Grinding and dry-type dust removal integrated cooperative system for washing machine tripod
By using an integrated system for grinding and dry dust removal for washing machine tripods, the geometric features of the grinding contact area are analyzed in real time and the dust escape trajectory is predicted. This solves the problem of low dust removal efficiency for workpieces with complex geometries and achieves efficient and precise dust capture and processing adaptability.
Patent Information
- Application Number
- CN202511627896.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-07
- Publication Date
- 2026-01-23
AI Technical Summary
Existing technologies are inefficient at removing dust when handling workpieces with complex geometries, such as washing machine tripods. They cannot effectively capture the complex and variable dust escape trajectories caused by the special geometry of the workpiece, leading to environmental pollution and product quality problems.
The system adopts an integrated collaborative system for grinding and dry dust removal for washing machine tripods. Through geometric feature recognition, trajectory prediction, visual feedback and collaborative control modules, it analyzes the geometric features of the grinding contact area in real time, predicts the dust escape trajectory, and controls the dust removal device to accurately intercept dust.
It enables proactive prediction and efficient capture of dust, improves dust removal efficiency, enhances the adaptability to processing workpieces with complex geometries, reduces dust pollution and the risk of secondary scratches, and improves product quality and production line automation.
Smart Images
Figure CN121374355A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of industrial automation control, in particular to a polishing and dry dust removal integrated collaborative system for a washing machine tripod. BACKGROUND
[0002] In the manufacturing industry of household appliances such as washing machines, the welding seam, edge and surface flatness of the washing machine tripod as a core load-bearing component directly affect the stability and service life of the whole machine. Therefore, using an industrial robot for automatic polishing is a key process to ensure the processing quality and consistency. During polishing, a large amount of metal dust is generated, which not only pollutes the production environment and harms the health of operators if not removed in time and effectively, but also scratches the workpiece surface due to secondary dust adhesion, affecting product quality.
[0003] To solve the dust problem, the existing technology usually adopts an integrated polishing and dust removal scheme. One is "cage dust removal", that is, a large negative pressure dust collection cover is arranged at the polishing station to cover the whole operation area. Although this scheme can collect most of the dust, it has problems such as large equipment size, high energy consumption, affecting the flexibility of the robot movement, and difficulty in handling local scattered dust.
[0004] Another is "accompanying dust removal", that is, the dust suction port is bound or simply servo-linked with the polishing head, so that the suction port follows the movement of the polishing head. However, this seemingly direct way is greatly discounted when dealing with workpieces such as washing machine tripods with complex geometric structures. The structural characteristics of the washing machine tripod are that it is not a simple plane, but is composed of multiple arc surfaces with different curvatures, grooves and convex ribs for reinforcement, and complex fillet transition zones at the branch arm intersection. When the high-speed rotating polishing head acts on these special geometric surfaces, the metal dust generated is not along the simple tangent direction, and the simple following dust removal port cannot predict and cover the complex and variable dust escape trajectories generated due to the special geometric shape of the workpiece, resulting in a large amount of dust leakage and low dust removal efficiency. Therefore, it is necessary to design a polishing and dry dust removal integrated collaborative system for a washing machine tripod with strong adaptability and high efficiency and precision. SUMMARY
[0005] The purpose of the present application is to provide a polishing and dry dust removal integrated collaborative system for a washing machine tripod to solve the problems raised in the background art.
[0006] In order to solve the above technical problems, the present application provides the following technical scheme: a polishing and dry dust removal integrated collaborative system for a washing machine tripod, comprising a polishing robot, a dust removal device, and a collaborative controller electrically connected with the polishing robot and the dust removal device, the system further comprising: A geometry feature recognition module is configured to store a digital twin model of the washing machine tripod and analyze geometric features and a surface normal vector of a contact area between the polishing head and the workpiece according to real-time pose data of the polishing robot. A trajectory prediction module is connected to the geometry feature recognition module and configured to calculate a prediction vector of a dust cloud core trajectory according to process parameters of the polishing robot, the geometric features of the contact area, and the surface normal vector. A visual feedback module is configured to identify an actual position of the dust cloud and compare the actual position with the prediction vector to calculate a prediction error. A cooperative control module is connected to the trajectory prediction module and the visual feedback module and configured to calculate an optimal spatial interception point for intercepting the dust cloud and control a suction port of the dust removal device to move to the optimal spatial interception point.
