Rail conveying and drying production system for remotely identifying hub baking varnish production line

Abstract

The invention belongs to the technical field of hub stoving varnish production lines, and particularly relates to a production system for remotely identifying hub stoving varnish production line track conveying and drying, which comprises the following steps of: 1, production binding: feeding hubs needing to be dried onto a carrier at a hub feeding station, taking a carrier electronic identification module on the carrier as a carrier electronic unique identification, and taking the carrier electronic identification module as a carrier electronic unique identification module; meanwhile, the hub model bar code scanning module scans the bar code of the hub to identify the hub model; the hub information is bound with the carrier electronic identification module, and unique identity association of the hub is established; and the binding information is transmitted to the data storage and association module for persistent storage through the data transmission module, and the data storage and association module is provided with a remote access interface to support a remote terminal to query hub binding data and production progress. According to the production system for remotely identifying the conveying and drying of the hub baking varnish production line track, the identity binding of hubs can be precise by setting the production system for remotely identifying the conveying and drying of the hub baking varnish production line track.

Description

Technical Field

[0001] This invention relates to the field of wheel hub painting production line technology, and in particular to a production system for remotely identifying the track conveying and drying of wheel hub painting production lines. Background Technology

[0002] A wheel hub painting production line is an automated or semi-automated industrial system specifically designed for coating the surface of automobile wheels. It consists of a series of devices arranged in a strict process sequence, which, through key processes such as pretreatment, spraying, and curing (baking), form a high-performance coating on the metal surface of the wheel hub that is decorative, corrosion-resistant, weather-resistant, and scratch-resistant.

[0003] Existing wheel painting production lines typically use overhead chains to transport wheel hubs, which then undergo pretreatment, spraying, leveling, and drying processes. However, traditional systems have the following drawbacks:

[0004] Identification relies on manual labor or a single tag: most use barcodes or low-frequency RFID, which are easily damaged or fail to be read in high-temperature, dusty paint-baking environments, and have limited information capacity;

[0005] Delayed quality inspection: The quality of the paint film (such as orange peel, color difference, sagging, particles, etc.) is usually inspected manually or by offline equipment after drying. By the time the problem is discovered, it is too late, resulting in a batch of scraps and high rework costs.

[0006] Production process black box: It is impossible to accurately know the real-time position, residence time, and temperature curve experienced by each wheel hub in a specific drying oven, making it difficult to trace the root cause when quality problems occur;

[0007] Lack of automated visual inspection system: This makes it impossible to make a comprehensive and objective quality judgment on the dried wheel hubs, and there are blind spots in the process quality control of the production line, resulting in poor product quality consistency and an increased risk of defective products being released. Summary of the Invention

[0008] Addressing the technical problems of existing wheel hub painting production lines, such as reliance on manual identification or single tags, lagging quality inspection, lack of production process records, and lack of automated visual inspection, this invention proposes a remote identification production system for wheel hub painting production lines that uses track conveying and drying.

[0009] This invention proposes a remote identification production system for track conveying and drying in a wheel hub painting production line, comprising the following steps: Step 1: Production binding step: At the wheel hub loading station, the wheel hubs to be dried are loaded onto a carrier. The carrier's electronic identification module serves as the carrier's unique electronic identifier. Simultaneously, the wheel hub model is identified by scanning the wheel hub's own barcode using a wheel hub model barcode scanning module. The wheel hub information is bound to the carrier's electronic identification module to establish a unique wheel hub identity association. Then, the binding information is transmitted to the data storage and association module for persistent storage via a data transmission module. The data storage and association module is configured with a remote access interface, supporting remote terminals to query wheel hub binding data and production progress.

[0010] Step 2: Drying monitoring steps: The actual temperature of the wheel hub is recorded by the furnace temperature acquisition module inside the drying oven and the wheel hub temperature recording module on the carrier. The position trajectory of the wheel hub in the oven and the time of entering and leaving the drying oven are recorded by the position and time recording module. The surface temperature data of the wheel hub and the air temperature inside the oven are collected at a frequency of 1-5Hz to generate a temperature-time curve. The curve and the raw data are synchronously transmitted to the data storage and association module.

[0011] Step 3: Quality Inspection and Judgment: After drying, the track conveyor line drives the wheel hub into the inspection component. The carrier arrives at the inspection station, and the RFID information reading module triggers the vision system. The image acquisition module simultaneously acquires multi-angle image data of the wheel hub, obtaining images of surface defects, color differences, orange peel, and multi-angle reflected light intensity data. The paint film thickness acquisition module acquires the paint film thickness data of the wheel hub. The quality judgment module inspects and judges the quality of the wheel hub, automatically determining whether the wheel hub is qualified based on the preset quality standards, and identifying the defect type and location of the unqualified product.

[0012] Step 4: Data Association and Recording Steps: Use the SPC statistical process control analysis module to analyze the correlation between process parameters and quality, use the defect diagnosis module to automatically diagnose defect causes, generate a quality report containing qualified status, defect causes and related process data, and link and store the quality report with the binding relationship established in Step 1.

[0013] Step 5: Information Query and Sorting Steps: The track conveyor line transports the wheel hubs that have completed the inspection to the unloading station. By scanning the RFID information reading module on the carrier, the quality report of the corresponding wheel hub is retrieved from the data storage and association module. The sorting execution module sorts the wheel hubs according to the qualified status and defect reasons in the report.

[0014] Preferably, in step one, the vehicle electronic identification module is an RFID tag, the RFID information reading module is an RFID reader / writer, the wheel hub model identification module is a barcode scanning module, the data transmission module is a PLC, and the data storage and association module is the MES system central database;

[0015] In step one, a binding success rate formula is used to evaluate the binding effect;

[0016] The formula for calculating the binding success rate is: ,in, To increase the success rate of binding, The successful binding count refers to the number of records in the data storage and association module that have persistently stored the association between vehicle ID and wheel hub ID, excluding temporary cache and unconfirmed association data. The total number of binding attempts refers to the total number of single operations from the vehicle ID of the RFID information reading module carrier electronic tag module to the end of the operator clicking the binding confirmation, including invalid attempts such as barcode scanning module scanning failure, data transmission module transmission timeout, and association storage failure.

[0017] Preferably, the furnace temperature acquisition module in step two is a K-type thermocouple temperature sensor, the hub temperature recording module is an integrated temperature sensor, and the position and time recording module is a positioning sensor.

[0018] Step two uses a temperature uniformity index. Formula and equivalent drying time Formulas are used to assess the degree of drying;

[0019] The formula for calculating the temperature uniformity index is: ,in, For the first The real-time temperature of each temperature measuring point is collected by the oven temperature acquisition module arranged in sections inside the drying oven.

[0020] The regional average temperature refers to the arithmetic mean of the temperatures at all valid temperature measurement points, i.e. ,in, The number of effective temperature measurement points refers to the number of K-type thermocouple temperature sensors that normally output data.

[0021] The formula for calculating the equivalent drying time is: ,in, The equivalent drying time reflects the actual curing effect of the wheel hub paint film under temperature fluctuations. The time it takes for the wheel hub to enter the drying oven is recorded by the position and time recording module. The time it takes for the wheel hub to leave the drying oven is also triggered by the position and time recording module. The activation energy for film curing is determined based on the type of film, such as epoxy primer. ≈5.0×10 4 J / mol, polyester topcoat ≈6.0×10 4 J / mol, The value is an ideal gas constant, fixed at 8.314 J / (mol・K);

[0022] This is a real-time temperature function, referring to the temperature of the wheel hub inside the drying oven. The actual temperature at any given time is collected and output by the wheel hub temperature recording module, and needs to be calculated using the following formula: Perform unit conversion.

