A large data-based whole-process tracking management method for paintwork products

Abstract

The application discloses a kind of based on big data's lacquer piece product whole process traceability management method, belong to traceability management technical field, specifically include: input production instruction number including coating batch code and curing oven number;Retrieve resin characteristic peak sequence and pigment dispersity numerical value corresponding to coating batch code;With production instruction number as root node, build curing traceability chain, and the curing oven number, resin characteristic peak sequence and pigment dispersity numerical value are connected to root node;When lacquer film adhesion is lower than standard threshold, locate to resin characteristic peak sequence and pigment dispersity numerical value;Resin characteristic peak sequence is compared with standard library to generate peak deviation mark, and pigment dispersity numerical value is compared with standard interval to generate degree deviation mark, and original curing oven temperature curve is corrected according to the two;Output curing traceability atlas with peak deviation mark and degree deviation mark.The application realizes the closed loop traceability of coating characteristic quantization and process correction.

Description

Technical Field

[0001] This invention relates to the field of traceability management technology, specifically to a method for full-process traceability management of painted products based on big data. Background Technology

[0002] Painted products are widely used in industrial fields such as automobile manufacturing, home appliance production, and aerospace. Paint film adhesion is one of the core indicators for measuring product surface quality. The formation of paint film adhesion depends on the synergistic effect of the paint's own physicochemical properties and curing process conditions. Paint properties include intrinsic parameters such as the molecular structure stability of resin components and the uniformity of pigment particle dispersion, while curing process conditions mainly include external parameters such as the temperature profile of the curing oven, heating rate, and holding time. With the popularization of industrial big data technology, painted parts manufacturers generally use spectrometers to scan resin characteristic peaks and detect pigment dispersion for each batch of paint during the paint receiving process, and record the test results in the paint management system. During the curing process, temperature change data inside the curing oven is collected in real time using a curing oven temperature profile recorder to generate the original curing oven temperature profile. Finally, RFID tags are used to record the production equipment and process parameters that each product has passed through during the finished product delivery process.

[0003] In existing technologies, when the paint film adhesion test value of painted products is lower than the standard threshold during outbound inspection, quality management personnel need to retrieve the incoming inspection record of that batch of paint from the paint management system, the curing oven number and corresponding original curing oven temperature profile from the production execution system, and the temperature and humidity data during the curing process from the painting workshop environmental monitoring system. Then, they need to manually compare and analyze the cause of the adhesion abnormality. Some traceability systems use RFID tags to associate products with production instruction numbers, achieving one-way traceability of the production process.

[0004] However, when the coatings are received into the warehouse, the resin characteristic peak sequences and pigment dispersion values ​​collected by the spectrometer are only stored as static batch records in the coating management system. The original curing oven temperature curves collected by the curing oven temperature curve recorder are stored independently in the production execution system. There is no data mapping relationship between the two with the production instruction number as the primary key. During the storage process, the resin components may undergo slow cross-linking reactions, causing characteristic peaks to drift, and pigment particles may gradually decrease in dispersion values ​​due to sedimentation. These dynamic changes are not included in the traceability system linked to the curing temperature curves. When painted products show paint film adhesion test values ​​below the standard threshold during outbound inspection, quality management personnel need to retrieve the inbound inspection records of the batch of paint from the paint management system and the curing oven number and corresponding original curing oven temperature curve from the production execution system. Due to the lack of a structured correlation path between paint characteristic data and curing process data, it is impossible to distinguish whether the adhesion abnormality is caused by a deviation in paint characteristics or a deviation in curing temperature through data calculation. It is also impossible to quantify the actual degree of deviation in paint characteristics into correction parameters that act inversely on the curing process settings. As a result, the tracing of the cause of the abnormality relies on manual experience judgment and cannot form a closed loop from abnormality detection to process correction. Summary of the Invention

[0005] The purpose of this invention is to provide a big data-based method for full-process traceability management of painted products, addressing the following technical problems: In existing technologies, coating characteristic data and curing process data lack a structured correlation, making it impossible to distinguish whether abnormal adhesion is caused by deviations in coating characteristics or curing temperature, and also impossible to quantify characteristic deviations into correction parameters that can be applied in reverse to process settings.

[0006] The objective of this invention can be achieved through the following technical solutions: A big data-based method for full-process traceability management of painted products includes the following steps: S1. Input the production instruction number of the painted product. The production instruction number includes the paint batch code and curing oven number associated with the entire process from primer spraying to topcoat curing. S2. Retrieve the resin characteristic peak sequence corresponding to the paint batch code based on the production instruction number, and retrieve the pigment dispersion value corresponding to the same paint batch code. S3. Construct a curing traceability chain with the production instruction number as the root node, and connect the curing oven number, resin characteristic peak sequence and pigment dispersion value as chain nodes to the root node. S4. When the paint film adhesion is found to be lower than the standard threshold during the outbound inspection of painted products, the corresponding production instruction number is extracted and located from the curing traceability chain to the resin characteristic peak sequence and pigment dispersion value. S5. Compare the resin characteristic peak sequence with the standard resin characteristic peak library to generate peak deviation markers, compare the pigment dispersion value with the standard dispersion range to generate degree deviation markers, and correct the original curing oven temperature curve to a corrected curing oven temperature curve based on the peak deviation markers and degree deviation markers. S6. Output a curing traceability map of the painted parts with peak deviation markers and degree deviation markers. The curing traceability map of the painted parts includes the curing oven number, the paint batch code, and the corrected curing oven temperature curve.

[0007] As a further aspect of the present invention: In step S3, the process of constructing a curing traceability chain with the production instruction number as the root node, and connecting the curing oven number, resin characteristic peak sequence, and pigment dispersion value as chain nodes to the root node, is as follows: Extract the coating entry date carried in the production instruction number, and screen the resin characteristic peak sequences of the same coating batch code at multiple time points before entry from the historical detection records of the spectrometer. After arranging multiple resin characteristic peak sequences in chronological order, identify the characteristic peaks whose peak positions show unidirectional continuous drift over time as dynamic characteristic peaks. Extract the drift rate value and drift direction value of the dynamic characteristic peaks, and concatenate the drift rate value and drift direction value as sub-nodes of the resin characteristic peak sequence after the resin characteristic peak sequence. From the historical test records of the pigment dispersion tester, the pigment dispersion values ​​of the same paint batch code at multiple time points before warehousing are selected. After arranging the multiple pigment dispersion values ​​in chronological order, the difference between the values ​​at adjacent time points is calculated. The sequence segment with the most consecutive occurrences of the same sign in the difference is taken as the change trend segment of pigment dispersion. The difference between the starting value and the ending value of the change trend segment is taken as the child node of the pigment dispersion value and concatenated after the pigment dispersion value.

[0008] As a further aspect of the present invention: In step S4, the process of extracting the corresponding production instruction number and locating the resin characteristic peak sequence and pigment dispersion value from the curing traceability chain is as follows: Obtain the RFID tag number affixed to the surface of painted products with paint film adhesion below the standard threshold. Based on the RFID tag number, read the entry reading time point from the RFID reader at the curing oven entrance and the exit reading time point from the RFID reader at the curing oven exit. Filter out production instruction numbers from the production execution system that match the curing oven number in S1. Extract the painting completion time point and paint batch code corresponding to each production instruction number from the filtered production instruction numbers. Select production instruction numbers whose painting completion time point is earlier than the exit reading time point and whose time difference between the painting completion time point and the entry reading time point is within a set range as a candidate set. When there is only one production instruction number in the candidate set, that production instruction number is taken as the extraction result. When there are multiple production instruction numbers in the candidate set, the production instruction number whose paint batch code is the same as the paint batch code carried in the RFID tag number is selected as the extraction result.