[0007] According to the above technical solution, the geometry feature recognition module includes a workpiece digital twin sub-module, a real-time pose mapping sub-module, and a contact area geometry analysis sub-module. The workpiece digital twin sub-module is configured to import and store an accurate three-dimensional CAD model of the washing machine tripod. The real-time pose mapping sub-module is configured to read each axis encoder data of the polishing robot, map a real-time spatial position and attitude of the polishing head to the digital twin model, and the contact area geometry analysis sub-module is configured to calculate a contact point between the polishing head and the surface of the tripod in real time according to the mapped pose, and analyze key geometric features and a surface normal vector of the contact area, wherein the geometric features include a plane, a curved surface curvature, a groove, a reinforcing rib, and a weld.
[0008] According to the above technical solution, the trajectory prediction module includes an initial jet vector calculation sub-module and a geometric interference and reflection algorithm sub-module. The initial jet vector calculation sub-module is configured to receive process parameters, combine the surface normal vector of the contact area, and calculate an initial jet vector of dust departing from the workpiece. The geometric interference and reflection algorithm sub-module is configured to receive the contact area features output by the geometry feature recognition module, simulate and analyze deflection and reflection of dust flow after being interfered by a geometric structure, and calculate a prediction vector of a dust cloud core trajectory.
[0009] According to the above technical solution, the visual feedback module includes a high-speed image acquisition sub-module, a dust cloud centroid identification sub-module, a prediction error analysis sub-module, and a prediction model correction sub-module. The high-speed image acquisition sub-module is configured to deploy an industrial camera above a polishing area side to capture images of the dust cloud in real time. The dust cloud centroid identification sub-module is configured to calculate and determine an actual position and diffusion form of the dust cloud in real time. The prediction error analysis sub-module is configured to compare the output prediction position with the identified actual position to calculate a prediction error. The prediction model correction sub-module is configured to input long-term accumulated prediction error data as input, periodically fine-tune weight coefficients of algorithms in the trajectory prediction module through a machine learning algorithm.
[0010] According to the technical solution, the cooperative control module comprises an optimal interception point calculation submodule and a servo drive submodule, the optimal interception point calculation submodule is used to calculate the optimal spatial position for accurately intercepting the dust cloud and the advance amount of the suction port orientation, and the servo drive submodule is used to control the driving of the dust suction port to move quickly and smoothly to the optimal interception point and always maintain the optimal dust suction posture.
[0011] A working method of a polishing and dry dust removal integrated cooperative system for a washing machine tripod, the method comprising: Step S1: loading a digital twin model of the washing machine tripod; Step S2: during the polishing robot operation, real-time analysis of the geometric characteristics and surface normal vector of the polishing contact area; Step S3: according to the geometric characteristics of the polishing contact area and the process parameters, predicting the dust escape trajectory and generating a prediction vector; Step S4: calculating the optimal interception point and controlling the dust suction port to move to the point in advance; Step S5: capturing the actual dust trajectory through the vision system and comparing it with the generated prediction vector, calculating the prediction error, and adaptively correcting the trajectory prediction process in step S3 according to the accumulated prediction error.
[0012] According to the technical solution, the method of real-time analysis of the geometric characteristics and surface normal vector of the polishing contact area in step S2 specifically comprises: Step S21: high-frequency acquisition of the encoder data of each joint axis of the polishing robot, and solving the real-time spatial pose of the polishing head through the forward kinematics model; Step S22: mapping the real-time spatial pose to the digital twin model, and determining the instantaneous contact point through the collision detection algorithm; Step S23: taking the determined instantaneous contact point as the center, determining a neighborhood range with a preset radius on the digital twin model, and collecting the vertex coordinates and face normal vector data of all triangular facets in the neighborhood; Step S24: weighted average calculation of the normal vectors of the triangular facet where the instantaneous contact point is located and the adjacent triangular facets sharing the vertex; Step S25: calculation of the dispersion of the normal vectors of all triangular facets in the neighborhood range; if the dispersion is less than a first preset threshold, it is preliminarily determined that the geometric characteristics of the contact area are planar; if the dispersion is greater than the first preset threshold, it is preliminarily determined that the geometric characteristics are curved; Step S26: Establish a local coordinate system with the determined instantaneous contact point as the origin and the calculated accurate surface normal vector as the Z-axis; calculate the Z-axis coordinate values of all vertices in the neighborhood in the local coordinate system, i.e. the elevation difference; if the absolute values of the elevation differences of all vertices are less than a second preset threshold, it is finally determined that the geometric feature is a plane; if there is a set of continuous points in the neighborhood with negative elevation differences and the absolute values of which are greater than the second preset threshold, it is determined that the geometric feature is an inner groove; if there is a set of continuous points in the neighborhood with positive elevation differences and the values of which are greater than the second preset threshold, it is determined that the geometric feature is a reinforcing convex rib; Step S27: Encode the finally recognized geometric feature and package it together with the accurate surface normal vector calculated in step S24 for output.
