A quality management system and method for automotive parts coating

Abstract

This invention discloses a quality management system and method for automotive parts coating. The system includes a coating execution module, a camera module, a smart gateway, a server, a workshop temperature management module, a dust removal module, and a dust emission management module. Based on a matching graph of the relationship between workshop air supply volume and dust purification efficiency per unit time, this invention matches the optimal air supply volume and dust purification efficiency to obtain a first corrected predicted energy consumption value and a second corrected predicted energy consumption value. The working states of the workshop temperature management module and the dust emission management module are set according to the first and second corrected predicted energy consumption values, thereby reducing the overall power consumption of the coating workshop and lowering enterprise production costs while meeting environmental protection requirements.

Description

Technical Field

[0001] This invention relates to a coating production line management system, and more specifically, to a coating quality management system and method for automotive parts. Background Technology

[0002] With the continuous development of the automotive industry, rising labor costs, and increasing environmental awareness, automotive parts coating management systems need to be upgraded. Management is the backbone of the entire production line and has extraordinary significance for quality. To achieve a good coating on parts, coating management is fundamental.

[0003] Patent No. 201611252738.2 discloses a cloud-based intelligent automotive painting system, which includes a data management system, a 3D model database of automotive parts, a painting program database of automotive parts, a point cloud database of painted parts, a quality inspection database, and a painting path planning system for automotive parts, thereby solving the problems of high paint loss and poor quality consistency through dual-line access gigabit network.

[0004] However, all of the above management systems have the following problems when used in practice:

[0005] Existing coating quality management systems lack efficient temperature management within the coating workshop, resulting in high overall power consumption. Furthermore, workshop temperatures are often far below the specified values, hindering cost reduction and failing to meet green development requirements. Additionally, managing automotive parts painting operations necessitates dust control, requiring precise and timely monitoring of dust anomalies to ensure a safe working environment. This makes it difficult to simultaneously meet production standards and environmental protection / energy conservation requirements.

[0006] No effective solutions have yet been proposed to address the problems in the relevant technologies. Summary of the Invention

[0007] In view of the problems in the related technologies, the present invention proposes an automotive parts coating quality management system and quality management method to overcome the above-mentioned technical problems existing in the existing related technologies.

[0008] Therefore, the specific technical solution adopted by the present invention is as follows:

[0009] According to one aspect of the present invention, an automotive parts coating quality management system is provided, the system comprising a coating execution module, a camera module, an intelligent gateway, a server, a workshop temperature management module, a dust elimination module, and a dust emission management module;

[0010] The coating execution module is used to acquire the parameters required for the spraying process and start the spraying robot to complete the coating operation of the automotive parts.

[0011] The camera module captures and monitors various areas within the automotive parts painting workshop;

[0012] The smart gateway is used to send the daily production data and monitoring data obtained from the automotive parts painting workshop to the server;

[0013] The server is used to receive data sent by the smart gateway, issue production instructions, and send the production instructions to the automotive parts painting workshop through the smart gateway.

[0014] The workshop temperature management module is used to predict the cold load of the automotive parts painting workshop, accurately control the air supply in the workshop according to the predicted cold load, and calculate the first predicted energy consumption value.

[0015] The dust removal module is used to absorb the dust generated during the painting process in the automotive parts painting workshop and to purify the absorbed dust.

[0016] The dust emission management module is used to monitor the dust generated during the painting of automotive parts in the painting workshop, obtain the dust status of the workshop, and calculate the second predicted energy consumption value required for the dust elimination module to reach the preset standard based on the dust status of the workshop.

[0017] Based on the matching diagram of the relationship between workshop air supply and dust purification efficiency per unit time, the optimal air supply and dust purification efficiency are matched to obtain the first and second corrected predicted energy consumption values.

[0018] The working status of the workshop temperature management module and the dust emission management module are set according to the first and second corrected predicted energy consumption values, respectively.

[0019] Furthermore, when the workshop temperature management module predicts the cooling load of the automotive parts painting workshop and accurately controls the air supply in the workshop based on the predicted cooling load, it also predicts the internal and external cooling loads of the automotive parts painting workshop.

[0020] Furthermore, when predicting the internal cold load of the automotive parts painting workshop, for predicting the internal cold load a minutes later, the internal cold load at the same time last week is retrieved from the pre-built historical database as the predicted internal cold load value a minutes later, and the historical data is weighted and optimized at the same time.

