Track self-adaptive control method for furniture spraying process

Abstract

The invention relates to the technical field of spraying track planning, and provides a track self-adaptive control method for a furniture spraying process. The method comprises the following steps: establishing a parameterized furniture geometric model; obtaining a spraying demand, performing spraying area unit division in combination with the parameterized model, and establishing a division result; extracting geometric features of a division result, and establishing a geometric partition label set; a main direction is extracted based on the label set, and a basic spraying track is established; according to requirements and a label set, track layering is carried out, a path is optimized, and a following spraying track is established; and feedback data are obtained in real time during spraying, self-adaptive following optimization is carried out according to the feedback data and the following track, and track self-adaptive control is completed. The technical problems of uneven spraying and unstable spraying quality caused by path deviation of the spraying gun in the spraying process are solved, and the technical effect that the spraying precision and the spraying process stability of the complex furniture surface are improved through main direction path planning and layered following control driven by geometrical characteristics is achieved.

Description

Technical Field

[0001] The present application relates to the technical field of spray trajectory planning, and in particular to a trajectory adaptive control method for furniture spraying processes. Background Art

[0002] The furniture spraying process is an indispensable part of the furniture production process, and its purpose is to improve the aesthetics and durability of the furniture surface. Current furniture spraying operations generally have problems such as insufficient automation, unstable spraying quality on complex curved surfaces, and low paint utilization. Especially when faced with special-shaped structural parts or mixed-line production of multiple varieties, the path planning method that relies on manual teaching is difficult to ensure the uniformity of the spraying thickness, and the changeover and debugging cycle can be as long as several hours, which seriously restricts production efficiency. Traditional fixed-track spraying lacks the ability to dynamically respond to environmental disturbances and workpiece deformation. During the spraying process, factors such as robot arm vibration, workpiece placement errors, or paint viscosity fluctuations can easily cause trajectory deviations, resulting in defects such as sagging and orange peel. Summary of the Invention

[0003] This application provides a trajectory adaptive control method for furniture spraying process, aiming to solve the technical problems of uneven spraying and unstable spraying quality caused by spray gun path deviation during the spraying process.

[0004] The present application provides a trajectory adaptive control method for a furniture spraying process, the method comprising: establishing a parameterized furniture geometric model, the parameterized furniture geometric model being constructed by reading design data; obtaining the spraying requirements of the target furniture, dividing the spraying area into units according to the spraying requirements and the parameterized furniture geometric model, and establishing a spraying area unit division result; extracting geometric feature parameters of the spraying area unit division result, and establishing a geometric partition label set; extracting the main direction of the spraying area unit division result based on the geometric partition label set, and establishing a basic spraying trajectory according to the main direction extraction result; stratifying the basic spraying trajectory according to the spraying requirements and the geometric partition label set, optimizing the spraying path, and establishing a following spraying trajectory, wherein the following spraying trajectory is mapped with a spraying parameter label; when spraying is performed based on the following spraying trajectory, acquiring monitoring feedback data of the spray gun in real time, and performing adaptive following optimization according to the monitoring feedback data and the following spraying trajectory to complete trajectory adaptive control.

[0005] One or more technical solutions provided in this application have at least the following technical effects or advantages: The above-mentioned trajectory adaptive control method for furniture spraying process first establishes a parameterized furniture geometric model by reading the design data. Subsequently, the spraying area is divided according to the model and the spraying requirements, and the geometric features of each area are extracted to generate a geometric partition label set. After that, the main direction of each area is extracted based on these label sets, and the basic spraying trajectory is planned. Then, the trajectory is hierarchically optimized according to the spraying requirements to form the final follow-up spraying trajectory, and the spraying parameter label is assigned to it. Finally, during the spraying process, the feedback data of the spray gun is monitored in real time, combined with the follow-up spraying trajectory, and dynamic adjustments are made through the adaptive algorithm to ensure precise control of the spraying path.

[0006] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0008] Figure 1 Schematic diagram of a flow chart of a trajectory adaptive control method for a furniture spraying process in one embodiment.

[0009] Figure 2 The figure is a flow chart of establishing the result of spraying area unit division in a trajectory adaptive control method for a furniture spraying process in one embodiment. DETAILED DESCRIPTION

[0010] The embodiments of the present application solve the technical problems of uneven spraying and unstable spraying quality caused by spray gun path deviation during the spraying process by providing a trajectory adaptive control method for the furniture spraying process.

[0011] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only some of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0012] It should be noted that the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products or devices.

[0013] Examples, such as Figure 1 As shown, the present application provides a trajectory adaptive control method for a furniture spraying process, the method comprising: A parameterized furniture geometric model is established, wherein the parameterized furniture geometric model is constructed by reading design data.

