Intelligent glue scraping device in wind power generation blade manufacturing and control method of intelligent glue scraping device

Abstract

The invention provides an intelligent glue scraping device in wind power generation blade manufacturing and a control method thereof, and the method comprises the steps: calling a preset glue scraping device structure adjustment rule base for molded line gap change parameter values, constructing the rule base based on experimental verification and engineering practice, and obtaining an effective contact area adjustment scheme of the glue scraping device through matching calculation; according to the glue scraping device effective contact area adjustment scheme, generating a corresponding control instruction, transmitting the control instruction to a glue scraping device driving module, obtaining feedback adjustment state data from the driving module, and judging whether adjustment accords with a preset precision threshold range or not; and calling a flow channel design optimization algorithm based on fluid mechanics simulation according to the flow characteristic parameter values of the glue solution, analyzing and processing the original distribution data of the glue solution, and determining the position of a specific region where an edge accumulation phenomenon possibly exists in the flow channel.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to an intelligent adhesive scraping device and its control method in the manufacturing of wind turbine blades. Background Technology

[0002] Background of the problem: Wind turbine blade manufacturing is a core component of the clean energy sector, directly impacting the balance between turbine performance and production costs. High-quality blade bonding processes not only ensure blade structural stability but also determine production efficiency and economic benefits. However, traditional bonding processes have significant shortcomings in adhesive quantity control and thickness adjustment, making it difficult to meet the dual demands of high precision and low cost for modern wind turbine blades. Existing adhesive application devices mostly employ fixed structures, unable to flexibly adjust according to changes in blade profile clearance, resulting in uneven adhesive layer thickness, significant adhesive waste, and high production costs. Furthermore, traditional devices suffer from low adhesive flow efficiency, easily accumulating within the flow channels, affecting adhesive layer uniformity and bonding quality.

[0003] Against this backdrop, the core challenges of the bonding process have become increasingly apparent. Adhesive flow efficiency has become a key factor affecting bonding quality and cost. Due to the lack of optimized guidance for high-viscosity adhesives in traditional flow channel designs, the adhesive is unevenly distributed within the channel, easily leading to edge buildup and reduced coating efficiency. This inefficient flow characteristic further results in another technical challenge: difficulty in dynamically controlling the adhesive layer thickness. Current equipment cannot adjust the effective area of ​​the scraper in real time according to the profile gap, making it difficult to achieve precise control of the adhesive layer thickness, thus affecting bonding quality and material utilization. These two factors are interconnected: low adhesive flow efficiency limits the flexibility of thickness control, while the lack of dynamic control exacerbates adhesive waste and uneven quality.

[0004] Therefore, designing a scraping device that can optimize adhesive flow efficiency and dynamically control adhesive layer thickness has become a key issue in improving the quality of blade bonding process and reducing production costs. Summary of the Invention

[0005] This invention provides an intelligent adhesive scraping device and its control method for wind turbine blade manufacturing, mainly comprising: By collecting data on the changes in blade profile clearance in real time, and comparing and analyzing the data with a profile model database pre-built based on historical records and simulation results, the specific distribution characteristics of the changes in profile clearance, including the trend of changes in clearance width and position, are obtained, and the corresponding clearance change parameter values ​​are determined. For the parameter values ​​of profile gap variation, the preset rule library for adjusting the structure of the scraping device is called. This rule library is built based on experimental verification and engineering practice. The effective contact area adjustment scheme of the scraping device is obtained through matching calculation. Based on the effective contact area adjustment scheme of the scraping device, the corresponding control command is generated and transmitted to the scraping device drive module. The adjustment status data is obtained from the drive module to determine whether the adjustment meets the preset accuracy threshold range. If the adjusted state data meets the preset accuracy threshold range, the flow efficiency of the adhesive is monitored in real time. The original data of adhesive distribution is obtained from the flow channel sensor, and the flow characteristic parameters of the adhesive in the flow channel are analyzed, including the flow velocity distribution and pressure change. For the flow characteristic parameters of the adhesive, a flow channel design optimization algorithm based on fluid dynamics simulation is called to analyze and process the raw data of adhesive distribution and determine the specific location of the edge accumulation phenomenon that may exist in the flow channel. Based on the specific location of the edge accumulation phenomenon, a local adjustment command for the flow channel is generated and transmitted to the flow channel control unit. The adjusted adhesive distribution update data is obtained from the control unit to determine whether the preset flow uniformity threshold standard has been met. If the adjusted adhesive distribution update data reaches the preset flow uniformity threshold standard, then the target value of adhesive layer thickness control is calculated by combining the profile gap change parameter value and the adhesive flow characteristic parameter value through the thickness prediction model, and the dynamically adjustable thickness parameter value is determined. For the dynamically adjusted thickness parameter value, a real-time operation command for the glue scraping device is generated and transmitted to the execution unit. The execution unit obtains real-time monitoring data of the glue layer thickness and determines whether it is consistent with the target value and whether the error is within the preset range. If the real-time monitoring data of the adhesive layer thickness is consistent with the target value and the error is within the preset range, the data on the amount of adhesive used will be continuously recorded. The data will be analyzed and processed in conjunction with the material utilization efficiency evaluation model based on the ratio of usage to output to determine the solution for improving the efficiency of adhesive use, which will be used for subsequent process optimization and adjustment.