[0013] According to the above technical solution, the method for predicting the dust escape trajectory in step S3 specifically includes: Step S31: According to the polishing head linear velocity vector and the contact point surface normal vector , calculate the initial spray vector by the formula , wherein and are preset physical coefficients; Step S32: Obtain the recognized geometric feature code G and its key direction vector ; Step S33: Substitute the initial spray vector , the geometric feature code G, and the direction vector into the trajectory prediction function to calculate the final prediction vector , wherein is a deflection function based on the geometric feature, and is a weight coefficient that can be corrected by step S5.
[0014] According to the above technical solution, in step S4, the method for calculating the best interception point and controlling the dust removal suction port specifically includes: Step S41: Obtain the current polishing contact point position and set the total response delay time of the system ; Step S42: Calculate the best spatial interception point by the formula ; Step S43: Control the suction port of the dust removal device to move to the best spatial interception point and adjust its posture.
[0015] According to the above technical solution, in step S33, the final prediction vector is calculated by the trajectory prediction function The methods specifically include: Step S331: Obtain and determine the type of the geometric feature code G output in step S32 to determine the deflection model to be called; Step S332: If the geometric feature code G is a planar feature, then the deflection function... The output is set to a zero vector, making the final predicted vector... Directly equal to the initial injection vector ; Step S333: If the geometric feature code G is an inner groove feature, then the deflection function... The output is set to a vector that corresponds to the direction of the groove. Guide vectors in the same direction, and through the formula Vector superposition was performed to simulate the physical effect of dust flow being constrained by the sidewalls of the groove and accelerated outward along its direction; Step S334: If the geometric feature code G is a reinforcing rib feature, then the deflection function... The output is set to a surface normal vector at the contact point of the rib. Repulsive vectors in the same direction, and through the formula Vector superposition was performed to simulate the physical effect of dust flow being blocked by the windward side of the rib and reflected outward in the direction of its normal.
[0016] Compared with existing technologies, the beneficial effects achieved by this invention are as follows: By incorporating a geometric feature recognition module, a trajectory prediction module, a visual feedback module, and a collaborative control module, this invention can analyze the complex geometric features of the grinding contact area in real time based on the real-time pose data of the grinding robot and the stored digital twin model of the washing machine tripod. It can also analyze and calculate the optimal escape trajectory prediction vector for dust, achieving intelligent collaborative control of the dust removal device's suction port. This ensures that the movement of the dust removal suction port is no longer a simple, lagging follow of the grinding head, but rather, based on the predicted trajectory and system response delay, different interception strategies are formed, calculating and moving in advance to the optimal spatial interception point, allowing the suction port to accurately intercept dust upon arrival. This achieves the purpose of proactively predicting and efficiently capturing dust that escapes due to the complex geometry of the tripod. It is flexible in use and enhances the adaptability to processing irregularly shaped workpieces. Simultaneously, through continuous correction of the prediction model via visual feedback, it retains adaptive redundancy to changes in working conditions, achieving a highly adaptable and efficient collaborative dust removal effect. Attached Figure Description
[0017] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1This is a schematic diagram of the system module composition of the present invention. Detailed Implementation
