Bucket arm robot multi-scene grinding method and system based on adaptive parameter mapping

Abstract

The invention relates to a bucket rod robot multi-scene grinding method and system based on self-adaptive parameter mapping, and relates to the field of bucket rod robot grinding. According to the method, a technological parameter knowledge base is constructed, a Gaussian mixture model is used for parameter matching, it is ensured that the most suitable technological parameters can be rapidly selected in different scenes, the precision and consistency of polishing operation are greatly improved through the precise parameter matching and adjusting mechanism, and errors and rework caused by parameter mismatching are reduced; through application of an adaptive algorithm, process parameters can be dynamically adjusted according to real-time feedback and scene change, so that the requirement of manual intervention is reduced, the operation period is shortened, and the productivity is improved; by monitoring the grinding process in real time and adjusting parameters in time, it can be ensured that expected machining quality can be achieved under various conditions, the real-time monitoring and feedback mechanism is beneficial to finding and correcting problems in time, accumulation of quality problems is avoided, and the reliability and durability of products are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of boom robot grinding, and in particular to a boom robot multi-scenario grinding method and system based on adaptive parameter mapping. Background Art

[0002] In the field of precision machining in the manufacturing industry, robot grinding technology has become a key means to improve production efficiency, ensure product quality, and reduce labor costs. This technology uses industrial robots combined with special grinding tools to grind and polish the surface of workpieces to achieve the purpose of improving surface quality and precision. With the rapid development of science and technology, robot grinding technology has been widely used in many fields such as automobile manufacturing, aerospace, and shipbuilding.

[0003] However, existing boom robots suffer from errors and rework caused by parameter mismatch during multi-scenario grinding, long operation cycles, low productivity, and accumulation of quality problems, which reduce product reliability and durability. Summary of the invention

[0004] The purpose of the present invention is to provide a multi-scenario grinding method and system for a boom robot based on adaptive parameter mapping to solve the problems existing in the above-mentioned prior art.

[0005] To achieve the above object, the technical solution adopted by the present invention is:

[0006] In a first aspect, the present invention provides a multi-scenario grinding method of a stick robot based on adaptive parameter mapping, the method comprising:

[0007] S1. Construct a process parameter knowledge base K, wherein the process parameter knowledge base K is used to store process parameters of different end tools, wherein the process parameters include at least one of a force control parameter F, a speed parameter V or a position parameter P;

[0008] S2. Quickly match the process parameters through the mapping model M;

[0009] S3, before the polishing starts, dynamically adjusting the process parameters according to different scene features S in combination with the adaptive algorithm A;

[0010] S4. In response to the grinding process, adjust the parameter F′ based on real-time monitoring R and feedback to improve the grinding efficiency E and processing quality Q.

[0011] In a possible implementation, in step S1, the process parameter knowledge base K can automatically update and optimize parameters according to historical data and feedback, and the process parameter knowledge base K satisfies:

[0012] K = {F, V, P}.

[0013] In a possible implementation, in step S2, the mapping model M is a probability matching model based on a Gaussian mixture model GMM, which can provide accurate parameter matching even when the data is incomplete or noisy, and the mapping model M satisfies:

[0014] M(θ)=GMM(θ;π,μ,Σ);

[0015] Among them, θ is the process parameter of the end tool, π is the mixing coefficient, μ is the mean vector, and Σ is the covariance matrix.

[0016] In a possible implementation, in step S3, the adaptive algorithm A can predict the influence of parameter adjustment on the polishing effect and optimize the parameters accordingly, and the adaptive algorithm A satisfies:

[0017] A(K, S) = K′;

[0018] Wherein, K′ is the adjusted process parameter.

[0019] In a possible implementation, in step S4, the feedback adjustment parameter F′ can be adjusted in real time according to the monitoring data of the real-time monitoring R to optimize the grinding efficiency E and the processing quality Q.

[0020] In a possible implementation, the polishing efficiency E satisfies:

[0021] E = f(R, F′);

[0022] Where f is the efficiency improvement function.

[0023] In a possible implementation, the processing quality Q satisfies:

[0024] Q = g(R, F′);

[0025] Among them, g is the quality improvement function.