[0006] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects: This invention discloses an intelligent adhesive scraping device and its control method for wind turbine blade manufacturing, proposing an innovative solution to address the shortcomings of traditional bonding processes in adhesive quantity control and thickness adjustment. This invention acquires real-time data on blade profile gap changes and uses a pre-built model database for comparative analysis to obtain the distribution characteristics of gap changes. Based on these characteristics, this invention calls upon a preset rule base to calculate and determine the effective contact area adjustment scheme of the adhesive scraping device, achieving precise control of the adhesive layer thickness. Simultaneously, this invention introduces a flow channel design optimization algorithm, which analyzes adhesive flow characteristic parameters to identify and eliminate edge accumulation phenomena, improving the uniformity of adhesive distribution. Furthermore, this invention combines profile gap change parameters and flow characteristic parameters to dynamically calculate the target value for adhesive layer thickness control, and adjusts it through real-time operation commands to ensure that the adhesive layer thickness matches the target value. This intelligent dynamic control method not only significantly improves the quality of the bonding process but also effectively reduces production costs by optimizing material utilization efficiency, providing an efficient and precise bonding solution for the wind turbine blade manufacturing field. Attached Figure Description

[0007] Figure 1 This is a flowchart of an intelligent adhesive scraping device and its control method in the manufacturing of wind turbine blades according to the present invention.

[0008] Figure 2 This is a schematic diagram of an intelligent adhesive scraping device and its control method in the manufacturing of wind turbine blades according to the present invention.

[0009] Figure 3 This is another schematic diagram of an intelligent adhesive scraping device and its control method in the manufacturing of wind turbine blades according to the present invention. Detailed Implementation

[0010] The technical solutions of the present invention will be clearly and thoroughly described below with reference to the accompanying drawings. The described embodiments are merely some embodiments of the present invention.

[0011] like Figure 1-3 This embodiment of an intelligent adhesive scraping device and its control method in the manufacturing of wind turbine blades may specifically include: S101. By collecting data on the changes in blade profile clearance in real time, and comparing and analyzing the data with a profile model database pre-built based on historical records and simulation results, the specific distribution characteristics of the changes in profile clearance are obtained, including the trend of changes in clearance width and position, and the corresponding clearance change parameter values ​​are determined.

[0012] Real-time acquisition of blade profile clearance variation data is performed using a sensor system to obtain raw clearance variation signal data, which is stored in a temporary data buffer to obtain preliminary profile clearance variation records. Based on the acquired clearance variation records, corresponding historical records and simulation results are extracted from a pre-built profile model database for preliminary data matching to determine the reference model closest to the current variation data. A support vector machine algorithm is used to compare and analyze the matched reference model with the real-time acquired clearance variation records, extracting specific features of clearance width and position changes to obtain a detailed distribution feature description. For the extracted distribution feature description, the trend characteristics of clearance width and position changes are analyzed. If the trend characteristics deviate from a preset threshold range, anomalies are marked to identify potential clearance variation anomaly areas. Further data layering processing is performed on the marked anomaly areas to obtain the specific variation parameter values ​​corresponding to the anomaly areas, determining the mapping relationship between parameter values ​​and distribution features. Based on the determined mapping relationship, the blade clearance variation trend is dynamically monitored. If the parameter values ​​continuously deviate from the preset range during monitoring, a data update mechanism is triggered to obtain the latest blade clearance variation prediction results. By continuously comparing the predicted results with the real-time collected data, the reference values ​​in the profile model database are adjusted to obtain optimized model parameters, thus completing the accurate assessment of profile gap changes.

[0013] Specifically, in the process of realizing real-time monitoring and analysis of blade profile gap changes, high-precision sensors are first used to collect blade profile gap data in real time. Assuming the collection frequency is 100 times per second, the collected gap data ranges from 0.1mm to 2.5mm. The data is stored in the cloud database in the form of timestamps and position coordinates, with the format being (time, position X, position Y, gap value). Next, the collected data was compared with a pre-built profile model database. This database, built based on historical operating data and simulation results, contains 1000 profile models. Each model covers the distribution characteristics and corresponding parameters of gap widths from 0.05mm to 3.0mm. The matching degree between real-time data and database models was calculated using the Euclidean distance algorithm, with the formula D=√[(x1-x2)^2+(y1-y2)^2], where x and y represent position coordinates. The model with the highest matching degree was selected as the reference model. Assuming the matching result shows that the current gap data is closest to model number M-256, its characteristics are a mean gap width of 1.2mm and a standard deviation of 0.15mm. Subsequently, the specific distribution characteristics of gap changes were analyzed, and the trends of gap width and position changes were extracted. A trend line of gap change with position was fitted using a linear regression algorithm, yielding a slope k=0.02, indicating that the gap width gradually increases with position X. Combined with position data analysis, it was found that gap changes are mainly concentrated in the middle region of the blade, with a position range of X=50mm to X=80mm. Finally, based on the comparison results and trend analysis, the clearance change parameter values ​​were determined, and the average clearance change rate was calculated to be 0.01 mm per hour, with a maximum clearance deviation of 0.3 mm. These parameter values ​​were then automatically updated into the system prediction model for subsequent operation optimization and fault early warning, forming a complete closed-loop logic from data acquisition to parameter determination, ensuring that the system can respond to changes in blade status in real time.