[0009] As a further aspect of the present invention: in step S5, the process of comparing the resin characteristic peak sequence with the standard resin characteristic peak library to generate peak deviation markers is as follows: Retrieve the characteristic peak sequence of standard resin of the same coating type from the standard resin characteristic peak library. Convert the resin characteristic peak sequence and the standard resin characteristic peak sequence into two-dimensional point sets composed of peak position values ​​and peak height values, respectively. Calculate the Hausdorff distance between the two-dimensional point sets of the resin characteristic peak sequence and the standard resin characteristic peak sequence, and use the Hausdorff distance as the peak position deviation value. Calculate the ratio sequence of the peak height values ​​of all characteristic peaks in the resin characteristic peak sequence to the peak height values ​​of the corresponding characteristic peaks in the standard resin characteristic peak sequence. Extract the median of the ratio sequence as the peak height deviation value, and use the product of the peak position deviation value and the peak height deviation value as the peak deviation marker.

[0010] As a further aspect of the present invention: In step S5, the process of comparing the pigment dispersion value with the standard dispersion range to generate a deviation marker is as follows: Obtain the standard dispersion value recorded when the paint batch code is received. Retrieve the pigment dispersion re-inspection values ​​at different time points after the same paint batch code is received from the paint management system. Combine the pigment dispersion values ​​retrieved in S2 with each pigment dispersion re-inspection value to form a numerical sequence. Calculate the difference between every two adjacent values ​​in the numerical sequence. The number of times the positive and negative signs of the difference are flipped is taken as the stability number value of the pigment dispersion. The absolute value of the difference between the pigment dispersion value retrieved in S2 and the standard dispersion value is taken as the variation range value. The product of the variation range value and the stability number value is taken as the degree deviation mark.

[0011] As a further aspect of the present invention: In step S5, the process of correcting the original curing oven temperature curve to a corrected curing oven temperature curve based on the peak deviation marker and the degree deviation marker is as follows: Retrieve the original curing oven temperature curve corresponding to the production instruction number from the curing oven temperature curve recorder. The original curing oven temperature curve consists of temperature values ​​arranged in chronological order. Retrieve the standard curing temperature curve of the same coating type as the coating batch code from the coating process database. The standard curing temperature curve consists of temperature values ​​arranged at the same time points as the original curing oven temperature curve. Calculate the temperature difference between the original curing oven temperature curve and the standard curing temperature curve at the same time point to obtain a temperature difference value sequence. Multiply each temperature difference value in the temperature difference value sequence with the peak deviation mark and the degree deviation mark to obtain the first adjustment value sequence and the second adjustment value sequence. The first adjustment sequence and the second adjustment sequence are added together at the same time point to obtain the comprehensive adjustment sequence. The temperature value at each time point in the original curing oven temperature curve is added together with the comprehensive adjustment value at the corresponding time point in the comprehensive adjustment sequence to generate the corrected curing oven temperature curve.

[0012] As a further aspect of the present invention: in step S6, the process of outputting the curing traceability map of the painted product with peak deviation markers and degree deviation markers is as follows: The root node, curing oven number node, resin characteristic peak sequence node, and pigment dispersion value node in the curing traceability chain are converted into circular vertices in the graph. The serial relationship between nodes is converted into directed lines between circular vertices. The radius scaling ratio of the circular vertex corresponding to the resin characteristic peak sequence node is set according to the value of the peak deviation mark, and the radius scaling ratio of the circular vertex corresponding to the pigment dispersion value node is set according to the value of the dispersion mark. The line width of the directed lines is set to a fixed value. The corrected curing oven temperature curve is plotted as a line graph above the circular vertices and the directed lines. The horizontal time axis of the line graph extends from the production time corresponding to the root node to the re-inspection time corresponding to the pigment dispersion value node.

[0013] As a further aspect of the present invention: In step S6, the visual presentation process of the curing traceability map of the painted product, which includes the curing oven number, paint batch code, and corrected curing oven temperature curve, is as follows: Extract the temperature value corresponding to each time point in the temperature curve of the corrected curing oven, and map each time point to the horizontal coordinate axis in the circular vertex and directed line layout. The time range of the horizontal coordinate axis extends from the time point corresponding to the root node to the time point corresponding to the pigment dispersion value node. Convert the temperature value of each time point into a height value on the vertical coordinate axis, and connect the height values ​​of adjacent time points to form a polyline segment. Identify the inflection points in the polyline segment where the angle between adjacent line segments is less than a set angle threshold, and add a data point marker at the position corresponding to the inflection point. The data point marker contains the timestamp and temperature value at the inflection point, and the data point marker uses the same circular vertex fill pattern as the node corresponding to the time point of the inflection point.

[0014] The beneficial effects of this invention are: This invention establishes a structured correlation path between the original coating characteristic data and the curing process data by constructing a curing traceability chain with the production instruction number as the root node. This chain connects the resin characteristic peak sequence and pigment dispersion value of the coating upon warehousing to the curing oven number, creating a chain node that links them together. When the coating film adhesion is found to be below the standard threshold during outbound inspection, the production instruction number corresponding to the abnormal product is precisely located from the production execution system through dual matching of the RFID tag number and the inlet / outlet reading time points of the curing oven. The resin characteristic peak sequence and pigment dispersion value of that batch of coating are then retrieved from the curing traceability chain. The peak position deviation value is obtained by calculating the Hausdorff distance between the resin characteristic peak sequence and the standard resin characteristic peak library. A peak deviation marker is generated by combining this with the median peak height ratio. A degree deviation marker is generated by multiplying the pigment dispersion value by the variation amplitude and the number of times the value stabilized at each time point after warehousing. This achieves a quantitative expression of the degree of deviation in coating characteristics. Peak deviation and degree deviation markers are applied to the temperature difference sequence between the original curing oven temperature curve and the standard curing temperature curve, respectively. A corrected curing oven temperature curve is generated through multiplication and addition operations, allowing the corrected temperature curve to compensate for deviations in the actual characteristics of the current coating batch. A curing traceability map with peak and degree deviation markers is output, presenting the node relationships of the curing traceability chain with circular vertices and directed lines. The comparison between the corrected curve and the original curve is displayed as a superimposed line graph, with data point markers added at inflection points, achieving a complete traceability closed loop from quantitative analysis of coating characteristics to correction of curing process parameters. Attached Figure Description

[0015] The invention will now be further described with reference to the accompanying drawings.

[0016] Figure 1 This is a flowchart illustrating the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figure 1 As shown, this invention is a method for full-process traceability management of painted products based on big data, including the following steps: S1. Enter the production instruction number for the painted product. The production instruction number includes the paint batch code and curing oven number associated with the entire process from primer spraying to topcoat curing. The production instruction number is generated by the production execution system when the primer spraying begins on the painted product. The paint batch code is provided by the paint supplier and entered into the paint management system at the time of shipment. The curing oven number is determined by the equipment code of the curing oven temperature profile recorder.

[0019] S2. Retrieve the resin characteristic peak sequence corresponding to the paint batch code based on the production instruction number, and retrieve the pigment dispersion value corresponding to the same paint batch code. The resin characteristic peak sequence is collected by a spectrometer when the paint is received into the warehouse. The position and height of the characteristic peaks are obtained by scanning the resin components of the paint sample. The pigment dispersion value is collected by a pigment dispersion detector when the paint is received into the warehouse. The uniformity of pigment particle distribution in the paint is measured by laser scattering method.

[0020] S3. Construct a curing traceability chain with the production instruction number as the root node, and connect the curing oven number, resin characteristic peak sequence, and pigment dispersion value as chain nodes to the root node. The curing traceability chain uses a tree data structure for storage. The root node stores the production instruction number and the corresponding production time; the curing oven number node stores the curing oven cavity number and the corresponding curing start time and curing end time; the resin characteristic peak sequence node stores the characteristic peak position array and characteristic peak height array output by the spectrometer; and the pigment dispersion value node stores the dispersion value output by the pigment dispersion detector. The nodes are connected in series through pointers.

[0021] S4. When the paint film adhesion of a painted product is found to be lower than the standard threshold during the outbound inspection, the corresponding production instruction number is extracted, and the resin characteristic peak sequence and pigment dispersion value are located from the curing traceability chain. The paint film adhesion test is conducted using the cross-cut test or pull-out test. The test value is output by the adhesion tester. When it is lower than the standard threshold, the extraction operation is triggered. The entry and exit reading time points of the painted product when it passes through the curing oven are read through the RFID tag. The production instruction number with the same curing oven number and the painting completion time point falling between the entry and exit reading time points is selected from the production execution system as the extraction result. Based on the extraction result, the corresponding resin characteristic peak sequence node and pigment dispersion value node are traversed from the curing traceability chain.