[0014] In this embodiment, the design data of the target furniture, including length, width, height, and curvature of the surface, is first read from a library of furniture design drawings. This data is then input into computer-aided design (CAD) software. The CAD software uses this data to perform three-dimensional modeling, creating a digital model that reflects the actual form of the furniture. This digital model serves as a parametric furniture geometry model for subsequent spray control, allowing for flexible adaptation to different furniture designs.

[0015] The spraying requirements of the target furniture are obtained, and the spraying area unit division is performed according to the spraying requirements and the parameterized furniture geometric model, and the spraying area unit division result is established.

[0016] In one embodiment, according to the requirements of the current spraying business, the spraying requirements of the target furniture are obtained, such as spraying thickness requirements, spraying type requirements, etc. Subsequently, according to these requirements and the specific shape of the parameterized furniture geometric model, the spraying area is divided through clustering partitioning channels, thereby dividing the complex furniture surface into multiple spraying area units. These units may include different forms such as planes, curved surfaces or edges. Each unit represents an area that needs to be processed separately to ensure that different areas can obtain appropriate spraying methods and accuracy according to their characteristics. Afterwards, the multiple divided spraying area units are summarized to establish the spraying area unit division result. This spraying area unit division result can provide clear guidance for subsequent spraying path planning and spray gun movement, ensuring that the spraying process is more accurate.

[0017] Further, if Figure 2 As shown, the present application provides the method of dividing the spraying area unit according to the spraying requirements and the parameterized furniture geometric model, and establishing the spraying area unit division result, including: Analyze the spraying requirements and establish position requirement characteristics, which include spraying thickness requirements, spraying accuracy requirements, and spraying type requirements; establish position geometric shape characteristics based on the parameterized furniture geometric model; input the position requirement characteristics and the position geometric shape characteristics into the clustering division channel, perform area division, and establish the spraying area unit division result.

[0018] Preferably, the acquired target furniture's painting requirements are first parsed, and the required thickness, accuracy, and type of painting (e.g., coating type) are extracted according to pre-defined field names. These requirements determine the painting method and parameters for different areas. These requirements are then stored to establish a location requirement feature. Subsequently, the previously established parametric furniture geometry model is read to extract the furniture surface's geometric features. These geometric features, including surface curvature, angles, bends, and edge features, determine the difficulty and accuracy requirements of the painting path. The location requirement features and location geometry features are then input into a clustering pipeline to perform region segmentation. In this pipeline, the input data is normalized using the Z-score method to ensure that the influence of each feature is equal during the clustering process. An appropriate K value, representing the number of spray area units to be divided, is then determined through experimentation or based on specific requirements. Generally, the K value can be selected using the elbow rule: a K-means algorithm is run to calculate the sum of squared errors (SSE) for different K values, and the K value that results in a sharp decrease in SSE is selected. Then, K initial center points are randomly selected (i.e., K representative points of the spraying area, including the location requirement characteristics and location geometric shape characteristics of the area), and these center points will serve as the initial cluster centers. For each data point, its Euclidean distance from the K cluster centers is calculated, and it is assigned to the cluster center with the shortest distance. After the data point is assigned to the corresponding cluster center, the center point of each cluster is retrieved by calculating the average value within the cluster, and the updated cluster center is used as the basis for the next data point assignment. Repeat the above process until the cluster center no longer changes or changes very little (i.e., convergence is reached). Finally, each cluster generated by the cluster division channel is used as a spraying area unit. The areas in each spraying area unit have similar requirements and geometric characteristics. By storing these spraying area units, the spraying area unit division result is established. This spraying area unit division result provides the basis for subsequent spraying path planning and spray gun motion control, ensuring that each area can be accurately sprayed according to its specific needs.

[0019] Table 1: Example of location requirement characteristics ;

[0020] Table 1 above shows an example of location requirements. It lists the coating requirements for different areas of furniture. The coating thickness requirement specifies the coating thickness required for each area, the coating accuracy requirement specifies the required coating accuracy for each area (i.e., the allowable tolerance for coating thickness), and the coating type requirement describes the type of coating material selected based on the functional requirements of the area (e.g., general coating, anti-corrosion coating, waterproof coating, etc.). These data represent quantitative requirements for the coating area and provide a basis for subsequent coating route planning.

[0021] Table 2: Example table of position geometry features ; Table 2 above shows an example of positional geometry characteristics. This table illustrates the geometric characteristics of various areas on a furniture surface. Geometric types include flat areas, curved areas, and edge areas. The morphological characteristics of these areas determine the complexity of the spray path. Curvature indicates the degree of surface curvature; a larger value indicates a more pronounced curvature. Edge angle describes the angular difference between an edge area and adjacent areas. Surface curvature indicates the maximum deviation in curvature of a curved area. Based on these geometric characteristics, appropriate spray accuracy and spray paths can be selected for different areas.

