Polishing track real-time correction control method and system

Abstract

The invention relates to the technical field of track correction control, and discloses a polishing track real-time correction control method and system.The method comprises the steps that multi-source data of a curved surface to be polished is collected and preprocessed, and a real-time machining data set is obtained; deviation distribution characteristics are extracted, an initial deviation vector is determined, and a temporary machining path is generated; key parameters of the temporary machining path are obtained, matching verification is conducted on the key parameters and the curved surface to be polished, curvature optimization is conducted on the temporary machining path, and an optimized machining path is obtained; key path parameters are extracted, and a polishing track is determined; according to the polishing track, a polishing verification result is extracted, a supplementary machining instruction is generated and executed, and a polishing adjustment scheme is determined; and the moving track and pressure parameters of the mechanical arm are updated, and a polishing machining result is obtained in combination with real-time response verification and local defect detection. The method can solve the problem that an existing control method lacks a mechanism for sensing and responding to the dynamic change of the machining process.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of trajectory correction control, and in particular to a polishing trajectory real-time correction control method and system. BACKGROUND

[0002] At present, in modern manufacturing, especially in the manufacturing of industrial automatic control system devices, polishing process as an important link to improve the surface quality of products plays a crucial role in the precision and aesthetics of products, especially in the high-end manufacturing fields such as aerospace and automobile manufacturing, the accurate control of polishing trajectory is directly related to the performance and service life of parts.

[0003] In the prior art, a preset trajectory control or fixed parameter adjustment method is usually adopted, that is, the motion trajectory of the mechanical arm and the polishing parameters are planned in advance according to the shape of the processing object and the experience data. In the processing process, the system executes according to the preset path and does not have real-time adjustment capability.

[0004] However, the existing control method lacks the perception and response mechanism of dynamic changes in the processing process, so when the workpiece surface has complex curvature, uneven material or external disturbance, etc., the mechanical arm cannot correct the trajectory and parameters in time, resulting in deviation of the polishing trajectory from the ideal path. Therefore, the prior art has the problem of lacking the perception and response mechanism of dynamic changes in the processing process. SUMMARY

[0005] The present application provides a polishing trajectory real-time correction control method and system to solve the problem of the existing control method lacking the perception and response mechanism of dynamic changes in the processing process.

[0006] In a first aspect, to solve the above technical problems, the present application provides a polishing trajectory real-time correction control method, comprising:

[0007] Collecting multi-source data of a polishing surface, preprocessing the multi-source data to obtain a real-time processing data set;

[0008] According to the real-time processing data set, extracting deviation distribution features, determining an initial deviation vector, and generating a temporary processing path according to the initial deviation vector;

[0009] Obtaining key parameters of the temporary processing path and performing matching verification with the polishing surface, optimizing the curvature of the temporary processing path to obtain an optimized processing path;

[0010] According to the optimized processing path, extracting key path parameters, comparing the key path parameters with preset perception conditions, and determining a polishing trajectory;

[0011] According to the polishing track, a polishing verification result is extracted, a supplementary machining instruction is generated according to the polishing verification result and is executed, and a polishing adjustment scheme is determined;

[0012] According to the polishing adjustment scheme, the running track and the pressure parameter of the mechanical arm are updated, real-time response verification and local defect detection are combined, and a polishing machining result is obtained.

[0013] In an optional implementation, the multi-source data of the surface to be polished is collected, and the multi-source data is preprocessed to obtain a real-time machining data set, including:

[0014] Force feedback and surface roughness data of the surface to be polished are obtained by a sensor array to obtain multi-source data, and the multi-source data is denoised to obtain a denoised multi-source data set;

[0015] The denoised multi-source data set is standardized to obtain a fusion data set, and the fusion data set is filtered by a preset standard threshold to obtain a machining data set;

[0016] Real-time features are extracted from the machining data set to judge the stability of the machining environment to obtain a real-time machining data set.