[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0019] Please see Figure 1 This invention provides a technical solution: an integrated collaborative system for sanding and dry dust removal for washing machine tripods, comprising a sanding robot, a dust removal device, and a collaborative controller electrically connected to the sanding robot and the dust removal device. The system also includes: The geometric feature recognition module is used to store the digital twin model of the washing machine tripod and to parse the geometric features and surface normal vector of the contact area between the grinding head and the workpiece based on the real-time pose data of the grinding robot. The trajectory prediction module, connected to the geometric feature recognition module, is used to calculate the predicted vector of the core trajectory of the dust cloud based on the process parameters of the grinding robot, the geometric features of the contact area, and the surface normal vector. The visual feedback module is used to identify the actual location of the dust cloud and compare it with the prediction vector to calculate the prediction error. The collaborative control module, connected to the trajectory prediction module and the visual feedback module, is used to calculate the optimal spatial interception point for intercepting dust clouds and control the suction port of the dust removal device to move to the optimal spatial interception point. By incorporating geometric feature recognition, trajectory prediction, visual feedback, and collaborative control modules, the complex geometric features of the grinding contact area can be analyzed in real time based on the real-time pose data of the grinding robot and the stored digital twin model of the washing machine tripod. The optimal escape trajectory prediction vector for dust is calculated, enabling intelligent collaborative control of the dust removal device's suction port. This allows the suction port to move beyond simply lagging behind the grinding head; instead, it employs different interception strategies based on the predicted trajectory and system response delay, calculating and pre-moving to the optimal spatial interception point. This ensures precise interception of dust upon arrival, proactively predicting and efficiently capturing dust scattered due to the tripod's complex geometry. This approach is flexible and enhances adaptability to irregularly shaped workpieces. Furthermore, the continuous correction of the prediction model through visual feedback retains adaptive redundancy to changing working conditions, achieving highly adaptable and efficient collaborative dust removal.
[0020] The geometric feature recognition module includes a workpiece digital twin sub-module, a real-time pose mapping sub-module, and a contact area geometric analysis sub-module. The workpiece digital twin sub-module is used to import and store the accurate three-dimensional CAD model of the washing machine tripod. The real-time pose mapping sub-module is used to read the encoder data of each axis of the polishing robot, map the real-time spatial position and attitude of the polishing head to the digital twin model, and analyze the key geometric features and surface normal vectors of the contact area according to the mapped pose. The geometric features include planes, curved surfaces, grooves, reinforcing ribs, and welds.
[0021] The trajectory prediction module includes an initial jet vector calculation sub-module and a geometric interference and reflection algorithm sub-module. The initial jet vector calculation sub-module is used to receive process parameters and calculate the initial jet vector of the dust detached from the workpiece in combination with the surface normal vector of the contact area. The geometric interference and reflection algorithm sub-module is used to receive the contact area features output by the geometric feature recognition module and simulate and analyze the deflection and reflection of the dust flow after being interfered by the geometric structure to calculate the dust cloud core trajectory prediction vector.
[0022] The visual feedback module includes a high-speed image acquisition sub-module, a dust cloud centroid identification sub-module, a prediction error analysis sub-module, and a prediction model correction sub-module. The high-speed image acquisition sub-module is used to deploy an industrial camera above the polishing area to capture images of the dust cloud in real time. The dust cloud centroid identification sub-module is used to calculate and determine the actual position and diffusion form of the dust cloud in real time. The prediction error analysis sub-module is used to compare the predicted position with the actual position to calculate the prediction error. The prediction model correction sub-module is used to input the long-term accumulated prediction error data, adjust the weight coefficients of the algorithm in the trajectory prediction module periodically through machine learning algorithms such as gradient descent, and make the prediction model self-learn and evolve to adapt to the slight differences of different batches of workpieces or changes caused by new polishing consumables.
[0023] The collaborative control module includes an optimal interception point calculation sub-module and a servo drive sub-module. The optimal interception point calculation sub-module is used to calculate the optimal spatial position for accurately intercepting the dust cloud and the advance amount of the suction port orientation. The servo drive sub-module is used to control the driving of the dust suction port to move quickly and smoothly to the optimal interception point and always maintain the optimal dust suction attitude.