[0022] S5. The resin characteristic peak sequence is compared with the standard resin characteristic peak library to generate peak deviation markers. The pigment dispersion value is compared with the standard dispersion interval to generate degree deviation markers. Based on the peak deviation markers and degree deviation markers, the original curing oven temperature curve is corrected to a corrected curing oven temperature curve. The standard resin characteristic peak library stores the standard characteristic peak positions and heights corresponding to different coating types. The peak position deviation value is obtained by calculating the waveform deviation between the resin characteristic peak sequence and the standard resin characteristic peak sequence using Hausdorff distance. The peak height deviation value is calculated by using the median of the characteristic peak height ratio. The product of the two is used as the peak deviation marker. The standard dispersion interval stores the upper and lower limits of standard dispersion corresponding to different coating types. The variation amplitude value is obtained by calculating the absolute value of the difference between the pigment dispersion value and the standard dispersion value. The stability frequency value is obtained by calculating the number of times the sign of the difference between the re-examination values ​​at each time point after entering the library. The product of the variation amplitude value and the stability frequency value is used as the degree deviation marker. The original curing oven temperature curve is collected sequentially by the curing oven temperature curve recorder during the curing process. The corrected curing oven temperature curve is generated by multiplying the temperature difference between the original curing oven temperature curve and the standard curing temperature curve at the same time point by the peak deviation mark and the degree deviation mark, respectively, and then adding them to the temperature value of the corresponding time point of the original curing oven temperature curve.

[0023] S6. Output a curing traceability map of the coated product with peak deviation and degree deviation markers. The curing traceability map includes the curing oven number, paint batch code, and corrected curing oven temperature curve. The curing traceability map is presented through a graphical interface. The root node, curing oven number node, resin characteristic peak sequence node, and pigment dispersion value node in the curing traceability chain are converted into circular vertices. The serial relationship between nodes is converted into directed lines between circular vertices. The corrected curing oven temperature curve is plotted as a line graph overlaid on the circular vertices and directed lines. Peak deviation and degree deviation markers are added to data points in the line graph that differ from the original curing oven temperature curve.

[0024] In a preferred embodiment of the present invention, in step S3, the process of constructing a curing traceability chain with the production instruction number as the root node, and connecting the curing oven number, resin characteristic peak sequence, and pigment dispersion value as chain nodes to the root node is as follows: When generating production instruction numbers, the production execution system embeds the paint warehousing date information into the number field. This date is recorded by the paint management system when the paint completes its warehousing inspection. During paint warehousing inspection, the spectrometer collects a sequence of resin characteristic peaks. This sequence contains multiple characteristic peaks, each composed of two parameters: peak position and peak height. The peak position indicates the location of the absorption peak at a specific wavenumber, and the peak height indicates the intensity of that absorption peak. The spectrometer stores the resin characteristic peak sequences from each test in the historical test record in chronological order. After the paint is stored, the manufacturer will re-inspect the same batch of paint multiple times at set time intervals during storage. During each re-inspection, the spectrometer re-collects the resin characteristic peak sequence and appends the re-inspection results to the historical test record.

[0025] Taking paint batch code PC-20231015 as an example, the resin characteristic peak sequence collected upon warehousing of this batch of paint contained three characteristic peaks: characteristic peak A, characteristic peak B, and characteristic peak C, with peak position values ​​of 1710, 1735, and 1760, and peak height values ​​of 0.85, 0.92, and 0.78, respectively. Upon re-inspection on the 3rd day after warehousing, the peak position values ​​of the three characteristic peaks changed to 1712, 1738, and 1763, and the peak height values ​​changed to 0.82, 0.89, and 0.75. Upon re-inspection on the 7th day after warehousing, the peak position values ​​of the three characteristic peaks changed to 1715, 1742, and 1767, and the peak height values ​​changed to 0.79, 0.85, and 0.71. Upon re-inspection on the 10th day after warehousing, the peak position values ​​of the three characteristic peaks changed to 1718, 1746, and 1771, and the peak height values ​​changed to 0.76, 0.81, and 0.67. Extract the coating's warehousing date from the production instruction number; this date corresponds to the inspection time upon warehousing. From the historical testing records of the spectrometer, filter out resin characteristic peak sequences for the same coating batch code, PC-20231015, at multiple time points prior to warehousing. These prior-warehousing testing records include supplier factory inspection records and third-party sampling inspection records before entry into the factory; these records are also stored in the historical testing records in chronological order. Arrange the multiple resin characteristic peak sequences prior to warehousing with the resin characteristic peak sequence at the time of warehousing in chronological order to form a time series group of resin characteristic peak sequences.

[0026] Characteristic peaks exhibiting a unidirectional, continuous drift in peak position over time are identified from time series data sets as dynamic characteristic peaks. Taking characteristic peak A as an example, its peak position value is 1705 during factory inspection, 1710 during warehousing inspection, 1712 on the 3rd day after warehousing, 1715 on the 7th day after warehousing, and 1718 on the 10th day after warehousing. The peak position value continuously increases over time, exhibiting a unidirectional, continuous drift trend. Characteristic peaks B and C also show a continuously increasing peak position value trend. For characteristic peaks exhibiting unidirectional, continuous drift, the drift rate and drift direction values ​​are calculated. The drift rate value is obtained by dividing the change in peak position by the corresponding change over time. Taking characteristic peak A as an example, from warehousing inspection to the 10th day after warehousing, the peak position value changes from 1710 to 1718, a change of 8, with a time span of 10 days. The drift rate value is 0.8 per day, and the drift direction value is positive, indicating that the peak position moves towards higher wavenumbers. The drift rate value and drift direction value are used as sub-nodes of the resin characteristic peak sequence and concatenated to the resin characteristic peak sequence node. This allows the resin characteristic peak sequence node to store the original characteristic peak data while also carrying dynamic characteristic information that characterizes the changing trend of the resin composition of this batch of coatings.

[0027] A pigment dispersion analyzer collects pigment dispersion values ​​during the initial inspection of paint upon warehousing. This value indicates the uniformity of pigment particle distribution in the paint; a higher value indicates more uniform dispersion, and a lower value indicates poorer dispersion. After the paint is stored, the manufacturer conducts multiple re-inspections of the same batch of paint at set time intervals during storage. Each time a re-inspection is conducted, the pigment dispersion analyzer re-collects the pigment dispersion value, and the re-inspection results are stored in the historical inspection record. Taking paint batch code PC-20231015 as an example, the pigment dispersion value collected upon warehousing is 92.5; the re-inspection on the 3rd day after warehousing is 91.8; the re-inspection on the 7th day after warehousing is 90.6; the re-inspection on the 10th day after warehousing is 89.2; and the re-inspection on the 14th day after warehousing is 87.5. Pigment dispersion values ​​for the same paint batch code PC-20231015 were selected from historical test records of the pigment dispersion analyzer at multiple time points before warehousing. These pre-warehousing test records included supplier factory inspection records and third-party sampling inspection records. The multiple pre-warehousing pigment dispersion values ​​were then arranged chronologically with the pigment dispersion values ​​at the time of warehousing, forming a time series of pigment dispersion values.

[0028] Calculate the differences between adjacent time points in the time series. The difference between the inspection upon entry into storage and the 3rd day after entry is 91.8 minus 92.5, resulting in -0.7. The difference between the 3rd day and the 7th day after entry is 90.6 minus 91.8, resulting in -1.2. The difference between the 7th day and the 10th day after entry is 89.2 minus 90.6, resulting in -1.4. The difference between the 10th day and the 14th day after entry is 87.5 minus 89.2, resulting in -1.7. These differences are all negative, indicating that the pigment dispersion value continuously decreases over time. The time segment with the most consecutive occurrences of the same sign in the differences is taken as the trend segment of pigment dispersion. In this example, all differences are negative, and there are 4 consecutive occurrences of the same sign; therefore, the time period from the inspection upon entry into storage to the 14th day after entry is taken as the trend segment. Calculate the difference between the starting and ending values ​​of the trend segment. The starting value is 92.5 when the product is received into the warehouse, and the ending value is 87.5 on the 14th day after receiving the product. The difference is -5.0, and the absolute value of 5.0 is taken as the difference of the trend segment. This difference is used as a child node of the pigment dispersion value and concatenated after the pigment dispersion value node. This allows the pigment dispersion value node to store the original dispersion data while also carrying information representing the trend of pigment dispersion change in this batch of coatings.