[0017] In an optional implementation, the deviation distribution feature is extracted from the real-time machining data set, the initial deviation vector is determined, and the temporary machining path is generated according to the initial deviation vector, including:

[0018] The real-time machining data set is matched with the preset track data point by point to determine a preliminary deviation set;

[0019] The deviation values of the preliminary deviation set are spatially distributed to eliminate abnormal points of the distribution feature to obtain a deviation distribution feature;

[0020] The deviation distribution feature is comprehensively processed to determine an initial deviation vector, and the initial deviation vector is matched with a preset threshold range one by one to obtain a deviation out-of-limit point;

[0021] The angle parameter and the speed parameter of the deviation out-of-limit point are extracted to obtain parameter adjustment data, the abnormal fluctuations of the parameter adjustment data are corrected, key control parameters are extracted, and an adjustment amplitude range is obtained;

[0022] The adjustment amplitude range is simulated dynamically to obtain a temporary machining path.

[0023] In an optional implementation, the key parameters of the temporary machining path are obtained and matched with the surface to be polished for matching verification, the curvature of the temporary machining path is optimized to obtain an optimized machining path, including:

[0024] extracting key parameters from the temporary machining path, comparing the key parameters with a preset parameter range, marking and obtaining a mismatching area;

[0025] detecting a polishing pressure distribution of the mismatching area, calibrating the polishing pressure distribution and analyzing curvature variation of the surface to be polished to obtain curvature variation data;

[0026] According to the curvature variation data, dynamically correcting the temporary machining path to correct path deviation to obtain an optimized machining path.

[0027] In an optional embodiment, the extracting key path parameters according to the optimized machining path and comparing the key path parameters with a preset sensing condition to determine a polishing trajectory comprises:

[0028] extracting key path parameters according to the optimized machining path to obtain control point data;

[0029] determining feedback distribution and response matching degree according to the control point data to obtain response data;

[0030] obtaining parameter set data according to the response data in combination with the characteristics of the surface to be polished;

[0031] comparing the parameter set data with a preset sensing condition, calibrating operation parameters of a mechanical arm, and determining a polishing trajectory.

[0032] In an optional embodiment, the extracting polishing verification results according to the polishing trajectory, generating supplementary machining instructions and executing according to the polishing verification results, and determining a polishing adjustment scheme comprises:

[0033] obtaining surface roughness according to the polishing trajectory, comparing the surface roughness with a preset roughness threshold to obtain an initial distribution result;

[0034] obtaining quality distribution information and performing trend analysis according to the initial distribution result to determine polishing verification results;

[0035] classifying characteristics of the surface to be polished according to the polishing verification results to obtain classification area information;

[0036] generating supplementary machining instructions and executing according to the classification area information to obtain a polishing adjustment scheme.

[0037] In an optional embodiment, the updating operation trajectory and pressure parameters of the mechanical arm according to the polishing adjustment scheme, and combining real-time response verification and local defect detection to obtain polishing machining results comprises:

[0038] According to the polishing adjustment scheme, a mechanical arm control instruction is updated and executed to obtain running track data;

[0039] According to the running track data, pressure adjustment is performed to determine a pressure parameter configuration;

[0040] According to the pressure parameter configuration, defect recognition is performed to obtain defect distribution details;

[0041] According to the defect distribution details, a track correction instruction is generated, the mechanical arm control instruction is dynamically updated, a final running track and a final pressure parameter configuration are determined, and a polishing processing result is obtained.

[0042] In a second aspect, the present application provides a polishing track real-time correction control system, comprising:

[0043] A data acquisition and processing module is configured to acquire multi-source data of a surface to be polished, pre-process the multi-source data, and obtain a real-time processing data set;

[0044] A deviation analysis module is configured to extract deviation distribution characteristics from the real-time processing data set, determine an initial deviation vector, and generate a temporary processing path according to the initial deviation vector;

[0045] A path optimization module is configured to obtain key parameters of the temporary processing path, perform matching verification with the surface to be polished, perform curvature optimization on the temporary processing path, and obtain an optimized processing path;

[0046] An execution correction module is configured to extract key path parameters from the optimized processing path, compare the key path parameters with preset sensing conditions, and determine a polishing track;

[0047] A supplementary processing module is configured to extract a polishing verification result from the polishing track, generate a supplementary processing instruction according to the polishing verification result, and execute the supplementary processing instruction to determine a polishing adjustment scheme;

[0048] A dynamic updating module is configured to update a running track and pressure parameters of a mechanical arm according to the polishing adjustment scheme, combine real-time response verification and local defect detection, and obtain a polishing processing result.