[0023] Preferably, in step three, the image acquisition module is a plurality of industrial cameras, the paint film thickness acquisition module is a thickness sensor, and the quality determination module is a CNN network defect classification algorithm module based on deep learning.

[0024] The third step uses color difference. Formula, Orange Peel Grading Formula and defect density Formulas for determining wheel hub appearance quality;

[0025] The formula for calculating color difference is: ,in, The color difference index is a comprehensive color difference index based on the CIELAB color space. The smaller the value, the closer the color of the tested wheel hub is to the standard color swatch.

[0026] For differences in brightness, The value range is 0-100, where 0 represents pure black and 100 represents pure white. ,in, To represent the difference in brightness, it is an index in the CIE LAB color space used to quantify the difference between the lightness and darkness of an object's surface and a standard color swatch. Its value can be positive or negative, and it intuitively reflects the direction and degree of deviation of the wheel hub's brightness from the standard. The lightness value of the standard color swatch is a pre-set target lightness benchmark, which is clearly specified in the production process documents;

[0027] A value greater than 0 indicates that the tested wheel hub is brighter than the standard color. When the value is less than 0, it indicates that the wheel hub being tested is darker than the standard color.

[0028] For the difference in red and green hues, There is no absolute upper limit to the value. ,in, To represent the red-green tint value of the wheel hub, To represent the red and green hue values ​​of a standard color chart;

[0029] A value greater than 0 indicates that the tested wheel hub has a reddish tint. A value less than 0 indicates that the tested wheel hub has a greenish tint.

[0030] Due to the difference in yellow and blue tint, There is no absolute upper limit to the value. ,in, To indicate the yellow-blue tint value of the wheel hub, To represent the yellow-blue tint value of a standard color swatch;

[0031] A value greater than 0 indicates that the tested wheel hub has a yellowish tint. A value less than 0 indicates that the tested wheel hub has a bluish tint.

[0032] Orange Peel Grade The calculation formula is: , Orange peel grading value = ,in, The orange peel coefficient is a long-wavelength corrugation that reflects large, easily visible corrugations on the wheel hub surface. It is characterized by wavelengths greater than 1 mm. A higher value indicates more pronounced long-wavelength corrugations. The orange peel effect is a short-wavelength coefficient that reflects the fine ripples on the wheel hub surface, with a wavelength of 0.1-1 mm. A higher value indicates more pronounced short-wavelength ripples. No. The intensity of reflected light at each observation angle is collected by a multi-angle optical sensor mounted on the image acquisition module. , No. The cosine and sine values ​​of each observation angle are used to distinguish between long-wave and short-wave reflected signals.

[0033] The above color data was acquired by the image acquisition module under standardized illumination provided by the D65 standard light source;

[0034] Defect density formula: ,in, Total number of defective pixels refers to the number of pixels corresponding to defective areas identified by the image acquisition module through the quality judgment module. The total number of pixels in the effective detection area refers to the total number of pixels in the area of ​​the wheel hub that needs to be detected, excluding the inner non-exterior surfaces, mounting holes, and marking areas. This number is determined by the image acquisition module based on the resolution and detection range calibration.

[0035] Preferably, the defect diagnosis module in step four is a decision tree algorithm module;

[0036] Correlation coefficient between process parameters and quality Formulas and weighted formulas for defect causes are used to optimize processes and diagnose defects;

[0037] Formula for the correlation coefficient between process parameters and quality: ,in, The Pearson correlation coefficient, with a value range of [-1, 1], is calculated by the SPC statistical process control analysis module and is used to quantify the degree of linear correlation between process parameters and quality indicators. A value greater than 0.7 indicates a strong correlation, while a value less than 0.3 indicates a strong correlation. <0.7 indicates moderate correlation. <0.3 indicates a weak correlation. No. The measured values ​​of the secondary process parameters include data on furnace temperature and drying time collected by the furnace temperature acquisition module and the position and time recording module. It is the average value of process parameters over a certain period of time, i.e. ;

[0038] For the first The corresponding wheel hub quality index value includes the color difference output by the quality assessment module. Defect density Orange peel grade ;

[0039] The average value of the quality index over a certain period of time, i.e. ;

[0040] The sample size refers to the number of times process parameters and quality indicators are measured simultaneously within a certain period of time.

[0041] The formula for calculating the weight of defect causes is: ,in, No. The weight of each defect cause, ranging from [0,1], is calculated by the defect diagnosis module. A larger weight indicates a higher contribution of the cause to the defect, and it should be addressed first. No. The frequency of occurrence of a particular defect cause refers to the proportion of defects caused by that cause to the total number of defects within a certain period. No. The severity of the cause of the defect is quantified on a scale of 1 to 5: 1 indicates minor impact, repairable; 2 indicates minor impact, requiring simple handling; 3 indicates moderate impact, requiring partial rework; 4 indicates significant impact, requiring complete rework; 5 indicates severe impact, requiring immediate scrapping. The total number of defect causes refers to all defect cause types that the defect diagnosis module can identify.

[0042] Preferably, in the information query and sorting step of step five, the sorting accuracy formula, system availability formula, and information query response time formula are used to evaluate system performance;

[0043] The formula for sorting accuracy is: ,in, To ensure accurate quantity sorting, the sorting execution module sorts the wheel hubs to the corresponding areas based on the quality judgment results. The number of wheel hubs whose actual destination matches the system's judgment result is categorized as either the qualified parts removal area, the unqualified rework area, or the unqualified scrap area. Total sorting quantity refers to the total number of wheel hubs that need to be sorted at the unloading station within a certain period of time, including wheel hubs that are correctly sorted and incorrectly sorted;

[0044] The system availability formula is: ,in, Planned operating time refers to the pre-set normal production time of the production line. The downtime refers to the time during which the entire production system cannot operate normally. The entire production system includes RFID readers, vision inspection systems, PLCs, MES databases, and track positioning sensors, excluding planned downtime for maintenance and equipment calibration.

[0045] The information query response formula is: ,in, The total response time for information retrieval refers to the total time from the start of scanning the vehicle's RFID tag to the operator seeing the wheel hub quality report on the query terminal. Scan time refers to the time it takes for the RFID information reading module to scan the electronic tag module on the carrier and identify the tag. ≤0.5s, Network transmission time, measured in seconds, refers to the time it takes for the data transmission module to transmit RFID tags to the data storage and association module, and for the database to send the quality report back to the terminal. Requirements: ≤1.0s, Database query time refers to the time it takes for the data storage and association module to retrieve the corresponding wheel hub quality report based on the RFID tag. ≤0.3s, Interface rendering time refers to the time it takes for the query terminal to render and display the quality report, and the requirements are... ≤0.2s.

[0046] Preferably, the detection component in step three is disposed on the outer surface of the track conveyor line, the carrier is fixedly installed on the outer surface of the slider of the track conveyor line, the RFID tag is fixedly installed on the outer surface of the carrier, the drying oven is fixedly installed on the outer surface of the track conveyor line, the K-type thermocouple temperature sensor is fixedly installed on the inner wall of the track conveyor line and distributed in a rectangular array, the integrated temperature sensor is fixedly installed on the outer surface of the carrier, and the two sets of positioning sensors are fixedly installed on both sides of the inner wall of the drying oven.