[0029] In another preferred embodiment of the present invention, the process of extracting the corresponding production instruction number and locating the resin characteristic peak sequence and pigment dispersion value from the curing traceability chain in step S4 is as follows: After the topcoat has cured, painted products enter the outgoing inspection stage. Inspectors use either the cross-cut adhesion test or the pull-off test to test the adhesion of the paint film on the product surface. The cross-cut adhesion test uses a cross-cutting tool to create a grid on the paint film surface. Adhesive tape is applied to the grid area and then peeled off. The area of ​​paint film peeling off is observed, and the adhesion level is assessed according to a standard grading table. If the adhesion level is below the standard threshold, the product is considered unqualified. The pull-off test uses a pull-off adhesion tester. A test column is attached to the paint film surface, and a vertical pull is applied until the paint film separates from the substrate. The pull force value at separation is read as the adhesion test value. If the test value is below the standard threshold, the product is considered unqualified. Before the primer coating process begins, each painted product is affixed with an RFID tag. The RFID tag stores a unique tag number for that painted product. At the start of the primer coating process, the Production Execution System (MES) binds the RFID tag number to the production instruction number and stores the binding relationship in the MES product traceability database.

[0030] RFID readers are installed at the entrance and exit of the curing oven, with their antennas covering the entrance and exit channels of the oven cavity. When painted parts enter the curing oven via the conveyor belt, the RFID reader at the entrance reads the RFID tag number on the surface of the part and records the reading time as the entrance reading time point. The production execution system associates and stores the entrance reading time point with the RFID tag number. When the painted parts leave the curing oven via the conveyor belt, the RFID reader at the exit reads the same RFID tag number and records the reading time as the exit reading time point. The production execution system also associates and stores the exit reading time point with the RFID tag number. For example, with RFID-8972, the RFID reader at the entrance reads the tag number at 09:32:17, and the RFID reader at the exit reads the tag number at 10:05:43.

[0031] When the paint adhesion test value is lower than the standard threshold, the inspector obtains the RFID tag number RFID-8972 affixed to the surface of the painted product using a scanning device. The production instruction number bound to the RFID tag number RFID-8972 is retrieved from the product traceability database of the production execution system. Simultaneously, the entry read time (09:32:17) and exit read time (10:05:43) associated with this RFID tag number are retrieved. All production instruction numbers matching the curing oven number included in the production instruction number input in step S1 are selected from the production execution system; the curing oven number is F-03. The painting completion time and paint batch code corresponding to each production instruction number are extracted from the selected production instruction numbers. The painting completion time is recorded by the topcoat spraying robot when completing the topcoat spraying and stored in the production execution system.

[0032] Taking the selected production instruction numbers PO-241015-001, PO-241015-002, and PO-241015-003 as examples, the curing oven number corresponding to all three production instruction numbers is F-03. The painting completion time for PO-241015-001 is 09:28:05, and the paint batch code is PC-20231015. The painting completion time for PO-241015-002 is 09:35:22, and the paint batch code is PC-20231015. The painting completion time for PO-241015-003 is 09:41:48, and the paint batch code is PC-20231018. Production instruction numbers whose painting completion time is 10:05:43 seconds earlier than the exit reading time and whose time difference between the painting completion time and the entry reading time 09:32:17 is within a set range are included in the candidate set. The set range is predetermined based on the conveyor chain speed and curing chamber length of the curing oven. For curing oven F-03, the set range is a time difference between the painting completion time and the entry reading time between 3 minutes and 8 minutes. For PO-241015-001, the time difference between the painting completion time 09:28:05 and the entry reading time 09:32:17 is -4 minutes and 12 seconds. Since the painting completion time is earlier than the entry reading time, it does not meet the condition of a time difference between 3 minutes and 8 minutes, and therefore is not included in the candidate set. The time difference between the painting completion time (09:35:22) and the entry reading time (09:32:17) of PO-241015-002 is 3 minutes and 5 seconds, which is within the range of 3 to 8 minutes. Furthermore, the painting completion time is earlier than the exit reading time (10:05:43), therefore it is included in the candidate set. The time difference between the painting completion time (09:41:48) and the entry reading time (09:32:17) of PO-241015-003 is 9 minutes and 31 seconds, exceeding the upper limit of the set range of 8 minutes, therefore it is not included in the candidate set.

[0033] The candidate set contains one production instruction number, PO-241015-002, which is selected as the extraction result. If the candidate set contains multiple production instruction numbers, for example, if both PO-241015-002 and PO-241015-003 meet the time difference condition, then the production instruction number whose paint batch code matches the paint batch code carried in the RFID tag number is selected as the extraction result. The paint batch code bound to the RFID tag number RFID-8972 in the production execution system is PC-20231015; therefore, the production instruction number PO-241015-002 with paint batch code PC-20231015 is selected as the extraction result from the candidate set. After extracting the production instruction number, the corresponding resin characteristic peak sequence node and pigment dispersion value node are located in the solidification traceability chain constructed in step S3 according to the production instruction number. The drift rate value and drift direction value of the resin characteristic peak sequence and its sub-nodes stored under the node are obtained, and the change trend segment value of the pigment dispersion value and its sub-nodes is obtained, which are used for deviation mark generation and temperature curve correction in subsequent steps.

[0034] In another preferred embodiment of the present invention, the process of comparing the resin characteristic peak sequence with the standard resin characteristic peak library to generate peak deviation markers in step S5 is as follows: The standard resin characteristic peak library is pre-stored in the coating process database. This database stores the standard resin characteristic peak sequences corresponding to different coatings, categorized by coating type. The standard resin characteristic peak sequence for each coating type is determined by the coating supplier during the coating R&D stage by averaging multiple samples taken using a spectrometer. This sequence includes the number of characteristic peaks, the reference values ​​for the peak position and peak height of each characteristic peak. Taking acrylic polyurethane paint as an example, the standard resin characteristic peak sequence for this coating type includes three characteristic peaks: P1, P2, and P3. The reference value for the peak position of characteristic peak P1 is 1710, and its peak height is 0.85. The reference value for the peak position of characteristic peak P2 is 1735, and its peak height is 0.92. The reference value for the peak position of characteristic peak P3 is 1760, and its peak height is 0.78.

[0035] The resin characteristic peak sequence retrieved in step S2 is the actual test data corresponding to the paint batch code PC-20231015. This sequence contains three characteristic peaks: peak A has a peak position value of 1718 and a peak height value of 0.76; peak B has a peak position value of 1746 and a peak height value of 0.81; and peak C has a peak position value of 1771 and a peak height value of 0.67. The actual resin characteristic peak sequence and the standard resin characteristic peak sequence are converted into two-dimensional point sets composed of peak position and peak height values, respectively. For the standard resin characteristic peak sequence, the three characteristic peaks are converted into three two-dimensional points: point 1 has coordinates of 1710 and 0.85; point 2 has coordinates of 1735 and 0.92; and point 3 has coordinates of 1760 and 0.78. For the actual resin characteristic peak sequence, the three characteristic peaks are converted into three two-dimensional points: point A has coordinates of 1718 and 0.76; point B has coordinates of 1746 and 0.81; and point C has coordinates of 1771 and 0.67.