[0049] Compared with the prior art, the present application has the following beneficial effects:

[0050] (1) The present application determines a preliminary deviation set by point-by-point matching of the real-time processing data set and the preset track data, further calculates the average deviation direction and the average deviation amplitude of all remaining effective deviation points, and synthesizes an initial deviation vector with direction and size information, so that the track correction can truly reflect the actual deviation of the tool in the three-dimensional space, laying a solid foundation for subsequent accurate spatial path compensation.

[0051] (2) The present application introduces a curved surface geometric model, obtains the average curvature value of each point on the complex curved surface to be polished by calling the curved surface curvature analysis function in CGAL or OpenCASCADE, and adopts a nonlinear least squares method for iterative calculation to minimize the attitude change rate of the curvature fluctuation area as the objective function, thereby optimizing the path. This method ensures that the generated polishing trajectory can be highly matched with the geometric shape of the complex curved surface, effectively avoiding machining defects such as overcutting and underpolishing caused by the mismatch between the path and the curved surface.

[0052] (3) The present application obtains force feedback and surface roughness data of the curved surface to be polished through a sensor array, generates a force feedback and roughness distribution trend chart during the machining process in combination with the SIMATIC WinCC OA system, and constructs a multi-source perception system. The system can capture the dynamic changes of the machining state in real time, provide comprehensive and reliable data support for control decisions, and thus realize overall monitoring and rapid response to the machining process.

[0053] (4) The present application designs a complete quality control closed loop, compares the surface roughness with the preset roughness threshold value, and calculates the percentage of points with roughness values below the upper limit of the qualified value among all sampling points in the entire polishing area. If the qualified rate does not meet the standard, supplementary machining instructions are generated for the high-priority area, and linear interpolation or Bezier curve optimization is used for detailed adjustment. This dynamic compensation mechanism based on real-time verification significantly improves the consistency and reliability of the final polished surface quality.

[0054] In summary, the present application collects force feedback and surface roughness data through a sensor array, performs high-precision fusion denoising standardization to generate real-time machining data sets, compares and analyzes the deviations point by point based on the data sets, extracts the initial deviation vector, generates a temporary trajectory if the threshold is exceeded, and optimizes the curvature change to determine the optimized trajectory. Further simulation of the motion behavior extracts a parameter set, updates the mechanical arm parameters to generate a correction trajectory, collects roughness trend verification results, generates supplementary instructions for the insufficient area, and finally adjusts the path and pressure parameters to achieve lean polishing. This method effectively solves the dynamic deviation and quality unevenness problems in complex curved surface machining, improves the polishing precision and efficiency, and ensures high-quality surface treatment effect. BRIEF DESCRIPTION OF DRAWINGS

[0055] Figure 1 is a polishing trajectory real-time correction control method flowchart provided by the first embodiment of the present application;

[0056] Figure 2 is a polishing trajectory real-time correction control system structure schematic diagram provided by the second embodiment of the present application. DETAILED DESCRIPTION

[0057] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0058] With reference to Figure 1 The first embodiment of the present application provides a polishing track real-time correction control method, comprising the following steps:

[0059] S11, collecting multi-source data of a polishing surface, pre-processing the multi-source data to obtain a real-time processing data set;

[0060] S12, extracting deviation distribution characteristics according to the real-time processing data set, determining an initial deviation vector, and generating a temporary processing path according to the initial deviation vector;

[0061] S13, obtaining key parameters of the temporary processing path, and performing matching verification with the polishing surface, performing curvature optimization on the temporary processing path to obtain an optimized processing path;

[0062] S14, extracting key path parameters according to the optimized processing path, comparing the key path parameters with preset perception conditions, and determining a polishing track;

[0063] S15, extracting a polishing verification result according to the polishing track, generating a supplementary processing instruction according to the polishing verification result and performing the supplementary processing instruction to determine a polishing adjustment scheme;

[0064] S16, updating a running track and a pressure parameter of a mechanical arm according to the polishing adjustment scheme, combining real-time response verification and local defect detection to obtain a polishing processing result.

