that the internal density state of the adhesive layer is a warning state, and triggers the density detection module to determine whether there is density degradation in the warning area based on the density attenuation characterization value inside the adhesive layer. The warning area is the area enclosed by sampling points whose absolute temperature gradient value is greater than a preset gradient threshold, and the sampling points are two-dimensional grid nodes obtained by dividing the adhesive layer surface into equal intervals.

[0009] Preferably, the thermal impedance determination module includes: The sliding calculation unit is used to set the sliding sub-window to traverse the surface temperature sequence of the adhesive layer along the time axis, calculate the ratio of the standard deviation to the mean of the temperature sequence in each sliding sub-window, and record it as the sub-window thermal resistance fluctuation value of each sliding sub-window. The instantaneous value determination unit is used to determine the maximum value among the thermal impedance fluctuation values ​​of each sub-window as the instantaneous thermal impedance fluctuation characterization value; The ratio determination unit is used to determine the ratio of the instantaneous thermal impedance fluctuation characterization value to the preset reference value as the thermal impedance fluctuation characterization value.

[0010] Preferably, the density detection module determines that there is density degradation in the warning area when the density attenuation characterization value inside the adhesive layer is greater than or equal to the preset density attenuation characterization value, and reduces the coating rate according to the difference between the density attenuation characterization value and the preset density attenuation characterization value.

[0011] Preferably, the density attenuation determination module is used to determine the density attenuation characterization value inside the adhesive layer based on the warning area. The density attenuation determination module obtains the energy of a preset characteristic frequency band based on the contact force signal corresponding to the warning area, calculates the ratio of the energy of the preset characteristic frequency band to the energy of the preset total frequency band, obtains the high-frequency energy ratio of each warning area, and determines the maximum value of each high-frequency energy ratio as the density attenuation characterization value inside the adhesive layer.

[0012] Preferably, the density detection module is equipped with several rate adjustment methods to address the reduction in the coating rate, and each rate adjustment method reduces the coating rate by a different amount.

[0013] Preferably, the sound field evaluation module includes: A frequency determination unit is used to compare the frequency value of the acoustic emission event with a preset frequency threshold. A strategy determination unit is used to determine an adjustment strategy for the adhesive layer based on the comparison results of the frequency determination unit, wherein... If the frequency value of the acoustic emission event is less than the preset frequency threshold, the adjustment strategy is determined to be to increase the coating pressure based on the difference between the preset frequency threshold and the frequency value of the acoustic emission event. If the frequency value of the acoustic emission event is greater than or equal to the preset frequency threshold, the adjustment strategy is determined to be to output a shutdown command.

[0014] Preferably, the acoustic emission determination module includes: The spectrum analysis unit is used to perform spectrum analysis on the acoustic emission signal of the adhesive layer, extract the signal components in the preset high frequency band, and obtain the high frequency acoustic emission signal. An event recognition unit is used to compare the amplitude of the high-frequency acoustic emission signal with a preset background noise threshold to identify an acoustic emission event. Specifically, when the amplitude of the high-frequency acoustic emission signal rises from less than the preset background noise threshold to greater than the preset background noise threshold, it is determined as the start of an acoustic emission event. When the amplitude of the high-frequency acoustic emission signal falls back to less than the preset background noise threshold and remains locked for a preset duration, it is determined as the end of the acoustic emission event. An event counting unit is used to count the total number of acoustic emission events that occur within a preset acquisition time. The frequency calculation unit is used to divide the total number of acoustic emission events by the preset acquisition duration to obtain the frequency value of the acoustic emission events.

[0015] Preferably, the increase in coating pressure is positively correlated with the difference between the preset frequency threshold and the frequency value of the acoustic emission event.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention simultaneously acquires the surface temperature, contact force signal, and acoustic emission signal of the adhesive layer, enabling parallel sensing of three different physical dimensions—thermal field, force field, and acoustic field—during the coating process of silicone adhesive. A state-splitting module categorizes the internal density state of the adhesive layer into three differentiated processing paths: normal state, warning state, and abnormal state, based on thermal impedance fluctuation characterization values. In the warning state, a density attenuation determination module and a density detection module perform cross-modal secondary cross-validation using the frequency domain characteristics of the contact force signal within the warning area located by the thermal field. In the abnormal state, an acoustic emission determination module and an acoustic field evaluation module determine the repair status of the adhesive layer based on the frequency value of acoustic emission events and output a graded adjustment strategy, such as increasing the coating pressure or issuing a shutdown command. This constructs a progressive judgment and graded control closed-loop architecture, achieving real-time differentiated control of non-destructive testing and coating process parameters, thereby improving the coating efficiency of silicone adhesive.