[0036] Calculate the Hausdorff distance between the two-dimensional point set of the actual resin characteristic peak sequence and the two-dimensional point set of the standard resin characteristic peak sequence. The calculation process for the Hausdorff distance is as follows: For each point in the actual point set, calculate the Euclidean distance from that point to all points in the standard point set, and take the minimum value as the one-way distance from that point to the standard point set; take the maximum value among all the one-way distances corresponding to the actual point as the Hausdorff distance from the actual point set to the standard point set. Taking the actual point A with coordinates 1718 and 0.76 as an example, the Euclidean distance from it to standard point 1 is calculated as √(1718 - 1710²) + 0.76 - 0.85², which is √(64² + 0.0081), approximately equal to 8.00. The Euclidean distance from it to standard point 2 is calculated as √(1718 - 1735²) + 0.76 - 0.92², which is √(289² + 0.0256), approximately equal to 17.00. The Euclidean distance from point A to standard point 3 is calculated as √(1718 - 1760²) + 0.76 - 0.78², which equals √(1764 + 0.0004), approximately 42.00. The minimum Euclidean distance from actual point A to the standard point set is 8.00. The coordinates of actual point B are 1746, 0.81. Its distance to standard point 1 is calculated as √(1746 - 1710²) + 0.81 - 0.85², which equals √(1296 + 0.0016), approximately 36.00. Its distance to standard point 2 is calculated as √(1746 - 1735²) + 0.81 - 0.92², which equals √(121 + 0.0121), approximately 11.00. The distance from point B to standard point 3 is √(1746 - 1760)² + 0.81 - 0.78², which is √(196 + 0.0009), approximately 14.00. The minimum Euclidean distance from actual point B to the standard point set is 11.00. The coordinates of actual point C are 1771, 0.67. The distance from actual point C to standard point 1 is √(1771 - 1710)² + 0.67 - 0.85², which is √(3721 + 0.0324), approximately 61.00. The distance from actual point C to standard point 2 is √(1771 - 1735)² + 0.67 - 0.92², which is √(1296 + 0.0625), approximately 36.00. The distance to the standard point 3 is the square root of 1771 minus the square of 1760 plus the square root of 0.67 minus the square of 0.78, which is the square root of 121 plus 0.0121, approximately equal to 11.00. The minimum Euclidean distance from the actual point C to the standard point set is 11.00. Taking the maximum value of the three minimum distances (8.00, 11.00, and 11.00), 11.00, as the Hausdorff distance, this value is used as the peak position deviation.

[0037] This involves calculating the ratio of the peak height values ​​of all characteristic peaks in the resin characteristic peak sequence to the peak height values ​​of the corresponding characteristic peaks in the standard resin characteristic peak sequence. The determination of the corresponding positions is based on the matching of peak position values. The actual characteristic peaks are arranged in ascending order of their peak position values, and the standard characteristic peaks are also arranged in ascending order of their peak position values, ensuring a one-to-one correspondence. In the actual resin characteristic peak sequence, characteristic peaks A, B, and C, arranged in ascending order of their peak position values, are: A (peak position value 1718), B (peak position value 1746), and C (peak position value 1771). Similarly, in the standard resin characteristic peak sequence, characteristic peaks P1, P2, and P3, arranged in ascending order of their peak position values, are: P1 (peak position value 1710), P2 (peak position value 1735), and P3 (peak position value 1760). Therefore, characteristic peak A corresponds to characteristic peak P1, characteristic peak B corresponds to characteristic peak P2, and characteristic peak C corresponds to characteristic peak P3. The peak height of characteristic peak A is 0.76, and the peak height of characteristic peak P1 is 0.85. The ratio 0.76 divided by 0.85 equals 0.894. The peak height of characteristic peak B is 0.81, and the peak height of characteristic peak P2 is 0.92. The ratio 0.81 divided by 0.92 equals 0.880. The peak height of characteristic peak C is 0.67, and the peak height of characteristic peak P3 is 0.78. The ratio 0.67 divided by 0.78 equals 0.859. The ratio sequence is 0.894, 0.880, and 0.859. The median of this sequence is extracted, and the three values ​​are arranged in ascending order as 0.859, 0.880, and 0.894. The median, 0.880, is located in the middle and is used as the peak height deviation value. Multiply the peak position deviation value of 11.00 by the peak height deviation value of 0.880, the product is 9.68, and this value is used as the peak deviation mark.

[0038] In another preferred embodiment of the present invention, the process of comparing the pigment dispersion value with the standard dispersion range to generate a deviation marker in step S5 is as follows: The standard dispersion range is pre-stored in the coating process database, which stores the standard dispersion values ​​and allowable fluctuation ranges for different coatings according to coating type. Taking acrylic polyurethane paint as an example, the standard dispersion value is 90.0, and the standard dispersion range is 85.0 to 95.0. The pigment dispersion value retrieved in step S2 is 92.5, collected when coating batch code PC-20231015 was received into the warehouse. This value falls within the standard dispersion range of 85.0 to 95.0.

[0039] Retrieve the pigment dispersion retest values ​​for the same paint batch code PC-20231015 at different time points after warehousing from the paint management system. After the paint is put into storage, the quality inspection department conducts retests on the 3rd, 7th, 10th, and 14th days after warehousing, with retest values ​​of 91.8, 90.6, 89.2, and 87.5 respectively. Combine the pigment dispersion value of 92.5 retrieved in step S2 at the time of warehousing with each pigment dispersion retest value in chronological order to form a numerical sequence: 92.5 at the time of warehousing, 91.8 on the 3rd day after warehousing, 90.6 on the 7th day after warehousing, 89.2 on the 10th day after warehousing, and 87.5 on the 14th day after warehousing.

[0040] Calculate the difference between any two adjacent values ​​in the numerical sequence. The difference between the initial storage date and the 3rd day after storage is 91.8 minus 92.5, which equals -0.7. The difference between the 3rd day and the 7th day after storage is 90.6 minus 91.8, which equals -1.2. The difference between the 7th day and the 10th day after storage is 89.2 minus 90.6, which equals -1.4. The difference between the 10th day and the 14th day after storage is 87.5 minus 89.2, which equals -1.7. All four differences are negative, the signs are not reversed, and the number of sign reversals is 0. This value is taken as the stability index of pigment dispersion. The smaller the stability index, the more uniform the trend of pigment dispersion during storage, i.e., a continuous decrease or increase. The larger the stability index, the more repeated fluctuations the pigment dispersion exhibits during storage.

[0041] The absolute value of the difference between the pigment dispersion value of 92.5 obtained in step S2 and the standard dispersion value of 90.0 is taken as the variation range value. The variation range value is 92.5 minus 90.0, which equals 2.5. The variation range value of 2.5 is multiplied by the stability count value of 0, and the product is 0. This value is taken as the degree deviation mark. If another batch of paint shows fluctuations during the re-inspection after warehousing, for example, it is 91.0 when it is put into storage, 90.2 on the 3rd day after storage, 91.5 on the 7th day after storage, 90.8 on the 10th day after storage, and 89.5 on the 14th day after storage, with adjacent differences of -0.8, +1.3, -0.7, and -1.3, and the sign changes from negative to positive and then back to negative, with a reversal count of 2, the variation range value is 91.0 minus 90.0, which equals 1.0. The degree deviation mark is 1.0 multiplied by 2, which equals 2.0.

[0042] In another preferred embodiment of the present invention, in step S5, the process of correcting the original curing oven temperature curve to a corrected curing oven temperature curve based on the peak deviation marker and the degree deviation marker is as follows: Retrieve the original curing oven temperature curve corresponding to production instruction number PO-241015-002 from the curing oven temperature curve recorder. The curing oven temperature curve recorder collects the oven temperature value every 10 seconds as the painted product passes through the curing oven. Data collection begins at 09:32:17 when the painted product enters the curing oven and ends at 10:05:43 when the painted product leaves the curing oven, collecting a total of 201 temperature values, forming a chronological temperature value sequence. Taking the heating stage of curing oven F-03 as an example, the temperature values ​​of the original curing oven temperature curve at the first 30 time points during the heating stage are 85.2, 88.5, 92.1, 96.0, 100.2, 104.5, 108.9, 113.2, 117.5, 121.8, 126.0, 130.2, 134.3, 138.5, 142.6, 146.7, 150.8, 154.9, 158.9, 162.9, 166.9, 170.9, 174.8, 178.7, 182.6, 186.4, 190.2, 194.0, 197.8, and 201.5.