[0065] In step S11, multi-source data of a polishing surface is collected, and the multi-source data is pre-processed to obtain a real-time processing data set, comprising:

[0066] Force feedback and surface roughness data of the polishing surface are obtained by a sensor array to obtain multi-source data, and the multi-source data is denoised to obtain a denoised multi-source data set;

[0067] The denoised multi-source data set is standardized to obtain a fusion data set, and the fusion data set is filtered by a preset standard threshold to obtain a processing data set;

[0068] Real-time characteristics are extracted from the processing data set to judge the stability of the processing environment to obtain a real-time processing data set.

[0069] The sensor array collects force feedback signals and surface roughness distribution data of the surface to be polished, and denoises the original signals to obtain a preliminary cleaned force feedback data set and a roughness data set, i.e., a denoised multi-source data set. Based on the denoised multi-source data set, dimension-uniform processing is performed to obtain a standardized fusion data set. According to the material properties of the surface to be polished, the process parameter range of the polishing tool, and the normal fluctuation range in the historical processing data, a preset threshold is comprehensively set. For the standardized fusion data set, the preset threshold range is used for screening. If an abnormal value is detected, the abnormal data points outside the allowed range are removed to determine a processing data set. Real-time features are extracted from the processing data set, including the fluctuation amplitude and frequency of the force feedback signal, and the local variation gradient and statistical distribution variance of the surface roughness data. A SIMATIC WinCC OA SCADA system is used to generate a force feedback and roughness distribution trend chart during processing, including a two-dimensional trend chart of force feedback and roughness changing with time, and a three-dimensional distribution cloud chart of roughness on the surface. The stability of the processing environment is evaluated, and a real-time processing data set is finally obtained.

[0070] Exemplarily, when polishing the curved surface part of an aero-engine blade, the sensor array is arranged in the contact area between the polishing tool and the workpiece. A torque sensor with a sampling frequency of 1 kHz (i.e., 1000 times per second) is used to collect force feedback data. A confocal white light interferometer with a sampling frequency of 100 Hz is used to collect surface roughness data. A laser scattering roughness sensor can also be used to continuously collect optical roughness signals, which are converted into equivalent surface roughness values through a pre-calibrated model. Mechanical vibration or environmental noise is often mixed in the original signal, affecting data accuracy. Therefore, a Kalman filter algorithm tool is used for denoising. The motion state (position, speed) of the mechanical arm is taken as the system state quantity, and the sensor readings are taken as the observation quantity. The state space model of the system is established. The "prediction-update" two steps of Kalman filtering are used for recursive calculation: in the prediction step, the state at the current time is predicted according to the state estimation at the last time and the mechanical arm motion model; in the update step, the actual sensor observation value at the current time is compared with the predicted value, and the state estimation is optimally corrected based on the Kalman gain, so that random high-frequency noise is effectively removed, the main trend signal is retained, and a preliminary cleaned data set is obtained.

[0071] In a possible implementation, when extracting real-time features from the processing data set, attention can be paid to the fluctuation frequency of the force feedback and the local variation rate of the roughness. The local variation rate is the absolute value of the difference between the current sampling value and the previous sampling value, divided by the sampling time interval. The calculation formula is wherein is a sampling value, is a time, is a sampling period, the force feedback data is 1 ms (sampling frequency 1 kHz), and the roughness data is 10 ms (sampling frequency 100 Hz). A trend chart is generated according to the characteristics, the fluctuation of the curve is judged, and if the standard deviation of the force feedback signal in the set time window is less than the stability threshold set according to the process requirement, it is determined that the force feedback signal is a smooth fluctuation. It is worth noting that the stability threshold is set to be the expected force value range, such as 5% of 2N-5N, i.e. 0.1-0.25N.