[0047] Preferably, the detection component includes a detection housing, which is fixedly installed on the outer surface of the track conveyor line. An RFID reader is fixedly installed on the inner wall of the detection housing. A moving component is fixedly installed on the inner wall of the detection housing. The outer surfaces of the sliders of the moving component located on the top wall, one inner wall, and the bottom wall of the detection housing are respectively fixedly installed on the outer surfaces of a plurality of industrial cameras. Two thickness sensors are respectively fixedly installed on the outer surfaces of the sliders of the moving component located on the top wall and one inner wall of the detection housing.

[0048] Preferably, a pushing hydraulic cylinder is fixedly installed on the outer surface of the slider of the moving component located on the inner wall of the other side of the detection housing. A notched annular moving guide rail is fixedly installed on the outer surface of the piston rod of the pushing hydraulic cylinder. A mounting housing is fixedly installed on the outer surface of the slider of the notched annular moving guide rail. An adjusting hydraulic cylinder is fixedly installed on the inner wall of the mounting housing. A sliding plate is fixedly installed at one end of the piston rod of the adjusting hydraulic cylinder. The outer surface of the sliding plate is slidably inserted into the inner wall of the mounting housing. The outer surface of the sliding plate is fixedly installed on the outer surface of an industrial camera.

[0049] Preferably, a fixing sleeve is fixedly installed on the outer surface of the slide plate, and a turntable with a rotating ring is slidably inserted into the inner wall of the fixing sleeve. The outer surface of the turntable is fixedly installed on the outer surface of another industrial camera. A rotating groove is formed in the inner wall of the fixing sleeve, and the rotating ring of the turntable is slidably connected to the inner wall of the rotating groove. A spiral groove is formed at one end of the turntable. A push sleeve with a rack is slidably inserted into the inner wall of the fixing sleeve. The inner wall of the push sleeve is slidably connected to the inner wall of the spiral groove through a column. An adsorption electromagnetic ring is fixedly installed on the inner wall of the push sleeve. The inner wall of the adsorption electromagnetic ring is magnetically connected to the outer surface of the turntable. A drive motor is fixedly installed on the outer surface of the fixing sleeve through a support base. One end of the output shaft of the drive motor meshes with the rack of the push sleeve through a gear.

[0050] The beneficial effects of this invention are as follows:

[0051] 1. By setting up a track-based conveyor drying system for the wheel hub painting production line, precise identification of each wheel hub can be achieved. This is accomplished through dual identification using RFID tags and barcode scanning, combined with persistent storage in the MES system's central database. This allows for unique traceability of each wheel hub from production line entry to exit. Furthermore, the production process is visualized and monitored. Data is collected from K-type thermocouple temperature sensors and integrated temperature sensors, simultaneously recording the location trajectory and entry / exit times. This generates a temperature-time curve, which, combined with an equivalent drying time formula, precisely quantifies the actual impact of temperature fluctuations on paint film curing, avoiding insufficient curing or other issues caused by traditional fixed-time drying methods. The system addresses the issue of excessive defects by enabling intelligent quality assessment of wheel hubs. During the inspection phase, data is collected using multi-angle industrial cameras and thickness sensors. Based on a CNN network defect classification algorithm, combined with quantitative assessments of color difference, orange peel grade, and defect density, the system replaces the subjectivity of manual visual inspection. Simultaneously, SPC statistical process control is used to analyze the correlation between process parameters and quality, and decision tree algorithms are used to diagnose defect causes. This forms a closed loop of parameter monitoring, quality inspection, defect diagnosis, and process optimization, thereby reducing the defect rate and solving the technical problems of existing wheel hub painting production lines, such as reliance on manual identification or single tags, lagging quality inspection, and lack of production process records.

[0052] 2. By setting up detection components, the wheel hub can be scanned from multiple angles, reducing blind spots. A moving component drives an industrial camera to move synchronously with the wheel hub, enabling high-speed, multi-angle scanning of the outer side of the wheel hub during movement. On the other side, an industrial camera is propelled into the inner cavity of the wheel hub by a hydraulic cylinder, scanning all internal surfaces in synchronous motion. The camera scanning the inner cavity of the wheel hub can rotate, performing a comprehensive scan of the wheel hub frame components. This achieves online inspection without altering the existing production line cycle. Through multi-station collaboration of the outer, inner, top, and bottom sides, it ensures complete scanning of all surfaces of the wheel hub, even in complex suspension postures, solving the technical problem of the lack of automated visual inspection in existing wheel hub painting production lines. Attached Figure Description

[0053] Figure 1 This is a schematic diagram of a remote identification wheel hub painting production line track conveyor drying system proposed in this invention;

[0054] Figure 2 This is a perspective view of the drying furnace structure of a remote identification wheel hub painting production line track conveyor drying system proposed in this invention;

[0055] Figure 3 This is a perspective view of the guide rail conveyor structure of a remote identification wheel hub paint production line track conveyor drying system proposed in this invention.

[0056] Figure 4This is a perspective view of the moving component structure of a remote identification wheel hub painting production line track conveyor drying system proposed in this invention;

[0057] Figure 5 This is a perspective view of the notched annular guide rail structure of a remote identification wheel hub painting production line track conveying and drying production system proposed in this invention;

[0058] Figure 6 This is a perspective view of the industrial camera structure of a remote identification production system for track conveying and drying of a wheel hub paint production line, as proposed in this invention.

[0059] Figure 7 This is a perspective view of the turntable structure of a remote identification wheel hub painting production line track conveyor drying system proposed in this invention;

[0060] Figure 8 This is a perspective view of the spiral groove structure of a remote identification wheel hub painting production line track conveyor drying system proposed in this invention;

[0061] Figure 9 This is a perspective view of the vehicle structure of a remote identification wheel hub painting production line track conveyor drying system proposed in this invention.

[0062] In the diagram: 1. Track conveyor line; 11. Carrier; 12. Drying oven; 13. K-type thermocouple temperature sensor; 14. Integrated temperature sensor; 15. Positioning sensor; 16. RFID tag; 2. Detection housing; 21. RFID reader / writer; 22. Moving part; 23. Industrial camera; 24. Thickness sensor; 3. Pushing hydraulic cylinder; 31. Notched annular moving guide rail; 32. Mounting housing; 33. Adjusting hydraulic cylinder; 34. Slide plate; 4. Fixing sleeve; 41. Turntable; 42. Rotating groove; 43. Spiral groove; 44. Pushing sleeve; 45. Adsorption electromagnetic ring; 46. Drive motor. Detailed Implementation

[0063] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0064] Example 1

[0065] Reference Figure 1A remote identification production system for wheel hub painting production line track conveyor drying includes step one: production binding step: at the wheel hub loading station, the wheel hubs to be dried are loaded onto carrier 11. The electronic identification module on carrier 11 serves as the unique electronic identifier of carrier 11. At the same time, the wheel hub model barcode scanning module scans the barcode of the wheel hub itself to identify the wheel hub model, binding the wheel hub information with the electronic identification module of carrier 11 to establish a unique identity association for the wheel hub. Then, the binding information is transmitted to the data storage and association module for persistent storage through the data transmission module. The data storage and association module is configured with a remote access interface to support remote terminals to query wheel hub binding data and production progress.

[0066] In step one, the electronic identification module uses an 860-960MHz RFID tag 16 (high temperature resistance -40℃-200℃, protection level IP68, suitable for dusty and high temperature environments in paint baking workshops), with a reading distance of 0.5-3m, to meet the non-contact identification requirements next to the track conveyor line 1.