[0026] In a possible implementation, the method further includes:

[0027] In response to the start of grinding, the stick robot is modeled using DH parameters.

[0028] In a possible implementation, the method further includes:

[0029] In response to the grinding process, inverse kinematics is used to solve the robot joint angles.

[0030] In a possible implementation, the method further includes:

[0031] In response to the grinding being completed, the position of the end tool of the boom robot is corrected to compensate for the position change of the end tool caused by wear or thermal expansion.

[0032] In a second aspect, the present invention provides a multi-scenario grinding system of a stick robot based on adaptive parameter mapping, the system is applied to the multi-scenario grinding method of a stick robot based on adaptive parameter mapping as described above, and the system comprises:

[0033] A construction module, used to construct a process parameter knowledge base K, wherein the process parameter knowledge base K is used to store process parameters of different end tools, wherein the process parameters include at least one of a force control parameter F, a speed parameter V or a position parameter P;

[0034] A matching module, used for quickly matching the process parameters through a mapping model M;

[0035] An adjustment module, for dynamically adjusting the process parameters in response to different scene features S and in combination with an adaptive algorithm A before the polishing starts;

[0036] The control module is used to respond to the grinding process, based on real-time monitoring R and feedback adjustment parameters F', to improve the grinding efficiency E and processing quality Q.

[0037] The beneficial effects brought about by the technical solution provided by the present invention include at least:

[0038] By building a process parameter knowledge base and using Gaussian mixture models for parameter matching, it is ensured that the most appropriate process parameters can be quickly selected in different scenarios. This precise parameter matching and adjustment mechanism greatly improves the accuracy and consistency of grinding operations and reduces errors and rework caused by parameter mismatches. The application of adaptive algorithms enables process parameters to be dynamically adjusted according to real-time feedback and scene changes, thereby reducing the need for manual intervention. This automated parameter adjustment mechanism significantly improves the efficiency of grinding operations, shortens the operation cycle, and improves productivity. By real-time monitoring of the grinding process and timely adjustment of parameters, it can ensure that the expected processing quality can be achieved under various conditions. This real-time monitoring and feedback mechanism helps to discover and correct problems in a timely manner, avoid the accumulation of quality problems, and improve product reliability and durability. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0040] Figure 1 A schematic flow chart of a multi-scenario grinding method of a stick robot based on adaptive parameter mapping provided by an exemplary embodiment of the present invention is shown.

[0041] Figure 2 A structural block diagram of a boom robot multi-scenario grinding system based on adaptive parameter mapping provided by an exemplary embodiment of the present invention is shown. DETAILED DESCRIPTION

[0042] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0044] Figure 1 A schematic flow chart of a multi-scenario grinding method of a stick robot based on adaptive parameter mapping provided by an exemplary embodiment of the present invention is shown. The multi-scenario grinding method of a stick robot based on adaptive parameter mapping comprises the following steps:

[0045] Step S1, constructing a process parameter knowledge base K, where the process parameter knowledge base K is used to store process parameters of different end tools, and the process parameters include at least one of a force control parameter F, a speed parameter V or a position parameter P.

[0046] Specifically, the process parameter knowledge base K satisfies: K = {F, V, P}. This means that for each specific task or application, the corresponding force control settings, movement speed and target position information are recorded. The force control parameter F refers to the magnitude and direction of the force applied to the workpiece during the operation; the speed parameter V determines the speed at which the end effector moves; and the position parameter P specifies the specific coordinate point where the end effector should reach. These three types of parameters are key factors in ensuring that the robot can accurately complete the task. For example, in application scenarios such as grinding, assembly or welding, precise control of these parameters can help achieve high-quality work results.

[0047] Furthermore, the process parameter knowledge base K can automatically update and optimize parameters based on historical data and feedback; that is, the process parameter knowledge base is optimized using a machine learning algorithm, and the machine learning algorithm should be able to identify and utilize the complex relationships between process parameters to optimize the knowledge base. Specifically, this function relies on machine learning algorithms to process large amounts of historical data and real-time feedback information. By analyzing past operation records, including success stories and failure lessons, the machine learning model can identify which parameter combinations are most conducive to improving work efficiency and product quality. In addition, the algorithm can also capture potential connections between different parameters, such as how the force control parameter F affects the final product quality, or how the speed parameter V changes the processing time, etc. Such complex relationships are often not obvious, but they are crucial to optimizing the entire production process.