[0072] In step S12, according to the real-time machining data set, the deviation distribution characteristics are extracted, the initial deviation vector is determined, and the temporary machining path is generated according to the initial deviation vector, including:

[0073] The real-time machining data set is matched with the preset trajectory data point by point to determine the preliminary deviation set;

[0074] The deviation values of the preliminary deviation set are processed in space distribution to eliminate abnormal points of the distribution characteristics to obtain the deviation distribution characteristics;

[0075] The deviation distribution characteristics are comprehensively processed to determine the initial deviation vector, and the initial deviation vector is matched with the preset threshold range one by one to obtain the deviation out-of-limit point;

[0076] The angle parameters and speed parameters of the deviation out-of-limit point are extracted to obtain the parameter adjustment data, the abnormal fluctuation of the parameter adjustment data is corrected, the key control parameters are extracted, and the adjustment amplitude range is obtained;

[0077] Dynamic simulation is performed according to the adjustment amplitude range to obtain the temporary machining path.

[0078] It should be noted that before polishing, a series of path point sequences containing three-dimensional space coordinates and tool posture are generated in advance by professional computer-aided manufacturing software (such as Siemens NXCAM, Dassault System CATIA or Powermill), and are stored in the mechanical arm controller, which represents the preset trajectory data that the polishing tool needs to follow in the case of no interference. The real-time processing data set is matched with the preset trajectory data point by point, and the system clock is used as the synchronization reference. The actual three-dimensional coordinates of the tool collected by the pose sensor in the real-time processing data set are subtracted from the corresponding target three-dimensional coordinates in the preset trajectory data at the same timestamp to obtain the position deviation of the point in the X, Y and Z directions. The position deviations at all time points collectively constitute a preliminary deviation set. Using a three-dimensional point cloud processing library, the position deviation values are mapped to a three-dimensional curved surface model defined by the preset trajectory as point cloud attributes to generate a deviation distribution graph that represents the deviation size by color depth. If there are abnormal points in the deviation distribution graph that exceed the preset threshold, the abnormal points are removed to obtain the deviation distribution feature. For the above deviation distribution feature, the average deviation direction and the average deviation amplitude of all remaining valid deviation points are calculated, and the two average values are combined into an initial deviation vector with direction and size information.

[0079] The initial deviation vector is matched with the preset threshold to identify deviation overrun points. For the deviation overrun points, the angle parameters of each point are extracted and the speed parameters are collected to form parameter adjustment data. If there are abnormal fluctuations, the parameter adjustment data is corrected, the key control parameters are extracted, and the adjustment amplitude range is determined. Finally, the motion trajectory correction is dynamically simulated to generate a path adjustment strategy that meets the requirements, and a temporary processing path is obtained.

[0080] Exemplarily, for the point-by-point matching processing of the real-time processing data set and the preset trajectory data, a specific scenario can be imagined, such as processing the curved surface area of an aero-engine blade. The real-time collected processing path points are compared with the preset trajectory points one by one to obtain the deviation value at each time point, such as a certain time point deviating from the preset position by 0.2 mm, and another time point deviating by 0.1 mm, finally forming a preliminary deviation set containing all deviation values. The principle of this point-by-point matching is to ensure the accuracy of the processing path through time sequence comparison, providing basic data for subsequent analysis.

[0081] For the spatial distribution processing of the preliminary deviation set, a three-dimensional point cloud visualization software (such as CloudCompare or ParaView) can map each deviation value as a data point to the corresponding spatial position of the three-dimensional digital model of the blade imported from the CAD model file, thereby generating a deviation distribution graph. Assuming that in a certain area of the blade surface, the deviation values are concentrated between 0.1 and 0.3 mm, and in another area, the deviation values are as high as 0.5 mm, the graph will intuitively display these distribution characteristics in terms of color depth or height difference.

[0082] For example, if the parameter adjustment data has abnormal fluctuations, such as frequent jumps in speed parameters within a short period of time, a one-dimensional digital signal filter can perform correction processing, set the cutoff frequency or moving window size of the filter to filter out high-frequency jump components in the signal caused by noise, and retain low-frequency signals representing the true motion trend. Assuming that the speed values of a certain segment of data vary irregularly between 1.5 and 2.5 mm / s, the smoothing process stabilizes them at around 2 mm / s, and key control parameters such as speed stability and angle consistency are extracted, determining that the appropriate adjustment range is a speed increase of 0.2 mm / s and an angle adjustment of 2 degrees or less.