[0024] A working method of a polishing and dry dust removal integrated collaborative system for washing machine tripods, the method comprising: Step S1: loading a digital twin model of a washing machine tripod; Step S2: during the operation of the polishing robot, real-time analysis of the geometric features and surface normal vectors of the polishing contact area; Step S3: According to the geometric characteristics and process parameters of the polishing contact area, the dust escape trajectory is predicted, and a prediction vector is generated; Step S4: Calculate the optimal interception point, and control the dust removal suction port to move to the point in advance; Step S5: Capture the actual dust trajectory through the vision system, compare it with the generated prediction vector, calculate the prediction error, and adaptively correct the trajectory prediction process in step S3 according to the accumulated prediction error.
[0025] The method for real-time analyzing the geometric characteristics of the polishing contact area and the surface normal vector in step S2 specifically includes: Step S21: High-frequency acquisition of encoder data of each joint shaft of the polishing robot, and solving of the real-time spatial pose of the polishing head through the forward kinematics model; Step S22: Mapping the real-time spatial pose to the digital twin model, and determining the instantaneous contact point through the collision detection algorithm; In the embodiment of the present application, the collision detection algorithm checks whether the geometric surface (usually composed of a large number of triangular facets) of the polishing head model intersects or overlaps with the geometric surface of the washing machine tripod static model fixed in the virtual space and the polishing head dynamic model constantly updated in the virtual space at a high frequency, and outputs the specific position of intersection when the two models intersect, thereby providing an accurate target position for subsequent steps; Step S23: Determining a neighborhood range with a preset radius centered on the determined instantaneous contact point on the digital twin model, and collecting the vertex coordinates and face normal vector data of all triangular facets in the neighborhood; Step S24: Weighted average calculation of the normal vector of the triangular facet where the instantaneous contact point is located and the normal vector of the adjacent triangular facets sharing the vertexes thereof, to obtain an accurate surface normal vector that is smoothed and more representative of the true surface normal of the point; Step S25: Calculation of the dispersion of the normal vectors of all triangular facets in the neighborhood; if the dispersion is less than a first preset threshold, it is preliminarily determined that the geometric characteristics of the contact area are planar; if the dispersion is greater than the first preset threshold, it is preliminarily determined that the geometric characteristics are curved; Step S26: Establishing a local coordinate system with the determined instantaneous contact point as the origin and the calculated accurate surface normal vector as the Z-axis; calculating the Z-axis coordinate values, i.e. the elevation differences, of all vertices in the neighborhood in the local coordinate system; if the absolute values of the elevation differences of all vertices are less than a second preset threshold, it is finally determined that the geometric characteristics are planar; if there is a set of continuous points in the neighborhood with negative elevation differences and absolute values greater than the second preset threshold, it is determined that the geometric characteristics are an inner groove; if there is a set of continuous points in the neighborhood with positive elevation differences and values greater than the second preset threshold, it is determined that the geometric characteristics are a reinforcing convex rib; Step S27: Encode the identified geometric features and pack them together with the accurate surface normal vector calculated in step S24 for output.
[0026] The method of predicting the dust escape trajectory in step S3 specifically includes: Step S31: Calculate the initial spray vector according to the polishing head linear velocity vector and the contact point surface normal vector by the formula wherein and are preset physical coefficients; Step S32: Obtain the identified geometric feature code G and its key direction vector ; Step S33: Substitute the initial spray vector , the geometric feature code G, and the direction vector into the trajectory prediction function to calculate the final prediction vector , wherein is a geometric feature-based deflection function, and is a weight coefficient that can be corrected by step S5.
[0027] The method of calculating the optimal interception point and controlling the dust removal suction port in step S4 specifically includes: Step S41: Obtain the current polishing contact point position and set the total response delay time of the system ; Step S42: Calculate the optimal spatial interception point by the formula ; Step S43: Control the suction port of the dust removal device to move to the optimal spatial interception point and adjust its posture; The method of controlling the suction port of the dust removal device to move to and adjusting its posture in step S43 specifically includes: Step S431: Input the coordinate values of the optimal spatial interception point as the target point into the cooperative controller, and the cooperative controller plans the optimal motion trajectory from the current position to the target point and drives the joint servo motor to control the suction port to move along the trajectory to the position of the center point of the optimal spatial interception point ; Step S432: Obtain the final prediction vector calculated in step S33, and calculate The posture vector in the opposite direction; taking the posture vector as a target posture, instructing the end wrist joint of the dust removal device servo mechanical arm to adjust until the opening normal direction of the suction port aligns with the posture vector, thereby realizing the optimal interception posture of the suction port directly facing the predicted airflow direction.