[0022] After extracting geometric feature parameters from the spraying area unit division result, a geometric partition label set is established.

[0023] In one embodiment, after the division of the spraying area units is completed, the geometric feature parameters of each area will be extracted. In this process, the spraying area units in the spraying area unit division results will be traversed, and the traversed position geometric shape features will be screened according to the geometric feature parameter name to determine the specific features of each spraying area unit, such as curvature, bending, edge angle, surface type features, etc. After these geometric features are extracted, these features will be organized into a geometric partition label set, each label represents an area with specific geometric features, and these labels are used to identify the geometric types of different areas to ensure that subsequent spray trajectory planning and spray path optimization can be properly processed for different types of areas. For example, plane areas and curved surface areas may have different spraying accuracy requirements, so through these labels, each area can be effectively identified and assigned a suitable spraying strategy.

[0024] The main direction of the spraying area unit division result is extracted based on the geometric partition label set, and a basic spraying trajectory is established according to the main direction extraction result.

[0025] In one embodiment, based on the established geometric partition label set, the main direction of each spraying area will be further analyzed. The extraction of the main direction is mainly to identify the main spraying direction or path of each area. Specifically, first, based on the geometric features of each area in the geometric label set, the shape characteristics of each area are analyzed to identify the direction most suitable for spraying. For example, for a flat area, the main direction of spraying may be a straight line direction, while for a curved area, the main direction may be along the tangent direction of the curved surface. By analyzing these geometric features, a suitable spraying direction can be determined for each spraying area. Subsequently, based on the extracted main direction, the system will establish a preliminary spraying trajectory for each spraying area. These basic spraying trajectories are paths optimized for the geometric shape and main direction of each area. The purpose is to ensure the coverage uniformity and accuracy during the spraying process and provide a reference for subsequent path optimization and adjustment.

[0026] Furthermore, the present application provides the main direction extraction of the spraying area unit division result based on the geometric partition label set, including: Obtain the plane area label and the surface area label in the geometric partition label; establish the main direction of the geometric surface feature according to the plane area label and the surface area label respectively; perform edge area and corner feature recognition on the spraying area unit division result to establish the recognition result; use the recognition result to configure edge deflection compensation and corner main direction multi-axis fusion respectively; use the edge deflection compensation and corner main direction multi-axis fusion to compensate for the main direction of the geometric surface feature, and generate the main direction extraction result.

[0027] Preferably, a planar region label and a curved region label are first extracted from the geometric partition label set for each region. Planar region labels identify areas with flat surface geometry, while curved region labels identify areas with varying curvature. Subsequently, for regions with planar region labels, the principal dimension of the planar region (the larger of width or length) is used as the principal direction of the geometric surface feature of the planar region; for regions with curved region labels, the surface tangent is used as the principal direction of the geometric surface feature of the curved region, ensuring uniform spray coverage across the entire curved surface. Next, by analyzing the connectivity between adjacent cells, surface edges connecting to surrounding areas are identified. Edge regions are typically located where a furniture surface meets other surfaces (such as adjacent faces or seams). A triangular patch method can be used to determine the number of adjacent triangles associated with a vertex. If the number of adjacent triangles is less than three, the vertex is considered an edge point. By clustering these edge points into lines, edge regions can be identified. Alternatively, each patch normal can be calculated. If the rate of change of the normal exceeds a set threshold, the region is also considered an edge region. After these areas are marked, the marked areas and corresponding geometric features (such as straightness, curvature, width, etc.) are added to the recognition results. In addition, by identifying places where the angle has a sharp geometric change (such as angles between 30° and 150°), for example, the intersection of two planes, or the intersection of a curved surface and a plane, the corner areas are marked, and the marked areas and corresponding geometric features (such as the angle, the relative inclination angle between the two surfaces, etc.) are added to the recognition results. Then, for the edge areas in the recognition results, edge deflection compensation is performed, that is, the current main direction of the geometric surface feature and the edge direction are used for weighted calculation, where the weight of the edge direction is the deflection strength control factor, which is usually between 0.1 and 0.3, and the main direction of the geometric surface feature is 1 minus the deflection strength control factor. For the corner areas in the recognition results, multi-axis fusion of the main directions of the corners will be performed. That is, for each corner, the normal vectors of the two adjacent surfaces are calculated, and then the main direction interpolation method is used to determine the movement direction of the spray gun by weighted combination of the two main directions. The weight of the main direction is usually set to 0.5, and the weight can also be adjusted according to the curvature. Finally, after using edge compensation to compensate for the main direction of the geometric surface features and completing the compensation using multi-axis fusion of corners, a final main direction extraction result can be obtained. This main direction extraction result can ensure that the spray gun sprays accurately along the edge area, avoiding paint overflow or uneven coverage, and can also ensure that the spray gun path sprays in multiple directions in the corner area, avoiding dead corners and ensuring uniform spraying.

