Container paint spraying method and system, intelligent terminal and storage medium

Abstract

The invention relates to the technical field of container paint spraying, in particular to a container paint spraying method and system, an intelligent terminal and a storage medium, and the method comprises the steps that box body reference coordinate data of a container final assembly are obtained through a three-dimensional laser scanner, and corner fitting geometric feature data are synchronously obtained; on the basis of the box reference coordinate data, the boundary of the to-be-sprayed area is recognized; generating a spraying area topological graph based on the boundary of the to-be-sprayed area and the geometric feature data of the corner fitting; based on the spraying area topological graph, a spraying track is generated through an intelligent path planning algorithm; based on the spraying track, a multi-axis spraying system is driven to execute high-pressure airless spraying; and through the industrial Internet of Things platform, the operation time sequence of the main spraying unit and the auxiliary spraying unit is coordinated, and the operation time sequence interval is controlled to be at least equal to a preset interval threshold value. The container paint spraying device has the effects that the situation that spraying leakage or cross contamination is likely to occur during mechanical paint spraying is improved, and the completeness of container paint spraying is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of container painting, and in particular to a container painting method, a container painting system, an intelligent terminal and a storage medium. BACKGROUND

[0002] A container is a standardized cargo transport unit commonly used in global ports, which realizes seamless connection of sea transportation, railway transportation and highway transportation through unified size (20 / 40 feet are mainstream), strength (compression resistance ≥480 MPa) and sealing performance according to ISO standards. A corner fitting of a container is a key load-bearing component of the container and is also the only component directly contacting external equipment (cranes, forklifts and ships). The weight of the container body, the pressure of the cargo and the transportation vibration are all transmitted to the ground or the transportation tool through the corner fitting. The top hole, the bottom hole and the side hole on the corner fitting are all standardized interfaces globally to ensure the compatibility of global logistics chain automation.

[0003] In related technologies, the concave-convex structure between the corner fittings enables the seamless stacking of containers, and the bottom corner fitting bears the full weight of the upper container. Therefore, the corner fitting needs to have a large structural strength so as to be able to bear a large stacking pressure (maximum load of a single corner fitting > 20 tons). During sea transportation, the corner fitting needs to directly contact rainwater, salt mist (salt concentration in sea transportation environment > 3%) and ultraviolet rays. If rust occurs, the structural strength will decrease. Therefore, when painting the container, the corner fitting needs to be painted without dead angle inside and outside, so as to form a corrosion barrier, reduce the risk of surface micro-crack expansion of the corner fitting and avoid fatigue fracture of the corner fitting, thereby affecting the service life.

[0004] In the related technologies described above, during the painting of the container, mechanical painting can usually only deal with large-area painting of the surface of the container body. For fine painting of the corner fitting position of the container, the following situations usually occur: the gap between the corner fitting and the container body is not painted, the inside of each hole of the corner fitting is not painted, or cross contamination occurs between the surface of the container body and the corner fitting when the corner fitting is painted. SUMMARY

[0005] In order to improve the situation that mechanical painting is prone to missing painting or cross contamination and improve the integrity of container painting, the present application provides a container painting method, a container painting system, an intelligent terminal and a storage medium.

[0006] In a first aspect, the present application provides a container painting method, which adopts the following technical solution: A container painting method, comprising: The container body reference coordinate data is acquired by a three-dimensional laser scanner, and the corner piece geometric feature data is synchronously acquired, the corner piece geometric feature data including hole diameters, depths, and spatial coordinate geometric parameters of top holes, bottom holes, and side holes of the corner piece; Based on the container body reference coordinate data, a to-be-sprayed region boundary is identified; Based on the to-be-sprayed region boundary and the corner piece geometric feature data, a spraying region topology graph is generated, the spraying region topology graph including planes, curved surfaces, and special-shaped structures; Based on the spraying region topology graph, a spraying trajectory is generated by an intelligent path planning algorithm; Based on the spraying trajectory, a multi-axis spraying system including a main spraying unit, an auxiliary spraying unit, and a dynamic adjustment module is driven to perform high-pressure airless spraying; Through an industrial Internet of Things platform, the operation time sequence of the main spraying unit and the auxiliary spraying unit is coordinated, and the operation time sequence interval is controlled to be at least equal to a preset interval threshold.

[0007] By adopting the technical solution, the corner piece geometric feature data (hole diameter, depth, and coordinate) is acquired by three-dimensional laser scanning, and the spraying topology graph including special-shaped structures is generated in combination with the container body reference coordinate, so that the gaps and holes between the corner piece and the container body can be precisely covered, and the problem of missed spraying in traditional mechanical spraying can be avoided. The multi-axis spraying system cooperates with the dynamic adjustment module, so that the complex spatial structure of the corner piece can be adapted, and uniform coverage of the anticorrosive coating on the inner walls of the top hole, the bottom hole, and the side hole is ensured. The industrial Internet of Things platform coordinates the time sequence interval of the main spraying unit and the auxiliary spraying unit, and avoids the overlapping of paint mist between the corner piece spraying and the container body surface spraying through time sequence operation, so that the problem of uneven color or coating thickness at the junction of the corner piece and the container body in the traditional process is solved. The high-pressure airless spraying technology in combination with the intelligent path planning can form a paint layer with uniform thickness on the surface of the corner piece, so as to block the penetration of salt mist. The optimized spraying trajectory generated based on the topology graph improves the spraying efficiency of the multi-axis system, and compared with the traditional manual corner piece dead angle re-spraying mode, the rework rate is reduced. The closed-loop feedback mechanism of the three-dimensional scanning data and the spraying trajectory realizes real-time monitoring of the coating thickness, and ensures that the anticorrosive layer thickness of the key parts of the corner piece meets the standard requirements.

[0008] Optionally, the method for identifying the missed spraying region includes: Point cloud data of an easy-to-miss spraying region is acquired by a three-dimensional laser scanner, the easy-to-miss spraying region including the inside of a corner piece hole and the edge of a corner piece-container gap; High-resolution images are synchronously captured by a vision system to obtain a point cloud-image registration data set; Based on the point cloud-image registration data set, a missed spraying region data set is output, the missed spraying region data set including xyz three-axis missed spraying coordinate data and area, depth, and curvature geometric feature data; A vortex thickness gauge and a Doppler ultrasonic sensor are arranged in the easy-to-leak spraying area to obtain the film thickness value of the easy-to-leak spraying area; When the film thickness value is detected to be lower than a preset threshold value, a re-spraying amount is calculated through a PID algorithm, and a re-spraying signal is triggered to the multi-axis spraying system; The position of the leak-spraying area and the film thickness deviation value are recorded synchronously to the MES system.

[0009] By adopting the above technical solution, millimeter-level point cloud data of the inside of the hole of the corner piece and the edge of the gap are obtained through three-dimensional laser scanning, and the xyz three-axis coordinates and geometric features are accurately positioned in combination with the high-resolution image registration of the vision system, compared with traditional manual visual detection, the leak-spraying identification accuracy is improved. The vortex thickness gauge and the Doppler ultrasonic sensor are arranged, the film thickness values of the key areas such as the corner piece rotation locking surface and the weld are obtained in real time. When the film thickness is detected to be lower than the standard requirement, the PID algorithm is triggered to calculate the re-spraying amount, and the film thickness after re-spraying is ensured to meet the standard. The topology map generated based on the point cloud-image registration data set can identify the small gap between the corner piece and the box, and the multi-axis spraying system can realize full coverage spraying of the curved surface and special-shaped structure, compared with the traditional spraying gun operation, the leak-spraying rate inside the corner piece is reduced. The leak-spraying area position and the film thickness deviation value are uploaded to the MES system in real time, historical data can be queried according to the box number or batch, the spraying flow is dynamically adjusted through the PID algorithm, and the paint utilization rate is improved.

[0010] Optionally, the re-spraying path planning method of the leak-spraying area comprises: Based on the leak-spraying area data set, the leak-spraying area surface is reconstructed, and the defect boundary is extracted; Based on the defect boundary and the film thickness value of the leak-spraying area, a three-dimensional defect model containing geometric features, material properties and process parameters is constructed; Based on the three-dimensional defect model, re-spraying path information is generated; The re-spraying amount is calculated synchronously through the PID algorithm, and re-spraying amount information is generated; The re-spraying path information and the re-spraying amount information are synchronously output to the multi-axis spraying system.

[0011] By adopting the technical scheme, based on the curved surface reconstruction technology and the defect boundary extraction algorithm, the missed spraying area of the complex curved surface such as the inside of the hole of the corner piece and the edge of the gap is accurately identified. Combined with the geometric characteristics (such as the radius of curvature) in the three-dimensional defect model, the material properties (such as the elastic modulus of the epoxy zinc-rich primer), and the process parameters (spraying pressure), the re-spraying path covering the missed spraying area is generated, and the spraying path planning efficiency is improved. The PID algorithm dynamically adjusts the re-spraying amount according to the film thickness deviation value of the missed spraying area, for example, when the corner piece is missed in the lock surface, the error of the re-spraying amount can be controlled within a small range. Combined with the positioning accuracy of the multi-axis spraying system, it can avoid excessive accumulation or insufficient coverage of the coating. For the special-shaped gap between the corner piece and the box body and the inner wall of the hole, the three-dimensional defect model can generate adaptive path trajectories. For example, a spiral spraying path is used inside the top hole to ensure that the paint mist uniformly covers the inner wall, which improves the coating uniformity compared with traditional linear spraying. Through the real-time feedback of the re-spraying path and the amount information, the multi-axis system completes the re-spraying operation at one time, avoiding repeated spraying. The re-spraying path and amount data (XYZ coordinates, film thickness value, spraying time) are uploaded to the MES system in real time, supporting the query of historical records by box number or batch, and providing data support for preventive maintenance.

[0012] Optionally, the re-measuring and compensation method after re-spraying includes: Based on the point cloud-image registration data set, the missed spraying point cloud data V0 of the missed spraying area is output; After the re-spraying is completed, the missed spraying area is defined as a repaired area; The repaired area is three-dimensionally re-measured by the online detection module integrated in the system to obtain repaired point cloud data V of the repaired area; The point cloud difference value AV of the missed spraying point cloud data V0 and the repaired point cloud data V is obtained; When AV / V is less than a preset percentage, the output is that the re-measuring is up to standard; When AV / V is greater than or equal to the preset percentage, the output is that the re-measuring is not up to standard, and a secondary re-spraying signal is triggered to the multi-axis spraying system until the point cloud difference value AV converges to a preset percentage tolerance band.