[0048] In one example, considering that continuous variables (such as force value, speed, etc.) need to be processed and it is desired to establish a mapping relationship between input and output, regression analysis or support vector machine (SVM) algorithm is preferably used.

[0049] Step S2: quickly matching process parameters through mapping model M.

[0050] Specifically, the mapping model M satisfies: M(θ)=GMM(θ; π, μ, Σ); where θ is the process parameter of the end tool, such as force control, speed or position; π is the mixing coefficient, which reflects the importance of each component; μ is the mean vector, indicating the center point of each Gaussian distribution; Σ is the covariance matrix, which describes the degree to which the data is distributed around its mean.

[0051] Furthermore, the mapping model M is a probability matching model based on the Gaussian mixture model GMM, which can provide accurate parameter matching even when the data is incomplete or noisy. In addition, the Gaussian mixture regression model is applied to predict the polishing quality during the polishing process, and the Gaussian mixture regression model should be able to handle multivariate inputs, provide statistical predictions about the polishing quality, and automatically adjust the parameters of the Gaussian mixture model according to the polishing effect. Specifically, as a probability matching model, GMM can handle data sets containing uncertainty and noise, and can provide relatively accurate parameter predictions even in the case of missing data. This is because GMM assumes that the observed data is generated by a mixture of multiple different Gaussian distributions, each of which has its own shape (determined by the covariance matrix) and position (determined by the mean vector), and the mixing coefficient determines the proportion of each distribution's contribution to the overall data. When new process parameters need to be matched, GMM calculates the probability that the new data point belongs to each Gaussian distribution, and selects the one with the highest probability as the best matching result.

[0052] In addition, GMM can be extended to Gaussian Mixture Regression (GMR). In this case, not only input variables (such as process parameters) are considered, but also output variables (such as polishing quality). GMR combines the advantages of GMM and is able to establish complex nonlinear relationships between input variables and output variables. It is particularly suitable for situations where the input-output relationship is complex and can be better represented by a mixture of multiple Gaussian distributions. Applying GMR in the polishing process means that the surface finish or other quality indicators after polishing can be predicted based on the process parameters currently used. More importantly, since GMR allows for the processing of multivariate inputs, it can consider the influence of multiple factors at the same time, thereby giving a more comprehensive and accurate quality prediction. In addition, as the polishing process proceeds, GMR can also automatically adjust the parameters in GMM based on the actual polishing effect to ensure that the prediction remains accurate at all times. For example, in one example, the data is first clustered using the GaussianMixture model, and then a regression model (such as linear regression) is trained for each cluster; for new samples, the cluster to which it belongs is first determined, and then the corresponding regression model is used for prediction; this method not only captures the clustering structure of the data, but also improves the regression performance, especially in scenarios where the data distribution is complex and presents nonlinear relationships. Therefore, the mapping model M is not just a static selector, but a dynamic learning system that uses the powerful ability of GMM to adapt to changes in various working conditions, and by integrating GMR, this enables the stick robot to always maintain a high level of performance in different types of grinding tasks, no matter how complex or changing the operating environment is.

[0053] Step S3: before the polishing starts, dynamically adjust the process parameters according to different scene features S in combination with the adaptive algorithm A.

[0054] Specifically, the adaptive algorithm A satisfies: A(K, S) = K′; where K′ is the adjusted process parameter. The adaptive algorithm can automatically adjust the processing method, processing sequence, processing parameters, boundary conditions or constraints according to the input data characteristics to match the statistical distribution characteristics of the processed data. In order to achieve the transformation from K to K′, the adaptive algorithm usually goes through the following key steps: initialization stage, first set a set of preliminary process parameters K for the upcoming polishing task based on known information (such as historical data, experience values, etc.); feature extraction, then analyze the current scene feature S, and extract key information that helps guide subsequent adjustments; parameter adjustment, then use the adaptive algorithm A, combined with the extracted feature S, to make necessary adjustments to the original parameter K to form a new parameter configuration K′. In this process, the algorithm will consider the influence of multiple factors, such as the gradient and second-order derivative of the objective function, to ensure that the adjusted parameters can better adapt to the needs of specific scenarios; verification and iteration, finally test the adjusted parameter K′ to evaluate whether it has achieved the expected effect. If necessary, the above process can be further iterated until the most suitable parameter combination is found.