[0014] S102. For the parameter value of the profile gap change, call the preset rule library of the scraper device structure adjustment. This rule library is built based on experimental verification and engineering practice. The effective contact area adjustment scheme of the scraper device is obtained through matching calculation.

[0015] By obtaining varying parameters from profile gap data, the collected raw data is cleaned and formatted using preprocessing methods to obtain a preliminary parameter set. Based on this preliminary parameter set, a pre-defined rule base is used for parameter analysis, extracting features related to the correlation between varying parameters and profile gaps to identify key influencing factors. If key influencing factors exceed a pre-defined threshold, a matching calculation method is used to retrieve corresponding structural adjustment strategies from the rule base to obtain a preliminary adjustment direction for the scraping device. Based on this preliminary adjustment direction and the requirements for effective contact, finite element analysis is used to simulate and calculate the contact area, obtaining optimized area adjustment data. Using this optimized area adjustment data, a specific structural adjustment scheme is generated, and the geometric parameters of the scraping device are precisely calibrated to determine the final adjustment scheme. If the final adjustment scheme deviates from the experimental verification data, a secondary correction is performed using engineering practice records in the rule base to obtain a corrected adjustment scheme result. Based on the corrected adjustment scheme result, a digital instruction file is generated and transmitted to the scraping device's control system to complete the automated execution of the structural adjustment.

[0016] S103. Based on the effective contact area adjustment scheme of the scraping device, generate corresponding control commands and transmit them to the scraping device drive module. Obtain the feedback adjustment status data from the drive module and determine whether the adjustment meets the preset accuracy threshold range.

[0017] By using real-time monitoring data of the contact area, an initial adjustment requirement analysis is generated. A pre-established mapping model is used to determine the corresponding control command content. Based on the generated control commands, the command data is sent to the drive module using a transmission protocol. During transmission, the transmission status is recorded, and a confirmation signal indicating completion is obtained. Feedback adjustment status data is obtained from the drive module and formatted to obtain structured status information. If deviation data exists in the structured status information, it is compared with a preset accuracy threshold range to determine if the deviation exceeds the allowable range. If the deviation exceeds the allowable range, a new adjustment requirement is generated based on the deviation data, and the control command is recalculated using a regression analysis model and transmitted to the drive module. The adjusted status data is obtained again from the drive module, and a second comparison is performed to determine if the adjustment result meets the preset accuracy threshold range. If the adjustment result still does not meet the preset accuracy threshold range, a loop processing mechanism is used to repeatedly generate and transmit new control commands until the required status data is obtained.

[0018] Specifically, regarding the adjustment scheme for the effective contact area of ​​the scraping device, the system first generates specific control commands. For example, setting the target contact area to 50 square centimeters, and based on the current area of ​​40 square centimeters, it calculates the required adjustment of 10 square centimeters. Using a proportional control algorithm, the adjustment command is to increase the drive motor speed to 1200 revolutions per minute for 2 seconds. This command is transmitted to the scraping device drive module via the Modbus protocol through the communication interface. Next, the drive module receives the command, executes the adjustment, and provides real-time status data. For example, the sensor collects the adjusted contact area as 49.8 square centimeters, and simultaneously returns a motor speed feedback value of 1198 revolutions per minute. This data is uploaded to the control center via the system interface at a frequency of 10 times per second. Subsequently, the system analyzes the feedback data. The preset accuracy threshold range is ±0.5 square centimeters of the target area, i.e., 49.5 to 50.5 square centimeters. The deviation between the current area of ​​49.8 and the target of 50 is calculated to be 0.2 square centimeters, which meets the threshold range, thus the adjustment is deemed successful. If the deviation exceeds the threshold, for example, if the feedback area is 49.2 square centimeters and the deviation is 0.8 square centimeters, the system automatically triggers a secondary adjustment algorithm, calculates a new speed increment of 200 revolutions per minute, lasts for 1 second, and regenerates and issues the command, forming a closed-loop control logic. To ensure business continuity, the system is also linked to the pressure monitoring module of the scraping device. If the adjusted pressure value exceeds the safe range (e.g., exceeding 5 MPa), the adjustment is paused and logged. Once the pressure recovers to below 4.5 MPa, the adjustment process is automatically restarted to ensure stable equipment operation. Through the above methods, data-driven and automated processing at each stage forms a complete technical chain from command generation to status verification.

[0019] S104. If the adjusted state data meets the preset accuracy threshold range, the flow efficiency of the adhesive is monitored in real time, the original data of adhesive distribution is obtained from the flow channel sensor, and the flow characteristic parameters of the adhesive in the flow channel are analyzed, including the flow velocity distribution and pressure change.