[0083] It is worth noting that the cutoff frequency or moving window size of the filter is set, for example, the cutoff frequency is set to 10 Hz for mechanical vibration noise, or the moving window size is set to 5 sampling points to filter out high-frequency jump components in the signal caused by noise.

[0084] In step S13, the key parameters of the temporary machining path are obtained, and matching verification is performed with the surface to be polished. The curvature of the temporary machining path is optimized to obtain an optimized machining path, including:

[0085] The key parameters are extracted from the temporary machining path, and the regions where the key parameters do not match the preset parameter range are marked and obtained as mismatched regions.

[0086] The polishing pressure distribution of the mismatched region is detected, the polishing pressure distribution is calibrated, and the curvature change of the surface to be polished is analyzed to obtain curvature change data.

[0087] According to the curvature change data, the temporary machining path is dynamically corrected, and the path deviation is corrected to obtain an optimized machining path.

[0088] In one implementation, the key parameters are extracted from the temporary machining path, and whether these parameters match the preset machining area coverage range is analyzed. The mismatched regions are marked. For the mismatched regions, the polishing pressure parameters are extracted. The pressure distribution is adjusted to balance the pressure value to the target value to obtain pressure distribution data.

[0089] Based on the adjusted pressure distribution data, the average curvature value of each point on the complex surface to be polished is obtained by calling the surface curvature analysis function in CGAL or OpenCASCADE, thus obtaining the curvature change data of the surface. The standard deviation of the curvature data (e.g., a 5×5 dot matrix window) within the moving window is calculated. If this standard deviation exceeds the curvature change threshold set according to the workpiece surface finish requirements (e.g., the threshold is...), then... If the curvature fluctuation in the region is large, path optimization is required to obtain curvature change data. A nonlinear least squares method is used for iterative calculation, with minimizing the attitude change rate in the curvature fluctuation region as the objective function. By fine-tuning the 3D coordinates of the path points, the attitude change rate in the curvature fluctuation region is minimized, thus generating a smoother trajectory. A geometric constraint solver is used to fine-tune the deviation between the dynamically corrected path and the original preset trajectory on key geometric features (such as blade edge contours). While ensuring the smoothness of the corrected path, distance constraints are applied to control the path deviation within a preset tolerance range (such as ±0.05mm), thus determining the optimized processing path.

[0090] For example, for the labeled area distribution data, a thin-film pressure sensor array (e.g., Tekscan's I-Scan system) installed at the end of the polishing tool is used to extract the polishing pressure parameters at each point in real time. The standard deviation of the pressure values ​​of all valid measuring points within a set time window (typically 100 ms) is calculated. If this standard deviation exceeds the uniformity threshold set according to process requirements (e.g., the threshold is set to 0.4N), it is determined that there is uneven pressure distribution in that area. To address this unevenness, the impedance controller or admittance controller in the robotic arm force control system will be adjusted. Using the target pressure value (e.g., 2.5N) as input and real-time pressure feedback as the feedback signal, the contact force between the tool and the workpiece is dynamically corrected by adjusting the position offset of the robotic arm end in the normal direction of the curved surface. This gradually balances the pressure value to near the target value, ensuring that the force exerted by the tool on the curved surface is consistent during processing.

[0091] In step S14, based on the optimized processing path, key path parameters are extracted, and the key path parameters are compared with preset sensing conditions to determine the polishing trajectory, including:

[0092] Based on the optimized processing path, key path parameters are extracted to obtain control point data;

[0093] Based on the control point data, the feedback distribution and response matching degree are determined, and the response data is obtained;

[0094] Based on the response data and the characteristics of the surface to be polished, parameter set data is obtained;

[0095] The parameter set data is compared with a preset perception condition, the operation parameters of the robot arm are calibrated, and the polishing track is determined.