[0017] 2. This invention achieves a three-level classification of the internal density state of the adhesive layer by comparing the thermal impedance fluctuation characterization value with the first preset impedance fluctuation threshold and the second preset impedance fluctuation threshold, and respectively corresponding to different processing paths such as maintaining the current process parameters, triggering the density detection module for secondary verification, and triggering the sound field evaluation module to determine the adjustment strategy. This realizes a graded response mechanism in the allocation of detection resources, thereby avoiding the problem of high false alarm rate caused by single threshold judgment.

[0018] 3. This invention determines the adhesive layer with thermal impedance fluctuation characterization value between the first preset impedance fluctuation threshold and the second preset impedance fluctuation threshold as a warning state, and defines the area enclosed by sampling points with an absolute temperature gradient value greater than a preset gradient threshold as the warning area. This allows the density detection module to perform directional analysis of the contact force signal within the warning area without having to perform a full-area scan of the entire adhesive layer, thereby improving the pertinence of the determination of density degradation in the warning area.

[0019] 4. This invention determines the presence of densification degradation in the adhesive layer by identifying a pre-set densification degradation value within the adhesive layer when the density attenuation characterization value is greater than or equal to the pre-set value, and then reduces the coating rate. This achieves a quantitative assessment of the severity of densification degradation within the adhesive layer and precise matching of control measures. Reducing the coating rate prolongs the residence time of the adhesive layer in the open environment, slowing down the surface skinning process. This provides a more sufficient time window for the gas inside the adhesive layer to escape before the surface is completely sealed, allowing for process intervention before the microbubble clusters are permanently encapsulated.

[0020] 5. This invention achieves effective differentiation between repairable severe anomalies and irreparable fatal defects by comparing the frequency value of acoustic emission events with a preset frequency threshold under abnormal conditions and outputting a differentiated adjustment strategy based on this comparison to increase coating pressure or shut down the machine. Attached Figure Description

[0021] The present application will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0022] Figure 1 This is a schematic diagram of the module connection of the intelligent coating system for silicone adhesives based on multi-source data fusion according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating how the internal compactness of the adhesive layer is determined based on the thermal resistance fluctuation characterization value, according to an embodiment of the present invention. Figure 3 This is a flowchart illustrating how an embodiment of the present invention determines whether a pre-warning region exhibits density degradation based on the density attenuation characterization value. Figure 4 This is a flowchart illustrating the process of determining the adjustment strategy of the adhesive layer based on the frequency value of the acoustic emission event, according to an embodiment of the present invention. Detailed Implementation

[0023] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0024] Please see Figure 1 , Figure 2 , Figure 3 as well as Figure 4 The diagrams shown are: a schematic diagram of the module connection of the intelligent coating system for silicone adhesive based on multi-source data fusion according to an embodiment of the present invention; a flowchart of determining the internal density state of the adhesive layer based on the thermal impedance fluctuation characterization value according to an embodiment of the present invention; a flowchart of determining whether there is density degradation in the warning area based on the density attenuation characterization value according to an embodiment of the present invention; and a flowchart of determining the adjustment strategy of the adhesive layer based on the acoustic emission event frequency value according to an embodiment of the present invention.

[0025] This invention provides an intelligent coating system for silicone adhesives based on multi-source data fusion, comprising: The data acquisition module is used to collect multi-source data during the silicone adhesive coating process. The multi-source data includes the current coating rate, the current coating pressure, the surface temperature of the adhesive layer, the contact force signal of the adhesive layer, and the acoustic emission signal inside the adhesive layer. It can be understood that the current coating rate is provided in real time by the robot controller of the coating system, in mm / s; the current coating pressure is provided in real time by the pressure sensor of the adhesive supply line of the coating system, in MPa. A thermal impedance determination module, which is connected to the data acquisition module, is used to calculate the thermal impedance fluctuation characterization value of the adhesive layer based on the surface temperature; A state diversion module, which is connected to the thermal impedance determination module, is used to determine the internal density state of the adhesive layer based on the thermal impedance fluctuation characterization value, wherein the internal density state includes a normal state, a warning state, and an abnormal state. A density attenuation determination module, which is connected to the data acquisition module, is used to determine the density attenuation characterization value inside the adhesive layer based on the contact force signal corresponding to the warning area. A density detection module, which is connected to the state diversion module and the density attenuation determination module respectively, is used to determine whether there is density degradation in the warning area based on the density attenuation characterization value when the internal density state is a warning state. An acoustic emission determination module, which is connected to the data acquisition module, determines the frequency value of acoustic emission events based on the acoustic emission signal; The acoustic field evaluation module is connected to the acoustic emission determination module, the data acquisition module, and the state diversion module, respectively. It is used to determine the adjustment strategy of the adhesive layer based on the frequency value of the acoustic emission event when the internal density state is abnormal. The adjustment strategy is to increase the coating pressure or output a shutdown command.

[0026] In this embodiment, the specific structure of the state splitting module, the compactness detection module, and the sound field evaluation module is not limited. They themselves and each unit therein can be composed of logic components, including field-programmable components, computers, or microprocessors in computers.