[0020] The flow efficiency of the adhesive is monitored in real time using flow channel sensors, collecting raw data on the adhesive distribution to obtain an initial flow information dataset. Based on this initial dataset, data preprocessing methods are used to remove noise and outliers, resulting in a clean dataset. If data points in the clean dataset deviate from the preset normal range, anomaly verification is performed by comparing with historical data to determine if any flow anomalies exist. The results of the anomaly verification are obtained, and a support vector machine algorithm is used to extract features from the velocity distribution and pressure changes under abnormal conditions, obtaining key parameters of the flow characteristics. Based on these key parameters, the uniformity of the adhesive distribution within the flow channel is analyzed to determine the specific area and degree of distribution deviation. By analyzing the specific area and degree of distribution deviation, combined with historical flow characteristic data, potential flow blockage points are predicted, yielding prediction results. If the prediction results indicate a risk of blockage, an automatic adjustment command for the flow channel parameters is triggered, optimizing the adhesive flow efficiency.

[0021] Specifically, after the status data is adjusted to meet the preset accuracy threshold range, the system will automatically trigger a real-time monitoring mechanism for the flow efficiency of the adhesive. Assuming the preset accuracy threshold is a deviation range of ±0.05mm, when the current status data deviation is detected to be within 0.03mm, the system will collect raw data of adhesive distribution 1000 times per second through the flow channel sensor. These data include the pressure value (unit Pa) and flow rate value (unit m / s) at each point in the flow channel. For example, if the pressure at a certain point is 500Pa and the flow rate is 0.2m / s. Next, the system uses fluid dynamics analysis algorithms to process the raw data and employs the Navier-Stokes equations for numerical simulation to calculate the velocity distribution and pressure change characteristics of the adhesive in the flow channel. In the specific calculation, it is assumed that the viscosity of the adhesive is 0.01 Pa·s and the density is 1000 kg / m³. Using the finite element analysis method, the flow channel is divided into 10,000 grid elements, and the velocity distribution map is obtained by iterative solution. It shows that the highest velocity in the center of the flow channel is 0.25 m / s, and the lowest velocity at the edge is 0.05 m / s. At the same time, the pressure change curve shows that the inlet pressure is 600 Pa and the outlet pressure drops to 400 Pa. Furthermore, the system compares the analysis results with a standard flow characteristic database. The database specifies a standard flow velocity range of 0.1-0.3 m / s and a pressure drop range of 150-250 Pa. If the current pressure drop is detected to be 200 Pa, which meets the standard, the system automatically records it as normal. If it exceeds the range, an early warning mechanism is triggered and linked to the adhesive ratio adjustment module. The system automatically optimizes the ratio parameters to ensure flow efficiency. For example, if the ratio is adjusted from 1:1 to 1:1.2, the algorithm simulates and predicts that the flow velocity can be increased to 0.22 m / s and the pressure drop can be stabilized at 180 Pa after the adjustment. This forms a closed-loop control logic to ensure the stability of adhesive flow during the production process.

[0022] S105. For the flow characteristic parameter values ​​of the adhesive, call the flow channel design optimization algorithm based on fluid dynamics simulation to analyze and process the original data of adhesive distribution, and determine the specific location of the edge accumulation phenomenon that may exist in the flow channel.

[0023] Using fluid dynamics simulation technology, the characteristic parameters of the adhesive flow are modeled to obtain an initial flow distribution data model. An optimization algorithm is used to iteratively calculate this initial flow distribution data model, identifying potential areas in the flow channel design that may lead to edge buildup. Based on the distribution data within these potential areas, finite element analysis is used to calculate the local pressure and velocity variation trends of the adhesive flow within the flow channel. If the local pressure and velocity variation trends exceed preset thresholds, the geometric parameters of the flow channel design are adjusted to obtain optimized flow channel structure data. The fluid dynamics simulation is then rerun using the optimized flow channel structure data to obtain updated adhesive flow distribution data. Based on the updated adhesive flow distribution data, the existence of edge buildup is analyzed to determine the final potential area location. Based on the final potential area location, further parameter fine-tuning is performed on local areas of the flow channel design to obtain the final optimized flow channel scheme.

[0024] Specifically, regarding the analysis of adhesive flow characteristic parameters and optimization of flow channel design, the flow characteristic data of the adhesive, such as viscosity of 0.5 Pa·s, density of 1.2 g / cm³, and flow velocity of 0.1 m / s, were first collected. Combining this with fundamental fluid dynamics equations such as the Navier-Stokes equations, numerical simulations were performed using computational fluid dynamics software to generate velocity and pressure field distribution data within the flow channel. The maximum velocity value (0.15 m / s) was observed at the center of the flow channel, while the minimum value (0.02 m / s) was observed at the edge, initially indicating potential flow stagnation in the edge region. Next, a flow channel design optimization algorithm based on a genetic algorithm was applied, adjusting the flow channel width from the initial 5 mm to an optimized 6.2 mm and the radius of curvature from 10 mm to 12.5 mm. Through 100 iterative calculations, the uniformity of the flow velocity distribution within the flow channel was improved by 15%, and the pressure loss was reduced by 8%. The optimized data was stored in the system database. Subsequently, the raw data on adhesive distribution were analyzed and processed. Grid cells with flow velocities below 0.03 m / s in the flow channel edge region were extracted. Using a particle tracking algorithm, the average residence time of the adhesive in the edge region was calculated to be 3.5 seconds, higher than the 1.2 seconds in the center region. This determined that edge accumulation was mainly concentrated on the left wall region 2 cm to 5 cm after the flow channel inlet, with an estimated accumulation thickness of 0.3 mm. Finally, a 3D visualization chart automatically generated by the system was used to mark the specific coordinate range of the accumulation area as (x: 2-5 cm, y: 0-1 cm, z: 0.5 cm). Comparison of the analysis results with historical data revealed a 10% reduction in accumulation in similar flow channel designs, validating the effectiveness of the optimization algorithm. The relevant data was automatically uploaded to the cloud backup system, forming a complete closed-loop logic from data acquisition to result verification.