[0013] By adopting the technical scheme, based on the dynamic threshold determination of the point cloud difference value AV / V, through the fusion detection of the three-dimensional scanner and the industrial camera, the residual defects after re-spraying can be accurately identified, and the coating integrity is improved. When the multi-axis spraying system performs secondary re-spraying, combined with the real-time feedback of AV, the spraying parameters (pressure and flow) can be dynamically adjusted. The compensation strategy driven by the point cloud difference value improves the success rate of single re-spraying and reduces the rework time. Through the MES system, the iterative convergence process of AV / V is recorded to form a full life cycle anti-corrosion data chain of the container corner piece. It can be traced back to the spraying equipment parameters, environmental temperature and humidity, and other process parameters.

[0014] Optionally, the dynamic adjustment method of the multi-axis spraying system includes: The spraying environment data is captured in real time by a distributed sensor network, and an environment parameter set is output in real time, wherein the distributed sensor network at least includes a high-precision temperature and humidity sensor and a laser salt fog detector, and the environment parameter set at least includes a salt fog concentration C; When it is detected that the salt fog concentration C exceeds a preset concentration threshold C0, a pressure compensation value ΔP is calculated; Based on a basic pressure value P of a main spraying unit, a target pressure value P+ΔP is calculated; A pressure adjustment instruction and the target pressure value P+ΔP are output to the main spraying unit.

[0015] By using the above technical solution, the laser salt fog detector is used to monitor the salt fog concentration C in real time, and when C>C0, the system automatically calculates the pressure compensation value ΔP. When the salt fog concentration suddenly increases, the spraying pressure is increased from the basic value P to P+ΔP, so as to ensure the coating coverage inside the hole of the corner piece. The distributed sensor network synchronously collects parameters such as temperature, humidity, and salt fog concentration, and constructs an environment parameter dynamic model. When the temperature rises, the system automatically compensates the pressure ΔP to offset the decrease in paint viscosity caused by high temperature. For the special-shaped gap between the corner piece and the box body, the dynamic pressure adjustment enables the spray gun to realize adaptive spraying in the Z-axis direction. The pressure is dynamically adjusted by the PID algorithm to improve the paint utilization rate. The mapping relationship between the pressure compensation value ΔP and the environment parameters is recorded in real time, which can be used to establish a digital twin model of the corner piece spraying in the later period, to predict the optimal spraying parameter combination (such as pressure-flow matching curve) under different salt fog concentrations, to provide data support for process optimization, and to shorten the new product development cycle.

[0016] Optionally, when the salt fog concentration C is abnormal, the dynamic adjustment method of the multi-axis spraying system comprises: In the multi-axis spraying system, the dynamic adjustment module includes a shielding baffle and an angle adjustment mechanism, and an initial opening and closing angle α of the shielding baffle is obtained; When the salt fog concentration C exceeds the preset concentration threshold C0, the angle adjustment mechanism is triggered and the shielding baffle is driven to adjust the opening and closing angle α of the shielding baffle to β, so as to block the salt fog particles from contacting the un-solidified coating surface; The deposition amount m of salt on the surface of the shielding baffle is obtained, and when the deposition amount m is greater than or equal to a deposition amount threshold m, the angle adjustment mechanism is triggered and the opening and closing angle of the shielding is adjusted to a cleaning angle θ; When the deposition amount m is greater than or equal to the deposition amount threshold m, high-pressure airflow is triggered for self-cleaning at the same time, and at least a self-cleaning period is maintained to be equal to a preset self-cleaning period T.

[0017] By adopting the technical scheme, the shielding baffle quickly adjusts the opening and closing angle through the angle adjusting mechanism when the salt mist concentration C > C0, and forms a physical barrier. The probability of salt mist particles contacting the uncured coating is reduced, and salt crystallization corrosion in key areas such as the corner piece rotation locking surface is effectively avoided. The deposition amount sensor monitors the salt deposition amount m on the surface of the baffle in real time, and triggers the cleaning program when m ≥ m0. Combined with the high-pressure airflow self-cleaning period T, the salt residue on the surface of the baffle is reduced, and the problem of uneven spraying caused by salt accumulation is avoided. The angle adjusting mechanism cooperates with the PID algorithm to quickly complete the angle switching, ensuring the continuity of the spraying operation. Real-time recording of salt mist concentration C, baffle angle α / β / θ, deposition amount m and other parameters can be used to establish a digital twin model for corner piece spraying, thereby predicting the optimal protection parameter combination in different marine environments and prolonging the corrosion resistance life of the key area of the corner piece.

[0018] Optionally, the method for building a spraying decision model based on an industrial Internet of Things platform comprises: Real-time acquisition of multi-source heterogeneous data through a distributed sensor network, and uploading the multi-source heterogeneous data to a cloud data lake, wherein the multi-source heterogeneous data at least includes environmental parameters, spraying equipment state parameters and spraying quality parameters; Feature extraction and correlation analysis of the multi-source heterogeneous data based on the cloud data lake to generate a spraying environment dynamic feature vector; Calling a preset decision rule library through the industrial Internet of Things platform, combining the spraying environment dynamic feature vector to generate time sequence coordination instructions for the main spraying unit and the auxiliary spraying unit; Based on the time sequence coordination instructions, dynamically adjusting the spraying pressure and flow parameters of the main spraying unit, and synchronously controlling the supplementary spraying path and operation mode of the auxiliary spraying unit; Real-time feedback of the spraying quality parameters to the decision model through the industrial Internet of Things platform to trigger parameter weight optimization and strategy library dynamic update of the decision rule.

[0019] By adopting the technical scheme, through a distributed sensor network, environmental temperature and humidity, salt mist concentration, equipment pressure and other multi-dimensional parameters are collected in real time to build a spraying environment dynamic feature vector. A cloud-trained neural network model is combined to dynamically predict the coating requirements of the key area of the corner piece, shorten the spraying parameter adjustment response time, and improve the coating uniformity. The industrial Internet of Things platform integrates the main spraying unit and auxiliary systems (supplementary spraying module, salt mist protection device) to generate time sequence coordination instructions based on the decision rule library. For example, when the salt mist concentration suddenly increases, the system automatically increases the main spraying pressure and triggers the angle adjustment of the shielding baffle of the auxiliary unit to ensure the coating coverage rate inside the hole of the corner piece. Real-time processing of the spraying quality parameters (film thickness deviation, adhesion level) through the edge computing node triggers dynamic optimization of the decision model parameters, improves the paint utilization rate, and reduces the single-box VOCs emission amount.

[0020] In a second aspect, the application provides a container paint spraying system, which adopts the following technical solution: A container paint spraying system comprises: An acquisition module is configured to acquire box reference coordinate data, corner fitting geometric feature data, a spraying area topology graph, and a spraying trajectory; A memory is configured to store a program of a control method of the container paint spraying method; A processor, and the program in the memory can be loaded and executed by the processor to implement the control method of the container paint spraying method.

[0021] By adopting the above technical solution, the acquisition module collects the box reference coordinate and the corner fitting geometric feature through high-precision laser scanning, and generates the optimal spraying trajectory in combination with the spraying area topology graph. The spraying coverage on the complex structure of the container corner fitting is improved, and the defect rate of the missed spraying is reduced. The memory pre-stores the intelligent control algorithm based on the industrial Internet of Things, and the processor analyzes and executes the instructions in real time. For example, when the salt mist concentration suddenly increases, the system automatically adjusts the spraying pressure and flow. Through the corner fitting geometric feature data modeling, the system can adaptively adjust the spraying gun posture (pitch angle and yaw angle). The processor dynamically adjusts the paint viscosity according to the box surface roughness, and cooperates with the high-pressure airless spraying technology to improve the paint utilization rate. The system stores the multi-dimensional process parameters through the data lake, and realizes the quality traceability in combination with the MES system.

[0022] In a third aspect, the application provides an intelligent terminal, which adopts the following technical solution: An intelligent terminal comprises a memory and a processor, and the memory stores a computer program of the container paint spraying method, which can be loaded and executed by the processor.

[0023] In a fourth aspect, the application provides a computer storage medium, which can store the program of the corresponding container paint spraying method, and adopts the following technical solution: A computer readable storage medium stores a computer program of any of the above container paint spraying methods, which can be loaded and executed by a processor.

[0024] In summary, the application has at least the following beneficial technical effects: 1. The hole diameter, depth, and spatial coordinates of the corner fitting top hole, bottom hole, and side hole are acquired by a three-dimensional laser scanner, and a spraying topology graph containing a special-shaped structure is generated in combination with the box reference coordinate. An intelligent path planning algorithm generates an optimal spraying trajectory, so that the spraying gun achieves micron-level positioning accuracy, and the paint layer coverage of the corner fitting dead angle is improved. The gap edge is identified in real time through point cloud-image registration technology, and the film thickness is monitored in combination with an eddy current thickness gauge, so as to reduce the defect rate of the missed spraying; 2. Multi-axis spraying system integrated with distributed sensor network, automatically increases main spraying pressure ΔP and adjusts the angle of the shielding baffle to block salt mist particles from contacting the unsolidified coating when the salt mist concentration C increases. The dynamic adjustment module cooperates with the PID algorithm to control the coating flow fluctuation, prolonging the corrosion resistance life of the key area of the corner piece. The self-cleaning system periodically removes salt from the baffle through high-pressure airflow, reducing coating pollution caused by salt crystallization; 3. Build a spraying decision model based on an industrial Internet of Things platform, real-time collect multi-dimensional process parameters (such as spraying gun voltage, current, environmental temperature and humidity), and realize millisecond-level response through edge computing nodes. In addition, the best spraying parameter combination (such as pressure-flow matching curve) under different salt mist concentrations can also be predicted through a digital twin model, improving the utilization rate of coatings. The MES system records the full life cycle data (such as ΔV / V iteration convergence process), supports tracing process parameters by box number, and reduces the rework rate. BRIEF DESCRIPTION OF DRAWINGS

[0025] Figure 1 is a flowchart of a container painting method according to an embodiment of the present application.