[0027] Specifically, the silicone adhesive described in this embodiment is preferably a one-component room temperature vulcanizing silicone rubber (RTV-1), and more preferably a deoxime-type or de-alcohol-type one-component room temperature vulcanizing silicone rubber. This type of adhesive is widely used in industrial coatings, such as sealing new energy vehicle battery packs and potting electronic components. Its typical physicochemical parameters are as follows: under standard environmental conditions (temperature 23℃±2℃, relative humidity 50%±5%), its initial dynamic viscosity is 30000mPa·s-80000mPa·s (it is a semi-fluid paste, and micro-air bubbles are easily trapped during coating due to shear); its density is 1.0 g / cm³. 3 -1.4g / cm 3 The surface drying time (i.e. the time it takes for the surface to completely lose its fluidity and form an elastic cured film) is usually 10-30 minutes. However, within 3-5 minutes after coating, the surface will undergo preliminary physical skinning (the surface viscosity increases sharply and loses its macroscopic fluidity).

[0028] Specifically, the state diversion module determines that the internal density of the adhesive layer is normal when the thermal impedance fluctuation characterization value is less than the first preset impedance fluctuation threshold of 1.3. The state diversion module, in response to the thermal impedance fluctuation characterization value being greater than or equal to the second preset impedance fluctuation threshold 2.0, determines that the internal density state of the adhesive layer is abnormal and triggers the sound field evaluation module to determine the adjustment strategy of the adhesive layer based on the frequency value of the acoustic emission event; wherein, the first preset impedance fluctuation threshold is less than the second preset impedance fluctuation threshold.

[0029] Specifically, the state diversion module, in response to the thermal impedance fluctuation characterization value being greater than or equal to the first preset impedance fluctuation threshold and less than the second preset impedance fluctuation threshold, determines the internal density state of the adhesive layer as a warning state and triggers the density detection module to determine whether there is density degradation in the warning area based on the density attenuation characterization value inside the adhesive layer. The warning area is the region enclosed by sampling points whose absolute temperature gradient value is greater than the preset gradient threshold of 0.09℃ / mm. The sampling points are two-dimensional mesh nodes obtained by dividing the adhesive layer surface into equally spaced nodes (0.5mm in this embodiment). It can be understood that the temperature gradient is calculated along the extension direction of the adhesive layer, i.e., the temperature difference between adjacent sampling points along the extension direction is divided by the sampling point spacing along the extension direction (i.e., 0.5mm).

[0030] Specifically, when the thermal impedance fluctuation value is less than the first preset impedance fluctuation threshold, it indicates that the dispersion of the surface temperature distribution of the adhesive layer is not significantly different from the temperature fluctuation level under normal coating conditions, the internal heat conduction of the adhesive layer is uniform, and there is no significant bubble cluster area. Therefore, it is judged as a normal state, and there is no need to start subsequent force field or acoustic detection. The monitoring of the current adhesive layer can be ended directly. When the thermal impedance fluctuation value is greater than or equal to the first preset impedance fluctuation threshold and less than the second preset impedance fluctuation threshold, it indicates that the dispersion of the surface temperature distribution of the adhesive layer has exceeded the normal fluctuation range, and local temperature anomalies have occurred. However, the degree of this anomaly is not sufficient to directly determine it as a serious defect. It may be caused by microbubble clusters inside the adhesive layer, or by external interference factors such as instantaneous fluctuations in ambient temperature and random noise from infrared sensors. Therefore, it is determined to be a warning state, and a secondary cross-verification is required in the temperature anomaly area using a force field sensor to confirm whether the thermal field anomaly truly originates from the deterioration of the density inside the adhesive layer, thereby avoiding false alarms caused by environmental interference due to a single thermal field judgment. When the thermal impedance fluctuation characterization value is greater than or equal to the second preset impedance fluctuation threshold, it indicates that the dispersion of the surface temperature distribution of the adhesive layer has deviated significantly from the normal range, the degree of heat conduction is severely hindered, and there is a large area of ​​significant microbubble clusters inside the adhesive layer. At this time, there is no need to perform force field confirmation, and it is directly determined to be an abnormal state and enters the acoustic rating to assess whether the serious defect can still be repaired through process adjustments.