[0067] The RFID information reading module is the ImpinjR2000 UHF RFID reader 21, with an adjustable output power of 1-33dBm;

[0068] The wheel hub model identification module uses an industrial-grade two-dimensional barcode scanner, model Zebra DS2200. It is fixed by a robotic arm and automatically scans to mark DataMatrix codes or QR codes on the mounting surface of the wheel hub (the surface in contact with the axle), around the bolt holes, and the inside of the rim. These codes are small in size, have a large information capacity, and are highly resistant to dirt.

[0069] The data transmission module uses a Siemens S7-1200 PLC to realize the real-time transmission of vehicle ID, wheel hub model, temperature data, etc., to avoid transmission timeout. The timeout threshold is set to 1 second, and automatic retransmission is performed after the timeout.

[0070] The data storage and association module is the central database of the MES system. It uses Oracle 19c, supports distributed storage, is configured with a WebService remote access interface, supports remote terminals, and can be queried through account permissions. The query content includes: single wheel hub binding data (ID of vehicle 11, wheel hub model, loading time), production progress (quantity of work-in-process at each station, load rate of drying oven 12), and quality reports (daily pass rate, defect type distribution).

[0071] Step one uses a binding success rate formula to evaluate the binding effect;

[0072] The formula for calculating the binding success rate is: ,in, The successful binding count refers to the number of records in the data storage and association module that have persistently stored the association between the ID of vehicle 11 and the wheel hub ID, excluding temporary cache: such as temporary storage records where the PLC has not received a confirmation signal from the MES database, and unconfirmed association data: binding requests that have not been confirmed. The total number of binding attempts refers to the total number of single operations from the ID of the carrier 11 of the RFID information reading module carrier 11 to the end of the operator clicking the binding confirmation, including invalid attempts such as barcode scanning module scanning failure, data transmission module transmission timeout, and association storage failure.

[0073] The binding success rate is automatically calculated every hour. When the efficiency is less than 98%, the system automatically pushes an alarm to the equipment maintenance terminal, prompting troubleshooting for issues such as barcode scanner lens contamination, insufficient power of RFID reader / writer 21, or abnormal MES database connection, and generates a troubleshooting guide.

[0074] Step 2: Drying monitoring steps: The actual temperature of the wheel hub is recorded by the furnace temperature acquisition module in the drying oven 12 and the wheel hub temperature recording module on the carrier 11. The position trajectory of the wheel hub in the furnace and the time of entering and leaving the drying oven 12 are recorded by the position and time recording module. The surface temperature data of the wheel hub and the air temperature in the furnace are collected at a frequency of 1-5Hz to generate a temperature-time curve. The curve and the raw data are synchronously transmitted to the data storage and association module.

[0075] The furnace temperature acquisition module in step two is a K-type thermocouple temperature sensor 13, model OMEGAKMQXL-121-GG, which uses a stainless steel protective tube and is fixed to the inner wall of the drying furnace 12. It is connected to the temperature acquisition module, model Advantech ADAM-4018, through a compensation wire. The sampling accuracy is 16 bits to ensure that the temperature data is drift-free.

[0076] The hub temperature recording module is an integrated temperature sensor 14, and the position and time recording module is a positioning sensor 15, model Keyence LV-H32. The two sets are installed at the inlet and outlet of the drying oven 12, respectively. When the carrier (11) passes through, the sensor sends a rising edge signal to the PLC, and the PLC records the current time. =Entry time, =Exit time), and at the same time, the position of the carrier 11 in the furnace is calculated by the encoder (mounted on the track motor shaft, resolution 1000 lines), position = encoder pulse number × pulse equivalent, pulse equivalent = 0.1mm / pulse.

[0077] Step 2 uses the temperature uniformity index Formula and equivalent drying time Formulas are used to assess the degree of drying;

[0078] The formula for calculating the temperature uniformity index is: ,in, For the first The real-time temperature of each temperature measuring point is collected by the furnace temperature acquisition module arranged in sections inside the drying oven 12.

[0079] The regional average temperature refers to the arithmetic mean of the temperatures at all valid temperature measurement points, i.e. ,in, The effective number of temperature measurement points refers to the K-type thermocouple temperature sensor 13 that outputs normal data.

[0080] The formula for calculating the equivalent drying time is: ,in, The equivalent drying time reflects the actual curing effect of the wheel hub paint film under temperature fluctuations. The time it takes for the wheel hub to enter the drying oven 12 is recorded by the position and time recording module. The time it takes for the wheel hub to leave the drying oven 12 is also triggered by the position and time recording module. The activation energy for film curing is determined based on the type of film, such as epoxy primer. ≈5.0×10 4 J / mol, polyester topcoat ≈6.0×10 4 J / mol, The value is an ideal gas constant, fixed at 8.314 J / (mol·K).

[0081] This is a real-time temperature function, referring to the temperature of the wheel hub inside the drying oven 12. The actual temperature at any given time is collected and output by the wheel hub temperature recording module, and needs to be calculated using the following formula: Convert to Kelvin temperature.

[0082] Step 3: Quality Inspection and Judgment: After drying, the track conveyor line 1 drives the wheel hub into the inspection component, and the carrier 11 arrives at the inspection station. The RFID information reading module triggers the vision system, and the image acquisition module simultaneously acquires multi-angle image data of the wheel hub, obtaining images of surface defects, color differences, orange peel, and multi-angle reflected light intensity data. The paint film thickness acquisition module acquires the paint film thickness data of the wheel hub, and the quality judgment module inspects and judges the quality of the wheel hub. Based on the preset quality standards, it automatically determines whether the wheel hub is qualified and identifies the defect type and location of the unqualified product.

[0083] In step three, the image acquisition module consists of multiple 5-megapixel industrial cameras 23, model Hikvision MV-CA050-10GM, with a frame rate of 30fps;

[0084] The paint film thickness acquisition module is a thickness sensor 24, model Testo 244, with a measurement range of 0-200μm. It moves along the surface of the wheel hub and collects one data point every 1cm, for a total of 20 points. The average value is taken as the final paint film thickness. The quality judgment module is a defect classification algorithm module based on deep learning CNN network.

[0085] The third step uses color difference. Formula, Orange Peel Grading Formula and defect density Formulas for determining wheel hub appearance quality;

[0086] The formula for calculating color difference is: ,in, The color difference index is a comprehensive color difference index based on the CIELAB color space. The smaller the value, the closer the color of the tested wheel hub is to the standard color swatch.

[0087] For differences in brightness, The value range is 0-100, where 0 represents pure black and 100 represents pure white. ,in, To represent the difference in brightness, it is an index in the CIE LAB color space used to quantify the difference between the lightness and darkness of an object's surface and a standard color swatch. Its value can be positive or negative, and it intuitively reflects the direction and degree of deviation of the wheel hub's brightness from the standard. The lightness value of the standard color swatch is a pre-set target lightness benchmark, which is clearly specified in the production process documents;

[0088] A value greater than 0 indicates that the tested wheel hub is brighter than the standard color. When the value is less than 0, it indicates that the wheel hub being tested is darker than the standard color.

[0089] For the difference in red and green hues, There is no absolute upper limit to the value. ,in, To represent the red-green tint value of the wheel hub, To represent the red and green hue values ​​of a standard color chart;

[0090] A value greater than 0 indicates that the tested wheel hub has a reddish tint. A value less than 0 indicates that the tested wheel hub has a greenish tint.

[0091] Due to the difference in yellow and blue tint, There is no absolute upper limit to the value. ,in, To indicate the yellow-blue tint value of the wheel hub, To represent the yellow-blue tint value of a standard color swatch;

[0092] A value greater than 0 indicates that the tested wheel hub has a yellowish tint. A value less than 0 indicates that the tested wheel hub has a bluish tint.