[0055] Furthermore, before the grinding process begins, it is crucial to understand and identify specific scene features S. These features may include but are not limited to material type, workpiece shape, environmental conditions, etc. Different scene features will directly affect the final grinding effect and efficiency. For example, when processing different materials, different grinding strategies need to be adopted to ensure the best quality output; for workpieces with complex geometries, more delicate operations may be required to avoid damaging the surface or causing unnecessary wear.

[0056] Step S4: In response to the grinding process, adjust the parameter F′ based on real-time monitoring R and feedback to improve the grinding efficiency E and processing quality Q.

[0057] Specifically, the feedback adjustment parameter F′ can be adjusted in real time according to the monitoring data of the real-time monitoring R to optimize the grinding efficiency E and the processing quality Q. Furthermore, the grinding efficiency E satisfies: E = f(R, F′); where f is the efficiency improvement function; this means that the grinding efficiency E depends on the data obtained by the real-time monitoring R and the feedback adjustment parameter F′ generated thereby. When it is detected that certain parameters deviate from the preset optimal range, measures will be taken immediately to correct them so that the grinding process always runs in an efficient state. The processing quality Q satisfies: Q = g(R, F′); where g is the quality improvement function; similarly, the processing quality Q is also affected by the real-time monitoring R and is optimized by the feedback adjustment parameter F′. If it is found that the grinding effect of a certain area is not ideal, the relevant parameters can be adjusted immediately, such as increasing or decreasing the applied pressure, changing the moving speed, etc., until a satisfactory quality standard is reached.

[0058] In the embodiment of the present application, real-time monitoring (Real-time Monitoring, R) plays a vital role. It provides a basis for subsequent feedback adjustment by collecting various data in the grinding process, such as force control parameters, speed parameters, position parameters, etc. Real-time monitoring R can not only capture all the details in the current operating state, but also promptly detect potential problems or abnormal conditions to ensure the safety and stability of the entire production process. For example, in the process of grinding and weight reduction of the robot in the semi-enclosed cabin cavity, a six-dimensional force sensor is used to monitor the milling force changes in real time, and overcutting is identified online by setting thresholds, thereby avoiding unnecessary material removal and ensuring the accuracy and surface finish of the product. In addition, for contact-type operating robots, their main body must be in contact with the external environment to complete the specified work tasks, so effective perception and control technology is essential. The robot needs to continuously adjust its behavior according to environmental feedback to adapt to different task requirements.

[0059] In a preferred embodiment, the multi-scenario grinding method of the stick robot based on adaptive parameter mapping is not limited to pre-set path planning, but also responds dynamically during the actual grinding process, and optimizes the grinding effect through real-time calculation and adjustment. Specifically, during the grinding process, the inverse kinematics solution is continuously used to determine the angles required for each joint of the robot. This process is not a simple mathematical conversion, but a comprehensive consideration of the physical limitations of the robot, including but not limited to the maximum and minimum angles of each joint, the allowed rotation range, and possible mechanical dead zones. For example, when facing a complex surface, in order to ensure that the grinding head can accurately contact the target surface and maintain a constant pressure, the inverse kinematics algorithm needs to accurately calculate the positions of all active joints at each time point, while ensuring that these positions do not exceed the capabilities of the hardware itself. In addition, considering that the working environment of the stick robot is often complex, there may be obstacles or narrow spaces, so any posture that may cause collisions must be avoided during the solution process to ensure the safety and effectiveness of the operation. In this way, the stick robot can maintain efficient and high-quality grinding operations even under variable working conditions. In this process, inverse kinematics solving plays a vital role, which allows the robot to flexibly respond to different grinding task requirements while respecting its own structural constraints. For high-degree-of-freedom robots with six or more degrees of freedom, the inverse kinematics problem becomes particularly complex because they can achieve almost infinite different pose combinations in three-dimensional space. However, it is this complexity that gives the stick robot a strong adaptability, enabling it to find the optimal solution under various working conditions, thereby providing a more refined and smoother surface treatment result.