[0031] In this embodiment, the first preset impedance fluctuation threshold ranges from [1.0, 1.5], and the second preset impedance fluctuation threshold ranges from [1.8, 2.3]. Preferably, the first preset impedance fluctuation threshold is 1.3, and the second preset impedance fluctuation threshold is 2.0. In 50 groups of defect-free coating tests, the mean value of the thermal impedance fluctuation characterization value was 1.00, the standard deviation was 0.04, and the mean plus three times the standard deviation was 1.12. This indicates that the fluctuation of the thermal impedance fluctuation characterization value within the range of 1.0 to 1.12 is a normal statistical fluctuation due to the differences in the micro-uniformity of the material and the slight fluctuations in the ambient temperature. The first preset impedance fluctuation threshold is set to a range of [1.0, 1.5]. The lower limit of 1.0 corresponds to the mean level of defect-free samples, while the upper limit of 1.5 provides sufficient tolerance space to accommodate the material dispersion of different batches of silicone adhesives and normal fluctuations under different ambient temperatures. Preferably, it is set to 1.3, which provides a safety margin above the mean plus three standard deviations (1.12). This not only avoids the normal adhesive layer being falsely triggered as a warning state, but also allows for timely detection of anomalies when the temperature dispersion deviates significantly. Fifty samples confirmed as having severe defects (i.e., dense bubbles forming large cavities) by cross-sectional microscopic observation after curing had thermal impedance fluctuation values ​​ranging from 1.8 to 2.3. The lower limit of 1.8 corresponds to the lowest value of severe defect samples, effectively distinguishing between warning and abnormal states. Preferably, it is set to 2.0, which is in the low to medium range of severe defect sample distribution, enabling high sensitivity in triggering abnormal state detection and avoiding missed detection of severe defects due to excessively high threshold settings.

[0032] Specifically, the thermal impedance determination module includes: For acquiring the surface temperature of the adhesive layer, an infrared thermal imaging sensor was used to continuously acquire the temperature sequence of the adhesive layer surface at a sampling frequency of 50Hz. The infrared thermal imaging sensor was installed behind the dispensing head, with a field of view covering the entire adhesive layer surface. The adhesive layer surface was divided into equally spaced two-dimensional grid nodes as sampling points. The spacing between each sampling point was 0.5mm in the adhesive layer extension direction and 0.5mm in the adhesive layer width direction. The preset acquisition time was 2.5 seconds, and a total of 125 frames of temperature data were obtained. Each frame contained the temperature value of each sampling point, in °C.

[0033] The sliding calculation unit has a sliding sub-window width set to 500ms (corresponding to 25 frames of temperature data) and a step size of 1 frame. It slides along the time axis from the start to the end of the acquisition duration, generating a total of 101 sliding sub-windows. For each sliding sub-window, the ratio of the standard deviation to the mean of the temperature in each of the 25 frames within the sub-window is taken and recorded as the sub-window thermal impedance fluctuation value of each sliding sub-window. The instantaneous value determination unit is used to determine the maximum value among the thermal impedance fluctuation values ​​of each sub-window as the instantaneous thermal impedance fluctuation characterization value; The ratio determination unit is used to determine the ratio of the instantaneous thermal resistance fluctuation characterization value to the preset benchmark value of 0.08 as the thermal resistance fluctuation characterization value. The preset benchmark value is determined as follows: before the formal coating construction, an environment with the same construction conditions as the target conditions is selected, namely, temperature 23℃±2℃, relative humidity 50%±5% and process parameters. The same batch of silicone adhesive is used, and multiple sets of defect-free coating tests are carried out at a coating rate of 100mm / s and a coating pressure of 0.3MPa. The instantaneous thermal resistance fluctuation characterization value of each coating is collected, and the arithmetic mean of the results of multiple tests is taken as the preset benchmark value.

[0034] Specifically, the density detection module determines that there is density degradation in the warning area when the density attenuation characterization value inside the adhesive layer is greater than or equal to the preset density attenuation characterization value of 0.11, and reduces the coating rate according to the difference between the density attenuation characterization value and the preset density attenuation characterization value; the density detection module determines that there is no density degradation in the warning area when the density attenuation characterization value inside the adhesive layer is less than the preset density attenuation characterization value.

[0035] Specifically, the density attenuation characterization value reflects the degree of local stiffness anomaly in the adhesive layer within the warning area caused by microbubble clusters. When the density attenuation characterization value is greater than or equal to the preset density attenuation characterization value, it indicates that at least one region within the warning area has a significantly higher high-frequency energy ratio. The stiffness of this local region exhibits an anomaly that can be detected by the force field sensor, and this anomaly region spatially coincides with the temperature anomaly region located by the first layer of thermal field. The thermal field detects temperature anomalies through impaired thermal conduction, while the force field detects stiffness anomalies through contact mechanical response. Two completely different physical principles arrive at the same conclusion in the same spatial location, providing double confirmation of the physical evidence for the existence of internal defects. This eliminates external interference factors such as ambient temperature fluctuations or instantaneous noise from the infrared sensor, thus determining that there is density degradation in the warning area. At this time, reducing the coating rate can prolong the residence time of the adhesive layer in the open environment, slow down the surface skinning speed, and provide a more sufficient time window for the gas inside the adhesive layer to escape before the surface is completely sealed. When the density attenuation characterization value is less than the preset density attenuation characterization value, it indicates that the force field sensor in the warning area has not detected a significant stiffness anomaly. The thermal field warning may be caused by external factors such as instantaneous fluctuations in ambient temperature and random noise from the infrared sensor. Therefore, it is determined that there is no density degradation in the warning area.