[0028] Furthermore, the present application provides the method of establishing a basic spraying trajectory according to the main direction extraction result, including: The real main direction of the main direction extraction result is identified for each of the spraying area unit division results to establish a real main direction identifier; the initial trajectory of the corresponding spraying area unit is planned according to the real main direction identifier to establish a planning result; the local compensation optimization of the planning result is performed using the main direction extraction result to establish the basic spraying trajectory.

[0029] Optionally, after obtaining the main direction extraction result, for each spraying area unit, the most important spraying direction of the spraying area unit will be identified from the main direction extraction result, and this direction will be defined as the true main direction. For a plane area, the true main direction is generally along the direction of the main dimension, and for a curved surface area, the true main direction is generally along the direction of the surface tangent. After determining the true main direction, a true main direction identifier will be established for each area to indicate the most important spraying path direction during the spraying process. Subsequently, based on the true main direction identifier, a preliminary spraying trajectory plan is generated for each spraying area unit. The trajectory planning will be carried out according to the true main direction of each area to ensure that the spray gun sprays accurately along these main directions during the spraying process. During the planning, adjacent main directions will be connected in sequence to determine the spray gun movement path for each spraying area unit and add it to the planning result. Afterward, based on the initial trajectory planning, local optimization is performed based on the principal direction extraction results. During this process, the principal directions compensated for edge and corner areas are incorporated into the planned results, and the principal directions at corresponding locations in the planned results are fine-tuned to accommodate the specific morphology of these areas. These local compensations ensure a more precise spray path, ultimately generating a basic spray trajectory that adapts to the complex geometry of furniture and meets spraying requirements, ensuring that the spray gun applies the coating accurately and evenly across the furniture surface.

[0030] After the basic spraying trajectory is layered according to the spraying requirements and the geometric partition label set, the spraying path is optimized and a follow-up spraying trajectory is established, wherein the follow-up spraying trajectory is mapped with a spraying parameter label.

[0031] In one embodiment, after obtaining a base spray trajectory, it is stratified based on spray requirements (such as spray thickness and accuracy) and a set of geometric partition labels (including labels for plane and curved surfaces). This stratification divides regions with different spray requirements into multiple layers, each representing a specific spray requirement. After trajectory stratification, the spray path for each layer is optimized. During this process, curve fitting or spline interpolation is used to smooth the spray path, avoiding sharp turns or unnecessary repetitions, thereby improving spray efficiency and reducing paint waste. After path optimization, a follow-up spray trajectory is generated. This is the actual trajectory that the spray gun will follow during spraying, ensuring uniform coverage. The follow-up spray trajectory not only considers the smoothness of the spray path but also includes the adjustment of spray parameters. Each spray trajectory is appropriately labeled and adjusted based on the required spray parameters (such as spray thickness, coating type, and spray gun angle). When establishing the follow-up spray trajectory, the required spray parameters are mapped into spray parameter labels for each trajectory. These tags contain the spray parameters required for each spray path. Through the above steps, the spray path can be optimized while ensuring the spray quality, ensuring that the spray gun accurately follows the planned trajectory during the spraying process, and adjusting the spray parameters according to the needs of each area to achieve an efficient and accurate spraying process.

[0032] Furthermore, the present application provides the step of layering the basic spraying trajectory according to the spraying requirements and the geometric partition label set, including: The basic spraying trajectory is annotated with labels using the spraying requirements and the geometric partition label set, and hierarchical classification is performed to construct a trajectory layering result, which includes a spraying function layer, an edge compensation layer, a surface fitting layer, and a parameter transition layer.