[0028] In step S33, the final prediction vector is calculated by the trajectory prediction function The method specifically comprises: Step S331: Obtain and judge the type of the geometric feature code G output in step S32 to determine the called deflection model; Step S332: If the geometric feature code G is a plane feature, set the output of the deflection function to a zero vector, so that the final prediction vector is directly equal to the initial spraying vector ; Step S333: If the geometric feature code G is an inner groove feature, set the output of the deflection function to a guide vector in the same direction as the groove trend vector , and perform vector superposition through the formula to simulate the physical effect that the dust flow is constrained by the side wall of the groove and sprayed along the trend; Step S334: If the geometric feature code G is a reinforcing rib feature, set the output of the deflection function to a repulsion vector in the same direction as the surface normal vector of the contact point of the reinforcing rib , and perform vector superposition through the formula to simulate the physical effect that the dust flow is blocked by the windward surface of the reinforcing rib and reflects and splashes to the outside of the normal direction thereof; The application solves the two core technical contradictions of dust removal when polishing a workpiece with a complex geometric structure such as a washing machine tripod, namely: first, the inner groove, reinforcing rib, and other structures of the tripod will cause the high-speed splashing metal dust to produce unpredictable reflection and turning, resulting in complex and variable dust trajectories; second, the dust suction port as a mechanical execution mechanism has physical delays in response speed and movement speed, and can never catch up with the dust cloud that has been generated and dispersed in milliseconds in real time.
[0029] In view of the above contradictions, the present application can analyze the geometric feature currently contacted by the polishing head from the digital twin model in real time by setting a geometric feature recognition module; at the same time, the present application introduces the analysis of the kinematics of the polishing robot itself and the mechanical delay of the dust removal device. When the trajectory prediction module calculates the predicted trajectory of the dust sprayed along the notch at high speed according to the geometric feature (such as identifying as “inner groove”), the cooperative control module does not immediately instruct the suction port to chase the trajectory. On the contrary, the system will cooperatively calculate in combination with the known system response delay time, and when it is identified that the delay time is long and the dust prediction speed is extremely fast, the cooperative control module will dynamically calculate a best spatial interception point with an advance. Conversely, if the system delay time is short and the dust speed is slow, the advance will be shortened accordingly. Further, the present application achieves a dynamic balance between more accurate and timely dust trajectory prediction and overcoming the inherent physical delay of the mechanical system. Compared with the traditional follow-up dust removal that simply binds the suction port with the polishing head, the present application can additionally increase the cooperative judgment standard for the complex geometric features of the workpiece and the delay time of the system itself. When the system is polishing a groove or other special parts, no matter how fast the polishing head moves, the cooperative control module can be triggered to preferentially execute the advance interception instruction to drive the suction port to move to the meeting point in advance and adjust the posture of the suction port to be directly opposite the dust inflow direction. Thus, the problem of dust pollution in the workshop is greatly reduced, the risk of secondary scratches of the processed surface by dust is effectively reduced, and the automation level of the production line and the product excellent rate are improved.
[0030] The present application is described with reference to flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as combinations of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing apparatus to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing apparatus generate a means for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in a flow or multiple flows and / or blocks Figure 1 The functions specified in a flow or multiple flows and / or blocks
[0031] These computer program instructions can also be stored in a computer-readable memory capable of guiding a computer or other programmable data processing apparatus to work in a specific manner, so that the instructions stored in the computer-readable memory produce a product including instruction means, which implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in a flow or multiple flows and / or blocks Figure 1 The functions specified in a flow or multiple flows and / or blocks
[0032] These computer program instructions can also be loaded into a computer or other programmable data processing devices, so that a series of operational steps are generated to realize the computer-implemented processes, and the instructions executed on the computer or other programmable devices provide a process for implementing the functions specified in the flowchart Figure 1 one flow or multiple flows and / or the functions specified in the block Figure 1 one flow or multiple flows and / or the functions specified in the block
[0033] The embodiments of the present application are described above with reference to the accompanying drawings, but the present application is not limited to the specific embodiments described above, and the specific embodiments described above are merely illustrative, but not restrictive, and those of ordinary skill in the art can make many forms under the inspiration of the present application without departing from the purpose of the present application and the scope protected by the claims, which are all within the protection of the present application.