[0021] Furthermore, when predicting the external cold load of the automotive parts painting workshop, the temperature and humidity parameters outside the workshop are collected and stored in real time, and the temperature and chilled water flow rate of the chilled water pipes in the air conditioning system inside the workshop are collected in real time. The real-time load of the air conditioning system is calculated and stored.

[0022] The real-time load of the air conditioning system is compared with the collected external temperature and humidity parameters of the workshop. Since there is a time difference between the real-time load of the air conditioning system and the peak values ​​of the collected external temperature and humidity parameters, this time difference is the conduction time T. Based on the conduction time T, the external cold load can be predicted. The predicted external cold load and the internal cold load values ​​are added together to obtain the predicted cold load of the automotive parts painting workshop.

[0023] Furthermore, after predicting the cooling load of the automotive parts painting workshop, the total predicted air supply volume of the air handling units in the air conditioning system of the automotive parts painting workshop is calculated:

[0024]

[0025] In the formula, Q represents the predicted cooling load, and C... p T is the specific heat capacity of air. n Indoor temperature, T s The supply air temperature is used, and the final total predicted supply air volume is obtained by correcting the first corrected predicted energy consumption value.

[0026] Furthermore, after obtaining the final total predicted air volume of the air handling unit, the rotation speed of the air handling unit and the return and exhaust fans of the air conditioning system in the workshop is changed according to the final total predicted air volume, that is, the air supply volume is adjusted.

[0027] Furthermore, the dust emission management module is used to monitor the dust generated during the painting of automotive parts in the painting workshop, and when unexpected dust situations occur in the workshop, a monitoring camera is used to obtain dust images, and the dust images are input into the improved YOLOv4-tiny model to determine whether the dust is abnormal.

[0028] Furthermore, the improved YOLOv4-tiny model includes a backbone network and a neck network;

[0029] In the backbone network of the improved YOLOv4-tiny model, the CSP structure composed of Res residual components and route layers in the original YOLOv4-tiny network is replaced with SERes modules, the second Max pool layer in the original YOLOv4-tiny network is replaced with a 3×3 convolution with a stride of 2, and an XRes module is added after the second Max pool layer.

[0030] Add an SPP module to the neck network and combine the PRN module into the FPN structure of the original YOLOv4-tiny network;

[0031] The structural formula for the PRN module is:

[0032] P i =F i +Upsample(F i+1 )

[0033] F out =Concat(P2, P3, ..., P i );

[0034] In the formula, F i F represents the feature map at the i-th scale in the FPN structure. i+1 Indicates that F i A feature map at a larger scale;

[0035] P i This represents the feature map at the i-th scale after adjustment by the PRN module;

[0036] Upsample represents the upsampling operation, and Concat represents the concatenation operation.

[0037] F out This represents the multi-scale feature map that is ultimately output by the PRN module.

[0038] Furthermore, the SERes module is used to enhance information exchange between network feature channels;

[0039] The XRes module is used to separate channel correlation and spatial correlation to the greatest extent possible; the SPP module is used to extract features from different angles through pooling layers of different sizes to form feature maps with different receptive fields, and to stitch the feature maps along the channel dimension through the route layer to extract multi-scale information.

[0040] The PRN module is used to divide the input dust image into two parts. One part is convolved, and the other part is fused with the result of the convolution operation to obtain a feature map containing more semantic information.

[0041] According to another aspect of the present invention, a method for quality management of automotive parts coating is provided, the method comprising the following steps:

[0042] Obtain the parameters required for the painting process, start the painting robot to complete the painting operation of automotive parts, and take pictures and monitor various areas in the automotive parts painting workshop.

[0043] Data transmission between the automotive parts painting workshop and the server is conducted via a smart gateway.

[0044] Predict the cooling load of the automotive parts painting workshop, accurately control the air supply in the workshop based on the predicted cooling load, and calculate the first predicted energy consumption value.

[0045] The dust generated during the painting process in the automotive parts painting workshop is absorbed and purified. At the same time, the dust generated during the painting process in the automotive parts painting workshop is monitored, the dust status in the workshop is obtained, and the second predicted energy consumption value required for the dust elimination module to reach the preset standard is calculated based on the dust status in the workshop.

[0046] Based on the matching diagram of the relationship between workshop air supply and dust purification efficiency per unit time, the optimal air supply and dust purification efficiency are matched to obtain the first and second corrected predicted energy consumption values.