[0093] A color difference of less than 3 is considered acceptable.

[0094] Orange Peel Grade The calculation formula is: , Orange peel grading value = ,in, The orange peel coefficient is a long-wavelength corrugation that reflects large, easily visible corrugations on the wheel hub surface. It is characterized by wavelengths greater than 1 mm. A higher value indicates more pronounced long-wavelength corrugations. The orange peel effect is a short-wavelength coefficient that reflects the fine ripples on the wheel hub surface, with a wavelength of 0.1-1 mm. A higher value indicates more pronounced short-wavelength ripples. No. The intensity of reflected light at each observation angle is collected by a multi-angle optical sensor mounted on the image acquisition module. , No. The cosine and sine values ​​of each observation angle are used to distinguish between long-wave and short-wave reflected signals.

[0095] The aforementioned color data was acquired by the image acquisition module under standardized illumination provided by the D65 standard light source. D65 standard light sources (color temperature 5000K±200K, illuminance 1000lux±100lux) were fixed on the inner top wall, inner bottom wall, and inner walls on both sides of the inner wall of the detection component. The distance between the light source and the surface of the wheel hub was uniformly set to 30cm. The reflected light intensity at 5 observation angles (0°, 45°, 90°, 135°, 180°) was acquired by integrating the optical sensor with the industrial camera 23.

[0096] Defect density formula: ,in, Total number of defective pixels refers to the number of pixels corresponding to defective areas identified by the image acquisition module through the quality judgment module. The total number of pixels in the effective detection area refers to the total number of pixels in the area of ​​the wheel hub that needs to be detected, excluding the inner non-exterior surfaces, mounting holes, and marking areas. This number is determined by the image acquisition module based on the resolution and detection range calibration.

[0097] Step 4: Data Association and Recording Steps: Use the SPC statistical process control analysis module to analyze the correlation between process parameters and quality, use the defect diagnosis module to automatically diagnose defect causes, generate a quality report containing qualified status, defect causes and related process data, and link and store the quality report with the binding relationship established in Step 1.

[0098] The SPC statistical process control analysis module uses XR control charts (subgroup size n=5, sampling interval 15 minutes) to calculate the average and range of process parameters (such as furnace temperature and drying time). When a parameter exceeds the control limit, it is automatically marked as an outlier and associated with the corresponding quality indicators (such as color difference) for that time period. ), and analyze the correlation.

[0099] Step four's defect diagnosis module is a decision tree algorithm module, based on the C4.5 decision tree algorithm. The training samples include 1000 process parameters and quality defect data. Common defect causes include: temperature fluctuations (above ±5℃) resulting in color difference; insufficient drying time... <250s) indicates orange peel; uneven paint film thickness (deviation >10μm) indicates sagging; dust inside the furnace indicates particulate defects, and the algorithm diagnosis accuracy is ≥95%.

[0100] Correlation coefficient between process parameters and quality Formulas and weighted formulas for defect causes are used to optimize processes and diagnose defects;

[0101] Formula for the correlation coefficient between process parameters and quality: ,in, The Pearson correlation coefficient, with a value range of [-1, 1], is calculated by the SPC statistical process control analysis module and is used to quantify the degree of linear correlation between process parameters and quality indicators. A value greater than 0.7 indicates a strong correlation, while a value less than 0.3 indicates a strong correlation. <0.7 indicates moderate correlation. <0.3 indicates a weak correlation. No. The measured values ​​of the secondary process parameters include data on furnace temperature and drying time collected by the furnace temperature acquisition module and the position and time recording module. It is the average value of process parameters over a certain period of time, i.e. ;

[0102] For the first The corresponding wheel hub quality index value includes the color difference output by the quality assessment module. Defect density Orange peel grade ;

[0103] The average value of the quality index over a certain period of time, i.e. ;

[0104] For the first The corresponding wheel hub quality index value includes color difference. Defect density Orange peel grade ;

[0105] The sample size refers to the number of times process parameters and quality indicators are measured simultaneously within a certain period of time.

[0106] The formula for calculating the weight of defect causes is: ,in, No. The weight of each defect cause, ranging from [0,1], is calculated by the defect diagnosis module. A larger weight indicates a higher contribution of the cause to the defect, and it should be addressed first. No. The frequency of occurrence of a particular defect cause refers to the proportion of defects caused by that cause to the total number of defects within a certain period. No. The severity of the cause of the defect is quantified on a scale of 1 to 5: 1 indicates minor impact, repairable; 2 indicates minor impact, requiring simple handling; 3 indicates moderate impact, requiring partial rework; 4 indicates significant impact, requiring complete rework; 5 indicates severe impact, requiring immediate scrapping. The total number of defect causes refers to all defect cause types that the defect diagnosis module can identify.

[0107] Step 5: Information Query and Sorting Steps: The track conveyor line 1 transports the wheel hubs that have completed the inspection to the unloading station. By scanning the RFID information reading module on the carrier 11, the corresponding wheel hub's quality report is retrieved from the data storage and association module. The sorting execution module sorts the wheel hubs according to the qualified status and defect reasons in the report.

[0108] The sorting execution module can use a pneumatic sorting robot arm, model ABBIRB120. According to the quality report instructions, qualified products are sent to the next conveyor belt, unqualified rework products are sent to the rework station, and scrap products are sent to the waste bin. The sorting action response time is ≤1s.

[0109] In step five, the information query and sorting steps, the sorting accuracy formula, system availability formula, and information query response time formula are used to evaluate system performance.

[0110] The formula for sorting accuracy is: ,in, To ensure accurate quantity sorting, the sorting execution module sorts the wheel hubs to the corresponding areas based on the quality judgment results. The number of wheel hubs whose actual destination matches the system's judgment result is categorized as either the qualified parts removal area, the unqualified rework area, or the unqualified scrap area. Total sorting quantity refers to the total number of wheel hubs that need to be sorted at the unloading station within a certain period of time, including wheel hubs that are correctly sorted and incorrectly sorted;

[0111] The system availability formula is: ,in, Planned operating time refers to the pre-set normal production time of the production line. The downtime refers to the time during which the entire production system cannot operate normally. The entire production system includes RFID readers 21, vision inspection system, PLC, MES database, and track positioning sensors, excluding planned downtime for maintenance and equipment calibration.

[0112] The information query response formula is: ,in, The total response time for information queries refers to the total time from the start of scanning the RFID tag 16 on the vehicle 11 to the operator seeing the wheel hub quality report on the query terminal. Scan time refers to the time it takes for the RFID information reading module to scan the electronic tag module on the carrier 11 and identify the tag. ≤0.5s, Network transmission time, measured in seconds, refers to the time it takes for the data transmission module to transmit RFID tags to the data storage and association module, and for the database to send the quality report back to the terminal. Requirements: ≤1.0s, Database query time refers to the time it takes for the data storage and association module to retrieve the corresponding wheel hub quality report based on the RFID tag. ≤0.3s, Interface rendering time refers to the time it takes for the query terminal to render and display the quality report, and the requirements are... ≤0.2s.

[0113] The information query terminal uses an industrial tablet PC with an MES client installed. It supports barcode scanning (scanning the RFID tag 16 of the carrier 11) and number-based query. The query response time is displayed on the terminal interface in real time.