[0060] In a preferred embodiment, the multi-scenario grinding method of the arm robot based on adaptive parameter mapping automatically performs the position correction step of the end tool after the grinding process is completed. This process is intended to compensate for the end tool position offset caused by long-term operation, material removal, and environmental factors (such as thermal expansion caused by temperature changes). Specifically, when the grinding task is completed, a precise position detection program will be started to evaluate the difference between the actual position of the current end effector and the expected ideal position through a sensor network integrated on the robot, such as a six-dimensional force sensor or other high-precision measuring device. Once the deviation value is determined, a preset algorithm model will be applied for analysis and the necessary adjustment amount will be calculated. Subsequently, these adjustment instructions will be sent to each joint servo motor to fine-tune its angle to ensure that the end tool can be restored to the precise position initially set. In addition, in order to improve the accuracy of the correction, external measurement equipment, such as a laser tracker or an optical coordinate measuring machine, can also be used in combination for independent verification and further correction. In this way, not only can the accumulated errors caused by physical wear or thermal effects be effectively eliminated, but also the consistency and reliability of subsequent processing tasks can be guaranteed, thereby maintaining a high level of product quality standards. This position correction mechanism is essential to maintaining the consistency of the stick robot's grinding operations, especially when the same or similar tasks need to be repeated. It can help reduce the need for human intervention, reduce production costs and improve work efficiency. At the same time, regular such corrections can also help extend the service life of the equipment, because precise operation can avoid unnecessary stress concentration and reduce excessive friction and wear between components. In short, by meticulously managing the position of the stick robot's end tool, this method provides a solid technical guarantee for achieving efficient and stable automated grinding.

[0061] In a preferred embodiment, the multi-scene grinding method of the stick robot based on adaptive parameter mapping not only relies on preset paths and mechanical control, but also cleverly integrates image recognition technology to assist scene recognition. During the grinding process, the camera or other visual sensors equipped by the robot will continuously collect image information of the working area, and these image data are then sent to the image recognition system for analysis and processing. The image recognition technology must be intelligent enough to cope with complex scene changes, such as fluctuations in lighting conditions, cluttered backgrounds, or partial occlusion of target objects. In order to provide accurate scene features, image recognition algorithms usually use deep learning frameworks, such as convolutional neural networks (CNNs), and train models with a large number of annotated data sets to enable them to distinguish different types of surface materials, textures, shapes and other detailed characteristics. In addition, for special needs in grinding operations, knowledge in specific fields, such as professional rules for metal surface treatment, can also be introduced to further enhance recognition accuracy. When encountering complex scenes, not only can the exact position and posture of the object to be processed be identified, but also the environmental factors around it, such as the distance to nearby obstacles, can be judged, thereby providing guarantees for safe and effective operations. More importantly, the application of image recognition technology allows robots to adjust their behavioral strategies in real time in a dynamic environment; for example, after detecting unexpected defects or shape changes on the surface of a workpiece, the robot can quickly update the polishing plan based on the latest visual feedback to ensure that the quality of the final product is not affected.

[0062] In a preferred embodiment, the multi-scenario grinding method of the stick robot based on adaptive parameter mapping not only relies on preset paths and mechanical control, but also integrates advanced force feedback control technology to ensure high efficiency and accuracy in various grinding scenarios. When the robot performs the grinding task, the force sensor monitors the force or torque information generated when it contacts the workpiece in real time, and feeds back this data to the corresponding grinding system of the robot. Through this force perception ability, the changes in material properties and surface conditions faced by the current operation can be identified. The core of the force feedback control technology is that it can dynamically adjust the grinding force according to the actual measured force signal, so as to ensure that the force applied to the workpiece remains within an ideal range. For example, when processing a metal surface with higher hardness, the grinding force will be increased to ensure that enough material is removed; when facing softer or fragile materials such as wood or composite materials, the grinding force will be automatically reduced to avoid damaging the workpiece. In addition, for workpieces with complex geometric shapes, such as curved surfaces, edges, etc., force control technology can maintain a constant pressure distribution by fine-tuning the robot's movements, which helps to improve the quality and consistency of grinding. To achieve this, force feedback control usually uses a series of complex algorithms for signal processing and control decisions: first, the original force signal is filtered to remove noise interference, then the expected force value is calculated based on the predefined task model, and the next action adjustment is determined by comparing the difference between the actual measured force value and the target value. Such a closed-loop control mechanism allows the robot to respond flexibly under different materials and surface conditions, providing a more refined and smoother operating experience.