Claims
1. A polishing and dry dust removal integrated collaborative system for a washing machine tripod, comprising a polishing robot, a dust removal device, and a collaborative controller electrically connected with the polishing robot and the dust removal device, characterized in that, The system further comprises: a geometric feature recognition module for storing a digital twin model of the washing machine tripod and analyzing geometric features and surface normal vectors of a contact area of the polishing head and the workpiece according to real-time pose data of the polishing robot; a trajectory prediction module connected with the geometric feature recognition module, for calculating a prediction vector of a dust cloud core trajectory according to process parameters of the polishing robot, geometric features of the contact area, and surface normal vectors; a visual feedback module for identifying an actual position of the dust cloud and comparing the actual position with the prediction vector to calculate a prediction error; a collaborative control module connected with the trajectory prediction module and the visual feedback module, for calculating an optimal spatial interception point for intercepting the dust cloud and controlling a suction port of the dust removal device to move to the optimal spatial interception point.
2. The polishing and dry dusting integrated collaborative system for washing machine tripod according to claim 1, characterized in that: The geometric feature recognition module comprises a workpiece digital twin sub-module, a real-time pose mapping sub-module, and a contact area geometric analysis sub-module. The workpiece digital twin sub-module is configured to import and store an accurate three-dimensional CAD model of the washing machine tripod. The real-time pose mapping sub-module is configured to read data of each axis encoder of the polishing robot, map a real-time spatial position and attitude of the polishing head to the digital twin model, and calculate a contact point of the polishing head and the surface of the tripod in real time, and analyze key geometric features and surface normal vectors of the contact area. The geometric features include a plane, a curved surface curvature, a groove, a reinforcing rib, and a weld.
3. The polishing and dry dusting integrated collaborative system for washing machine tripod according to claim 1, characterized in that: The trajectory prediction module comprises an initial jet vector calculation sub-module and a geometric interference and reflection algorithm sub-module. The initial jet vector calculation sub-module is configured to receive process parameters, combine the surface normal vectors of the contact area, and calculate an initial jet vector of dust departing from the workpiece. The geometric interference and reflection algorithm sub-module is configured to receive the contact area features output by the geometric feature recognition module, simulate and analyze deflection and reflection of dust flow after the dust flow is interfered by a geometric structure, and calculate a prediction vector of a dust cloud core trajectory.
4. The polishing and dry dusting integrated collaborative system for washing machine tripod according to claim 1, characterized in that: The visual feedback module comprises a high-speed image acquisition sub-module, a dust cloud centroid identification sub-module, a prediction error analysis sub-module, and a prediction model correction sub-module. The high-speed image acquisition sub-module is configured to deploy an industrial camera above a polishing area side to capture images of the dust cloud in real time. The dust cloud centroid identification sub-module is configured to calculate and determine an actual position and diffusion form of the dust cloud in real time. The prediction error analysis sub-module is configured to compare the output prediction position with the identified actual position to calculate a prediction error. The prediction model correction sub-module is configured to input long-term accumulated prediction error data, periodically fine-tune weight coefficients of algorithms in the trajectory prediction module through a machine learning algorithm.
5. The polishing and dry dusting integrated collaborative system for washing machine tripod according to claim 1, characterized in that: The collaborative control module comprises an optimal interception point calculation sub-module and a servo drive sub-module. The optimal interception point calculation sub-module is configured to calculate an optimal spatial position for accurately intercepting the dust cloud and an advance amount of a suction port orientation. The servo drive sub-module is configured to control the dust removal suction port to move quickly and smoothly to the optimal interception point and always maintain an optimal dust suction posture.
6. A method for using the polishing and dry dust removal integrated collaborative system for the washing machine tripod according to any one of claims 1-5, characterized in that: The method comprises: Step S1: load the digital twin model of the washing machine tripod; Step S2: during the polishing robot operation, real-time analyze the geometric characteristics and surface normal vector of the polishing contact area; Step S3: according to the geometric characteristics of the polishing contact area and the process parameters, predict the dust escape trajectory, and generate a prediction vector; Step S4: calculate the optimal interception point, and control the dust removal suction port to move to the point in advance; Step S5: capture the actual dust trajectory through the vision system, compare it with the generated prediction vector, calculate the prediction error, and adaptively correct the trajectory prediction process in step S3 according to the accumulated prediction error.