[0033] Preferably, each basic spray trajectory is first annotated with labels using the spraying requirements and geometric partitioning label sets. These labels identify the spraying requirements and regional characteristics corresponding to each trajectory. For example, for areas requiring high-precision spraying, high-precision labels will be added to the relevant trajectories, while for edge areas, edge compensation labels may be added to facilitate subsequent compensation adjustments. Subsequently, the spray trajectories are hierarchically classified based on different label annotations. The purpose of hierarchical classification is to divide the spray trajectories into different levels according to different spraying requirements for more precise control. After completing the hierarchical classification, the trajectory layering results will be generated, including the spraying function layer, edge compensation layer, surface fitting layer, and parameter transition layer. Among them, the spraying function layer contains the main spraying function paths, which are planned according to the basic requirements of spraying (such as thickness, coating type, etc.). The spray gun will perform regular spraying operations along these trajectories; the edge compensation layer corresponds to the edge area or corner area of ​​the furniture surface. The spraying trajectory will be compensated and adjusted at this level to ensure that the spraying process covers the edge and avoids missing spraying or uneven coating; the surface fitting layer corresponds to the curved surface area, and the spraying trajectory will be optimized at this level to ensure that the spraying path closely follows the shape of the curved surface to avoid spraying errors on curved surfaces; the parameter transition layer is used to process the transition area in the spraying path, especially where the spraying parameters change (such as adjustment of spray thickness, spray gun angle, and spray amount). These changes are smoothed through the transition layer to avoid uneven or abrupt spraying effects. Through this process, the spraying trajectory can be finely divided into multiple levels according to the spraying requirements and geometric characteristics. Each level has a targeted control strategy to ensure accurate and uniform spraying effects.

[0034] When spraying is performed based on the following spraying trajectory, monitoring feedback data of the spray gun is acquired in real time, and adaptive tracking optimization is performed according to the monitoring feedback data and the following spraying trajectory to complete trajectory adaptive control.

[0035] In one embodiment, when spraying is performed, feedback data of the spray gun is collected in real time through monitoring sensors. These feedback data may include information such as spraying angle and coating thickness, which can help understand the current working status of the spray gun and ensure that the parameters and paths during the spraying process match the preset trajectory. Subsequently, the monitoring feedback data obtained in real time is compared with the previously planned follow-up spraying trajectory to check whether the spray gun has deviated from the predetermined trajectory or whether the spraying parameters have deviated. If there is a deviation, a multi-dimensional deviation grading model will be used to automatically adjust the path and motion control of the spray gun, correct the error in real time, and ensure that the spraying trajectory remains in the optimal state. By continuously obtaining feedback data and making adaptive adjustments, it is possible to adapt to changes in the working environment in real time (such as changes in the surface of furniture during spraying or the offset of the spray gun position) to ensure that the spraying quality meets the predetermined requirements.

[0036] Furthermore, the present application provides a method for obtaining monitoring feedback data of the spray gun in real time when spraying is performed based on following the spraying trajectory, including: A monitoring sensor group is configured. After the spray gun is started, the monitoring sensor group is synchronously activated. The monitoring sensor group includes a distance sensor, a thickness sensor, a posture sensor, and a visual sensor. Data monitoring of the spray gun posture trajectory and spraying effect is performed according to the monitoring sensor group to establish the monitoring feedback data.

[0037] Preferably, a set of monitoring sensors is configured, including distance sensors, thickness sensors, posture sensors, and visual sensors. These sensors are installed in the spray gun or spraying work area to collect relevant data in real time during the spraying process. Among them, the distance sensor is used to measure the distance between the spray gun and the target surface to ensure that the spray gun is spraying in the correct position; the thickness sensor is used to detect the thickness of the sprayed layer to ensure that the coating meets the predetermined thickness requirements; the posture sensor is used to monitor the angle and direction of the spray gun to ensure that the spray gun sprays according to the predetermined trajectory and angle; the visual sensor is used to obtain a real-time image of the spraying area, detect the spraying effect, and identify any defects or uneven spraying. After the spray gun is started, these monitoring sensors will be activated synchronously to continuously collect and record real-time data of the spray gun, including data such as the movement state of the spray gun, the thickness of the sprayed layer, the spraying angle, and the spraying effect. This data will be used for subsequent adaptive control and path optimization to adjust the behavior of the spray gun in real time to ensure the accuracy and quality of the spraying process.

[0038] Furthermore, the present application provides the adaptive tracking optimization based on the monitoring feedback data and the following spraying trajectory, including: Based on the monitoring feedback data and the tracking spray trajectory, the offset feature analysis is performed to establish the spray trajectory offset feature, the surface spray thickness offset feature, and the posture angle error feature; a state offset vector is established for the spray trajectory offset feature, the surface spray thickness offset feature, and the posture angle error feature; an area recognition constraint is established, and after configuring a multidimensional deviation grading model according to the area recognition constraint, the state offset vector is analyzed using the multidimensional deviation grading model to generate a tracking optimization result.