[0043] Retrieve the standard curing temperature profile of acrylic polyurethane paint of the same type as paint batch code PC-20231015 from the paint process database. The standard curing temperature profile is determined by the paint supplier through process experiments during the paint R&D stage and stored in the paint process database. This profile contains temperature values ​​arranged at the same time points as the original curing oven temperature profile, with the time points also spaced 10 seconds apart, and the time range is consistent with the time range of the original curing oven temperature profile. The standard curing temperature profile shows the following temperature values ​​at the first 30 time points during the heating phase: 86.0, 89.5, 93.0, 97.0, 101.0, 105.0, 109.0, 113.0, 117.0, 121.0, 125.0, 129.0, 133.0, 137.0, 141.0, 145.0, 149.0, 153.0, 157.0, 161.0, 165.0, 169.0, 173.0, 177.0, 181.0, 185.0, 189.0, 193.0, 197.0, and 201.0.

[0044] The temperature difference sequence is obtained by calculating the temperature difference between the original curing oven temperature curve and the standard curing temperature curve at the same time point. Taking the first 30 time points of the heating stage as an example, the temperature difference at the first time point is 85.2 minus 86.0, which equals -0.8; the temperature difference at the second time point is 88.5 minus 89.5, which equals -1.0; the temperature difference at the third time point is 92.1 minus 93.0, which equals -0.9; the temperature difference at the fourth time point is 96.0 minus 97.0, which equals -1.0; the temperature difference at the fifth time point is 100.2 minus 101.0, which equals -0.8; and the temperature difference at the sixth time point is 1. 04.5 minus 105.0 equals -0.5. The temperature difference at the seventh time point is 108.9 minus 109.0 equals -0.1. The temperature difference at the eighth time point is 113.2 minus 113.0 equals 0.2. The temperature difference at the ninth time point is 117.5 minus 117.0 equals 0.5. The temperature difference at the tenth time point is 121.8 minus 121.0 equals 0.8. The temperature differences at subsequent time points are calculated in this manner, forming a temperature difference sequence containing 201 temperature differences.

[0045] Each temperature difference in the temperature difference sequence is multiplied by the peak deviation marker and the degree deviation marker respectively to obtain the first adjustment sequence and the second adjustment sequence. The peak deviation marker is 9.68, and the degree deviation marker is 0. Taking the first ten time points of the warming phase as an example, the first adjustment in the first adjustment sequence is -0.8 multiplied by 9.68, which equals -7.744; the second adjustment is -1.0 multiplied by 9.68, which equals -9.68; the third adjustment is -0.9 multiplied by 9.68, which equals -8.712; the fourth adjustment is -1.0 multiplied by 9.68, which equals -9.68; the fifth adjustment is -0.8 multiplied by 9.68, which equals -7.744; the sixth adjustment is -0.5 multiplied by 9.68, which equals -4.84; the seventh adjustment is -0.1 multiplied by 9.68, which equals -0.968; the eighth adjustment is 0.2 multiplied by 9.68, which equals 1.936; the ninth adjustment is 0.5 multiplied by 9.68, which equals 4.84; and the tenth adjustment is 0.8 multiplied by 9.68, which equals 7.744. All adjustments in the second adjustment sequence are 0 because the degree deviation is marked as 0.

[0046] The combined adjustment sequence is obtained by adding the adjustments of the first adjustment sequence and the second adjustment sequence at the same time point. Since all values ​​in the second adjustment sequence are 0, the combined adjustment sequence is the same as the first adjustment sequence. The combined adjustments for the first ten time points are -7.744, -9.68, -8.712, -9.68, -7.744, -4.84, -0.968, 1.936, 4.84, and 7.744, respectively.

[0047] The corrected curing oven temperature curve is generated by adding the temperature value at each time point in the original curing oven temperature curve to the corresponding comprehensive adjustment value in the comprehensive adjustment sequence. Taking the first ten time points of the heating stage as an example, the corrected temperature at the first time point is 85.2 + -7.744 = 77.456; the corrected temperature at the second time point is 88.5 + -9.68 = 78.82; the corrected temperature at the third time point is 92.1 + -8.712 = 83.388; the corrected temperature at the fourth time point is 96.0 + -9.68 = 86.32; and the corrected temperature at the fifth time point is 100.2 + -7.744 = 92. The corrected temperature for the sixth time point is 104.5 + -4.84 = 99.66; for the seventh time point, it is 108.9 + -0.968 = 107.932; for the eighth time point, it is 113.2 + 1.936 = 115.136; for the ninth time point, it is 117.5 + 4.84 = 122.34; and for the tenth time point, it is 121.8 + 7.744 = 129.544. The temperature values ​​for subsequent time points are calculated sequentially in the same way to form a complete corrected curing oven temperature curve.

[0048] In another preferred embodiment of the present invention, the process of outputting the curing traceability map of the painted product with peak deviation markers and degree deviation markers in step S6 is as follows: The graph generation module reads the solidification traceability chain data structure constructed in step S3 from memory. The solidification traceability chain starts with a root node, which stores the production instruction number PO-241015-002 and its associated production time, which is 09:28:05, the start time of the primer spraying process. The root node is connected to the curing oven number node via a pointer. This node stores the curing oven number F-03 and the entry and exit times of the coated part (09:32:17 and 10:05:43 respectively). The curing oven number node is connected to the resin characteristic peak sequence node via a pointer. This node stores the peak position and height values ​​of characteristic peaks A, B, and C, as well as the drift rate value (0.8 per day) and the positive drift direction value stored in the child nodes. The resin characteristic peak sequence node is connected to the pigment dispersion value node via a pointer. This node stores the pigment dispersion value (92.5) at the time of entry into the database, and the trend difference value (5.0) stored in the child nodes.

[0049] The four nodes in the solidification traceability chain are converted into circular vertices in the graph. The circular vertex corresponding to the root node is located at the far left of the graph, with its center coordinates set to 100 pixels x, 300 pixels y, and a radius of 30 pixels. The circular vertex corresponding to the curing oven number node is located to the right of the root node, with its center coordinates set to 300 pixels x, 300 pixels y, and a radius of 30 pixels. The circular vertex corresponding to the resin characteristic peak sequence node is located to the right of the curing oven number node. Its center x-coordinate is calculated based on the value of the peak deviation marker. The baseline offset is set to 200 pixels, and the peak deviation marker is 9.68. Multiplying this value by a coefficient of 2 gives 19.36. Adding the baseline offset of 200 pixels to 19.36 pixels gives 219.36 pixels. Therefore, the x-coordinate of the center of the resin characteristic peak sequence node is the curing oven number node's x-coordinate of 300 pixels plus 219.36 pixels, which equals 519.36 pixels. The y-coordinate remains 300 pixels. The radius scaling of this node is also set according to the peak deviation mark. The base radius is set to 30 pixels, and the scaling factor is 1 plus the peak deviation mark divided by 100, that is, 1 plus 9.68 divided by 100 equals 1.0968. After scaling, the radius is 30 pixels multiplied by 1.0968, which equals 32.9 pixels.

[0050] The circular vertex corresponding to the pigment dispersion value node is located to the right of the resin characteristic peak sequence node. The x-coordinate of its center is calculated based on the magnitude of the degree deviation marker. With a degree deviation marker of 0, the offset is the base offset of 200 pixels multiplied by 1, equaling 200 pixels. Therefore, the x-coordinate of the pigment dispersion value node's center is the resin characteristic peak sequence node's x-coordinate of 519.36 pixels plus 200 pixels, equaling 719.36 pixels, while the y-coordinate remains at 300 pixels. The radius scaling of this node is set according to the degree deviation marker; the scaling factor is 1 plus the degree deviation marker divided by 100, i.e., 1 plus 0 divided by 100 equals 1, while the radius remains unchanged at 30 pixels.

[0051] The serial connections between nodes are converted into directed lines between circular vertices. A straight line with an arrow is drawn between the root node's circular vertex and the curing oven number node's circular vertex. The line starts at the right edge of the root node's circular vertex and ends at the left edge of the curing oven number node's circular vertex, with the arrow pointing towards the curing oven number node. The line width is set to 2 pixels. A directed line of the same specification is drawn between the curing oven number node's circular vertex and the resin characteristic peak sequence node's circular vertex, and between the resin characteristic peak sequence node's circular vertex and the pigment dispersion value node's circular vertex.