[0026] Figure 2 is a flowchart of a method for identifying a missed spraying area according to an embodiment of the present application.

[0027] Figure 3 is a flowchart of a method for planning a re-spraying path of a missed spraying area according to an embodiment of the present application.

[0028] Figure 4 is a flowchart of a method for re-measuring and compensating after re-spraying according to an embodiment of the present application.

[0029] Figure 5 is a flowchart of a dynamic adjustment method of a multi-axis spraying system according to an embodiment of the present application.

[0030] Figure 6 is a flowchart of a dynamic adjustment method of a multi-axis spraying system when the salt mist concentration C is abnormal according to an embodiment of the present application.

[0031] Figure 7 is a flowchart of a method for building a spraying decision model based on an industrial Internet of Things platform according to an embodiment of the present application.

[0032] Figure 8 is a module diagram of a container painting method according to an embodiment of the present application. DETAILED DESCRIPTION

[0033] The present application will be further described in detail below with reference to the accompanying drawings.

[0034] The specific embodiments are merely illustrative of the present application and are not intended to limit the present application. Those skilled in the art can make modifications to the embodiments without creative contribution, according to the needs, and as long as the modifications are within the scope of the claims of the present application, they are protected by the patent law.

[0035] To make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the following will combine the accompanying drawings of the embodiments of the present application to further describe the present application in detail. Figures 1-8 The technical solutions in the embodiments of the present application are described clearly and completely. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative contribution are within the scope of protection of the present application.

[0036] The embodiments of the present application disclose a container paint spraying method. Referring to Figure 1 , the container paint spraying method comprises: Step 100: obtaining container assembly body reference coordinate data by a three-dimensional laser scanner, and synchronously obtaining corner fitting geometric feature data, the corner fitting geometric feature data including hole diameters, depths and spatial coordinate geometric parameters of top holes, bottom holes and side holes of the corner fitting.

[0037] The container assembly body reference coordinate data is a three-dimensional coordinate set taking the overall structure of the container assembly after assembly as a reference, contains key turning points and facade boundary coordinates in length, width and height directions, and is a spatial positioning reference for subsequent operations. The corner fitting geometric feature data describes geometric properties of the corner fitting and the top holes, bottom holes and side holes, including hole diameters, depths and spatial coordinates, and reflects the internal and peripheral structures of the holes.

[0038] The obtaining method is that the three-dimensional laser scanner scans around the container in all directions, captures massive point cloud information of the surface and the corner fitting, and synchronously extracts the container assembly body reference coordinate data and the corner fitting geometric feature data through denoising, splicing, fitting and the like.

[0039] For example, after scanning a 20-foot container, the container assembly body reference coordinate data contains coordinates of the lower left corner of the front end face (0, 0, 0) and the upper right corner of the rear end face (12.192, 2.438, 2.591); in the corner fitting geometric feature data, the top hole of a corner fitting has a hole diameter of 80 mm and a depth of 60 mm, and a spatial coordinate of (0, 0, 2.591).

[0040] Step 101: identifying a to-be-sprayed region boundary based on the container assembly body reference coordinate data.

[0041] The boundary of the area to be sprayed refers to the demarcation line between the area to be sprayed and the non-sprayed area (such as the pre-installed connecting piece), including the planar boundary (the rectangular edge of the side surface of the box), the curved surface boundary (the circular arc transition surface edge of the corner piece and the box), and the special-shaped boundary (the orifice edge of the corner piece hole).

[0042] The acquisition method is to extract key coordinate points (such as edge turning points and non-sprayed area contour points) from the box reference coordinate data, connect adjacent points through a three-dimensional space polygon fitting algorithm to form a closed boundary line to determine the three-dimensional range to be sprayed.

[0043] For example, the boundary of the left side surface of a 20-foot container to be sprayed is a rectangle: (0, 0, 0) → (0, 2.438, 0) → (0, 2.438, 2.591) → (0, 0, 2.591) → (0, 0, 0); the top hole orifice of the upper right corner piece is fitted as a circular boundary by 8 coordinate points.

[0044] Step 102, based on the boundary of the area to be sprayed and the geometric feature data of the corner piece, a spraying area topology graph is generated, which contains planar, curved surface and special-shaped structure.

[0045] The spraying area topology graph is a digital model with topology relationship (connection, adjacency, inclusion) as the core, integrating the boundary of the area to be sprayed and the features of the corner piece, and presenting the spatial correlation of planar (box side surface), curved surface (circular arc transition surface) and special-shaped structure (corner piece hole).

[0046] The acquisition method is to import the boundary of the area to be sprayed and the geometric feature data of the corner piece into a three-dimensional modeling system and align them in space, divide them into three categories according to planarity and curvature, construct a topology relationship matrix (such as the adjacent edges of planar and curved surfaces), and present them in the form of three-dimensional grid with topology attributes.

[0047] For example, in the left side surface topology graph, the rectangular planar area is adjacent to the circular arc curved surface area of the upper right corner piece (the connection line is the right upper corner edge of the planar surface); the orifice edge of the corner piece top hole (special-shaped structure) is connected with the curved surface area, and the topology relationship is marked as "special-shaped-curved surface: edge connection".

[0048] Step 103, based on the spraying area topology graph, a spraying trajectory is generated through an intelligent path planning algorithm.

[0049] The spraying trajectory refers to the movement path of the spraying unit, composed of consecutive spatial coordinate points, containing position and corresponding spraying parameters (angle, flow), and needs to cover all the areas to be sprayed.

[0050] The generation method is to use reciprocating scanning (interval 50mm) for planar areas, generate adaptive arc trajectories for curved surface areas, and use spiral progression (spiral downward from the orifice to the hole bottom, pitch 20mm) for special-shaped structures (such as corner piece holes), to ensure that the trajectory connection is without repetition or interruption.

[0051] For example, the left side vertical plane area adopts a transverse reciprocating trajectory: from (0, 0, 0) to (0, 2.438, 0) along the Y axis, and then moves back along the X axis after moving 50 mm; the corner piece top hole adopts a clockwise spiral trajectory, and moves from the hole center (0, 2.438, 2.591) to the hole bottom (0, 2.438, 2.651).

[0052] In step 104, based on the spraying trajectory, a multi-axis spraying system including a main spraying unit, an auxiliary spraying unit and a dynamic adjustment module is driven to perform high-pressure airless spraying.

[0053] The multi-axis spraying system refers to that the main unit is responsible for large-area regions (large-flow spray guns), the auxiliary unit is responsible for special-shaped structures (small-caliber spray guns), and the dynamic adjustment module real-time calibrates position, angle, pressure and other parameters.

[0054] The execution mode is that after the trajectory is converted into a control instruction, the main unit moves along the path and starts high-pressure spraying (pressure 20 MPa, speed 0.5 m / s), the auxiliary unit moves into the hole along the spiral trajectory, and the dynamic adjustment module adjusts the parameters (such as avoiding collision and correcting pressure) according to the sensor feedback.

[0055] For example, the main unit moves along the reciprocating trajectory of the left side vertical plane, and the spray gun is perpendicular to the plane; the auxiliary unit spray gun with a diameter of 5 mm moves into the hole along the spiral trajectory of the corner piece top hole, and the dynamic adjustment module adjusts the angle by 3° when detecting the hole protrusion to avoid collision.

[0056] In step 105, through an industrial Internet of Things platform, the operation time sequence of the main spraying unit and the auxiliary spraying unit is coordinated, and the operation time sequence interval is controlled to be at least equal to a preset interval threshold.

[0057] The operation time sequence refers to the action sequence and time interval of the main unit and the auxiliary unit; the preset interval threshold (such as 3 s) is the minimum interval set to avoid cross contamination and ensure that the paint is preliminarily solidified.

[0058] The coordination mode is that the industrial Internet of Things platform real-time receives the states (position, operation progress) of the two, formulates a time sequence table: after the main unit completes a certain region, the auxiliary unit starts the operation of the adjacent region after the interval is greater than the threshold, and the platform synchronously issues the instructions.

[0059] For example, the main unit completes the planar region close to the corner piece of the left side vertical plane (end time T1), and the platform sends a start instruction to the auxiliary unit at T1+3 s to make it start the corner piece top hole spraying, thereby avoiding cross contamination caused by the un-solidified paint.

[0060] Referring to Figure 2 The method for identifying the missed spraying region comprises: Step 200, obtain point cloud data of the easy-to-miss spraying area by a three-dimensional laser scanner, the easy-to-miss spraying area including the inside of the corner hole and the edge of the gap between the corner and the box.

[0061] The point cloud data of the easy-to-miss spraying area is a high-density three-dimensional point coordinate set of the high-miss spraying area captured by the laser scanner, such as the inside of the corner hole and the edge of the gap, which can present the hole wall profile, gap width and edge shape, and provide original geometric data for identifying miss spraying.

[0062] The acquisition method is to switch the scanner to a high-precision focusing mode (laser frequency 200 kHz, resolution 0.05 mm), and to scan the hole (cooperate with the telescopic probe to scan layer by layer) and the edge of the gap (multi-angle scanning ≥3) of the 8 corners, with a point cloud density ≥100 points / mm², and to form a complete data set after denoising and splicing.

[0063] For example, the top hole (diameter 80 mm, depth 60 mm) of the right upper corner of the front end of a 20-foot container, the scanner scans every 0.5 mm into the hole (360 points per layer), generating 43200-point point cloud, presenting the subtle concave-convex of the hole wall; and the gap (length 150 mm, width 0.8-1.5 mm) between the corner and the box is scanned at 3 angles and spliced, and the point cloud clearly reflects the irregular protrusions on the edge.

[0064] Step 201, synchronously capture high-resolution images by a vision system to obtain point cloud-image registration data set.

[0065] The high-resolution image refers to the image of the easy-to-miss spraying area taken by an industrial camera (≥20 million pixels), which presents color, texture and coating coverage, making up for the lack of optical information of point cloud. The point cloud-image registration data set refers to the data set that maps the image pixel coordinates to the point cloud coordinate system through a spatial transformation matrix by matching feature points (such as hole corner points) through SIFT algorithm, and fuses three-dimensional geometry and two-dimensional optical information.

[0066] The acquisition method is to synchronize the vision system and the scanner in time (error ≤1 ms), to synchronously take multi-angle images, to realize registration through feature matching and spatial transformation, and to make each point cloud coordinate correspond to image pixel information.