[0047] The working status of the workshop temperature management module and the dust emission management module are set according to the first and second corrected predicted energy consumption values, respectively.

[0048] The beneficial effects of this invention are as follows: The automotive parts painting workshop of this invention transmits data with a remote server through an intelligent gateway, meeting the needs of remote management. By precisely controlling the air supply volume within the workshop, the temperature inside the painting workshop can be managed efficiently, reducing the overall power consumption of the painting workshop, lowering enterprise production costs, and meeting the requirements of green development. Simultaneously, the dust generated during the painting process is absorbed and purified, reducing adverse environmental impacts. Furthermore, the improved YOLOv4-tiny model can detect abnormal dust conditions in a timely and accurate manner, determining whether an anomaly has occurred in the painting workshop based on these abnormalities. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 This is a structural block diagram of an automotive parts coating quality management system according to an embodiment of the present invention.

[0051] 1. Materials Management Module; 2. Parts Database; 3. Painting Execution Module; 4. Camera Module; 5. Smart Gateway; 6. Server; 7. Workshop Temperature Management Module; 8. Dust Elimination Module; 9. Dust Emission Management Module. Detailed Implementation

[0052] To further illustrate the various embodiments, the present invention provides accompanying drawings, which are part of the disclosure of the present invention. These drawings are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these drawings, those skilled in the art should be able to understand other possible implementation methods and the advantages of the present invention. The components in the drawings are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0053] According to embodiments of the present invention, an automotive parts coating quality management system and a quality management method are provided.

[0054] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments, such as... Figure 1 As shown, according to one aspect of the present invention, an automotive parts coating quality management system is provided. The system includes a materials management module 1, a parts database 2, a coating execution module 3, a camera module 4, a smart gateway 5, a server 6, a workshop temperature management module 7, a dust elimination module 8, and a dust emission management module 9.

[0055] The material management module 1 is used to manage the materials required in the process of painting automotive parts, obtain the types, quantities, dates and usage plans of the required materials, and send material information to the automotive parts painting workshop.

[0056] The component database 2 is used to store information and models of automotive components that need to be painted, and to output the parameters required for the painting process to the automotive component painting workshop.

[0057] The coating execution module 3 is used to acquire the parameters required for the spraying process and start the spraying robot to complete the coating operation of the automotive parts.

[0058] The camera module 4 is used to photograph and monitor various areas within the automotive parts painting workshop.

[0059] The intelligent gateway 5 is used to send the daily production data and monitoring data obtained from the automotive parts painting workshop to the server;

[0060] The server 6 is used to receive data sent by the smart gateway, issue production instructions, and send the production instructions to the automotive parts painting workshop through the smart gateway.

[0061] The workshop temperature management module 7 is used to predict the cold load of the automotive parts painting workshop, accurately control the air supply in the workshop according to the predicted cold load, and calculate the first predicted energy consumption value.

[0062] Among them, the workshop temperature management module 7 predicts the cold load of the automotive parts painting workshop and accurately controls the air supply in the workshop according to the predicted cold load, predicting the internal and external cold load of the automotive parts painting workshop.

[0063] The internal cold load of the automotive parts painting workshop is related to the equipment and personnel flow in the workshop. The working time and number of equipment and personnel in the workshop tend to be stable. At the same time, a historical database is constructed to store the workshop's past internal cold load data.

[0064] When predicting the internal cold load of the automotive parts painting workshop, for the predicted internal cold load a minutes later, the internal cold load at the same time last week is retrieved from the pre-built historical database as the predicted internal cold load value a minutes later, and the historical data is weighted and optimized.

[0065] When predicting the external cold load of the automotive parts painting workshop, since it takes a certain amount of time for the external temperature to transfer to the workshop, the time can be calculated to know how long it will take for the external temperature to take effect on the load inside the workshop; the external temperature and humidity parameters of the workshop are collected and stored in real time, and the temperature and chilled water flow of the chilled water pipes in the air conditioning system inside the workshop are collected in real time, and the real-time load of the air conditioning system is calculated and stored.

[0066] The real-time load of the air conditioning system is compared with the collected external temperature and humidity parameters of the workshop. Since there is a time difference between the peak values ​​of the real-time load of the air conditioning system and the peak values ​​of the collected external temperature and humidity parameters, this time difference is the conduction time T. The external cooling load can be predicted based on the conduction time T. The predicted external cooling load and the internal cooling load are then added together to obtain the predicted cooling load of the automotive parts painting workshop. The relationship between the external cooling load and temperature changes can be obtained from historical data.