[0052] The corrected curing oven temperature curve generated in step S5 is plotted as a line graph above the circular vertex and the directed line. The horizontal axis of the line graph extends from the production time corresponding to the root node to the re-inspection time corresponding to the pigment dispersion value node. The production time corresponding to the root node is the start time of the primer spraying process at 09:28:05, and the re-inspection time corresponding to the pigment dispersion value node is the re-inspection time on the 14th day after warehousing, which is recorded in the paint management system as 14:30:00. The range of the horizontal axis is from 09:28:05 to 14:30:00, with a total duration of 5 hours, 1 minute, and 55 seconds, which is equivalent to 18,115 seconds. The width of the graph canvas is set to 1200 pixels. The time axis is mapped to the horizontal axis of the canvas, with a horizontal axis of 0 pixels corresponding to 09:28:05 and a horizontal axis of 1200 pixels corresponding to 14:30:00. The horizontal axis position corresponding to each time point is calculated by linear interpolation.

[0053] In another preferred embodiment of the present invention, in step S6, the visual presentation process of the curing traceability map of the painted product, which includes the curing oven number, paint batch code, and corrected curing oven temperature curve, is as follows: Extract the temperature value corresponding to each time point in the corrected curing oven temperature curve. The corrected curing oven temperature curve contains 201 time points, each corresponding to a sampling time and a temperature value. Taking the first ten time points of the heating stage as an example, the first time point corresponds to the sampling time 09:32:17, with a corrected temperature of 77.456 degrees; the second time point corresponds to the sampling time 09:32:27, with a corrected temperature of 78.82 degrees; the third time point corresponds to the sampling time 09:32:37, with a corrected temperature of 83.388 degrees; the fourth time point corresponds to the sampling time 09:32:47, with a corrected temperature of 86.32 degrees; and the fifth time point corresponds to the sampling time 09:32:57, with a corrected temperature of 92 degrees. The temperature was 456 degrees. The sixth time point corresponds to the acquisition time of 09:33:07, with a corrected temperature of 99.66 degrees. The seventh time point corresponds to the acquisition time of 09:33:17, with a corrected temperature of 107.932 degrees. The eighth time point corresponds to the acquisition time of 09:33:27, with a corrected temperature of 115.136 degrees. The ninth time point corresponds to the acquisition time of 09:33:37, with a corrected temperature of 122.34 degrees. The tenth time point corresponds to the acquisition time of 09:33:47, with a corrected temperature of 129.544 degrees.

[0054] Each time point is mapped to a horizontal coordinate axis in a circular vertex and directed connection layout. The time range of the horizontal coordinate axis extends from the production time (09:28:05) corresponding to the root node to the re-inspection time (14:30:00) corresponding to the pigment dispersion value node. The first time point (09:32:17) is offset by 4 minutes and 12 seconds (252 seconds) from the starting point (09:28:05), with a total duration of 18115 seconds. Its horizontal coordinate position is 252 divided by 18115 multiplied by 1200 pixels, which equals 16.7 pixels. The second time point (09:32:27) is offset by 4 minutes and 22 seconds (262 seconds), with its horizontal coordinate position being 262 divided by 18115 multiplied by 1200 pixels, which equals 17.4 pixels. The horizontal coordinate positions of subsequent time points are calculated in the same way.

[0055] The temperature value at each time point is converted into a height value on the vertical coordinate axis. The temperature range of the vertical coordinate axis is determined based on the minimum and maximum values ​​of the modified curing oven temperature curve. The minimum value is 77.456 degrees Celsius at the beginning of the heating phase, and the maximum value is 205.3 degrees Celsius after the temperature stabilizes during the holding phase. The canvas height is set to 400 pixels. The bottom of the vertical coordinate axis corresponds to the minimum temperature of 77.456 degrees Celsius, and the top corresponds to the maximum temperature of 205.3 degrees Celsius, with a temperature span of 127.844 degrees Celsius. Each pixel corresponds to a temperature value of 0.3196 degrees Celsius. The temperature value of 77.456 degrees Celsius at the first time point corresponds to 0 pixels at the bottom of the vertical coordinate axis. The temperature value of 78.82 degrees Celsius at the second time point minus the minimum value of 77.456 degrees Celsius equals 1.364 degrees Celsius. Dividing this by 0.3196 degrees Celsius gives 4.3 pixels per pixel. Therefore, the height value of the second time point is 4.3 pixels. The height value of the third time point is 83.388 degrees Celsius minus 77.456 degrees Celsius equals 5.932 degrees Celsius. Dividing this by 0.3196 degrees Celsius gives 18.6 pixels. Calculate the height value at each time point in sequence.

[0056] Connect the height values ​​of adjacent time points sequentially to form a broken line segment. The first time point has an x-coordinate of 16.7 pixels and a height of 0 pixels; the second time point has an x-coordinate of 17.4 pixels and a height of 4.3 pixels; the third time point has an x-coordinate of 18.1 pixels and a height of 18.6 pixels; the fourth time point has an x-coordinate of 18.8 pixels and a height of 27.8 pixels; the fifth time point has an x-coordinate of 19.5 pixels and a height of 46.9 pixels; the sixth time point has an x-coordinate of 20.2 pixels and a height of 69.4 pixels; the seventh time point has an x-coordinate of 20.9 pixels and a height of 95.2 pixels; the eighth time point has an x-coordinate of 21.6 pixels and a height of 117.9 pixels; the ninth time point has an x-coordinate of 22.3 pixels and a height of 140.5 pixels; and the tenth time point has an x-coordinate of 23.0 pixels and a height of 162.9 pixels. Connect these points sequentially with straight line segments to form a broken line segment.

[0057] Identify inflection points in a polyline segment where the angle between adjacent segments is less than a set angle threshold. The set angle threshold is 120 degrees. Taking three consecutive time points—the seventh, eighth, and ninth—as an example, the coordinates of the seventh time point are 20.9 pixels and 95.2 pixels, the eighth time point is 21.6 pixels and 117.9 pixels, and the ninth time point is 22.3 pixels and 140.5 pixels. Calculate the vector from the seventh point to the eighth point as 0.7 pixels and 22.7 pixels, and the vector from the eighth point to the ninth point as 0.7 pixels and 22.6 pixels. The dot product of the two vectors is 0.7 multiplied by 0.7 plus 22.7 multiplied by 22.6, which equals 0.49 plus 513.02, which equals 513.51. The magnitude of the first vector is √0.7² + 22.7² = √0.49 + 515.29 = √515.78 = 22.71. The magnitude of the second vector is √0.7² + 22.6² = √0.49 + 510.76 = √511.25 = 22.61. The cosine of the angle is the product of the dot product and the magnitudes, i.e., 513.51 divided by 22.71 multiplied by 22.61 equals 513.51 divided by 513.47, approximately equal to 1.00. The angle is 0 degrees, which is less than the set angle threshold of 120 degrees. Therefore, the eighth time point is identified as an inflection point.

[0058] Data point markers are added at the locations corresponding to the inflection points. The acquisition time corresponding to the eighth time point is 09:33:27, and the corrected temperature is 115.136 degrees Celsius. Circular data point markers are drawn at coordinates of 21.6 pixels and 117.9 pixels at the eighth time point, with a radius of 5 pixels. The pattern filled inside the circle is the same as the circular vertex filling pattern of the node corresponding to the inflection point. The node corresponding to the eighth time point is the resin characteristic peak sequence node. Because this time point falls within the coating characteristic analysis time period represented by the resin characteristic peak sequence node, the circular vertex of the resin characteristic peak sequence node is filled with a dot pattern. Therefore, the data point marker is also filled with a dot pattern. A text label is displayed above the data point marker, with the label content "09:33:27 115.1℃".