[0067] For example, the images of the top hole and the gap of the corner are taken from the front of the hole, 30° obliquely above and from the side, presenting the light-colored non-coating area (suspected miss spraying) in the hole and the dark coating accumulation in the gap; after registration with the 43200-point point cloud of the top hole, the hole wall point (0, 2.44, 2.62) corresponds to the light gray pixel (no coating) in the image, and the gap point (0, 2.43, 2.59) corresponds to the dark gray pixel (with coating).

[0068] Step 202, based on the point cloud-image registration dataset, output the missed spraying area dataset, which includes xyz three-axis missed spraying coordinate data and area, depth, curvature geometric feature data.

[0069] The missed spraying area dataset is a data collection that quantifies the missed spraying part, including xyz coordinates (positioning spatial position), area, depth (vertical distance from normal coating), and curvature (surface bending degree), providing parameters for the repair spraying plan.

[0070] The output method is to identify suspected missed spraying areas (color, texture abnormalities) in the image through deep learning semantic segmentation, combine point cloud conversion into three-dimensional coordinates, determine the boundary through cluster analysis, calculate geometric features, and then integrate and output.

[0071] For example, the corner piece top hole registration dataset shows that the image light gray area (missed spraying) is mapped to the three-dimensional coordinate range X (0.002-0.01mm), Y (2.440-2.445mm), Z (2.620-2.630mm), with an area of 150mm², a depth of 0.3mm, and a curvature radius of 40mm (consistent with the inner wall of the hole); the gap edge missed spraying area coordinate range X (0.001-0.005mm), Y (2.430-2.432mm), Z (2.590-2.593mm), area 30mm², depth 0.2mm, approximately flat (curvature radius >1000mm).

[0072] Step 203, deploy eddy current thickness gauge and Doppler ultrasonic sensor in the easy-to-miss spraying area to obtain the film thickness value of the easy-to-miss spraying area.

[0073] The eddy current thickness gauge measures the coating thickness (range 0-500μm, accuracy ±1μm) through electromagnetic induction and is suitable for various structures. The Doppler ultrasonic sensor analyzes the internal structure of the coating through echo and assists in verifying the accuracy of the film thickness, especially suitable for complex holes. The film thickness value is the core indicator for judging whether the coating meets the standard, and if it is lower than the preset threshold (such as 30μm), it is considered as missed spraying.

[0074] The acquisition method is that the sensor is integrated into the auxiliary spraying unit, moves along the preset path, the eddy current thickness gauge collects data every 0.5mm, the ultrasonic sensor synchronously collects echo, and after filtering and calibration, it is associated with the coordinates to form a film thickness distribution dataset.

[0075] For example, in the corner piece top hole missed spraying area, the eddy current thickness gauge measures 15μm at (0.002, 2.440, 2.620) and 12μm at (0.005, 2.442, 2.625) (both <30μm); the ultrasonic echo intensity is higher than that of the normal area (verifying the thin film). The gap missed spraying area measures 8-10μm, and the ultrasonic sensor detects the base metal reflection (confirming the missed spraying).

[0076] Step 204, when the film thickness value is detected to be lower than the preset threshold, calculate the replenishing amount by PID algorithm, and trigger the replenishing signal to the multi-axis spraying system.

[0077] The preset threshold is the minimum coating thickness required by the process (such as 30 μm), and the replenishing is needed when the value is lower than this. The PID algorithm is to calculate the replenishing amount by combining the film thickness deviation, regional characteristics and process parameters through the proportional (calculate the basic quantity), integral (eliminate the deviation) and differential (adjust the speed) links. The replenishing signal is a control instruction containing the replenishing region coordinates, amount and path, which drives the multi-axis system to work.

[0078] The calculation and triggering method is that when the film thickness value is lower than the threshold, the deviation value is calculated and input into the PID algorithm, the replenishing amount (μm or ml) is output, the signal containing the coordinate range, amount and speed is generated, and the signal is transmitted to the multi-axis system through the industrial Internet of Things platform.

[0079] For example, the preset threshold of the top hole missing spraying area of the corner piece is 30 μm, the actual measurement is 12-15 μm (deviation 15-18 μm), the PID algorithm combines the area of 150 mm², the curvature of 40 mm and the atomization efficiency of 85%, calculates the average replenishing amount of 16 μm (0.02 ml), and the replenishing signal clearly indicates the coordinate range and spiral path, which drives the auxiliary unit to prepare for work.

[0080] Step 205, record the missing spraying region position and film thickness deviation value to the MES system synchronously.

[0081] The MES system is a manufacturing execution system connecting the management layer and the workshop, which stores the missing spraying data for quality traceability and process optimization. The recording content includes the missing spraying region coordinate range, center point, film thickness maximum / minimum / average deviation, and is associated with the container identification, batch and time.

[0082] The recording method is that the data is transmitted to the MES through the industrial Ethernet after format conversion, is associated and stored, and generates the missing spraying heat map and deviation trend chart, which supports the management personnel to analyze the reasons (such as improper equipment angle).

[0083] For example, the missing spraying area coordinate range of the top hole of the corner piece is X (0.002-0.01 mm), Y (2.440-2.445 mm), Z (2.620-2.630 mm), the center point is (0.006, 2.4425, 2.625 mm), the film thickness deviation is 15-18 μm (average 16.5 μm), the data is associated with the container number storage, and the MES generates the top hole missing spraying heat map and high deviation trend, which assists in adjusting the equipment parameters.

[0084] Reference Figure 3 The missing spraying region replenishing path planning method comprises: Step 300, based on the missing spraying region data set, reconstruct the missing spraying region surface, and extract the defect boundary.

[0085] The spray-missing area surface reconstruction is to use the point cloud coordinates in the spray-missing area data set to construct a three-dimensional surface model (accuracy ± 0.05 mm) reflecting the surface morphology of the spray-missing area by algorithm fitting, and to intuitively present its spatial structure (plane, curved surface, special shape). The defect boundary refers to the demarcation line between the spray-missing area and the normally sprayed area, which is composed of spatial coordinate points of curvature mutation positions, and is used to clearly define the spray-repairing range.

[0086] The reconstruction and extraction method is to screen the point cloud from the spray-missing area data set, generate a three-dimensional surface by Poisson surface reconstruction algorithm fitting; identify the curvature mutation position of the surface by Canny edge detection algorithm, and extract the defect boundary composed of spatial coordinate points (at least 1 boundary point per millimeter).

[0087] For example, the spray-missing area (area 80 mm², depth 0.2 mm) of the 20-foot container corner fitting side hole, the Poisson algorithm fits the point cloud to generate an arc-shaped surface model consistent with the inner wall of the hole; edge detection finds the curvature mutation of the spray-missing area edge, extracts 30 coordinate points to form a closed boundary, and the inside is the spray-repairing area.

[0088] Step 301, based on the defect boundary and the film thickness value of the spray-missing area, a three-dimensional defect model containing geometric features, material properties and process parameters is constructed.

[0089] The three-dimensional defect model refers to a digital model integrating the geometric features (shape, size, curvature) of the spray-missing area, the material properties (substrate roughness, paint viscosity) and the process parameters (spraying pressure, gun distance), and is associated with the box reference coordinate system.

[0090] The construction method is to generate a basic geometric model with the defect boundary as the framework, label the film thickness distribution to form a geometric feature layer; integrate the substrate steel roughness, paint solid content, etc. to form a material property layer; preset the spraying pressure, flow, etc. to form a process parameter layer, and integrate them into a three-dimensional defect model through layer fusion.

[0091] For example, the spray-missing area of the corner fitting side hole (arc-shaped surface, 30 boundary points), the geometric model fits the inner wall of the hole, and the film thickness at (X1, Y1, Z1) is labeled as 12 μm and the film thickness at (X2, Y2, Z2) is labeled as 10 μm; the material properties include a steel roughness of 1.6 μm and a paint viscosity of 800 mPa・s; the process parameters are preset to a gun distance of 20 mm and a pressure of 18 MPa, and the integrated model clearly presents the comprehensive information required for spray-repairing.

[0092] Step 302, based on the three-dimensional defect model, generate spray-repairing path information.

[0093] The supplementary spraying path information refers to the motion trajectory parameters of the auxiliary spraying unit, including the starting point, the ending point, the coordinates of the passing points, and the corresponding motion speed and spraying gun angle, to ensure uniform coverage of the missed spraying area and no collision.

[0094] The generation method is that the planar area adopts a parallel reciprocating path (uniform interval), and the curved surface / irregular shaped area (such as a hole) adopts a spiral path (advancing from the edge to the center or from the hole opening to the hole bottom, and the path interval is dynamically adjusted according to the film thickness deviation); combined with process parameters (such as spraying gun distance), the coordinate sequence and motion parameter table are output through collision detection algorithm verification.

[0095] For example, the missed spraying area of the side hole of the corner piece (arc-shaped curved surface, curvature radius 35 mm), taking the edge point (X0, Y0, Z0) as the starting point, spirally moving along the tangent direction (advancing 0.5 mm every 30°, interval 0.3 mm), the coordinates of the passing points are (X1, Y1, Z1)…(Xn, Yn, Zn) in turn, the spraying gun angle is adjusted in real time (maintaining perpendicularity), the speed is 0.2 m / s, and the path is 10-15 mm away from the inner wall of the hole (no collision).

[0096] In step 303, the supplementary spraying amount is calculated synchronously by a PID algorithm to generate supplementary spraying amount information.

[0097] The supplementary spraying amount information refers to the quantitative data of the supplementary spraying thickness (μm), the coating volume (ml), and the spraying time of each position in the missed spraying area, which is calculated by a PID algorithm (proportional algorithm based on quantity, integral deviation elimination, and differential speed adjustment).

[0098] The calculation method is to input the film thickness value, threshold value, and area in the three-dimensional defect model into the PID algorithm, calculate the supplementary spraying amount in combination with the area curvature and coating atomization efficiency, and output the supplementary spraying thickness of each sub-area, the total coating demand, and the spraying time corresponding to each path segment.