[0036] Specifically, an environment identical to the target construction conditions was selected: temperature 23℃±2℃, relative humidity 50%±5%, and process parameters (coating rate 100mm / s, coating pressure 0.3MPa). Fifty sets of defect-free coating tests were conducted using the same batch of silicone adhesive. Contact force signals corresponding to the warning areas of each adhesive layer were collected. The density attenuation characterization value of each adhesive layer was obtained according to the calculation method of the density attenuation determination module. The arithmetic mean of multiple test results was taken as the preset density attenuation characterization value. In this embodiment, the preset density attenuation characterization value was 0.11.

[0037] Specifically, a capacitive flexible tactile sensor is used to acquire the contact force signal of the adhesive layer. Since the silicone adhesive rapidly forms a skin after application, and remains viscous before skinning, the following measures are taken in this embodiment to prevent damage to the uncured adhesive surface or adhesion contamination during sensor sliding: the capacitive flexible tactile sensor is coated with a 10μm thick Teflon anti-stick coating, and the sensor is mounted behind the applicator head, ensuring that the adhesive surface has already formed a preliminary skin when it contacts the adhesive layer. The capacitive flexible tactile sensor slides against the adhesive surface with millinewton-level pre-pressure, making only elastic micro-contact with the pre-skinned surface without puncturing the skin. The capacitive flexible tactile sensor continuously acquires the contact force signal at a sampling frequency of 1000Hz, and simultaneously uses a robot or motor encoder to record the precise spatial position corresponding to each sampling point, establishing a mapping relationship between the force signal and the spatial position.

[0038] Specifically, a short-time Fourier transform is performed on the force signal segment corresponding to the warning area to extract the energy of the preset characteristic frequency band from 80Hz to 150Hz, which is recorded as the preset characteristic frequency band energy. This energy is then compared with the energy of the preset total frequency band from 0Hz to 500Hz, which is recorded as the preset total frequency band energy. The ratio of the two is calculated as the high-frequency energy ratio of the warning area. The maximum value of the high-frequency energy ratio in all warning areas is taken as the density attenuation characterization value of the adhesive layer.

[0039] Specifically, the density detection module has several rate adjustment methods for addressing the decrease in the coating rate, wherein, If the density attenuation deviation value is less than the first preset density attenuation deviation value of 0.10, then the coating rate is reduced to the corresponding value using the first adjustment coefficient of 0.95; If the density attenuation deviation value is greater than or equal to the first preset density attenuation deviation value and less than the second preset density attenuation deviation value of 0.20, then the coating rate is reduced to the corresponding value using the second adjustment coefficient of 0.90. If the density attenuation deviation value is greater than or equal to the second preset density attenuation deviation value, the coating rate is reduced to the corresponding value using a third adjustment coefficient of 0.85. The first preset dense attenuation deviation value is less than the second preset dense attenuation deviation value; the dense attenuation deviation value is the difference between the dense attenuation characterization value and the preset dense attenuation characterization value.

[0040] Specifically, the sound field evaluation module includes: A frequency determination unit is used to compare the frequency value of the acoustic emission event with a preset frequency threshold of 18 times / second; A strategy determination unit is used to determine an adjustment strategy for the adhesive layer based on the comparison results of the frequency determination unit, wherein... If the frequency value of the acoustic emission event is less than the preset frequency threshold, the adjustment strategy is determined to be to increase the coating pressure based on the difference between the preset frequency threshold and the frequency value of the acoustic emission event. If the frequency value of the acoustic emission event is greater than or equal to the preset frequency threshold, the adjustment strategy is determined to be to output a shutdown command.

[0041] Specifically, the acoustic emission event frequency reflects the frequency of bubble wall rupture events within the adhesive layer per unit time, and is a direct indicator of the density of microstructural damage within the adhesive layer. When the acoustic emission event frequency is less than the preset frequency threshold, it indicates that the frequency of bubble wall rupture events is still within a controllable range, the microbubble walls within the adhesive layer have not ruptured extensively, and the adhesive layer still maintains structural continuity. At this time, increasing the coating pressure can apply greater external extrusion force, partially offsetting the stretching of the bubble walls by the curing shrinkage stress, inhibiting further bubble expansion, and reducing the risk of continuous bubble wall rupture. When the acoustic emission event frequency is greater than or equal to the preset frequency threshold, it indicates that the frequency of bubble wall rupture events per unit time has exceeded the controllable range, and large-area, irreversible microstructural damage has occurred within the adhesive layer. After a large number of bubble walls rupture, they cannot reclose. At this time, any adjustment of process parameters cannot restore the adhesive layer to a dense state, and continued coating will produce an irreparable defective adhesive layer. Therefore, a stop command should be issued immediately to prevent the defective adhesive layer from entering subsequent processes, and manual intervention is required for troubleshooting.