[0063] Figure 2 A structural block diagram of a multi-scenario grinding system of a stick robot based on adaptive parameter mapping provided by an exemplary embodiment of the present invention is shown. The multi-scenario grinding system of a stick robot based on adaptive parameter mapping is applied to the multi-scenario grinding method of a stick robot based on adaptive parameter mapping as described above. The system includes:

[0064] A construction module 201 is used to construct a process parameter knowledge base K, where the process parameter knowledge base K is used to store process parameters of different end tools, where the process parameters include at least one of a force control parameter F, a speed parameter V or a position parameter P;

[0065] A matching module 202, used for quickly matching process parameters through a mapping model M;

[0066] An adjustment module 203 is used to dynamically adjust the process parameters in response to different scene features S and in combination with the adaptive algorithm A before the polishing starts;

[0067] The control module 204 is used to improve the grinding efficiency E and the processing quality Q in response to the grinding process, based on the real-time monitoring R and feedback adjustment parameter F′.

[0068] In a possible implementation, in step S1, the process parameter knowledge base K can automatically update and optimize parameters based on historical data and feedback, and the process parameter knowledge base K satisfies:

[0069] K = {F, V, P}.

[0070] In a possible implementation, in step S2, the mapping model M is a probability matching model based on a Gaussian mixture model GMM, which can provide accurate parameter matching even when the data is incomplete or noisy, and the mapping model M satisfies:

[0071] M(θ)=GMM(θ;π,μ,Σ);

[0072] Among them, θ is the process parameter of the end tool, π is the mixing coefficient, μ is the mean vector, and Σ is the covariance matrix.

[0073] In a possible implementation, in step S3, the adaptive algorithm A can predict the effect of parameter adjustment on the polishing effect and optimize the parameters accordingly, and the adaptive algorithm A satisfies:

[0074] A(K, S) = K′;

[0075] Wherein, K′ is the adjusted process parameter.

[0076] In a possible implementation, in step S4, the feedback adjustment parameter F′ can be adjusted in real time according to the monitoring data of the real-time monitoring R to optimize the grinding efficiency E and the processing quality Q.

[0077] In a possible implementation, the polishing efficiency E satisfies:

[0078] E = f(R, F′);

[0079] Where f is the efficiency improvement function.

[0080] In a possible implementation, the processing quality Q satisfies:

[0081] Q = g(R, F′);

[0082] Among them, g is the quality improvement function.

[0083] In a possible implementation, the method further includes:

[0084] In response to the start of grinding, the stick robot is modeled using DH parameters.

[0085] In a possible implementation, the method further includes:

[0086] In response to the grinding process, inverse kinematics is used to solve the robot joint angles.

[0087] In a possible implementation, the method further includes:

[0088] In response to the grinding being completed, the position of the end tool of the boom robot is corrected to compensate for the position change of the end tool caused by wear or thermal expansion.

[0089] It should be understood that the specific examples herein are only intended to help those skilled in the art to better understand the present disclosure, rather than to limit the scope of the present invention.

[0090] It is understood that in the various implementations of this specification, the size of the sequence number of each process does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the present disclosure.

[0091] It can be understood that the various implementation modes described in this specification can be implemented individually or in combination, and the present disclosure is not limited to this.

[0092] Unless otherwise stated, all technical and scientific terms used in this disclosure have the same meaning as those generally understood by those skilled in the art of the technical field of this specification. The terms used in this specification are only for the purpose of describing specific embodiments and are not intended to limit the scope of this specification. The term "and / or" used in this specification includes any and all combinations of one or more related listed items. The singular forms of "a", "above", and "the" used in this disclosure and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings.