[0039] Optionally, after obtaining the monitoring feedback data, the monitoring feedback data and the spray trajectory will be analyzed for offset characteristics to analyze the situation in which the spray gun deviates from the predetermined trajectory or has errors during the spraying process. Specifically, the distance difference between the actual spraying trajectory and the target trajectory will be calculated to obtain the offset characteristics of the spraying trajectory; the difference between the actual spraying thickness and the target spraying thickness will be calculated to obtain the offset characteristics of the spraying thickness; the error between the current attitude angle of the spray gun and the target angle will be calculated to obtain the error characteristics of the attitude angle. Subsequently, the obtained spray trajectory offset characteristics, surface spray thickness offset characteristics, and attitude angle error characteristics are spliced ​​end to end according to the preset vector template to form a state offset vector. This state offset vector contains all the deviation information that occurred during the spraying process of the spray gun, including position error, thickness error, and angle error, providing necessary information for subsequent optimization and adjustment. Afterwards, regional identification constraints are established based on the geometric shapes and spraying requirements of different spraying areas. This regional identification constraint defines the deviation tolerance for different geometric shapes, spraying requirements, etc. By using this regional identification constraint, the spraying history data is screened to obtain sample data that meets current business needs, including sample offset vectors, sample tracking spraying trajectories, and sample tracking optimization trajectories. The sample data is then input into the multi-layer perceptron (MLP) framework and iteratively trained through steps such as forward propagation, loss calculation (such as mean squared error), backpropagation, and parameter optimization (such as Adam) until the maximum number of iterations is reached or the loss converges. After training is complete, the multi-layer perceptron is evaluated using validation data. If the evaluation results show that the accuracy meets the expected accuracy, the current multi-layer perceptron will be output as the final multi-dimensional deviation classification model. Otherwise, hyperparameters such as the learning rate and the number of training batches are adjusted to further improve the optimization effect of the multi-dimensional deviation classification model. After the multidimensional deviation grading model is configured, the current state offset vector and the current tracking spray trajectory are input into the multidimensional deviation grading model for offset analysis. This identifies any errors in the spraying process and makes adjustments based on the severity of the deviation to ensure that the spray gun can be restored to the correct trajectory and spray parameters as much as possible. Finally, the multidimensional deviation grading model generates a tracking optimization result, namely an optimized spray trajectory, which ensures that the spray gun sprays along the optimal path and can promptly correct any errors, thereby improving spraying accuracy and quality.

[0040] Furthermore, the present application provides that before performing the offset analysis on the state offset vector using the multi-dimensional deviation classification model, the method includes: Establishing an offset threshold of a state offset vector; performing offset warning identification on the state offset vector according to the offset threshold; and reporting an offset warning and generating a shutdown command if the offset warning identification result is a trigger result.

[0041] Optionally, first, set an offset threshold for the state offset vector. This threshold is defined based on the possible error range in the spraying process, which means that the spraying can still be considered qualified within the allowable error range. The offset threshold includes the maximum deviation of the spray trajectory, the maximum error of the spray thickness, and the maximum deviation of the attitude angle. Generally, stricter spraying requirements (such as high-precision detail areas) will have lower thresholds, while large-area flat areas may allow larger errors. Subsequently, the obtained state offset vector is compared and analyzed with the set offset threshold to check whether the behavior of the spray gun exceeds the allowable error range. If the trajectory offset, spray thickness deviation, or spray gun attitude angle error of the spray gun exceeds the set threshold, it will be identified as an offset warning, that is, the error in the spraying process has reached a level that requires attention. At this time, an offset warning will be triggered. This warning is usually displayed on the system interface to remind the operator that there is an abnormality in the spraying process. After the deviation warning is triggered, a shutdown command will be generated according to the preset control strategy. The shutdown command means that the spraying operation will be suspended to prevent further spraying errors and avoid situations that may affect the spraying quality or cause material waste, thereby ensuring the stability of the spraying quality and process.

[0042] Furthermore, the present application provides the adaptive tracking optimization based on the monitoring feedback data and the following spraying trajectory to complete trajectory adaptive control, including: An environmental state quantity vector is constructed; the environmental state quantity vector is input into a trajectory compensation controller as a disturbance observation vector; and the trajectory compensation controller is used to perform disturbance compensation on the tracking optimization result and update the tracking optimization result.

[0043] Preferably, first collect and integrate the environmental factors that may affect the spraying trajectory and effect during the spraying process. These factors are called environmental state quantities, which include temperature, humidity, airflow, vibration of the surrounding environment, etc. These environmental parameters are obtained through sensors and real-time monitoring equipment, and they are spliced ​​according to the template vector to generate a multidimensional vector, namely the environmental state quantity vector. This vector contains all environmental variables related to the spraying process and is used to describe the current working environment state. Subsequently, the environmental state quantity vector is input into the trajectory compensation controller as a disturbance observation vector. The trajectory compensation controller can also be constructed based on the aforementioned multi-layer perceptron framework. Its function is to dynamically adjust the tracking optimization result according to the current environmental state quantity. After the environmental state quantity vector is input into the trajectory compensation controller, the tracking optimization result will also be input into the trajectory compensation controller. By analyzing the impact of these environmental disturbances on the spraying trajectory, the spray gun path is adjusted to enable it to better adapt to the current environmental conditions. The compensated tracking spray trajectory will be used to update the tracking optimization result to make it more consistent with the current environmental conditions, thereby ensuring the accuracy and stability of the spraying process.