[0025] S106. Based on the specific location of the edge accumulation phenomenon, generate a local adjustment command for the flow channel and transmit it to the flow channel control unit. Obtain the adjusted adhesive distribution update data from the control unit and determine whether the preset flow uniformity threshold standard has been reached.

[0026] By analyzing the location information of the edge accumulation phenomenon, corresponding local flow channel adjustment commands are generated. A pre-established mapping model is used to determine the specific parameter values ​​for flow channel adjustment. The generated local flow channel adjustment commands are transmitted to the control unit, which provides feedback on the command execution status to determine if the command was successfully issued. If the command transmission fails, the adjustment command is regenerated. Adjusted adhesive distribution data is obtained from the control unit and preliminarily processed to obtain the characteristic value distribution of the adhesive distribution. Based on the characteristic value distribution of the adhesive distribution, a support vector machine algorithm is used to classify and evaluate the flow uniformity to determine if it meets a preset threshold standard. If the flow uniformity does not meet the preset threshold standard, new flow channel adjustment parameters are generated based on the classification evaluation results and transmitted to the control unit for secondary adjustment. The latest characteristic value distribution is obtained from the adhesive distribution data after the secondary adjustment to determine if it meets the flow uniformity requirements. If it still does not meet the requirements, the current adjustment data is recorded, and the next iteration of adjustment begins. Based on the adhesive distribution data after multiple iterations of adjustment, the final flow uniformity judgment result is determined, and an optimized record of the adjustment parameters is generated for subsequent flow channel control reference.

[0027] Specifically, to address edge buildup, the system first monitors the adhesive distribution within the flow channel in real time using a sensor array. Assuming a flow channel width of 50 cm, if the adhesive thickness detected at the edge (within 5 cm of the edge) is 3.2 mm, while the thickness at the center is only 1.8 mm, exceeding a preset threshold of 0.5 mm, the system automatically identifies this as edge buildup. Subsequently, based on the monitoring data, the system generates local adjustment commands and calculates adjustment parameters using a fluid dynamics model. For example, by adjusting the valve opening at the edge to 30% and at the center to 70%, the flow rate is balanced. The calculation formula is Q = kv * A, where Q is the flow rate, kv is the valve coefficient, and A is the opening ratio. The goal is to reduce the edge flow rate to 80% of the center flow rate. After the command is generated, it is transmitted to the flow channel control unit via industrial Ethernet. Upon receiving the command, the control unit executes the valve adjustment and provides real-time feedback on the adjusted adhesive distribution data. For example, after adjustment, the edge thickness decreases to 2.1 mm, the center thickness increases to 2.0 mm, and the difference is reduced to 0.1 mm. The system further analyzes the flow uniformity, with a preset uniformity threshold of a thickness difference of less than 0.2 mm. The current difference of 0.1 mm has met the standard, and the system records the adjusted parameters and updates the database. If the standard is not met, the system proceeds to the next iteration, optimizing the valve opening ratio based on historical data to ensure each adjustment gradually approaches the target value. This process forms a closed-loop control logic from monitoring, calculation, adjustment to feedback. It is also linked to the production batch management module to record the impact of each adjustment on product quality. For example, after improving the uniformity of the adhesive, the product defect rate decreased from 5% to 2%, providing data support for subsequent optimization.

[0028] S107. If the adjusted adhesive distribution update data reaches the preset flow uniformity threshold standard, then the target value of adhesive layer thickness control is calculated by combining the profile gap change parameter value and the adhesive flow characteristic parameter value through the thickness prediction model, and the dynamically adjustable thickness parameter value is determined.

[0029] By acquiring adhesive distribution data from sensors, it is determined whether the distribution meets a preset flow uniformity threshold, thus obtaining a preliminary uniformity assessment result. If the preliminary assessment result shows that the adhesive distribution meets the preset threshold, the profile gap variation parameters and adhesive flow characteristic parameters are retrieved from the database to determine the basic data combination for subsequent calculations. Based on the retrieved basic data combination, a pre-established thickness prediction model is used to calculate the target control value of the adhesive layer thickness. The calculated target control value of the adhesive layer thickness is compared with historical control records to determine if there is a deviation. If the deviation exceeds a preset range, a dynamic thickness parameter adjustment scheme is generated. Using the generated thickness parameter adjustment scheme, combined with real-time monitoring of adhesive distribution trends, the latest flow uniformity data is obtained to determine the applicability of the adjustment scheme. If the applicability of the adjustment scheme is confirmed, the parameters of the thickness prediction model are updated based on real-time monitoring data to obtain an optimized thickness control strategy. Based on the optimized thickness control strategy, the profile gap parameters and adhesive flow parameters are dynamically adjusted to determine the final adhesive layer thickness control result.