[0096] According to the optimized machining path, key control point data is extracted. According to the key control point data, force feedback data of the robot arm is obtained, and feedback force distribution is determined. At the same time, the time delay of the robot arm end effector from receiving a motion instruction to actually reaching the target position (i.e. step response time) is extracted, the measured response time is compared with the expected response time obtained based on the robot arm dynamics model simulation, and the relative error is calculated. If the relative error does not reach a preset threshold, adjustment is performed, the current loop gain of the servo driver is dynamically fine-tuned by the PID controller to reduce the error between the actual response time and the expected response time. Then, the force feedback data and the calibrated response value are integrated, a structured data set containing the average force feedback, response time, and their respective confidence weights is generated for each key control point by weighted average method, and parameter set data is formed. The confidence weight is determined based on sensor measurement accuracy or historical data statistical reliability, such as setting the weight of high-precision sensor to 0.9 and the weight of low-precision sensor to 0.6. The parameter set data is compared with a preset perception condition. The preset perception condition is a set of Boolean logic rules, for example:

[0097] “Is the average force feedback in the range of [2.5N, 3.5N]?” AND “Is the response time less than 50ms?” AND “Is the comprehensive confidence of the two higher than 90%?”.

[0098] If the condition is met, new position loop feedforward gain and speed loop proportional gain parameters are written to the underlying motion controller of the robot arm, combined with trajectory deviation analysis, to determine the polishing track. The comprehensive confidence is calculated by the arithmetic mean of the confidence weights of force feedback and response time, such as the sum of force feedback weight and response time weight, then divide by 2, if the value is higher than 90%, it is considered to meet the condition.

[0099] In an implementation, for the extraction of the trajectory path parameters, the key control points in the path can be analyzed through the API interface provided by the ROSMoveIt software running on the industrial computer. These control points are usually the core nodes in the trajectory path that determine the machining direction and precision. In the scenario of polishing the aero-engine blade, in the edge area of the blade, the turning angle between adjacent line segments in the calculated path is calculated. If the turning angle is greater than the angle threshold set according to the process smoothness requirement (for example, 15 degrees), it is determined that the turning angle at this point is relatively sharp. At this time, the analysis tool will extract these data to form detailed control point information, which provides the basis for subsequent adjustment. Based on the analyzed control point information, the Delta series sensor installed between the flange at the end of the mechanical arm and the polishing tool, as well as the matching signal conditioner and data acquisition card, are used to obtain the force feedback data in the machining process, so as to further understand the force condition when the tool contacts the curved surface.

[0100] The real-time force feedback value at a single control point is compared with the normal force range (for example, 2.0N to 3.0N) obtained based on historical machining data statistics. If the force feedback value is continuously higher than the upper limit of the range, it is determined that the force value at this point is high. In the middle area of the blade, the collected force feedback data shows that the force value at some control points is high, reaching 4.0N, while the surrounding area is only 2.5N. This uneven distribution will be recorded to form a data table of feedback distribution. The data table is a two-dimensional relationship table, with the control point ID as the row index, and the column fields including: three-dimensional coordinates, theoretical force value, measured force value, force deviation, and state flag. This data helps to identify potential risk points in the machining process. For the feedback distribution, the dynamic performance monitor built into the mechanical arm controller extracts the real-time response value (such as step response time) of the machining tool at each control point to judge the response matching degree. According to the theoretical response time calibrated in the factory specification book of the mechanical arm servo system, and combined with a certain safety margin (for example, 1.2 times of the theoretical value), the preset threshold is determined. Assuming that in the root area of the blade, the detected response value fluctuates greatly, and the response value of some points is only 70% of the preset threshold, which does not meet the expected standard. At this time, the online parameter self-tuner in the mechanical arm control system adjusts these areas. The response deviation identified by the tuner is updated online using the recursive least squares method to estimate the values of inertia and friction parameters in the system model, and then the pole placement method is used to recalculate the values of the feedforward gain and proportional gain of the servo driver. The newly calculated gain parameters are written into the servo driver in real time through the field bus (such as EtherCAT). Finally, the calibrated response data is obtained to ensure that the response value is close to 90% or more of the preset threshold.