[0067] After predicting the cooling load of the automotive parts painting workshop, the total predicted air volume of the air handling units in the air conditioning system of the automotive parts painting workshop is calculated:

[0068]

[0069] In the formula, Q represents the predicted cooling load, and C... p T is the specific heat capacity of air. n Indoor temperature, T s The supply air temperature is used as the reference, and the final total predicted supply air volume is obtained by correcting the first corrected predicted energy consumption value.

[0070] After obtaining the final total predicted air volume of the air handling unit, the speed of the air handling unit and the return and exhaust fans of the air conditioning system in the workshop are changed according to the final total predicted air volume, that is, the air supply volume is adjusted to reduce unnecessary energy waste and achieve the purpose of energy saving.

[0071] The dust removal module 8 is used to absorb the dust generated during the painting process in the automotive parts painting workshop and to purify the absorbed dust.

[0072] The dust emission management module 9 is used to monitor the dust generated during the painting of automotive parts in the painting workshop, obtain the dust status of the workshop, and calculate the second predicted energy consumption value required for the dust elimination module to reach the preset standard based on the dust status of the workshop.

[0073] Based on the matching diagram of the relationship between workshop air supply volume and dust purification efficiency per unit time, for example, when the workshop air supply volume per unit time is 0.75m / s, the dust only moves within a small range and can achieve a large purification efficiency, while when the workshop air supply volume per unit time is 2m / s, the dust only moves within a large range and the purification efficiency is poor. The optimal air supply volume and dust purification efficiency are matched to obtain the first and second corrected predicted energy consumption values.

[0074] The operating states of the workshop temperature management module and the dust emission management module are set according to the first and second corrected predicted energy consumption values, respectively; for example, the final total predicted air volume is obtained by correcting the first corrected predicted energy consumption value; furthermore, after the final total predicted air volume of the air handling unit is obtained, the speed of the air conditioning unit of the air conditioning system in the workshop is changed according to the final total predicted air volume, that is, the air supply is adjusted; and the operating power of the dust removal module is adjusted according to the second corrected predicted energy consumption value, so as to achieve a high-efficiency balance between air supply and dust purification efficiency.

[0075] The dust emission management module 9 is used to monitor the dust generated during the painting of automotive parts in the painting workshop. When an unexpected dust situation occurs in the workshop, a monitoring camera is used to obtain a dust image, and the dust image is input into the improved YOLOv4-tiny model to determine whether the dust is abnormal.

[0076] The improved YOLOv4-tiny model includes a backbone network and a neck network;

[0077] In the improved YOLOv4-tiny model, the backbone network replaces the CSP structure composed of Res residual components and route layers in the original YOLOv4-tiny network with SERes modules, replaces the second Maxpool layer in the original YOLOv4-tiny network with 3×3 convolutions with a stride of 2, and adds XRes modules after the second Maxpool layer (YOLOv4-tiny is a simplified version of YOLOv4 and belongs to the lightweight model).

[0078] An SPP module is added to the neck network, and the PRN module is combined with the FPN structure in the original YOLOv4-tiny network. The FPN structure is the feature pyramid structure in the original YOLOv4-tiny network.

[0079] The SERes module is used to enhance the information interaction between network feature channels. The SERes module is formed by embedding a lightweight SE module into the residual module. In this invention, it is used to replace the original three CSP structures in the original YOLOv4-tiny network.

[0080] The XRes module is used to separate channel correlation and spatial correlation to the greatest extent. The XRes module is a network structure obtained by adding residual edges to the basic Xception module. Xception is a convolutional structure that is a further improvement on the principle of the Inception-v3 module.

[0081] The SPP module is used to extract features from different angles through pooling layers of different sizes, forming feature maps with different receptive fields. The feature maps are then stitched together along the channel dimension through the route layer to extract multi-scale information. The main part of the SPP module consists of three parallel max pooling layers with pooling kernel sizes of 5×5, 9×9, and 13×13, and a stride of 1 for each layer.

[0082] The PRN module is used to divide the input dust image into two parts. One part is convolved, and the other part is fused with the result of the convolution operation to obtain a feature map containing more semantic information. The PRN module concatenates the two upsampling layers of the FPN with the 19th and 38th layers of the original YOLOv4-tiny network through a route layer, connects them to a 3×3 convolutional layer, and then directly connects them to the upsampling layer.