[0114] Production binding trigger conditions: When empty vehicle 11 arrives at the loading station, the UHF RFID reader (21) reads the fixed ID of its RFID tag 16. After vehicle 11 arrives at the loading station, it is automatically loaded by a robot arm. The positioning sensor at the loading station sends a position signal, and the PLC triggers the robot arm to drive the barcode scanning module to scan and obtain the unique information of the current wheel hub, such as barcode and model. After successful scanning, the binding is confirmed. After confirmation, the data is synchronized to the MES database, and a new binding record is established. In the binding table of the MES database, a record is inserted, which includes the ID of vehicle 11, wheel hub identity, status, and timestamp. If the binding fails (such as a blurry barcode or transmission timeout), an audible and visual alarm will be triggered.

[0115] Drying monitoring data linkage: The position and time recording module (positioning sensor 15 adopts photoelectric positioning sensor with response time ≤0.01s) records the time when the wheel hub enters the drying oven 12. When the wheel hub reaches each temperature measurement zone in the oven, the K-type thermocouple temperature sensor 13 and the integrated temperature sensor 14 synchronously collect data. The generated temperature-time curve is displayed in real time on the touch screen of the control cabinet of the drying oven 12. If the curve is abnormal, the power of the heating tube in the oven will be automatically adjusted.

[0116] Industrial camera (23) scanning and data calculation: After drying, the wheel hub enters the detection component through the carrier 11. The UHF RFID reader 21 takes the fixed ID of the RFID tag 16 on the carrier 11 to determine the wheel hub information. The upper camera vertically downwards to shoot the outer side of the upper half of the wheel hub, the side camera horizontally shoots the front of the wheel hub in a ring area, and the lower camera vertically upwards to shoot the outer side of the lower half of the wheel hub. The camera controller has a built-in edge computing module to process the scanned image in real time. The defect identification uses the Canny edge detection algorithm to first denoise and extract the edge of the wheel hub image taken by the industrial camera 23, and transform the blurred visual information into clear contour features to reduce invalid pixel interference.

[0117] Wheel hub quality assessment: The CNN network defect classification algorithm module receives the original image and edge feature map, calculates the quality of the wheel hub, and forms a structured dataset containing the ID of vehicle 11, defect type, defect level, drying temperature, drying time, paint film thickness, and conveyor linear speed. This dataset provides complete input features for the defect diagnosis module, which analyzes and diagnoses the occurrence of defects. The two modules are linked through the MES database. The defect diagnosis module uploads the defect cause, weight, and rectification suggestions to the MES database and associates them with the quality record ID to form a complete quality-cause traceability file.

[0118] Sorting linkage: The MES database synchronizes the qualified status determined by the CNN network defect classification algorithm module with the defect cause output by the decision tree to the sorting execution module for sorting according to the rules.

[0119] Example 2

[0120] Reference Figures 2-9 As shown, the detection components in step three are set on the outer surface of the track conveyor line 1, the carrier 11 is fixedly installed on the outer surface of the slider of the track conveyor line 1, the RFID tag 16 is fixedly installed on the outer surface of the carrier 11, the drying oven 12 is fixedly installed on the outer surface of the track conveyor line 1, the K-type thermocouple temperature sensor 13 is fixedly installed on the inner wall of the track conveyor line 1 and distributed in a rectangular array, the integrated temperature sensor 14 is fixedly installed on the outer surface of the carrier 11, and two sets of positioning sensors 15 are fixedly installed on both sides of the inner wall of the drying oven 12.

[0121] The detection component includes a detection housing 2, which is fixedly installed on the outer surface of the track conveyor line 1. An RFID reader 21 is fixedly installed on the inner wall of the detection housing 2. A moving part 22 is fixedly installed on the inner wall of the detection housing 2. It is driven by a servo motor and a ball screw. The slider stroke is 0-50cm and the moving speed is adjustable from 0.1-0.5m / s. A camera bracket is installed on the slider, which can rotate 360° to meet the shooting requirements of different angles. The outer surfaces of the sliders of the moving part 22 located on the inner top wall, one inner wall and the inner bottom wall of the detection housing 2 are fixedly installed on the outer surfaces of multiple industrial cameras 23, respectively. Two thickness sensors 24 are fixedly installed on the outer surfaces of the sliders of the moving part 22 located on the inner top wall and one inner wall of the detection housing 2, respectively.

[0122] A push hydraulic cylinder 3 is fixedly installed on the outer surface of the slider of the moving part 22 on the inner wall of the detection housing 2. A notched annular moving guide rail 31 is fixedly installed on the outer surface of the piston rod of the push hydraulic cylinder 3. The switching annular moving guide rail 31 drives the gear to rotate on the rack of the annular guide rail through the motor, thereby moving the moving seat. A mounting housing 32 is fixedly installed on the outer surface of the slider of the notched annular moving guide rail 31. An adjusting hydraulic cylinder 33 is fixedly installed on the inner wall of the mounting housing 32. A sliding plate 34 is fixedly installed on one end of the piston rod of the adjusting hydraulic cylinder 33. The outer surface of the sliding plate 34 is slidably inserted into the inner wall of the mounting housing 32. The outer surface of the sliding plate 34 is fixedly installed on the outer surface of an industrial camera 23.

[0123] A fixed sleeve 4 is fixedly installed on the outer surface of the slide plate 34. A turntable 41 with a rotating ring is slidably inserted into the inner wall of the fixed sleeve 4. The outer surface of the turntable 41 is fixedly installed on the outer surface of another industrial camera 23. A rotating groove 42 is opened on the inner wall of the fixed sleeve 4. The rotating ring of the turntable 41 is slidably connected to the inner wall of the rotating groove 42. A spiral groove 43 is opened at one end of the turntable 41. A push sleeve 44 with a rack is slidably inserted into the inner wall of the fixed sleeve 4. The inner wall of the push sleeve 44 is slidably connected to the inner wall of the spiral groove 43 through a column. An adsorption electromagnetic ring 45 is fixedly installed on the inner wall of the push sleeve 44. The inner wall of the adsorption electromagnetic ring 45 is magnetically connected to the outer surface of the turntable 41. A drive motor 46 is fixedly installed on the outer surface of the fixed sleeve 4 through a support base. One end of the output shaft of the drive motor 46 meshes with the rack of the push sleeve 44 through a gear.

[0124] Working principle: When inspection is required, the dried wheel hub enters the inspection housing 2. The RFID reader 21 reads the information of the wheel hub on the carrier 11. The moving part 22 is activated, which drives the camera and thickness sensor 24 located above and on one side of the wheel hub to move synchronously with the wheel hub to scan the outer side and front of the wheel hub. At the same time, the moving part 22 on the other side drives the hydraulic cylinder 3 to move synchronously. The hydraulic cylinder 3 pushes the notched annular moving guide rail 31 into the inner cavity of the wheel hub. The notched annular moving guide rail 31 drives the mounting housing 32 to move in a circle to scan the inner wall and side wall of the wheel hub.

[0125] As the mounting housing 32 returns, the drive motor 46 of the fixed sleeve 4 on the slide plate 34 starts, driving the push sleeve 44 to move within the fixed sleeve 4 through gear transmission. After the fixed sleeve 4 is magnetically connected to the turntable 41 via the adsorption electromagnetic ring 45, the push sleeve 44 pushes the turntable 41 outward, allowing an industrial camera 23 to enter the cavity between the two frames of the wheel hub. When the rotating ring of the turntable 41 pushed by the push sleeve 44 reaches the rotating groove 42, the movement of the turntable 41 is limited. After the adsorption electromagnetic ring 45 is de-energized, the push sleeve 44 continues to move, moving within the spiral groove 43 via the column. The turntable 41 rotates within the rotating groove 42, driving the industrial camera 23 to rotate, thereby completing the scanning of the outer surface of the wheel hub frame. After scanning, it resets and enters the cavity formed by the next frame. By adjusting the hydraulic cylinder 33, the slide plate 34 is moved within the mounting housing 32, allowing the position of the two industrial cameras 23 within the wheel hub cavity to be adjusted.