[0042] Specifically, samples with bubble defects of varying severity were prepared through preliminary experiments. The frequency values ​​of acoustic emission events for each sample were collected before complete curing. After complete curing, the adhesive layer was subjected to cross-sectional microscopic observation to count the number of bubble wall ruptures and the formation of through-cavities. The irreparable state, characterized by numerous bubble wall ruptures and the formation of through-cavities after curing, was used as the criterion for determination. The frequency range of acoustic emission events for the corresponding sample before curing was then calculated, and the lower limit of this range was taken as the preset frequency threshold. In this embodiment, after 50 sets of preliminary experiments, the frequency range of acoustic emission events for irreparable samples before curing was found to be 18 to 30 times per second. The lower limit was taken as the preset frequency threshold, i.e., the preset frequency threshold was set to 18 times per second.

[0043] Specifically, the acoustic emission determination module includes: For the acquisition of acoustic emission signals inside the adhesive layer, a piezoelectric acoustic emission sensor array is used, which is installed behind the adhesive applicator and coupled to the adhesive layer surface. Acoustic emission waveform data is continuously acquired at a sampling frequency of 1MHz, and the preset acquisition time is 2.5 seconds. The spectrum analysis unit is used to perform spectrum analysis on the acoustic emission signal of the adhesive layer, extract the signal components in the preset high frequency band of 60kHz to 300kHz, and obtain the high frequency acoustic emission signal. An event recognition unit is used to compare the amplitude of the high-frequency acoustic emission signal with a preset background noise threshold of 25dB to identify acoustic emission events. Specifically, an acoustic emission event is defined as the start of an event when the amplitude of the high-frequency acoustic emission signal rises from below the preset background noise threshold to above it, and the event ends when the amplitude of the high-frequency acoustic emission signal falls back below the preset background noise threshold and remains locked for a preset duration of 2ms. The preset background noise threshold is determined by continuously collecting acoustic emission signals for 10 seconds while the coating equipment is running but not dispensing adhesive, statistically analyzing their amplitude distribution, and using the mean plus three times the standard deviation as the preset background noise threshold. The preset lock duration is set based on the fact that the acoustic emission signal triggered by a bubble wall rupture event (i.e., an acoustic emission event) gradually decays after an initial rise, accompanied by aftershocks, and the signal amplitude may repeatedly cross the preset background noise threshold. If no lock duration is set, the aftershocks of the same rupture event will be repeatedly counted as multiple independent events, resulting in an artificially high frequency value for acoustic emission events. An event counting unit is used to count the total number of acoustic emission events that occur within a preset acquisition time of 2.5 seconds; The frequency calculation unit is used to divide the total number of acoustic emission events by the preset acquisition duration to obtain the frequency value of the acoustic emission events.

[0044] Specifically, the increase in coating pressure is positively correlated with the difference between the preset frequency threshold and the acoustic emission event frequency value. The increase in coating pressure = (increase in reference pressure × (preset frequency threshold - acoustic emission event frequency value) / preset frequency threshold), where the reference pressure increase is 0.05 MPa. The reference pressure increase is determined based on the following: when the acoustic emission event frequency value is close to the preset frequency threshold, the risk of bubble wall rupture is high, and the required increase in coating pressure is approximately 15% to 20% of the normal coating pressure. In this embodiment, the normal coating pressure is 0.3 MPa, and approximately 17% of this is taken as the reference pressure increase, i.e., 0.05 MPa. It can be understood that the larger the difference between the preset frequency threshold and the acoustic emission event frequency value, the greater the increase in coating pressure.

[0045] Example 1: Normal coating condition (no defects) In a sealing coating operation for a new energy vehicle battery pack, the system was set with an initial coating rate of 100 mm / s and a coating pressure of 0.3 MPa. The data acquisition module collected the surface temperature of the adhesive layer in real time. The thermal impedance determination module calculated the thermal impedance fluctuation value of each sliding sub-window within a 2.5-second acquisition period, and took the maximum value to obtain an instantaneous thermal impedance fluctuation characterization value of 0.084. The ratio determination unit compared this with a preset benchmark value of 0.08, obtaining a current thermal impedance fluctuation characterization value of 1.05. The state diversion module compared 1.05 with a first preset impedance fluctuation threshold of 1.3. Since 1.05 is less than 1.3, the coating system determined that the internal density of the current adhesive layer was normal. The system did not trigger subsequent force and sound field detection, maintaining the current coating rate of 100 mm / s and coating pressure of 0.3 MPa to continue operation.

[0046] Example 2: Warning status condition (minor bubble defect, triggering speed reduction) In another coating trajectory, slight pulsation in the adhesive supply line caused micro-air bubbles to become trapped inside the adhesive layer. The thermal resistance determination module calculated the current thermal resistance fluctuation characterization value to be 1.65.