[0030] Specifically, after adjusting the adhesive distribution, the system first collects adhesive distribution data using sensors. Assuming the collected distribution uniformity data is 0.85, and the preset flow uniformity threshold standard is 0.80, the system automatically determines that the current distribution data has reached the standard and proceeds to the next calculation step. Next, the system acquires the profile gap variation parameter value, assuming the measured value is 0.5 mm, and simultaneously extracts the adhesive flow characteristic parameter values, such as a viscosity of 300 centipoise and a flow resistance coefficient of 0.12. Database comparison confirms that these parameters are within the normal range. Then, the system inputs the above parameters into the thickness prediction model, which is based on a linear regression algorithm. The formula is: Thickness = Base Thickness + Gap Variation Value × 0.3 + Viscosity × 0.001 - Resistance Coefficient × 0.2. Substituting the data, the calculated thickness is 2.0 + 0.5 × 0.3 + 300 × 0.001 - 0.12 × 0.2 = 2.426 mm, which is used as the target value for adhesive layer thickness control. Furthermore, based on the difference of 0.126 mm between the target value and the current actual thickness of 2.3 mm, combined with historical control data analysis, the system determines the dynamically controlled thickness parameter value to be 0.13 mm, with the error controlled within 5%. Finally, the system transmits this parameter value to the execution module, automatically adjusting the equipment parameters to ensure stable adhesive layer thickness. To form a logical chain, the system also links the adhesive temperature data, assuming a temperature of 25 degrees Celsius, analyzing its slight impact on viscosity (viscosity decreases by 0.5 centipoise for every 1 degree Celsius increase), thereby correcting the model input and ensuring prediction accuracy. All of the above processes are completed automatically by the system, with data updated in real time to ensure the control effect.

[0031] S108. For the dynamically adjusted thickness parameter value, generate a real-time operation command for the adhesive scraping device and transmit it to the execution unit. Obtain real-time monitoring data of the adhesive layer thickness from the execution unit and determine whether it is consistent with the target value and whether the error is within the preset range.

[0032] To address the dynamic control requirements of thickness parameters, a pre-established mapping model generates corresponding operation commands, which are transmitted to the execution unit of the adhesive scraping device to obtain initial control results. Real-time monitored adhesive layer thickness data is acquired from the execution unit, and the raw data is filtered using a data processing module to determine the processed adhesive layer thickness value. Based on the comparison between the processed adhesive layer thickness value and the target value, if the detected value deviates from the target value and exceeds the preset error range, an adjusted operation command is generated and transmitted to the execution unit to obtain new control feedback. The feedback data returned by the execution unit continuously monitors the trend of adhesive layer thickness changes to determine if the trend is stabilizing and approaching the target value. If the trend has not reached a stable state, a regression analysis model is used to predict and adjust the thickness parameters, resulting in a new dynamic control scheme. For the predicted and adjusted dynamic control scheme, updated operation commands are generated and transmitted to the execution unit to obtain the latest adhesive layer thickness monitoring data. The latest adhesive layer thickness monitoring data is compared again with the target value to determine if the preset error range requirement is met, completing the dynamic control process of the thickness parameters.

[0033] Specifically, regarding the implementation of dynamically adjusted thickness parameters, the system first presets a target adhesive layer thickness of 5.0 mm and inputs this value as a reference to the control module. The control module calculates the initial operating parameters of the scraping device based on its built-in algorithm, such as setting the scraping speed to 0.5 m / s and adjusting the scraper angle to 30 degrees. The algorithm uses proportional-integral-derivative (PID) control, where the proportional coefficient Kp is set to 2.0, the integral coefficient Ki to 0.5, and the derivative coefficient Kd to 0.1. The parameters are dynamically adjusted by calculating the deviation value in real time to ensure accurate output commands. Subsequently, the generated real-time operating commands are transmitted to the execution unit of the scraping device via industrial Ethernet. The commands are sent in digital signal form, containing specific values ​​for speed and angle, ensuring that the execution unit can respond immediately and adjust the scraping action. Next, real-time monitoring data of the adhesive layer thickness is obtained from the execution unit. Using a laser thickness sensor installed on the scraping device, thickness data is collected 10 times per second. For example, if a collected value is 5.05 mm, the system transmits this data back to the control module for processing via a data acquisition card. Finally, the control module compares the monitored data with the target value of 5.0 mm, calculating an error of 0.05 mm. The preset error range is ±0.1 mm. The analysis results show that the error is within the allowable range, and the system determines that the current thickness meets the requirements, requiring no further parameter adjustments. If the error exceeds the range, for example, if the monitored value is 5.15 mm and the error is 0.15 mm, the PID algorithm is triggered to recalculate the adhesive scraping speed and angle, generating new instructions for regulation, forming a closed-loop control logic. To ensure system stability, an additional temperature compensation mechanism is introduced. If the ambient temperature exceeds 30 degrees Celsius, the system will correct according to the temperature's influence coefficient on the adhesive layer thickness of 0.01 mm / degree Celsius, ensuring the accuracy of the monitored data and maintaining the rigor and continuity of the entire regulation process.