7. The polishing and dry dusting integrated collaborative system for washing machine tripod according to claim 6, characterized in that: The method of real-time analyzing the geometric characteristics and surface normal vector of the polishing contact area in step S2 specifically includes: Step S21: high-frequency acquisition of the encoder data of each joint axis of the polishing robot, and solving the real-time spatial pose of the polishing head through the forward kinematics model; Step S22: mapping the real-time spatial pose to the digital twin model, and determining the instantaneous contact point through the collision detection algorithm; Step S23: taking the determined instantaneous contact point as the center, determining a neighborhood range with a preset radius on the digital twin model, and collecting the vertex coordinates and face normal vector data of all triangular facets in the neighborhood; Step S24: weighted average calculation of the normal vector of the triangular facet where the instantaneous contact point is located and the adjacent triangular facets sharing the vertex; Step S25: calculate the dispersion of the normal vector of all triangular facets in the neighborhood; if the dispersion is less than a first preset threshold, it is preliminarily determined that the geometric characteristics of the contact area are planar; if the dispersion is greater than the first preset threshold, it is preliminarily determined that it is a curved surface; Step S26: taking the determined instantaneous contact point as the origin and the calculated accurate surface normal vector as the Z-axis, a local coordinate system is established; calculate the Z-axis coordinate value of all vertices in the neighborhood in the local coordinate system, that is, the elevation difference; if the absolute value of the elevation difference of all vertices is less than a second preset threshold, it is finally determined that the geometric characteristics are planar; if there is a set of continuous points in the neighborhood with negative elevation difference and its absolute value greater than the second preset threshold, it is determined that the geometric characteristics are an inner groove; if there is a set of continuous points in the neighborhood with positive elevation difference and its value greater than the second preset threshold, it is determined that the geometric characteristics are a reinforcing convex rib; Step S27: encode the finally identified geometric characteristics, and package and output the accurate surface normal vector calculated in step S24.
8. The polishing and dry dusting integrated collaborative system for washing machine tripod according to claim 6, characterized in that: The method of predicting the dust escape trajectory in step S3 specifically includes: Step S31: Calculate initial spray vector according to polishing head linear velocity vector and contact point surface normal vector , by formula wherein and are preset physical coefficients; Step S32: Obtain the recognized geometric feature code G and its key direction vector ; Step S33: The initial spray vector , the geometric feature code G, the direction vector is substituted into the trajectory prediction function , the final prediction vector is calculated wherein is a geometric feature-based deflection function, is a weight coefficient that can be corrected by step S5.
9. The polishing and dry dusting integrated collaborative system for washing machine tripod according to claim 8, characterized in that: In step S4, the method of calculating the optimal interception point and controlling the dust removal suction port specifically includes: Step S41: Obtain the current polishing contact point position and set the total response delay time of the system ; Step S42: Calculate the optimal spatial intercept point by formula ; and ; Step S43: control the suction port of the dust removal device to move to the optimal spatial intercept point and adjust its attitude.
10. The polishing and dry dusting integrated collaborative system for washing machine tripod according to claim 8, characterized in that: In the step S33, the final prediction vector is calculated by the trajectory prediction function The method specifically comprises: Step S331: acquire and determine the type of geometric characteristic code G output in step S32 to determine the called deflection model; Step S332: if the geometric feature code G is a planar feature, then set the output of the deflection function to a zero vector, so that the final prediction vector is directly equal to the initial spray vector ; Step S333: If the geometric feature code G is an inner groove feature, the output of the deflection function is set to a guide vector in the same direction as the groove's tangent vector, and vector superposition is performed by the formula to simulate the physical effect that the dust flow is constrained by the side wall of the groove and ejected along the tangent direction of the groove. Step S334: If the geometric feature code G is a reinforced rib feature, set the output of the deflection function to a repulsion vector that is co-directional with the surface normal vector of the point of contact with the rib, and perform vector addition with the formula to simulate the physical effect of the dust stream being blocked by the windward face of the rib and reflecting and spilling out the side of the normal direction.