[0047] The state diversion module comparison revealed that the thermal impedance fluctuation characterization value was between the first preset impedance fluctuation threshold and the second preset impedance fluctuation threshold. It determined that the adhesive layer was in a warning state and locked the area with a temperature gradient greater than 0.09℃ / mm as the warning area.

[0048] At this time, the dense attenuation determination module extracts the contact force signal corresponding to the warning area, and after short-time Fourier transform, calculates the ratio of the energy of the preset characteristic frequency band (80Hz-150Hz) to the energy of the total frequency band (0Hz-500Hz), and obtains that the high-frequency energy ratio of the area is 0.15 (that is, the dense attenuation characterization value is 0.15).

[0049] The density detection module compares 0.15 with the preset density attenuation characterization value of 0.11. Since 0.15 is greater than or equal to the preset density attenuation characterization value of 0.11, it determines that there is density degradation in the warning area. The system further calculates the density attenuation deviation value: 0.15-0.11=0.04.

[0050] Since 0.04 is less than the first preset density attenuation deviation value of 0.10, the density detection module adopts the first adjustment coefficient of 0.95 to reduce the current coating rate from 100mm / s to 95mm / s (i.e., 100×0.95), thereby extending the open residence time of the adhesive layer surface and allowing sufficient time for the internal microbubbles to escape.

[0051] Example 3: Abnormal operating conditions (severe bubble defects, triggering pressurization or shutdown) When coating complex corner areas, large-area bubble accumulation occurred inside the adhesive layer. The thermal impedance determination module calculated the current thermal impedance fluctuation characterization value to be 2.25.

[0052] The state splitting module comparison found that 2.25 is greater than or equal to 2.0, directly determining that the adhesive layer is in an abnormal state and triggering the sound field evaluation module.

[0053] Scenario A (Repairable State): The acoustic emission characteristic determination module acquires high-frequency acoustic emission signals and identifies 20 valid acoustic emission events within 2.5 seconds, calculating the acoustic emission event frequency to be 8 times / second. The strategy determination unit compares the results and finds that 8 is less than the preset frequency threshold of 18, determining that the bubble walls have not yet ruptured significantly. The system calculates the difference as (18-8) = 10 and calculates the increase in coating pressure according to the formula: 0.05MPa × (10 / 18) ≈ 0.028MPa. The system automatically increases the coating pressure from 0.3MPa to 0.328MPa, suppressing bubble expansion by increasing the extrusion pressure.

[0054] Scenario B (Irreversible State): If the acoustic emission characteristic determination module identifies 55 valid acoustic emission events within 2.5 seconds, the calculated acoustic emission event frequency is 22 times / second. The strategy determination unit compares the results and finds that 22 is greater than or equal to the preset frequency threshold of 18, determining that irreversible microstructural damage has occurred inside the adhesive layer. The system immediately outputs a stop command.

[0055] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A smart coating system for silicone adhesives based on multi-source data fusion, characterized in that, include: The data acquisition module is used to collect multi-source data during the coating process of silicone adhesive. The multi-source data includes the current coating rate, the current coating pressure, the surface temperature of the adhesive layer, the contact force signal of the adhesive layer, and the acoustic emission signal inside the adhesive layer. A thermal impedance determination module, which is connected to the data acquisition module, is used to calculate the thermal impedance fluctuation characterization value of the adhesive layer based on the surface temperature; A state diversion module, which is connected to the thermal impedance determination module, is used to determine the internal density state of the adhesive layer based on the thermal impedance fluctuation characterization value, wherein the internal density state includes a normal state, a warning state, and an abnormal state. A density attenuation determination module, which is connected to the data acquisition module, is used to determine the density attenuation characterization value inside the adhesive layer based on the contact force signal corresponding to the warning area. A density detection module, which is connected to the state diversion module and the density attenuation determination module respectively, is used to determine whether there is density degradation in the warning area based on the density attenuation characterization value when the internal density state is a warning state. An acoustic emission determination module, which is connected to the data acquisition module, determines the frequency value of acoustic emission events based on the acoustic emission signal; The acoustic field evaluation module is connected to the acoustic emission determination module, the data acquisition module, and the state diversion module, respectively. It is used to determine the adjustment strategy of the adhesive layer based on the frequency value of the acoustic emission event when the internal density state is abnormal. The adjustment strategy is to increase the coating pressure or output a shutdown command.

2. The intelligent coating system for silicone adhesives based on multi-source data fusion according to claim 1, characterized in that, The state diversion module determines that the internal density of the adhesive layer is normal when the thermal impedance fluctuation characterization value is less than the first preset impedance fluctuation threshold. The state diversion module, in response to the thermal impedance fluctuation characterization value being greater than or equal to the second preset impedance fluctuation threshold, determines that the internal density state of the adhesive layer is abnormal and triggers the sound field evaluation module to determine the adjustment strategy of the adhesive layer based on the frequency value of the acoustic emission event; wherein, the first preset impedance fluctuation threshold is less than the second preset impedance fluctuation threshold.