[0126] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A remote identification production system for track conveying and drying in a wheel hub painting production line, characterized in that: Including step one: production binding step: at the wheel hub loading station, the wheel hubs that need to be dried are loaded onto the carrier (11). The electronic identification module on the carrier (11) serves as the electronic unique identifier of the carrier (11). At the same time, the wheel hub model barcode scanning module scans the barcode of the wheel hub itself to identify the wheel hub model, binds the wheel hub information with the electronic identification module of the carrier (11), establishes a unique identity association for the wheel hub, and then transmits the binding information to the data storage and association module for persistent storage through the data transmission module. The data storage and association module is configured with a remote access interface to support remote terminals to query wheel hub binding data and production progress. Step 2: Drying monitoring steps: The actual temperature of the wheel hub is recorded by the furnace temperature acquisition module in the drying oven (12) and the wheel hub temperature recording module on the carrier (11). The position trajectory of the wheel hub in the furnace and the time of entering and leaving the drying oven (12) are recorded by the position and time recording module. The surface temperature data of the wheel hub and the air temperature in the furnace are collected at a frequency of 1-5Hz to generate a temperature-time curve. The curve and the original data are synchronously transmitted to the data storage and association module. Step 3: Quality Inspection and Judgment Steps: After drying, the track conveyor (1) drives the wheel hub into the inspection component, the carrier (11) arrives at the inspection station, the RFID information reading module triggers the vision system, the image acquisition module synchronously collects multi-angle image data of the wheel hub, obtains images of surface defects, color difference, orange peel and multi-angle reflected light intensity data, the paint film thickness acquisition module collects the paint film thickness data of the wheel hub, the quality judgment module inspects and judges the quality of the wheel hub, automatically judges whether the wheel hub is qualified based on the preset quality standard, and identifies the defect type and location of the unqualified product; Step 4: Data Association and Recording Steps: Use the SPC statistical process control analysis module to analyze the correlation between process parameters and quality, use the defect diagnosis module to automatically diagnose defect causes, generate a quality report containing qualified status, defect causes and related process data, and link and store the quality report with the binding relationship established in Step 1. Step 5: Information Query and Sorting Steps: The track conveyor (1) transports the wheel hubs after inspection to the unloading station. By scanning the RFID information reading module on the carrier (11), the corresponding wheel hub quality report is retrieved from the data storage and association module. The sorting execution module sorts the wheel hubs according to the qualified status and defect reasons in the report.

2. The remote identification wheel hub painting production line track conveyor drying system according to claim 1, characterized in that: In step one, the electronic identification module is an RFID tag (16), the RFID information reading module is an RFID reader (21), the wheel hub model identification module is a barcode scanning module, the data transmission module is a PLC, and the data storage and association module is the MES system central database. In step one, a binding success rate formula is used to evaluate the binding effect: The formula for calculating the binding success rate is: ,in, To increase the success rate of binding, The number of successfully bound records refers to the number of records in the data storage and association module that have persistently stored the association between the vehicle (11) ID and the wheel hub ID, excluding temporary cache and unconfirmed associated data. The total number of binding attempts refers to the total number of single operations from the RFID information reading module carrier (11) electronic tag module carrier (11) ID to the end of the binding confirmation, including invalid attempts such as barcode scanning module scanning failure, data transmission module transmission timeout, and association storage failure.

3. The remote identification wheel hub painting production line track conveyor drying system according to claim 2, characterized in that: The furnace temperature acquisition module in step two is a K-type thermocouple temperature sensor (13), the hub temperature recording module is an integrated temperature sensor (14), and the position and time recording module is a positioning sensor (15). Step two uses a temperature uniformity index. Formula and equivalent drying time Formula for assessing drying degree: The formula for calculating the temperature uniformity index is: ,in, For the first The real-time temperature of each temperature measuring point is collected by the furnace temperature acquisition module arranged in a partitioned manner inside the drying oven (12); The regional average temperature refers to the arithmetic mean of the temperatures at all valid temperature measurement points, i.e. ,in, The number of effective temperature measurement points refers to the number of K-type thermocouple temperature sensors (13) that normally output data; The formula for calculating the equivalent drying time is: ,in, The equivalent drying time reflects the actual curing effect of the wheel hub paint film under temperature fluctuations. The time it takes for the wheel hub to enter the drying oven (12) is recorded by the position and time recording module. The time it takes for the wheel hub to leave the drying oven (12) is also triggered by the position and time recording module. The activation energy for film curing is determined based on the type of film, such as epoxy primer. ≈5.0×10 4 J / mol, polyester topcoat ≈6.0×10 4 J / mol, The value is an ideal gas constant, fixed at 8.314 J / (mol・K); The real-time temperature function refers to the temperature of the wheel hub inside the drying oven (12). The actual temperature at any given time is collected and output by the wheel hub temperature recording module, and needs to be calculated using the following formula: Perform unit conversion.

4. The production system for remote identification wheel hub painting production line track conveying and drying according to claim 3, characterized in that: In step three, the image acquisition module consists of multiple industrial cameras (23), the paint film thickness acquisition module consists of a thickness sensor (24), and the quality judgment module consists of a CNN network defect classification algorithm module based on deep learning. The third step uses color difference. Formula, Orange Peel Grading Formula and defect density Formulas for determining wheel hub appearance quality; The formula for calculating color difference is: ,in, The color difference index is a comprehensive color difference index based on the CIELAB color space. The smaller the value, the closer the color of the tested wheel hub is to the standard color swatch. For differences in brightness, The value range is 0-100, where 0 represents pure black and 100 represents pure white. ,in, To represent differences in brightness, this is an index in the CIE LAB color space used to quantify the difference between the lightness and darkness of an object's surface and a standard color swatch. Its value can be positive or negative, directly reflecting the direction and degree of deviation of the wheel hub's brightness from the standard. The lightness value of the standard color swatch is a pre-set target lightness benchmark, which is clearly specified in the production process documents; A value greater than 0 indicates that the tested wheel hub is brighter than the standard color. When the value is less than 0, it indicates that the wheel hub being tested is darker than the standard color. For the difference in red and green hues, There is no absolute upper limit to the value. ,in, To represent the red-green tint value of the wheel hub, To represent the red and green hue values ​​of a standard color chart; A value greater than 0 indicates that the tested wheel hub has a reddish tint. A value less than 0 indicates that the tested wheel hub has a greenish tint. Due to the difference in yellow and blue tint, There is no absolute upper limit to the value. ,in, To indicate the yellow-blue tint value of the wheel hub, To represent the yellow-blue tint value of a standard color swatch; A value greater than 0 indicates that the tested wheel hub has a yellowish tint. A value less than 0 indicates that the tested wheel hub has a bluish tint. Orange Peel Grade The calculation formula is: , Orange peel grading value = ,in, The orange peel coefficient is a long-wavelength corrugation that reflects large, easily visible corrugations on the wheel hub surface. It is characterized by wavelengths greater than 1 mm. A higher value indicates more pronounced long-wavelength corrugations. The orange peel effect is a short-wavelength coefficient that reflects the fine ripples on the wheel hub surface, with a wavelength of 0.1-1 mm. A higher value indicates more pronounced short-wavelength ripples. No. The intensity of reflected light at each observation angle is collected by a multi-angle optical sensor mounted on the image acquisition module. , No. The cosine and sine values ​​of each observation angle are used to distinguish between long-wave and short-wave reflected signals. The above color data was acquired by the image acquisition module under standardized illumination provided by the D65 standard light source; Defect density formula: ,in, Total number of defective pixels refers to the number of pixels corresponding to defective areas identified by the image acquisition module through the quality judgment module. The total number of pixels in the effective detection area refers to the total number of pixels in the area of ​​the wheel hub that needs to be detected, excluding the inner non-exterior surfaces, mounting holes, and marking areas. This number is determined by the image acquisition module based on the resolution and detection range calibration.