[0099] For example, the missed spraying area of the side hole of the corner piece (film thickness 10-15 μm, threshold value 30 μm, area 80 mm²), the PID algorithm calculates an average supplementary spraying of 17 μm in combination with a deviation of 15-20 μm and a curvature of 35 mm, and the total coating is 2.4 ml, of which the top of the curved surface is supplemented by 18 μm (0.8 ml, 2 seconds) and the bottom is supplemented by 16 μm (0.6 ml, 1.5 seconds), ensuring uniform supplementary spraying.

[0100] In step 304, the supplementary spraying path information and the supplementary spraying amount information are synchronously output to the multi-axis spraying system.

[0101] The synchronous output refers to converting the supplementary spraying path (coordinates, speed, and angle) and the supplementary spraying amount (thickness, volume, and time) into control instructions (such as G code) recognizable by the system, and synchronously transmitting them to ensure that the path matches the amount.

[0102] The output mode is to standardize the path and the supplementary spraying amount information, package and send to the multi-axis system controller through the industrial Internet of Things platform, generate a task list after the controller verifies (no collision, reasonable amount), and drive the auxiliary unit to execute the supplementary spraying according to the path and the amount.

[0103] For example, the corner piece side hole supplementary spraying path contains 50 points (speed 0.2 m / s, angle adjusted according to the surface), the supplementary spraying amount contains an average of 17 μm and a total of 2.4 ml (top 18 μm / 0.8 ml / 2 seconds, bottom 16 μm / 0.6 ml / 1.5 seconds), which is converted into G code and sent synchronously, and the controller drives the auxiliary unit to accurately supplementary spray after verification.

[0104] Referring to Figure 4 , the re-measuring and compensation method after the supplementary spraying includes: Step 400, based on the point cloud-image registration data set, output the missing spraying point cloud data V0 of the missing spraying area.

[0105] The missing spraying point cloud data V0 is a missing spraying area point cloud subset (after denoising) selected from the point cloud-image registration data set, which records the original geometric shape of the missing spraying area before the supplementary spraying, and serves as a re-measuring reference.

[0106] The output mode is to extract the point cloud subset by the region growing segmentation algorithm according to the missing spraying area boundary, retain the effective point cloud (three-dimensional coordinate format) after denoising, store it in association with the unique identifier of the missing spraying area, and keep it consistent with the box reference coordinate system.

[0107] For example, the missing spraying area of the top hole of the corner piece of a 20-foot container (X0.002-0.01 mm, Y2.440-2.445 mm, Z2.620-2.630 mm) selects 8000 points from 100,000 point cloud data, retains 7800 points after denoising to form V0, and presents the concave-convex shape of the inner wall of the hole before the supplementary spraying.

[0108] Step 401, after the supplementary spraying is completed, define the missing spraying area as a repair area.

[0109] The repair area refers to the re-labeling of the original missing spraying area after the supplementary spraying, which clearly defines the re-measuring object range, contains a unique identifier, a coordinate range, and a supplementary spraying completion time, and is marked in a specific color (such as blue) in the three-dimensional model for differentiation.

[0110] The definition mode is to mark the range as a repair area through the system automatic association function based on the missing spraying area boundary, record the identification information and visually mark it, and ensure that the re-measuring focuses on the target area.

[0111] For example, after the supplementary spraying of the missing spraying area of the top hole of the corner piece, the system marks its coordinate range as a repair area, assigns a unique identifier, fills it with blue in the three-dimensional model, and distinguishes it from the surrounding normal area to clearly define the re-measuring range.

[0112] Step 402, the repair area is three-dimensionally re-measured by the online detection module integrated in the system to obtain repair point cloud data V of the repair area.

[0113] The online detection module is a high-precision device (including a small laser scanner and a positioning sensor) integrated in the multi-axis system, which is used to quickly scan the repair area after the supplementary spraying. The repair point cloud data V is a set of point clouds (density ≥ 200 points / mm²) of the repair area after the supplementary spraying, which reflects the coating coverage and surface flatness and is used to evaluate the effect of the supplementary spraying.

[0114] After the supplementary spraying, the multi-axis system drives the detection module to the repair area, and the scanner scans at a high resolution (the auxiliary sensor calibrates the position), and V is generated after noise reduction and removal of abnormal light reflection points, which is associated with the identification of the repair area.

[0115] For example, for the repair area of the top hole of the corner piece, the detection module scans at 250 points / mm² to obtain 15,600 original data points, and 15,300 points are retained after noise reduction to constitute V, which presents the surface morphology of the inner wall of the hole after the supplementary spraying.

[0116] Step 403, obtain the point cloud difference value ΔV of the missed spraying point cloud data V0 and the repair point cloud data V.

[0117] The point cloud difference value ΔV is the coordinate deviation statistical value of the corresponding points after the registration (error ≤ 0.02 mm) of V and V0 by the ICP algorithm, which includes the Z-axis thickness increment (reflecting the coating thickening), the X / Y-axis deviation (reflecting the flatness), and the maximum / minimum / average difference value and the standard deviation.

[0118] The acquisition method is to calculate the Z-axis thickness increment and the X / Y-axis deviation of the corresponding points after the registration of V0 and V, and to obtain ΔV (the average difference value reflects the overall thickening, and the standard deviation reflects the uniformity).

[0119] For example, after the registration of V0 (7,800 points) and V (15,300 points), the maximum Z-axis thickening is 20 μm, the minimum is 14 μm, the average is 17 μm, and the standard deviation is 1.2 μm; the X / Y-axis deviation is ≤ 0.05 mm, and ΔV reflects that the coating thickening after the supplementary spraying is uniform and the flatness is good.

[0120] Step 404, when ΔV / V is less than a preset percentage, output the re-measurement as qualified.

[0121] ΔV / V refers to the ratio of the point cloud difference value ΔV to the repair point cloud V, which eliminates the influence of the area size and is used to judge whether the supplementary spraying is qualified (the preset percentage is, for example, 5%).

[0122] The determination method is to calculate the ratio of the average difference of ΔV to the average coordinate value of V. If < preset percentage, output "retest meets standard" and store to MES system; otherwise, output "not up to standard" and trigger alarm.

[0123] For example, the average difference of the repair area ΔV of the top hole of the corner piece is 17 μm, the average Z-axis coordinate V is 2625 μm, and ΔV / V ≈ 0.65% < 5%. It is determined that "retest meets standard", and the result is stored to the MES.

[0124] When ΔV / V is greater than or equal to the preset percentage, the output is "retest not up to standard", and a secondary spraying signal is triggered to the multi-axis spraying system until the point cloud difference value ΔV converges to the preset percentage tolerance band.

[0125] The secondary spraying signal is a control instruction containing the coordinates of the non-compliant area, the spraying amount, and the adjusted process parameters (such as the encrypted path and the increased pressure), which drives the system to spray again. Converging to the tolerance band means that ΔV / V is finally < preset percentage, ensuring that the repair area meets the standard.

[0126] The triggering and execution method is that when ΔV / V ≥ preset percentage, output "not up to standard", feedback data to the path planning module, recalculate the spraying amount, adjust the path and parameters, generate a secondary spraying signal, transmit it to the system through the industrial Internet of Things platform, and repeat the retest until it meets the standard.

[0127] For example, the repair area ΔV / V of the bottom hole of the corner piece is 7% ≥ 5%, and the output is "not up to standard". The system calculates that an additional 8-10 μm needs to be sprayed, the path spacing is adjusted to 0.2 mm, and the pressure is increased to 20 MPa. After secondary spraying, the retest ΔV / V = 4.5% < 5%, and it converges to the tolerance band.

[0128] Referring to Figure 5 , the dynamic adjustment method of the multi-axis spraying system comprises: Step 500, real-time capture spraying environment data through distributed sensor network, real-time output environment parameter set, wherein the distributed sensor network at least includes high-precision temperature and humidity sensor and laser salt spray detector, and the environment parameter set at least includes salt spray concentration C.

[0129] The distributed sensor network is composed of high-precision temperature and humidity sensors (±0.1℃, ±1%RH) and laser salt spray detectors (0-100mg / m³, ±0.5mg / m³) deployed at different positions in the workshop, which cooperatively monitor the spraying environment.

[0130] Environment parameter set: a set of integrated sensor data, including temperature T, humidity H, salt spray concentration C, etc., providing environmental basis for system dynamic adjustment.

[0131] The capture and output mode is that the sensors are deployed at an interval of 5 meters (top, side, near the work area), collect data at a frequency of 1 second / time, and after filtering, transmit to the central controller through LoRa / 5G, and output the environmental parameter set in real time.

[0132] For example, 3 temperature and humidity sensors (entrance, middle, exit) and 2 salt mist detectors (both sides of the work area) in the spray workshop output the parameter set at a certain time: T = 25.3°C, H = 60%, C = 12.8 mg / m³; when C suddenly rises to 20 mg / m³, the system triggers a warning.

[0133] Step 501, when the salt mist concentration C exceeds the preset concentration threshold C0, calculate the pressure compensation value ΔP.

[0134] The preset concentration threshold C0 refers to the critical value of salt mist concentration set according to the characteristics of the paint (usually 15 mg / m³), which needs to be adjusted when exceeded to offset the impact of salt mist on atomization and adhesion. The pressure compensation value ΔP is the pressure that needs to be supplemented based on the over-standard amount of salt mist, and the formula is ΔP = k × (C - C0) (k = 0.8-1.2 MPa / (mg / m³)), the more over-standard, the larger ΔP.

[0135] The calculation method is that when C > C0, calculate the over-standard amount (C - C0), multiply by the compensation coefficient k to get ΔP, which is used for subsequent pressure adjustment.

[0136] For example, C0 = 15 mg / m³, k = 1.0 MPa / (mg / m³), if C = 20 mg / m³ (over-standard by 5 mg / m³), then ΔP = 5 MPa; if C = 25 mg / m³ (over-standard by 10 mg / m³), then ΔP = 10 MPa.

[0137] Step 502, based on the base pressure value P of the main spraying unit, calculate the target pressure value P + ΔP.

[0138] The base pressure value P refers to the standard spraying pressure when the salt mist is normal (such as 18 MPa), which ensures uniform atomization and adhesion of the paint. The target pressure value P + ΔP refers to the actual spraying pressure when the salt mist is over-standard (base pressure + compensation value), which offsets the adverse effects of salt mist.

[0139] The calculation method is to directly call the current base pressure P of the main unit, add ΔP to get the target pressure value, and transmit it to the pressure control system as the basis for adjustment.