[0083] The structural formula for the PRN module is:

[0084] P i =F i +Upsample(F i+1 )

[0085] F out =Concat(P2, P3, ..., P i );

[0086] In the formula, F i F represents the feature map at the i-th scale in the FPN structure. i+1 Indicates that F i A feature map at a larger scale;

[0087] P i This represents the feature map at the i-th scale after adjustment by the PRN module;

[0088] Upsample represents the upsampling operation, and Concat represents the concatenation operation.

[0089] F out This represents the multi-scale feature map that is ultimately output by the PRN module.

[0090] According to another aspect of the present invention, a method for quality management of automotive parts coating is provided, the method comprising the following steps:

[0091] Obtain the parameters required for the painting process, start the painting robot to complete the painting operation of automotive parts, and take pictures and monitor various areas in the automotive parts painting workshop.

[0092] Data transmission between the automotive parts painting workshop and the server is conducted via a smart gateway.

[0093] Predict the cooling load in the automotive parts painting workshop and accurately control the air supply in the workshop based on the predicted cooling load.

[0094] The dust generated during the painting process in the automotive parts painting workshop is absorbed and purified. At the same time, the dust generated during the painting process in the automotive parts painting workshop is monitored, the dust status in the workshop is obtained, and the second predicted energy consumption value required for the dust elimination module to reach the preset standard is calculated based on the dust status in the workshop.

[0095] Based on the matching diagram of the relationship between workshop air supply and dust purification efficiency per unit time, the optimal air supply and dust purification efficiency are matched to obtain the first and second corrected predicted energy consumption values.

[0096] The working status of the workshop temperature management module and the dust emission management module are set according to the first and second corrected predicted energy consumption values, respectively.

[0097] In summary, the automotive parts painting workshop of this invention transmits data to a remote server via an intelligent gateway, meeting remote management requirements. By precisely controlling the air supply within the workshop, the temperature can be efficiently managed, reducing overall power consumption and production costs, thus aligning with green development requirements. Simultaneously, dust generated during the painting process is absorbed and purified, minimizing adverse environmental impacts. Furthermore, the improved YOLOv4-tiny model can promptly and accurately detect dust anomalies, allowing for the determination of any abnormalities in the painting workshop.

[0098] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A quality management system for automotive parts coating, characterized in that, The system includes a painting execution module, a camera module, a smart gateway, a server, a workshop temperature management module, a dust elimination module, and a dust emission management module; The coating execution module is used to acquire the parameters required for the spraying process and start the spraying robot to complete the coating operation of the automotive parts. The camera module captures and monitors various areas within the automotive parts painting workshop; The smart gateway is used to send the daily production data and monitoring data obtained from the automotive parts painting workshop to the server; The server is used to receive data sent by the smart gateway, issue production instructions, and send the production instructions to the automotive parts painting workshop through the smart gateway. The workshop temperature management module is used to predict the cold load of the automotive parts painting workshop, accurately control the air supply in the workshop according to the predicted cold load, and calculate the first predicted energy consumption value. The dust removal module is used to absorb the dust generated during the painting process in the automotive parts painting workshop and to purify the absorbed dust. The dust emission management module is used to monitor the dust generated during the painting of automotive parts in the painting workshop, obtain the dust status of the workshop, and calculate the second predicted energy consumption value required for the dust elimination module to reach the preset standard based on the dust status of the workshop. Based on the matching diagram of the relationship between workshop air supply and dust purification efficiency per unit time, the optimal air supply and dust purification efficiency are matched to obtain the first and second corrected predicted energy consumption values. The working status of the workshop temperature management module and the dust emission management module are set according to the first and second corrected predicted energy consumption values, respectively.

2. The automotive parts coating quality management system according to claim 1, characterized in that, The workshop temperature management module predicts the cooling load of the automotive parts painting workshop and accurately controls the air supply in the workshop based on the predicted cooling load, thereby predicting both the internal and external cooling loads of the automotive parts painting workshop.

3. The automotive parts coating quality management system according to claim 2, characterized in that, When predicting the internal cold load of the automotive parts painting workshop, for the predicted internal cold load a minutes later, the internal cold load at the same time last week is retrieved from the pre-built historical database as the predicted internal cold load value a minutes later, and the historical data is weighted and optimized.