5. The production system for remote identification wheel hub painting production line track conveying and drying according to claim 4, characterized in that: The defect diagnosis module in step four is a decision tree algorithm module; Using the correlation coefficient between process parameters and quality Formulas and weighted formulas for defect causes are used to optimize processes and diagnose defects; Formula for the correlation coefficient between process parameters and quality: ,in, The Pearson correlation coefficient, with a value range of [-1, 1], is calculated by the SPC statistical process control analysis module and is used to quantify the degree of linear correlation between process parameters and quality indicators. A value greater than 0.7 indicates a strong correlation, while a value less than 0.3 indicates a strong correlation. <0.7 indicates moderate correlation. <0.3 indicates a weak correlation. No. The measured values ​​of the secondary process parameters include data on furnace temperature and drying time collected by the furnace temperature acquisition module and the position and time recording module. It is the average value of process parameters over a certain period of time, i.e. ; For the first The corresponding wheel hub quality index value includes the color difference output by the quality assessment module. Defect density Orange peel grade ; The average value of the quality index over a certain period of time, i.e. ; The sample size refers to the number of times process parameters and quality indicators are measured simultaneously within a certain period of time. The formula for calculating the weight of defect causes is: ,in, No. The weight of each defect cause, ranging from [0,1], is calculated by the defect diagnosis module. A larger weight indicates a higher contribution of the cause to the defect, and it should be addressed first. No. The frequency of occurrence of a particular defect cause refers to the proportion of defects caused by that cause to the total number of defects within a certain period. No. The severity of the cause of the defect is quantified on a scale of 1 to 5: 1 is a minor impact, which can be repaired; 2 is a minor impact, which requires simple handling; 3 is a moderate impact, which requires partial rework; 4 is a major impact, which requires complete rework; and 5 is a severe impact, which requires immediate scrapping. The total number of defect causes refers to all defect cause types that the defect diagnosis module can identify.

6. The remote identification wheel hub painting production line track conveyor drying system according to claim 5, characterized in that: In step five, the information query and sorting steps are evaluated using the sorting accuracy formula, the system availability formula, and the information query response time formula to assess system performance. The formula for sorting accuracy is: ,in, To ensure accurate quantity sorting, the sorting execution module sorts the wheel hubs to the corresponding areas based on the quality judgment results. The number of wheel hubs whose actual destination matches the system's judgment result is categorized as either the qualified product unloading area, the unqualified rework area, or the unqualified scrap area. Total sorting quantity refers to the total number of wheel hubs that need to be sorted at the unloading station within a certain period of time, including wheel hubs that are correctly sorted and incorrectly sorted; The system availability formula is: ,in, Planned operating time refers to the pre-set normal production time of the production line. The downtime refers to the time during which the entire production system cannot work normally. The entire production system includes RFID readers (21), vision inspection system, PLC, MES database, and track positioning sensors, excluding planned downtime for maintenance and equipment calibration. The information query response formula is: ,in, The total response time for information queries refers to the total time from the start of scanning the RFID tag (16) on the vehicle (11) to the operator seeing the wheel hub quality report on the query terminal. The scanning time refers to the time it takes for the RFID information reading module to scan the electronic tag module of the carrier (11) and identify the tag. ≤0.5s, Network transmission time, measured in seconds, refers to the time it takes for the data transmission module to transmit RFID tags to the data storage and association module, and for the database to send the quality report back to the terminal. Requirements: ≤1.0s, Database query time refers to the time it takes for the data storage and association module to retrieve the corresponding wheel hub quality report based on the RFID tag. ≤0.3s, Interface rendering time refers to the time it takes for the query terminal to render and display the quality report, and the requirements are... ≤0.2s.

7. The remote identification wheel hub painting production line track conveyor drying system according to claim 6, characterized in that: The detection component in step three is set on the outer surface of the track conveyor line (1). The carrier (11) is fixedly installed on the outer surface of the slider of the track conveyor line (1). The RFID tag (16) is fixedly installed on the outer surface of the carrier (11). The drying oven (12) is fixedly installed on the outer surface of the track conveyor line (1). The K-type thermocouple temperature sensor (13) is fixedly installed on the inner wall of the track conveyor line (1) and distributed in a rectangular array. The integrated temperature sensor (14) is fixedly installed on the outer surface of the carrier (11). The two sets of positioning sensors (15) are fixedly installed on both sides of the inner wall of the drying oven (12).

8. The remote identification wheel hub painting production line track conveyor drying system according to claim 7, characterized in that: The detection component includes a detection housing (2), which is fixedly installed on the outer surface of the track conveyor line (1). An RFID reader (21) is fixedly installed on the inner wall of the detection housing (2). A moving component (22) is fixedly installed on the inner wall of the detection housing (2). The outer surface of the slider of the moving component (22) located on the inner top wall, one inner wall and the inner bottom wall of the detection housing (2) is fixedly installed on the outer surface of a plurality of industrial cameras (23). Two thickness sensors (24) are fixedly installed on the outer surface of the slider of the moving component (22) located on the inner top wall and one inner wall of the detection housing (2).

9. The remote identification wheel hub painting production line track conveyor drying system according to claim 8, characterized in that: A push hydraulic cylinder (3) is fixedly installed on the outer surface of the slider of the moving part (22) located on the inner wall of the other side of the detection housing (2). A notched annular moving guide rail (31) is fixedly installed on the outer surface of the piston rod of the push hydraulic cylinder (3). An installation housing (32) is fixedly installed on the outer surface of the slider of the notched annular moving guide rail (31). An adjusting hydraulic cylinder (33) is fixedly installed on the inner wall of the installation housing (32). A sliding plate (34) is fixedly installed at one end of the piston rod of the adjusting hydraulic cylinder (33). The outer surface of the sliding plate (34) is slidably inserted into the inner wall of the installation housing (32). The outer surface of the sliding plate (34) is fixedly installed on the outer surface of an industrial camera (23).

10. A remote identification wheel hub painting production line track conveyor drying system according to claim 9, characterized in that: A fixing sleeve (4) is fixedly installed on the outer surface of the slide plate (34). A turntable (41) with a rotating ring is slidably inserted into the inner wall of the fixing sleeve (4). The outer surface of the turntable (41) is fixedly installed on the outer surface of another industrial camera (23). A rotating groove (42) is opened on the inner wall of the fixing sleeve (4). The rotating ring of the turntable (41) is slidably connected to the inner wall of the rotating groove (42). A spiral groove (43) is opened at one end of the turntable (41). The inner wall of the fixing sleeve (4) slides... A push sleeve (44) with a rack is inserted. The inner wall of the push sleeve (44) is slidably connected to the inner wall of the spiral groove (43) through a column. An adsorption electromagnetic ring (45) is fixedly installed on the inner wall of the push sleeve (44). The inner wall of the adsorption electromagnetic ring (45) is magnetically connected to the outer surface of the turntable (41). A drive motor (46) is fixedly installed on the outer surface of the fixed sleeve (4) through a support seat. One end of the output shaft of the drive motor (46) meshes with the rack of the push sleeve (44) through a gear.

Images