[0140] For example, P = 18 MPa, ΔP = 5 MPa, target pressure = 23 MPa; ΔP = 10 MPa, target pressure = 28 MPa, which ensures the atomization effect in a high-salt-mist environment.

[0141] Step 503, output the pressure regulation instruction and the target pressure value P+ΔP to the main spraying unit.

[0142] The pressure regulation instruction refers to a control signal containing regulation direction (increase) and rate (≤0.5MPa / s), which ensures that the pressure is stable and meets the standard.

[0143] The output process is that after the system generates the instruction, the target pressure value is packaged and transmitted to the main unit pressure control module through industrial Ethernet, and the regulating valve is driven to act according to the instruction.

[0144] For example, the target pressure is 23MPa (current P=18MPa), the instruction is "increase by 0.3MPa / s", and it takes about 16.7 seconds to reach 23MPa; when the target pressure is 28MPa, the instruction rate is 0.5MPa / s, and it takes about 20 seconds to reach the standard, avoiding sudden changes in pressure affecting the coating.

[0145] Referring to Figure 6 When the salt mist concentration C is abnormal, the dynamic regulation method of the multi-axis spraying system includes: Step 600, in the multi-axis spraying system, the dynamic regulation module includes a shielding baffle and an angle adjustment mechanism, and the initial opening angle α of the shielding baffle is obtained.

[0146] The shielding baffle is a protective component in the dynamic regulation module for blocking the direct invasion of salt mist into the spraying area, and is made of corrosion-resistant metal material. The opening and closing angle of the shielding baffle directly affects the shielding effect of the salt mist. The angle adjustment mechanism is composed of a servo motor and a transmission assembly, which can precisely drive the shielding baffle to rotate and change its opening and closing angle to adapt to the protection requirements under different salt mist concentrations. The initial opening angle α refers to the reference angle of the shielding baffle when the salt mist concentration is within the normal range (does not exceed the preset concentration threshold C0). At this time, the baffle will not excessively shield the spraying operation, and can provide basic salt mist protection. Generally, it is preset to 30° according to the size and position of the spraying area.

[0147] The acquisition method is that after the multi-axis spraying system is started, the dynamic regulation module performs self-checking, the position sensor (such as an encoder) of the angle adjustment mechanism detects the current angle of the shielding baffle in real time, and compares the angle with the preset initial opening angle α. If there is a deviation, the servo motor drives the baffle to rotate to the initial angle. After confirming that the baffle is in the initial position, the position sensor transmits the value (such as 30°) of the initial opening angle α to the system control unit, which is stored as the reference value for subsequent angle adjustment.

[0148] For example, a certain multi-axis spraying system aims at the spraying area of the container side wall, and the initial opening angle a of the shielding baffle is set to 30°. After the system starts, the encoder of the angle adjusting mechanism detects that the current angle of the baffle is 28°, which deviates from the initial angle by 2°, and the servo motor immediately drives the baffle to rotate clockwise by 2°, so that the angle of the baffle reaches 30°. The position sensor transmits the initial opening angle a of 30° to the control unit, and the control unit records this value, preparing for the angle adjustment when the salt mist concentration is abnormal.

[0149] Step 601, when the salt mist concentration C exceeds the preset concentration threshold C0, the angle adjusting mechanism is triggered and the shielding baffle is driven to adjust the opening angle a of the shielding baffle to β, so as to block the salt mist particles from contacting the unhardened coating surface.

[0150] When the salt mist concentration C exceeds the preset concentration threshold C0, if the unhardened coating surface contacts the salt mist particles, it will cause defects such as pinholes and bubbles in the coating, affecting the corrosion resistance and adhesion of the coating. At this time, adjusting the opening angle of the shielding baffle can effectively block the salt mist particles from directly contacting the unhardened coating, and can gain time for the coating to harden.

[0151] The specific process is that after the system detects that the salt mist concentration C exceeds C0, it immediately sends a trigger signal to the angle adjusting mechanism. After receiving the signal, the angle adjusting mechanism drives the transmission assembly (such as gears, connecting rods) to drive the shielding baffle to rotate, and adjusts the initial opening angle a to the target angle β. The determination of the target angle β is related to the degree of exceeding the salt mist concentration, the higher the salt mist concentration, the larger the value of β (the maximum does not exceed 90° to avoid complete shielding affecting the spraying operation), for example, when the salt mist concentration exceeds 10 mg / m³, β can be set to 60°; when it exceeds 20 mg / m³, β can be set to 80°. During the adjustment process, the position sensor monitors the baffle angle in real time, and when it reaches β, the servo motor stops moving to ensure that the baffle is stable at the target angle.

[0152] For example, in a certain multi-axis spraying system, the preset concentration threshold C0 is 15 mg / m³, and the initial opening angle a of the shielding baffle is 30°. When it is detected that the salt mist concentration C is 30 mg / m³ (exceeding C0 by 15 mg / m³), the system triggers the angle adjusting mechanism, and the servo motor drives the shielding baffle to rotate, adjusting the angle from 30° to 70° (β=70°). At this time, the shielding range of the baffle is expanded, effectively blocking the salt mist particles from contacting the unhardened coating surface of the container side wall, and after the coating is hardened, the angle is adjusted back to the initial value according to the salt mist concentration.

[0153] Step 602, obtain the deposition amount m of salt on the surface of the shielding baffle, and when the deposition amount m is greater than or equal to the deposition threshold m, trigger the angle adjusting mechanism and drive the opening angle of the shielding to adjust to the cleaning angle θ.

[0154] In the process of blocking salt mist, the surface of the shielding baffle will gradually deposit salt. If the deposition amount is too much, not only will the baffle be corroded, but also the unhardened coating may be contaminated by the salt falling off, affecting the spraying quality. Obtaining the deposition amount m can timely grasp the degree of contamination of the baffle, and the deposition threshold m0 is a critical value (such as 5 g / m2) set according to the corrosion resistance of the baffle material and the spraying environment requirement. When m≥m0, the baffle needs to be adjusted to the cleaning angle θ for cleaning operation. The cleaning angle θ is a specific angle (usually set to 180°, i.e. the baffle is fully expanded) for the cleaning device to clean the surface of the baffle completely, ensuring that the salt can be completely removed.

[0155] The way to obtain the deposition amount m is to install a salt sensor (such as a conductivity salt sensor) on the surface of the shielding baffle. The sensor monitors the salt deposition amount on the surface of the baffle in real time and transmits the data to the system control unit. When the control unit determines that m≥m0, it immediately sends an adjustment instruction to the angle adjustment mechanism. The instruction drives the servo motor to rotate the baffle, so that the opening and closing angle is adjusted from the current angle (such as β=70°) to the cleaning angle θ. During the adjustment process, the position sensor feeds back the angle information in real time, and when θ is reached, the servo motor stops, and the baffle remains at this angle, waiting for the cleaning device to perform high-pressure flushing or wiping cleaning.

[0156] For example, the deposition threshold m0 of a certain shielding baffle is set to 5 g / m2, and the salt sensor detects that the salt deposition amount m on its surface is 6 g / m2 (6≥5). The system triggers the angle adjustment mechanism. At this time, the current opening and closing angle of the baffle is 70°, and the adjustment mechanism drives the servo motor to rotate the baffle until the angle reaches the cleaning angle θ=180°. After the cleaning device is started, it performs high-pressure water flushing on the fully expanded surface of the baffle to remove the salt deposition. After cleaning is completed, the sensor detects that m decreases to 2 g / m2 (less than m0), and the system adjusts the baffle angle back to the working angle suitable for the salt mist concentration.

[0157] Step 603, when the deposition amount m is greater than or equal to the deposition threshold m, trigger the high-pressure airflow for self-cleaning at the same time, and at least maintain the self-cleaning period equal to the preset self-cleaning period T.

[0158] When the salt deposition amount m on the surface of the shielding baffle reaches or exceeds the threshold m0, the high-pressure airflow self-cleaning is triggered at the same time, which can quickly remove the salt on the surface of the baffle, avoiding excessive salt accumulation to corrode the baffle or fall off to contaminate the coating. The high-pressure airflow has the characteristics of strong impact force and high cleaning efficiency, and can effectively remove the salt deposition without damaging the baffle. The preset self-cleaning period T is a time length determined according to the salt deposition speed and cleaning effect (such as 30 seconds), which ensures enough time to clean the baffle completely. The self-cleaning period is at least equal to T to ensure the cleaning effect.

[0159] The specific process is that the system detects that m≥m0 and triggers the angle adjusting mechanism to adjust the baffle to the cleaning angle θ, and sends a start signal to the high-pressure airflow generating device. The high-pressure airflow generating device (such as a high-pressure air pump) immediately works, and high-pressure air flow (the air pressure is usually 0.6-0.8 MPa) is sprayed through the nozzles installed around the baffle, and the air flow blows against the baffle surface at an inclined angle. After the self-cleaning starts, the system timing module starts to ensure that the self-cleaning duration is not shorter than the preset period T. During the cleaning process, the salt content sensor monitors the deposition amount on the baffle surface in real time. If m has decreased to below m0 within T, the cleaning continues until T ends; if m still does not meet the standard after T ends, the cleaning time can be extended until the standard is met.

[0160] For example, the m0 of a certain shielding baffle is 5 g / m2, and when m reaches 6 g / m2, the system triggers the high-pressure airflow self-cleaning simultaneously, and the preset self-cleaning period T is 30 seconds. After the high-pressure air pump is started, 0.7 MPa high-pressure air flow is sprayed through the four nozzles distributed around the baffle, and the baffle surface at the cleaning angle of 180° is blown. The timing module starts timing, and after 30 seconds, the salt content sensor detects that m decreases to 1.5 g / m2 (less than m0), and the self-cleaning stops. If in another case, m is still 3 g / m2 (still greater than or equal to m0) after 30 seconds, the system extends the cleaning time to 40 seconds until m decreases to 2 g / m2 and stops, ensuring that the baffle surface meets the cleaning standard.

[0161] Referring to Figure 7 , the method for building a spraying decision model based on an industrial Internet of Things platform coordination includes: Step 700, real-time collection of multi-source heterogeneous data through a distributed sensor network, and uploading the multi-source heterogeneous data to a cloud data lake, wherein the multi-source heterogeneous data at least includes environmental parameters, spraying equipment state parameters, and spraying quality parameters.