[0044] Furthermore, the present application provides the adaptive tracking optimization based on the monitoring feedback data and the following spraying trajectory to complete trajectory adaptive control, and also includes: The tracking optimization result is marked with a real spraying quality; the spraying demand and the demand satisfaction analysis of the real spraying quality mark are performed, real feedback is established, and trajectory adaptive control optimization management is performed according to the real feedback.

[0045] Preferably, during the spraying process, the actual spraying quality mark will be made for each spraying trajectory followed based on the actual spraying effect. This process uses sensors (such as visual sensors, thickness sensors, etc.) to check the actual quality of the spraying area and evaluate whether the spraying meets the predetermined standards. The quality mark includes the actual coating thickness, actual spraying accuracy, actual spraying type, etc. Subsequently, the actual spraying effect is compared and analyzed with the spraying requirements to verify whether the spraying meets the predetermined quality requirements. The spraying requirements include the thickness, accuracy, coating type, etc. of the spraying. Based on the actual spraying quality mark, the gap between the actual spraying effect and the requirements is analyzed. For example, if the actual coating thickness is insufficient or the spraying accuracy is not high, these deviations will be recorded and marked as unmet requirements. After the requirements are met analysis, a real feedback will be generated. This feedback reflects the gap between the spraying effect and the requirements. The real feedback includes specific deviation data. Afterwards, based on real feedback, the spray trajectory of the spray gun will be adaptively controlled and optimized. That is, the part that needs to be adjusted will be calculated again according to the multi-dimensional deviation grading model, which may be the movement path, spraying angle or spraying speed of the spray gun, so as to correct the movement trajectory and spraying parameters of the spray gun, ensure that the spraying quality gradually meets the predetermined requirements, and reduce paint waste and unnecessary re-spraying.

[0046] In summary, the embodiments of the present application have at least the following technical effects: The embodiment of the present application first establishes a parametric furniture geometric model, which is constructed by reading design data; then, the spraying requirements of the target furniture are obtained, and the spraying area unit division is performed according to the spraying requirements and the parametric furniture geometric model, and the spraying area unit division result is established; thereafter, the geometric feature parameters of the spraying area unit division result are extracted, and a geometric partition label set is established; further, the main direction of the spraying area unit division result is extracted based on the geometric partition label set, and a basic spray trajectory is established according to the main direction extraction result; then, the basic spray trajectory is layered according to the spraying requirements and the geometric partition label set, the spray path is optimized, and a following spray trajectory is established, wherein the following spray trajectory is mapped with a spray parameter label; finally, when spraying is performed based on the following spray trajectory, the monitoring feedback data of the spray gun is obtained in real time, and adaptive following optimization is performed according to the monitoring feedback data and the following spray trajectory to complete trajectory adaptive control. These technical effects jointly solve the technical problems of uneven spraying and unstable spraying quality caused by spray gun path deviation during the spraying process, and achieve the technical effect of improving the spraying accuracy and spraying process stability of complex furniture surfaces through main direction path planning and layered tracking control driven by geometric features.

[0047] It should be noted that the order in which the embodiments of the present application are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0048] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

[0049] This specification and drawings are merely illustrative of the present application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Obviously, those skilled in the art may make various modifications and variations to this application without departing from the scope of this application. Thus, this application is intended to include such modifications and variations as fall within the scope of this application and its equivalents.

Claims

1. A trajectory adaptive control method for furniture spraying process, characterized in that: The method comprises: Establishing a parametric furniture geometric model, wherein the parametric furniture geometric model is constructed by reading design data; Obtaining the spraying requirements of the target furniture, dividing the spraying area into units according to the spraying requirements and the parameterized furniture geometric model, and establishing a spraying area unit division result; After extracting geometric feature parameters from the spraying area unit division result, a geometric partition label set is established; Extracting the main direction of the spraying area unit division result based on the geometric partition label set, and establishing a basic spraying trajectory according to the main direction extraction result; After the basic spraying trajectory is layered according to the spraying requirements and the geometric partition label set, the spraying path is optimized to establish a follow-up spraying trajectory, wherein the follow-up spraying trajectory is mapped with a spraying parameter label; When spraying is performed based on the following spraying trajectory, monitoring feedback data of the spray gun is acquired in real time, and adaptive tracking optimization is performed according to the monitoring feedback data and the following spraying trajectory to complete trajectory adaptive control.