3. The intelligent coating system for silicone adhesives based on multi-source data fusion according to claim 2, characterized in that, The state diversion module responds to the thermal impedance fluctuation characterization value being greater than or equal to the first preset impedance fluctuation threshold and less than the second preset impedance fluctuation threshold by determining that the internal density state of the adhesive layer is in a warning state, and triggers the density detection module to determine whether there is density degradation in the warning area based on the density attenuation characterization value inside the adhesive layer. The warning area is the area enclosed by sampling points whose absolute temperature gradient value is greater than a preset gradient threshold, and the sampling points are two-dimensional grid nodes obtained by dividing the adhesive layer surface into equal intervals.

4. The intelligent coating system for silicone adhesives based on multi-source data fusion according to claim 3, characterized in that, The thermal impedance determination module includes: The sliding calculation unit is used to set the sliding sub-window to traverse the surface temperature sequence of the adhesive layer along the time axis, calculate the ratio of the standard deviation to the mean of the temperature sequence in each sliding sub-window, and record it as the sub-window thermal resistance fluctuation value of each sliding sub-window. The instantaneous value determination unit is used to determine the maximum value among the thermal impedance fluctuation values ​​of each sub-window as the instantaneous thermal impedance fluctuation characterization value; The ratio determination unit is used to determine the ratio of the instantaneous thermal impedance fluctuation characterization value to the preset reference value as the thermal impedance fluctuation characterization value.

5. The intelligent coating system for silicone adhesives based on multi-source data fusion according to claim 4, characterized in that, The density detection module responds to the fact that the density attenuation characterization value inside the adhesive layer is greater than or equal to the preset density attenuation characterization value, determines that there is density degradation in the warning area, and reduces the coating rate according to the difference between the density attenuation characterization value and the preset density attenuation characterization value.

6. The intelligent coating system for silicone adhesives based on multi-source data fusion according to claim 5, characterized in that, The dense attenuation determination module is used to determine the dense attenuation characterization value inside the adhesive layer based on the warning area. The dense attenuation determination module obtains the energy of a preset characteristic frequency band based on the contact force signal corresponding to the warning area, calculates the ratio of the energy of the preset characteristic frequency band to the energy of the preset total frequency band, obtains the high-frequency energy ratio of each warning area, and determines the maximum value of each high-frequency energy ratio as the dense attenuation characterization value inside the adhesive layer.

7. The intelligent coating system for silicone adhesives based on multi-source data fusion according to claim 6, characterized in that, The density detection module has several rate adjustment methods for reducing the coating rate, and each rate adjustment method reduces the coating rate by a different amount.

8. The intelligent coating system for silicone adhesives based on multi-source data fusion according to claim 7, characterized in that, The sound field evaluation module includes: A frequency determination unit is used to compare the frequency value of the acoustic emission event with a preset frequency threshold. A strategy determination unit is used to determine an adjustment strategy for the adhesive layer based on the comparison results of the frequency determination unit, wherein... If the frequency value of the acoustic emission event is less than the preset frequency threshold, the adjustment strategy is determined to be to increase the coating pressure based on the difference between the preset frequency threshold and the frequency value of the acoustic emission event. If the frequency value of the acoustic emission event is greater than or equal to the preset frequency threshold, the adjustment strategy is determined to be to output a shutdown command.

9. The intelligent coating system for silicone adhesives based on multi-source data fusion according to claim 8, characterized in that, The acoustic emission determination module includes: The spectrum analysis unit is used to perform spectrum analysis on the acoustic emission signal of the adhesive layer, extract the signal components in the preset high frequency band, and obtain the high frequency acoustic emission signal. An event recognition unit is used to compare the amplitude of the high-frequency acoustic emission signal with a preset background noise threshold to identify an acoustic emission event. Specifically, when the amplitude of the high-frequency acoustic emission signal rises from less than the preset background noise threshold to greater than the preset background noise threshold, it is determined as the start of an acoustic emission event. When the amplitude of the high-frequency acoustic emission signal falls back to less than the preset background noise threshold and remains locked for a preset duration, it is determined as the end of the acoustic emission event. An event counting unit is used to count the total number of acoustic emission events that occur within a preset acquisition time. The frequency calculation unit is used to divide the total number of acoustic emission events by the preset acquisition duration to obtain the frequency value of the acoustic emission events.

10. The intelligent coating system for silicone adhesives based on multi-source data fusion according to claim 9, characterized in that, The increase in coating pressure is positively correlated with the difference between the preset frequency threshold and the frequency value of the acoustic emission event.

Images