4. The automotive parts coating quality management system according to claim 2, characterized in that, When predicting the external cold load of the automotive parts painting workshop, the temperature and humidity parameters outside the workshop are collected and stored in real time, and the temperature and chilled water flow rate of the chilled water pipes in the air conditioning system inside the workshop are collected in real time. The real-time load of the air conditioning system is calculated and stored. The real-time load of the air conditioning system is compared with the collected external temperature and humidity parameters of the workshop. Since there is a time difference between the real-time load of the air conditioning system and the peak values ​​of the collected external temperature and humidity parameters, this time difference is the conduction time T. Based on the conduction time T, the external cold load can be predicted. The predicted external cold load and the internal cold load values ​​are added together to obtain the predicted cold load of the automotive parts painting workshop.

5. The automotive parts coating quality management system according to claim 4, characterized in that, After predicting the cooling load of the automotive parts painting workshop, the total predicted air volume of the air handling units in the air conditioning system of the automotive parts painting workshop is calculated: In the formula, Q represents the predicted cooling load, and C... p T is the specific heat capacity of air. n Indoor temperature, T s The supply air temperature is used, and the final total predicted supply air volume is obtained by correcting the first corrected predicted energy consumption value.

6. The automotive parts coating quality management system according to claim 5, characterized in that, After obtaining the final total predicted air volume of the air handling unit, the speed of the air handling unit and the return and exhaust fans of the air conditioning system in the workshop are changed according to the final total predicted air volume.

7. The automotive parts coating quality management system according to claim 1, characterized in that, The dust emission management module is used to monitor the dust generated during the painting process in the automotive parts painting workshop. When acquiring the dust status in the workshop, a monitoring camera is used to obtain dust images, which are then input into the improved YOLOv4-tiny model to determine whether the dust is abnormal.

8. The automotive parts coating quality management system according to claim 7, characterized in that, The improved YOLOv4-tiny model includes a backbone network and a neck network; In the backbone network of the improved YOLOv4-tiny model, the CSP structure composed of Res residual components and route layers in the original YOLOv4-tiny network is replaced with SERes modules, the second Max pool layer in the original YOLOv4-tiny network is replaced with a 3×3 convolution with a stride of 2, and an XRes module is added after the second Max pool layer. Add an SPP module to the neck network and combine the PRN module into the FPN structure of the original YOLOv4-tiny network; The structural formula for the PRN module is: P i =F i +Upsample(F i+1 ) F out =Concat(P2,P3,...,P i ) In the formula, F i F represents the feature map at the i-th scale in the FPN structure. i+1 Indicates that F i A feature map at a larger scale; P i This represents the feature map at the i-th scale after adjustment by the PRN module; Upsample represents the upsampling operation, and Concat represents the concatenation operation. F out This represents the multi-scale feature map that is ultimately output by the PRN module.

9. The automotive parts coating quality management system according to claim 8, characterized in that, The SERes module is used to enhance information exchange between network feature channels; The XRes module is used to separate channel correlation and spatial correlation to the greatest extent possible; the SPP module is used to extract features from different angles through pooling layers of different sizes to form feature maps with different receptive fields, and to stitch the feature maps along the channel dimension through the route layer to extract multi-scale information. The PRN module is used to divide the input dust image into two parts. One part is convolved, and the other part is fused with the result of the convolution operation to obtain a feature map containing more semantic information.

10. A method for managing the coating quality of automotive parts, applied to the automotive parts coating quality management system described in claims 1-9, characterized in that, This quality management method includes the following steps: Obtain the parameters required for the painting process, start the painting robot to complete the painting operation of automotive parts, and take pictures and monitor various areas in the automotive parts painting workshop. Data transmission between the automotive parts painting workshop and the server is conducted via a smart gateway. Predict the cooling load of the automotive parts painting workshop, accurately control the air supply in the workshop based on the predicted cooling load, and calculate the first predicted energy consumption value. The dust generated during the painting process in the automotive parts painting workshop is absorbed and purified. At the same time, the dust generated during the painting process in the automotive parts painting workshop is monitored, the dust status in the workshop is obtained, and the second predicted energy consumption value required for the dust elimination module to reach the preset standard is calculated based on the dust status in the workshop. Based on the matching diagram of the relationship between workshop air supply and dust purification efficiency per unit time, the optimal air supply and dust purification efficiency are matched to obtain the first and second corrected predicted energy consumption values. The working status of the workshop temperature management module and the dust emission management module are set according to the first and second corrected predicted energy consumption values, respectively.

Images