2. The trajectory adaptive control method for furniture spraying process according to claim 1, characterized in that: The step of dividing the spraying area into units according to the spraying requirements and the parameterized furniture geometric model and establishing the spraying area unit division result includes: Analyze the spraying requirements and establish position requirement characteristics, wherein the position requirement characteristics include spraying thickness requirements, spraying accuracy requirements, and spraying type requirements; Establishing positional geometric shape features according to the parameterized furniture geometric model; The position requirement features and the position geometric shape features are input into the cluster division channel, regional division is performed, and a spraying area unit division result is established.

3. The trajectory adaptive control method for furniture spraying process according to claim 1, characterized in that: The extracting the main direction of the spraying area unit division result based on the geometric partition label set includes: Get the plane area label and surface area label in the geometric partition label; Establish the main directions of geometric surface features according to the plane area labels and the surface area labels respectively; Performing edge area and corner feature recognition on the spraying area unit division result to establish a recognition result; Using the recognition results, respectively configure edge deflection compensation and corner main direction multi-axis fusion; After compensating the main direction of the geometric surface feature by utilizing the edge deflection compensation and the multi-axis fusion of the main direction of the corner, a main direction extraction result is generated.

4. The trajectory adaptive control method for furniture spraying process according to claim 3, characterized in that: The step of establishing a basic spraying trajectory according to the main direction extraction result includes: Performing true main direction identification on the main direction extraction result of each spraying area unit division result to establish a true main direction identifier; Performing initial trajectory planning for the corresponding spraying area unit according to the true main direction identifier and establishing a planning result; The main direction extraction result is used to perform local compensation optimization of the planning result to establish the basic spraying trajectory.

5. The trajectory adaptive control method for furniture spraying process according to claim 1, characterized in that: The basic spray trajectory layering is performed according to the spraying requirements and the geometric partition label set, including label annotation of the basic spray trajectory using the spraying requirements and the geometric partition label set, and performing hierarchical classification to construct a trajectory layering result, wherein the trajectory layering result includes a spraying function layer, an edge compensation layer, a surface fitting layer, and a parameter transition layer.

6. The trajectory adaptive control method for furniture spraying process according to claim 1, characterized in that: When spraying is performed based on following the spraying trajectory, real-time monitoring feedback data of the spray gun is obtained, including: A monitoring sensor group is configured, and after the spray gun is started, the monitoring sensor group is synchronously activated, and the monitoring sensor group includes a distance sensor, a thickness sensor, a posture sensor, and a visual sensor; The monitoring sensor group performs data monitoring of the spray gun posture trajectory and spraying effect to establish the monitoring feedback data.

7. The trajectory adaptive control method for furniture spraying process according to claim 6, characterized in that: The adaptive tracking optimization according to the monitoring feedback data and the following spraying trajectory includes: Based on the monitoring feedback data and the spray trajectory, an offset feature analysis is performed to establish a spray trajectory offset feature, a surface spray thickness offset feature, and an attitude angle error feature; Establishing a state offset vector for the spray trajectory offset feature, the surface spray thickness offset feature, and the attitude angle error feature; A region identification constraint is established, and after a multi-dimensional deviation classification model is configured according to the region identification constraint, the state deviation vector is subjected to deviation analysis using the multi-dimensional deviation classification model to generate a tracking optimization result.

8. The trajectory adaptive control method for furniture spraying process according to claim 7, characterized in that: Before performing the offset analysis on the state offset vector using the multi-dimensional deviation classification model, the method includes: Establishing a state offset threshold value of the offset vector; Performing offset warning identification on the state offset vector according to the offset threshold; If the deviation warning identification result is a trigger result, the deviation warning is reported and a shutdown command is generated.

9. The trajectory adaptive control method for furniture spraying process according to claim 1, characterized in that: The adaptive tracking optimization is performed according to the monitoring feedback data and the spraying trajectory to complete the trajectory adaptive control, including: Construct the environment state vector; Inputting the environmental state quantity vector as a disturbance observation vector into a trajectory compensation controller; The trajectory compensation controller is used to perform disturbance compensation on the tracking optimization result and update the tracking optimization result.

10. The trajectory adaptive control method for furniture spraying process according to claim 1, characterized in that: The method of performing adaptive tracking and optimization based on the monitoring feedback data and the spraying trajectory to complete trajectory adaptive control further includes: Mark the actual spraying quality of the optimization results; Perform demand satisfaction analysis between the spraying demand and the actual spraying quality indicator, establish real feedback, and perform trajectory adaptive control optimization management based on the real feedback.

Images