[0059] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A method for full-process traceability management of painted parts based on big data, characterized in that, Includes the following steps: S1. Input the production instruction number of the painted product. The production instruction number includes the paint batch code and curing oven number associated with the entire process from primer spraying to topcoat curing. S2. Retrieve the resin characteristic peak sequence corresponding to the paint batch code based on the production instruction number, and retrieve the pigment dispersion value corresponding to the same paint batch code. S3. Construct a curing traceability chain with the production instruction number as the root node, and connect the curing oven number, resin characteristic peak sequence and pigment dispersion value as chain nodes to the root node. S4. When the paint film adhesion is found to be lower than the standard threshold during the outbound inspection of painted products, the corresponding production instruction number is extracted and located from the curing traceability chain to the resin characteristic peak sequence and pigment dispersion value. S5. Compare the resin characteristic peak sequence with the standard resin characteristic peak library to generate peak deviation markers, compare the pigment dispersion value with the standard dispersion range to generate degree deviation markers, and correct the original curing oven temperature curve to a corrected curing oven temperature curve based on the peak deviation markers and degree deviation markers. S6. Output a curing traceability map of the painted parts with peak deviation markers and degree deviation markers. The curing traceability map of the painted parts includes the curing oven number, the paint batch code, and the corrected curing oven temperature curve.

2. The method for full-process traceability management of painted parts based on big data according to claim 1, characterized in that, In step S3, the process of constructing a curing traceability chain with the production instruction number as the root node, and connecting the curing oven number, resin characteristic peak sequence, and pigment dispersion value as chain nodes to the root node, is as follows: Extract the coating entry date carried in the production instruction number, and screen the resin characteristic peak sequences of the same coating batch code at multiple time points before entry from the historical detection records of the spectrometer. After arranging multiple resin characteristic peak sequences in chronological order, identify the characteristic peaks whose peak positions show unidirectional continuous drift over time as dynamic characteristic peaks. Extract the drift rate value and drift direction value of the dynamic characteristic peaks, and concatenate the drift rate value and drift direction value as sub-nodes of the resin characteristic peak sequence after the resin characteristic peak sequence. From the historical test records of the pigment dispersion tester, the pigment dispersion values ​​of the same paint batch code at multiple time points before warehousing are selected. After arranging the multiple pigment dispersion values ​​in chronological order, the difference between the values ​​at adjacent time points is calculated. The sequence segment with the most consecutive occurrences of the same sign in the difference is taken as the change trend segment of pigment dispersion. The difference between the starting value and the ending value of the change trend segment is taken as the child node of the pigment dispersion value and concatenated after the pigment dispersion value.

3. The method for full-process traceability management of painted parts based on big data as described in claim 1, characterized in that, In step S4, the process of extracting the corresponding production instruction number and locating the resin characteristic peak sequence and pigment dispersion value from the curing traceability chain is as follows: Obtain the RFID tag number affixed to the surface of painted products with paint film adhesion below the standard threshold. Based on the RFID tag number, read the entry reading time point from the RFID reader at the curing oven entrance and the exit reading time point from the RFID reader at the curing oven exit. Filter out production instruction numbers from the production execution system that match the curing oven number in S1. Extract the painting completion time point and paint batch code corresponding to each production instruction number from the filtered production instruction numbers. Select production instruction numbers whose painting completion time point is earlier than the exit reading time point and whose time difference between the painting completion time point and the entry reading time point is within a set range as a candidate set. When there is only one production instruction number in the candidate set, that production instruction number is taken as the extraction result. When there are multiple production instruction numbers in the candidate set, the production instruction number whose paint batch code is the same as the paint batch code carried in the RFID tag number is selected as the extraction result.

4. The method for full-process traceability management of painted parts based on big data as described in claim 1, characterized in that, In step S5, the process of comparing the resin characteristic peak sequence with the standard resin characteristic peak library to generate peak deviation markers is as follows: Retrieve the characteristic peak sequence of standard resin of the same coating type from the standard resin characteristic peak library. Convert the resin characteristic peak sequence and the standard resin characteristic peak sequence into two-dimensional point sets composed of peak position values ​​and peak height values, respectively. Calculate the Hausdorff distance between the two-dimensional point sets of the resin characteristic peak sequence and the standard resin characteristic peak sequence, and use the Hausdorff distance as the peak position deviation value. Calculate the ratio sequence of the peak height values ​​of all characteristic peaks in the resin characteristic peak sequence to the peak height values ​​of the corresponding characteristic peaks in the standard resin characteristic peak sequence. Extract the median of the ratio sequence as the peak height deviation value, and use the product of the peak position deviation value and the peak height deviation value as the peak deviation marker.

5. The method for full-process traceability management of painted parts based on big data according to claim 4, characterized in that, In step S5, the process of comparing the pigment dispersion value with the standard dispersion range to generate a deviation marker is as follows: Obtain the standard dispersion value recorded when the paint batch code is received. Retrieve the pigment dispersion re-inspection values ​​at different time points after the same paint batch code is received from the paint management system. Combine the pigment dispersion values ​​retrieved in S2 with each pigment dispersion re-inspection value to form a numerical sequence. Calculate the difference between every two adjacent values ​​in the numerical sequence. The number of times the positive and negative signs of the difference are flipped is taken as the stability number value of the pigment dispersion. The absolute value of the difference between the pigment dispersion value retrieved in S2 and the standard dispersion value is taken as the variation range value. The product of the variation range value and the stability number value is taken as the degree deviation mark.

6. The method for full-process traceability management of painted parts based on big data according to claim 5, characterized in that, In step S5, the process of correcting the original curing oven temperature curve to a corrected curing oven temperature curve based on the peak deviation marker and degree deviation marker is as follows: Retrieve the original curing oven temperature curve corresponding to the production instruction number from the curing oven temperature curve recorder. The original curing oven temperature curve consists of temperature values ​​arranged in chronological order. Retrieve the standard curing temperature curve of the same coating type as the coating batch code from the coating process database. The standard curing temperature curve consists of temperature values ​​arranged at the same time points as the original curing oven temperature curve. Calculate the temperature difference between the original curing oven temperature curve and the standard curing temperature curve at the same time point to obtain a temperature difference value sequence. Multiply each temperature difference value in the temperature difference value sequence with the peak deviation mark and the degree deviation mark to obtain the first adjustment value sequence and the second adjustment value sequence. The first adjustment sequence and the second adjustment sequence are added together at the same time point to obtain the comprehensive adjustment sequence. The temperature value at each time point in the original curing oven temperature curve is added together with the comprehensive adjustment value at the corresponding time point in the comprehensive adjustment sequence to generate the corrected curing oven temperature curve.

7. The method for full-process traceability management of painted parts based on big data according to claim 1, characterized in that, In step S6, the process of outputting the curing traceability map of the painted product with peak deviation markers and degree deviation markers is as follows: The root node, curing oven number node, resin characteristic peak sequence node, and pigment dispersion value node in the curing traceability chain are converted into circular vertices in the graph. The serial relationship between nodes is converted into directed lines between circular vertices. The radius scaling ratio of the circular vertex corresponding to the resin characteristic peak sequence node is set according to the value of the peak deviation mark, and the radius scaling ratio of the circular vertex corresponding to the pigment dispersion value node is set according to the value of the dispersion mark. The line width of the directed lines is set to a fixed value. The corrected curing oven temperature curve is plotted as a line graph above the circular vertices and the directed lines. The horizontal time axis of the line graph extends from the production time corresponding to the root node to the re-inspection time corresponding to the pigment dispersion value node.

8. The method for full-process traceability management of painted parts based on big data according to claim 7, characterized in that, In step S6, the visual presentation process of the curing traceability map of the painted parts, which includes the curing oven number, paint batch code, and corrected curing oven temperature curve, is as follows: Extract the temperature value corresponding to each time point in the temperature curve of the corrected curing oven, and map each time point to the horizontal coordinate axis in the circular vertex and directed line layout. The time range of the horizontal coordinate axis extends from the time point corresponding to the root node to the time point corresponding to the pigment dispersion value node. Convert the temperature value of each time point into a height value on the vertical coordinate axis, and connect the height values ​​of adjacent time points to form a polyline segment. Identify the inflection points in the polyline segment where the angle between adjacent line segments is less than a set angle threshold, and add a data point marker at the position corresponding to the inflection point. The data point marker contains the timestamp and temperature value at the inflection point, and the data point marker uses the same circular vertex fill pattern as the node corresponding to the time point of the inflection point.

Images