[0093] The above is only a specific implementation of this specification, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in this specification, which should be included in the protection scope of this specification. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A multi-scenario grinding method for a stick robot based on adaptive parameter mapping, characterized in that: The method comprises: S1. Construct a process parameter knowledge base K, wherein the process parameter knowledge base K is used to store process parameters of different end tools, wherein the process parameters include at least one of a force control parameter F, a speed parameter V or a position parameter P; S2. Quickly match the process parameters through the mapping model M; S3, before the polishing starts, dynamically adjusting the process parameters according to different scene features S in combination with the adaptive algorithm A; S4. In response to the grinding process, adjust the parameter F′ based on real-time monitoring R and feedback to improve the grinding efficiency E and processing quality Q.

2. The multi-scenario grinding method of a stick robot based on adaptive parameter mapping according to claim 1 is characterized in that: In step S1, the process parameter knowledge base K can automatically update and optimize parameters according to historical data and feedback. The process parameter knowledge base K satisfies: K = {F, V, P}.

3. The multi-scenario grinding method of a stick robot based on adaptive parameter mapping according to claim 1 is characterized in that: In step S2, the mapping model M is a probability matching model based on a Gaussian mixture model GMM, which can provide accurate parameter matching even when the data is incomplete or noisy. The mapping model M satisfies: M(θ)=GMM(θ;π,μ,Σ); Among them, θ is the process parameter of the end tool, π is the mixing coefficient, μ is the mean vector, and Σ is the covariance matrix.

4. The multi-scenario grinding method of a stick robot based on adaptive parameter mapping according to claim 1 is characterized in that: In step S3, the adaptive algorithm A can predict the influence of parameter adjustment on the polishing effect and optimize the parameters accordingly. The adaptive algorithm A satisfies: A(K, S) = K′; Wherein, K′ is the adjusted process parameter.

5. The multi-scenario grinding method of a stick robot based on adaptive parameter mapping according to claim 1 is characterized in that: In the step S4, the feedback adjustment parameter F' can be adjusted in real time according to the monitoring data of the real-time monitoring R to optimize the grinding efficiency E and the processing quality Q.

6. The multi-scenario grinding method of a stick robot based on adaptive parameter mapping according to claim 5 is characterized in that: The grinding efficiency E satisfies: E = f(R, F′); Where f is the efficiency improvement function.

7. The multi-scenario grinding method of a stick robot based on adaptive parameter mapping according to claim 5 is characterized in that: The processing quality Q satisfies: Q = g(R, F′); Among them, g is the quality improvement function.

8. The multi-scenario grinding method of a stick robot based on adaptive parameter mapping according to claim 1 is characterized in that: The method further comprises: In response to the start of grinding, the stick robot is modeled using DH parameters.

9. The multi-scenario grinding method of a stick robot based on adaptive parameter mapping according to claim 1, characterized in that: The method further comprises: In response to the grinding process, inverse kinematics is used to solve the robot joint angles.

10. The multi-scenario grinding method of a stick robot based on adaptive parameter mapping according to claim 1, characterized in that: The method further comprises: In response to the grinding being completed, the position of the end tool of the boom robot is corrected to compensate for the position change of the end tool caused by wear or thermal expansion.

11. A multi-scenario grinding system for a boom robot based on adaptive parameter mapping, the system being applied to the multi-scenario grinding method for a boom robot based on adaptive parameter mapping as claimed in any one of claims 1 to 10, characterized in that: The system comprises: A construction module, used to construct a process parameter knowledge base K, wherein the process parameter knowledge base K is used to store process parameters of different end tools, wherein the process parameters include at least one of a force control parameter F, a speed parameter V or a position parameter P; A matching module, used for quickly matching the process parameters through a mapping model M; An adjustment module, for dynamically adjusting the process parameters in response to different scene features S and in combination with an adaptive algorithm A before the polishing starts; The control module is used to respond to the grinding process, based on real-time monitoring R and feedback adjustment parameters F', to improve the grinding efficiency E and processing quality Q.

Images