[0034] S109. If the real-time monitoring data of the adhesive layer thickness is consistent with the target value and the error is within the preset range, the data on the amount of adhesive used will be continuously recorded. The data will be analyzed and processed in conjunction with the material utilization efficiency evaluation model based on the ratio of usage to output to determine the solution for improving the efficiency of adhesive use, which will be used for subsequent process optimization and adjustment.

[0035] Dynamic data on the adhesive layer thickness is acquired through a real-time monitoring system. The collected data is compared with a preset target value to determine if it falls within the error range, thus establishing the adhesive layer thickness compliance status. If the compliance status is consistent, adhesive usage data is continuously collected from the production equipment. Combined with output ratio information from production records, the original dataset of adhesive usage is determined. Based on this raw dataset, a pre-established material utilization efficiency assessment model is used to analyze the matching degree between usage and output ratio, yielding an evaluation result of utilization efficiency. Using the utilization efficiency evaluation result, combined with historical process adjustment data, key inefficiencies are analyzed to identify the main factors affecting adhesive usage efficiency. Based on these main factors, a logistic regression model is used to predict and analyze process parameters, identifying optimization directions for parameter adjustments and assessing their feasibility. If the feasibility of parameter adjustments is confirmed, a specific process adjustment plan is generated and verified using real-time monitoring data to obtain the adjusted adhesive usage trend. Based on the adjusted adhesive usage trend, the material utilization efficiency assessment model is continuously updated to optimize the data acquisition and analysis logic in subsequent production processes, determining the final process optimization path.

[0036] Specifically, during real-time monitoring of the adhesive layer thickness, the system uses a high-precision laser sensor to collect adhesive layer thickness data once per second. Assuming a target thickness of 2.00mm and a preset error range of ±0.05mm, if the real-time data is 2.03mm, the system determines that it is within the error range and automatically triggers the data recording module to continuously store the amount of adhesive used, for example, recording an hourly adhesive consumption of 5.2kg. Simultaneously, the system imports this data into a material utilization efficiency evaluation model based on the ratio of usage to output. The specific algorithm is: Material Utilization Efficiency = (Weight of Qualified Product / Total Adhesive Usage) × 100%. Assuming a batch of qualified products weighs 4.8kg and the total adhesive usage is 5.2kg, the calculated efficiency is 92.31%. Subsequently, the system analyzes the efficiency data and compares it with historical data. It finds that the current efficiency is below the average of 95%, likely due to uneven adhesive distribution leading to waste. The system automatically generates an optimization plan, suggesting adjusting the spraying equipment parameters and increasing the spraying pressure from 3.5 bar to 3.8 bar to improve adhesive adhesion uniformity. The system predicts that the efficiency will increase to 94% after the adjustment. This plan is stored in the process optimization database and linked to subsequent production batches. The system automatically adjusts equipment parameters and monitors the effect. If the efficiency does not meet expectations, it further analyzes environmental variables such as temperature and humidity (e.g., 25°C, 60% humidity) to determine whether the adhesive formulation ratio needs adjustment, ensuring continuous process optimization and forming a closed-loop control logic.

[0037] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A control method for an intelligent adhesive scraping device in the manufacturing of wind turbine blades, characterized in that, Includes the following steps: (1) Collect real-time data on the change of blade profile gap, compare and analyze it with the pre-built profile model database, obtain the trend of gap width and position change, and determine the gap change parameter value; (2) Based on the gap change parameter value, call the preset scraper device structure adjustment rule library, and obtain the effective contact area adjustment scheme of the scraper device through matching calculation; (3) Generate control instructions according to the effective contact area adjustment scheme and transmit them to the glue scraping device drive module, obtain the feedback adjustment status data, and determine whether the preset accuracy threshold range is met. (4) If the adjusted state data meets the accuracy threshold range, the flow efficiency of the adhesive is monitored in real time, the original data of the adhesive distribution is obtained, and the flow characteristic parameter values ​​of the adhesive, including the flow velocity distribution and pressure change, are analyzed. (5) Based on the flow characteristic parameter values ​​of the adhesive liquid, call the flow channel design optimization algorithm based on fluid dynamics simulation, analyze the original data of adhesive liquid distribution, and determine the specific location of the edge accumulation phenomenon in the flow channel; (6) Based on the specific location of the edge accumulation phenomenon, generate a local adjustment command for the flow channel and transmit it to the flow channel control unit, obtain the adjusted adhesive distribution update data, and determine whether the preset flow uniformity threshold standard has been reached. (7) If the updated data of adhesive distribution reaches the flow uniformity threshold standard, the target value of adhesive layer thickness control is calculated by combining the gap change parameter value and the adhesive flow characteristic parameter value through the thickness prediction model, and the dynamically controlled thickness parameter value is determined. (8) Based on the dynamically adjusted thickness parameter value, generate real-time operation instructions for the glue scraping device and transmit them to the execution unit, obtain real-time monitoring data of the glue layer thickness, and determine whether it is consistent with the target value and whether the error is within the preset range. (9) If the real-time monitoring data of the adhesive layer thickness is consistent with the target value and the error is within the preset range, the data of adhesive usage will be continuously recorded, and the material utilization efficiency evaluation model will be combined to analyze and process the data to determine the solution for improving the adhesive utilization efficiency.