[0162] The multi-source heterogeneous data is a collection of data from different sources, with different formats and structures. The environmental parameters cover temperature, humidity, salt mist concentration, and other spraying environment information; the spraying equipment state parameters include the pressure, flow rate, and motor speed of the main spraying unit, the position and speed of the auxiliary spraying unit, and the running state of the angle adjusting mechanism; and the spraying quality parameters include the coating thickness, missed spraying area position, and film thickness deviation value, which reflect the spraying effect. These data can comprehensively reflect each link of the spraying process and provide raw data for building a spraying decision model.

[0163] During the collection process, each sensor in the distributed sensor network (such as temperature and humidity sensors, pressure sensors, laser thickness gauges, etc.) acquires data in real time according to the preset sampling frequency (such as collecting environmental parameters every 1 second and device state parameters every 0.1 second). Due to different data sources, the formats differ (such as environmental parameters being numerical and device fault information being textual). The system will perform preliminary preprocessing on the data, including format conversion (unified conversion to JSON format) and data cleaning (removal of invalid and missing values). The preprocessed multi-source heterogeneous data is uploaded to the cloud data lake through the 5G industrial gateway or industrial Ethernet. The data lake uses a distributed storage architecture and can accommodate a large amount of multi-type data, supporting fast querying and calling of data to provide data support for subsequent model training and decision analysis.

[0164] For example, the distributed sensor network collects environmental parameters at a certain time: temperature 25°C, humidity 60%, and salt mist concentration 12 mg / m³; spraying device state parameters: main spraying unit pressure 18 MPa, flow rate 5 ml / min, and auxiliary spraying unit position coordinates (X1, Y1, Z1); spraying quality parameters: coating thickness in a certain area 30 μm, and no missed spraying. These data are converted to JSON format after preprocessing and uploaded to the cloud data lake through the 5G industrial gateway. The data lake stores these data in association with the collection time, device number, and other information. When building a spraying decision model, the required data can be quickly extracted for analysis.

[0165] Step 701, feature extraction and correlation analysis of multi-source heterogeneous data based on the cloud data lake to generate a spraying environment dynamic feature vector.

[0166] Feature extraction of multi-source heterogeneous data can filter out information that reflects the key characteristics of the spraying environment from massive data. Correlation analysis can uncover the internal relationships between different types of data (such as the correlation between salt mist concentration and coating adhesion). The combination of the two can convert complex data into feature indicators with clear physical meaning. The spraying environment dynamic feature vector is a multi-dimensional vector composed of these feature indicators, which can dynamically and comprehensively depict the real-time state of the spraying environment, provide input parameters for the spraying decision model, and improve the decision accuracy of the model.

[0167] Specifically, the cloud data lake calls a machine learning algorithm (such as principal component analysis, PCA) to extract features from the stored multi-source heterogeneous data: temperature rate of change, humidity fluctuation amplitude, salt mist concentration peak value, etc. from environmental parameters; pressure stability, flow fluctuation rate, etc. related to environmental adaptability from device state parameters; coating thickness uniformity, etc. affected by the environment from spraying quality parameters. Then, the association rule algorithm (such as Apriori algorithm) is used to analyze the correlation between the features, for example, for every 5mg / m³ increase in salt mist concentration, the coating thickness uniformity decreases by 2%. The extracted key features are sorted by time sequence to form a multi-dimensional array containing time stamps and feature values, i.e. a spraying environment dynamic feature vector, with the vector dimension determined by the number of features (such as an 8-dimensional vector containing 8 features).

[0168] For example, the cloud data lake extracts the following features from multi-source heterogeneous data in a certain period: temperature rate of change 0.5℃ / min, humidity fluctuation amplitude 5%, salt mist concentration peak value 18mg / m³, pressure stability 95%, flow fluctuation rate 2%, coating thickness uniformity 98%. Through correlation analysis, it is found that the salt mist concentration peak value is negatively correlated with the coating thickness uniformity. These features are integrated in chronological order to generate a spraying environment dynamic feature vector [0.5, 5, 18, 95, 2, 98, t1] (t1 is the time stamp), which fully reflects the dynamic characteristics of the spraying environment in this period. Over time, new feature data is continuously supplemented, and the vector is dynamically updated to [0.3, 4, 15, 96, 1, 99, t2], etc., providing real-time environmental state information for the decision model.

[0169] Step 702, call the preset decision rule library through the industrial Internet of Things platform, combine the spraying environment dynamic feature vector, and generate time sequence coordination instructions for the main spraying unit and the auxiliary spraying unit.

[0170] The decision rule library is a set of rules preset based on historical spraying data and process experience, containing the cooperation logic of the main and auxiliary spraying units under different spraying environments, such as "when the salt mist concentration peak value exceeds 15mg / m³ and the pressure stability is less than 90%, the main spraying unit preferentially increases the pressure to the target value, and the auxiliary spraying unit delays 5 seconds to start the supplementary spraying". The time sequence coordination instruction is an instruction that clearly defines the action sequence, interval and coordination method of the main and auxiliary spraying units in the time dimension, which can ensure that the two units work efficiently in complex environments, avoid operation conflicts, and improve the overall spraying efficiency and quality.

[0171] The generation process is that after the industrial Internet of Things platform receives the spraying environment dynamic feature vector, the feature values in the vector (such as salt mist concentration peak value, pressure stability) are matched with the conditions in the decision rule library. If the feature vector meets the triggering condition of a rule (such as salt mist concentration peak value 18 mg / m³>15 mg / m³, pressure stability 95%≥90%), the corresponding collaboration strategy of the rule is called, and the action time sequence is determined in combination with the timestamp of the feature vector. For example, the main spraying unit starts to adjust the pressure at t1, and the auxiliary spraying unit starts the supplementary spraying path planning at t1+3 seconds, so as to ensure that the auxiliary spraying can follow up in time after the main spraying completes the basic spraying. The finally generated time sequence coordination instruction contains the action type (such as pressure adjustment, path movement) of the main unit and the auxiliary unit, the start time, the duration and the like, which are synchronously issued to the controllers of the two units through the industrial Internet of Things platform.

[0172] For example, a spraying environment dynamic feature vector is [0.5, 5, 18, 95, 2, 98, t1], and the industrial Internet of Things platform matches to the rule “when the salt mist concentration peak value is greater than 15 mg / m³ and the pressure stability is greater than or equal to 90%, the main spraying unit adjusts the pressure to P+ΔP at t, and the auxiliary spraying unit starts the supplementary spraying at t+3 seconds” in the decision rule library. In combination with t1 in the vector, the time sequence coordination instruction is generated: the main spraying unit starts to adjust the pressure from 18 MPa to 23 MPa (for 16.7 seconds) at t1; and the auxiliary spraying unit starts the online detection module to scan the missed spraying area (for 5 seconds) at t1+3 seconds. After the instruction is issued, the main unit and the auxiliary unit execute according to the time sequence, which ensures that the main spraying pressure is stable and the auxiliary supplementary spraying follows up in time, thereby realizing efficient cooperation.

[0173] Step 703, based on the time sequence coordination instruction, dynamically adjusting the spraying pressure and flow parameters of the main spraying unit, and synchronously controlling the supplementary spraying path and operation mode of the auxiliary spraying unit.

[0174] The dynamic adjustment of the parameters of the main spraying unit and the synchronous control of the auxiliary spraying unit are to make the two units cooperate efficiently according to the time sequence coordination instruction, so as to ensure that in a complex spraying environment, the main spraying can provide a uniform basic coating, and the auxiliary spraying can accurately supplement the missed spraying area, thereby improving the overall spraying quality and efficiency. The pressure and flow of the main spraying unit directly affect the paint atomization and adhesion effect, and the path and operation mode of the auxiliary spraying unit determine the accuracy and timeliness of the supplementary spraying.

[0175] Specifically, after receiving the timing coordination instruction, the controller of the main spraying unit adjusts the pressure regulating valve and the flow control valve dynamically according to the pressure target value and the time requirement in the instruction through a PID adjustment algorithm, so that the spraying pressure is stabilized at the target value (such as 23 MPa) within the specified time, and the flow is adjusted synchronously in proportion to the change in pressure (the flow is appropriately increased when the pressure rises to ensure sufficient paint supply). At the same time, the controller of the auxiliary spraying unit generates a specific supplementary spraying path according to the path planning time and operation requirements in the instruction, and selects an operation mode according to the characteristics of the missed spraying area (such as reciprocating mode for flat areas and spiral mode for curved areas). During the adjustment and control process, the two units interact with real-time state information (such as whether the pressure of the main spraying unit meets the standard and whether the auxiliary spraying unit reaches the starting point) through the industrial Internet of Things platform to ensure synchronization.

[0176] For example, based on the timing coordination instruction, the main spraying unit starts to adjust the pressure from 18 MPa to 23 MPa at t1, and the controller uses the PID algorithm to stabilize the pressure at 0.3 MPa per second, while the flow is adjusted from 5 ml / min to 6.4 ml / min (proportional to the pressure), and after 16.7 seconds, the pressure is stabilized at 23 MPa, and the flow remains stable. The auxiliary spraying unit starts at t1+3 seconds, selects the spiral operation mode according to the supplementary spraying path planning, and performs supplementary spraying on the missed spraying area of the side hole of the container corner fitting, and the path spirals from the hole to the hole bottom. It synchronously receives the state feedback of the main spraying unit and starts supplementary spraying immediately after the main spraying pressure is stabilized, ensuring that the base coating is uniformly formed when supplementary spraying is performed, and improving the supplementary spraying effect. If the pressure of the main spraying unit fluctuates, the auxiliary spraying unit will temporarily pause the supplementary spraying, and then continue after the pressure stabilizes, ensuring consistency between the two.

[0177] Step 704, real-time feedback of spraying quality parameters to the decision model through the industrial Internet of Things platform, triggering parameter weight optimization of decision rules and dynamic updating of strategy library.

[0178] Real-time feedback of spraying quality parameters to the decision model is to enable the model to understand the actual effect of the current spraying operation in a timely manner, and to determine whether the executed timing coordination instruction meets the expected quality standard. These quality parameters (such as coating thickness uniformity, missed spraying rate, adhesion grade, etc.) are the direct basis for evaluating the effectiveness of the decision rules. If the parameters do not meet the standard, it means that the corresponding decision rules may need to be adjusted, thereby promoting parameter weight optimization and strategy library updating, so that the decision model can adapt to different spraying scenarios and continuously improve decision accuracy.