2. The control method according to claim 1, characterized in that, Step (1) specifically includes: The sensor system collects real-time data on changes in blade profile clearance and stores it in a temporary data cache. Historical data and simulation results are extracted from the profile model database, and data matching is performed to determine the reference model that is closest to the current data. The support vector machine algorithm is used to compare the reference model with real-time acquired data to extract the gap width and position change features; Analyze the characteristics of gap change trends, mark abnormal areas, and determine specific change parameter values; Based on the mapping relationship between parameter values ​​and distribution characteristics, the changing trend of profile gaps is dynamically monitored, triggering a data update mechanism to optimize model parameters.

3. The control method according to claim 1, characterized in that, Step (2) specifically includes: The collected raw data is cleaned and formatted to obtain a preliminary set of parameters; The system calls a pre-defined rule base to perform parameter analysis and extract key influencing factors. If the key influencing factors exceed the preset threshold range, the structure adjustment strategy will be retrieved from the rule base through matching calculation; By combining the finite element analysis method to simulate the contact area, optimized area adjustment data is generated. Based on the deviation between the adjustment plan and the experimental verification data, a second correction is performed through engineering practice records, and a digital instruction file is generated and transmitted to the glue scraping device control system.

4. The control method according to claim 1, characterized in that, Step (3) specifically includes: Based on real-time monitoring data of the contact area, an adjustment requirement analysis is generated to determine the content of the control instructions. The instructions are sent to the driver module via the transmission protocol, the transmission status is recorded, and an acknowledgment signal is obtained. Receive adjustment status data from the driver module and perform formatting processing; If the status data deviates from the allowable range, the control command is recalculated and transmitted until the adjustment result meets the accuracy threshold range.

5. The control method according to claim 1, characterized in that, Step (5) specifically includes: A data model of adhesive flow distribution was established using fluid dynamics simulation technology; An optimization algorithm is used for iterative calculations to identify potential areas within the flow channel that may lead to edge buildup. The finite element analysis method is used to calculate the local pressure and velocity variation trends. If the trend exceeds the preset threshold range, adjust the flow channel geometry parameters and re-simulate to determine the final location of the edge accumulation area.

6. The control method according to claim 1, characterized in that, Step (7) specifically includes: Data on adhesive distribution is obtained from sensors to determine whether the flow uniformity threshold standard has been met. The gap variation parameters and adhesive flow characteristic parameters were extracted from the database and used as the basic data for the thickness prediction model. The target control value of the adhesive layer thickness is calculated using a thickness prediction model; By comparing historical control records, a dynamic control scheme for adjusting thickness parameters is generated; The model parameters are updated based on real-time monitoring data to determine the final result of the adhesive layer thickness control.

7. The control method according to claim 1, characterized in that, Step (9) specifically includes: Collect dynamic data on the adhesive layer thickness and compare it with the target value to determine whether the error is within the preset range; Continuously collect data on rubber usage and output ratio information from production records to form a raw dataset; The matching degree between usage and output ratio is analyzed through a material utilization efficiency assessment model; Based on historical process adjustment data, the main factors affecting the efficiency of rubber compound use were identified; Logistic regression models are used to predict the direction of process parameter optimization, generate specific adjustment schemes, and verify them.

8. An intelligent adhesive scraping device for manufacturing wind turbine blades, characterized in that, include: The sensor system is used to collect real-time data on blade profile gap changes and adhesive flow efficiency. The data processing module is used to execute the control method according to any one of claims 1-7; The drive module is used to receive control commands and adjust the effective contact area of ​​the scraping device; The flow channel control unit is used to receive local flow channel adjustment commands and optimize adhesive distribution; The execution unit is used to receive real-time operation commands and adjust the adhesive layer thickness. The monitoring system is used to provide real-time feedback on adhesive layer thickness and adhesive usage.

9. The intelligent glue-scraping device according to claim 8, characterized in that, The data processing module includes: A profile model database stores historical records and simulation results. A rule library for adjusting the structure of the glue scraping device, built based on experimental verification and engineering practice; The fluid dynamics simulation module is used to optimize flow channel design. A thickness prediction model is used to calculate the target value for adhesive layer thickness control. A material utilization efficiency assessment model is used to analyze the efficiency of rubber compound utilization.

10. The intelligent glue-scraping device according to claim 8, characterized in that, The sensor system includes: High-precision laser sensors are used to monitor adhesive layer thickness; Flow channel sensor, used to collect raw data on adhesive distribution; Pressure sensor used to detect changes in adhesive pressure within the flow channel.

Images