[0179] In terms of parameter weight optimization, the decision model adopts a machine learning algorithm (such as a random forest algorithm), takes the feedback spraying quality parameters as target variables, and takes each parameter (such as salt mist concentration and pressure value) in the decision rule as a feature variable. The parameter weight is adjusted by calculating the feature importance score. For example, if multiple feedbacks show that the influence of salt mist concentration on coating adhesion is much greater than that of temperature, the weight of salt mist concentration in the decision rule is increased. Dynamic updating of the strategy library is to generate new collaborative strategies (such as adding the rule of "auxiliary spraying unit re-spraying interval shortened by 2 seconds under high humidity environment") based on new quality parameters and environmental characteristics when the optimized parameter weight does not significantly improve the decision effect of the original rule, and to eliminate ineffective rules, so as to ensure that the strategy library always contains efficient and applicable decision logic.

[0180] For example, the industrial Internet of Things platform feeds the spraying quality parameters of a certain period to the decision model: coating thickness uniformity 90% (lower than the target value 95%), leakage rate 2% (up to standard), adhesion level 4 (lower than the target 5). The model analysis finds that the weight of salt mist concentration in the current decision rule is 0.3, and the weight of humidity is 0.2, while the quality parameters show that humidity has a greater impact on uniformity. Through the random forest algorithm, the weight of humidity is optimized to 0.4, and the weight of salt mist concentration is adjusted to 0.25. At the same time, because the uniformity does not meet the standard in high humidity environment for multiple times, the strategy library adds the rule of "the flow fluctuation rate of the main spraying unit is controlled within 1% when the humidity is greater than 65%", and eliminates the original rule of "the flow fluctuation rate is less than or equal to 2% when the humidity is greater than 65%". The updated decision model can more accurately develop strategies according to the humidity parameter in subsequent spraying to improve the coating quality.

[0181] Reference Figure 8 Based on the same inventive concept, the embodiments of the present application provide a container paint spraying system, comprising: an acquisition module configured to acquire box reference coordinate data, corner piece geometric feature data, spraying area topology, and spraying trajectory.

[0182] a memory configured to store a program of a control method of the container paint spraying method; a processor, the program in the memory can be loaded and executed by the processor and implement the control method of the container paint spraying method.

[0183] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above functional modules is exemplified, and in actual application, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device and unit described above can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0184] The embodiment of the present application provides a computer readable storage medium, which stores a computer program capable of being loaded by a processor and executing a container paint spraying method.

[0185] The computer storage medium includes, for example, a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media capable of storing program codes.

[0186] Based on the same inventive concept, the embodiment of the present application provides an intelligent terminal, which comprises a memory and a processor, and the memory stores a computer program capable of being loaded by the processor and executing a container paint spraying method.

[0187] Those skilled in the art can clearly understand that, for the convenience and brevity, only the division of the above functional modules is taken as an example for description, and in actual application, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. The specific working process of the above-described system, device and unit can refer to the corresponding process in the foregoing method embodiments, which will not be described here.

[0188] The above are preferred embodiments of the present application, and are not intended to limit the protection scope of the present application. Any feature disclosed in the specification (including the abstract and the drawings) can be replaced by other equivalent or similar features unless specifically described. That is, each feature is only an example of a series of equivalent or similar features unless specifically described.

Claims

1. A method of painting a container, characterized by, The method comprises the following steps: Obtain the container body reference coordinate data of the container assembly by a three-dimensional laser scanner, and synchronously obtain the geometric feature data of the corner fittings, which includes the hole diameter, depth, and spatial coordinate geometric parameters of the top hole, bottom hole, and side hole of the corner fittings; Identify the boundary of the to-be-sprayed area based on the container body reference coordinate data; Generate a spraying area topology graph based on the boundary of the to-be-sprayed area and the geometric feature data of the corner fittings, wherein the spraying area topology graph contains planes, curved surfaces, and special-shaped structures; Generate a spraying trajectory by an intelligent path planning algorithm based on the spraying area topology graph; Drive a multi-axis spraying system to perform high-pressure airless spraying based on the spraying trajectory, wherein the multi-axis spraying system comprises a main spraying unit, an auxiliary spraying unit, and a dynamic adjustment module; Coordinate the operation timing of the main spraying unit and the auxiliary spraying unit through an industrial Internet of Things platform, and control the operation timing interval to be at least equal to a preset interval threshold.

2. A method of painting a container according to claim 1, wherein The method for identifying the missed spraying area comprises the following steps: Obtain the point cloud data of the easy-to-miss spraying area by a three-dimensional laser scanner, wherein the easy-to-miss spraying area includes the inside of the corner fitting hole and the edge of the gap between the corner fitting and the container body; Synchronously capture high-resolution images by a vision system to obtain a point cloud-image registration data set; Output a missed spraying area data set based on the point cloud-image registration data set, wherein the missed spraying area data set includes xyz three-axis missed spraying coordinate data and area, depth, and curvature geometric feature data; Deploy an eddy current thickness gauge and a Doppler ultrasonic sensor in the easy-to-miss spraying area to obtain the film thickness value of the easy-to-miss spraying area; When it is detected that the film thickness value is lower than a preset threshold, calculate the re-spraying amount by a PID algorithm, and trigger a re-spraying signal to the multi-axis spraying system; Synchronously record the position of the missed spraying area and the film thickness deviation value to a MES system.

3. A method of painting a container according to claim 2, wherein The re-spraying path planning method for the missed spraying area comprises the following steps: Reconstruct the surface of the missed spraying area based on the missed spraying area data set, and extract the defect boundary; Construct a three-dimensional defect model containing geometric features, material properties, and process parameters based on the defect boundary and the film thickness value of the missed spraying area; Generate re-spraying path information based on the three-dimensional defect model; Synchronously calculate the re-spraying amount by a PID algorithm to generate re-spraying amount information; Synchronously output the re-spraying path information and the re-spraying amount information to the multi-axis spraying system.

4. A method of painting a container according to claim 3, wherein The re-measurement and compensation method after the re-spraying is completed comprises the following steps: Output the missed spraying point cloud data V0 of the missed spraying area based on the point cloud-image registration data set; After the re-spraying is completed, define the missed spraying area as a repaired area; Perform three-dimensional re-measurement on the repaired area by an online detection module integrated in the system to obtain repaired point cloud data V of the repaired area; Obtain the point cloud difference value ΔV of the missed spraying point cloud data V0 and the repaired point cloud data V; When ΔV / V is less than a preset percentage, output that the re-measurement is up to standard; When ΔV / V is greater than or equal to the preset percentage, output that the re-measurement is not up to standard, trigger a secondary re-spraying signal to the multi-axis spraying system, and continue until the point cloud difference value ΔV converges to the preset percentage tolerance band.

5. The method of claim 1, wherein The dynamic adjustment method of the multi-axis spraying system comprises the following steps: Real-time capture spraying environment data by a distributed sensor network, and real-time output an environment parameter set, wherein the distributed sensor network at least includes a high-precision temperature and humidity sensor and a laser salt spray detector, and the environment parameter set at least includes a salt spray concentration C; When the salt mist concentration C is detected to exceed the preset concentration threshold C0, a pressure compensation value ΔP is calculated; Based on the base pressure value P of the main spraying unit, a target pressure value P+ΔP is calculated; The pressure adjustment instruction and the target pressure value P+ΔP are output to the main spraying unit.

6. A method of painting a container according to claim 5, wherein When the salt mist concentration C is abnormal, the dynamic adjustment method of the multi-axis spraying system comprises: In the multi-axis spraying system, the dynamic adjustment module comprises a shielding baffle and an angle adjustment mechanism, and the initial opening and closing angle α of the shielding baffle is obtained; When the salt mist concentration C exceeds the preset concentration threshold C0, the angle adjustment mechanism is triggered and the shielding baffle is driven to adjust the opening and closing angle α of the shielding baffle to β, so as to block the salt mist particles from contacting the unhardened coating surface; The deposition amount m of salt on the surface of the shielding baffle is obtained, and when the deposition amount m is greater than or equal to the deposition threshold m, the angle adjustment mechanism is triggered and the opening and closing angle of the shielding is adjusted to the cleaning angle θ; When the deposition amount m is greater than or equal to the deposition threshold m, the high-pressure airflow is triggered for self-cleaning at the same time, and the self-cleaning period is at least maintained equal to the preset self-cleaning period T.

7. A method of painting a container according to claim 5, wherein Based on the coordination of the industrial Internet of Things platform, the method for building a spraying decision model comprises: Real-time collection of multi-source heterogeneous data through a distributed sensor network, uploading the multi-source heterogeneous data to a cloud data lake, wherein the multi-source heterogeneous data at least includes environmental parameters, spraying equipment state parameters and spraying quality parameters; Based on the cloud data lake, feature extraction and correlation analysis are performed on the multi-source heterogeneous data to generate a spraying environment dynamic feature vector; Through the industrial Internet of Things platform, a preset decision rule library is called, and the spraying environment dynamic feature vector is combined to generate a timing coordination instruction for the main spraying unit and the auxiliary spraying unit; Based on the timing coordination instruction, the spraying pressure and flow parameters of the main spraying unit are dynamically adjusted, and the auxiliary spraying unit is synchronously controlled in the supplementary spraying path and the operation mode; Through the industrial Internet of Things platform, the spraying quality parameters are fed back to the decision model in real time, and the parameter weight optimization and strategy library dynamic update of the decision rule are triggered.

8. A container paint spraying system characterised in that, Comprise: The acquisition module is used for acquiring the box body reference coordinate data, the corner piece geometric feature data, the spraying area topology graph and the spraying track; The memory is used for storing the program of the control method of the container painting method according to any one of claims 1 to 7; The processor, the program in the memory can be loaded and executed by the processor and realizes the control method of the container painting method according to any one of claims 1 to 7.

9. A smart terminal, characterized by The memory and the processor are included, and the memory has stored the computer program which can be loaded and executed by the processor and realizes the container painting method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The memory has stored the computer program which can be loaded and executed by the processor and realizes the container painting method according to any one of claims 1 to 7.

Images