[0101] It is worth mentioning that in the field of high-precision processing such as polishing of aero-engine blades, the application of accurate sensing conditions is particularly crucial. For the acquisition of sensing data, high-sensitivity sensors can be arranged on the end tool of the mechanical arm to collect environmental information and contact state in real time during the processing. Assuming that in the edge area of the blade, the sensor detects that the surface roughness changes greatly, with data values fluctuating between 2.5 and 3.8 μm, this data will be transmitted to the data analysis tool for processing to extract key parameters such as roughness peak value and change frequency, providing a basis for subsequent comparison. For the comparison process of parameter information and the upper limit of the surface roughness tolerance specified in the aero-engine blade design drawings, a roughness threshold of 3.0 μm can be set. If the analysis result shows that some areas exceed this value, it is considered that the parameters meet the update condition, forming an adjustable sensing data set. In the specific implementation, the distribution of the areas exceeding the threshold can be displayed through a visual interface, such as finding an abnormal point with a roughness of 3.5 μm in the middle area of the blade, which is automatically marked and included in the adjustment range. This approach helps to quickly locate problem areas.

[0102] The force control servo loop of the mechanical arm is calibrated to adjust the polishing tool pressure and speed according to the sensing data set. Assuming that the target trajectory requires the polishing pressure to be stabilized at 3.2 N, while the current data is 3.5 N, the force control servo loop of the mechanical arm will take the target pressure value 3.2 N as the set point of the force control loop, and the real-time measurement value 3.5 N as the feedback. The PID controller of the force control loop calculates the required adjustment amount and outputs a correction signal to the servo driver, instructing the torque of the micro motor to gradually reduce the contact pressure and stabilize it at the target value of 3.2 N, while monitoring whether the trajectory conforms to the preset path. Through this calibration, the adjusted parameter set is obtained to ensure that the processing process better fits the blade surface characteristics.

[0103] In step S15, according to the polishing trajectory, the polishing verification result is extracted, and according to the polishing verification result, the supplementary processing instruction is generated and executed, and the polishing adjustment scheme is determined, including:

[0104] According to the polishing trajectory, the surface roughness is obtained, and compared with the preset roughness threshold to obtain the initial distribution result;

[0105] According to the initial distribution result, the quality distribution information is obtained and trend analysis is performed to determine the polishing verification result;

[0106] According to the polishing verification result, the feature classification of the surface to be polished is performed to obtain the classification region information;

[0107] According to the classification region information, the supplementary processing instruction is generated and executed to obtain the polishing adjustment scheme.

[0108] According to the polishing track, surface roughness data is obtained, a roughness threshold value is determined according to a surface roughness upper limit value specified in a product process specification, and the surface roughness data is compared with the roughness threshold value to obtain an initial distribution evaluation result. Multi-source feedback data is integrated, and the change rule of surface roughness is analyzed to obtain comprehensive trend distribution data. If the data does not meet the quality consistency requirement, secondary correction processing is performed to obtain adjusted quality distribution information. The quality consistency requirement refers to the percentage of sampling points (pass rate) with a roughness value lower than the upper limit value (such as 3.0 μm) in the entire polishing area must reach or exceed 98%. For example, if the total sampling points are 1000, at least 980 points need to meet the roughness ≤ 3.0 μm. Then, the number of points with a roughness value lower than the surface roughness upper limit value in all sampling points in the entire polishing area is counted, and the percentage of the total sampling points, i.e. the pass rate, is calculated. If the pass rate is greater than or equal to the minimum consistency standard required by the process (for example, 98%), it is determined that the compliance meets the standard, otherwise it is determined as not meeting the standard. Finally, the polishing verification result is obtained.

[0109] According to the polishing verification result, the local details of insufficient area coverage or uneven roughness distribution are obtained, and feature classification is performed thereon, such as being divided into high, medium and low priority according to the percentage of deviation values exceeding the threshold value. For the high priority area, supplementary processing instructions are generated, the processing range is determined, the preliminary supplementary processing path is determined, and the linear interpolation or Bezier curve optimization is refined and adjusted to determine the polishing adjustment scheme.

[0110] Exemplarily, in the field of polishing of aero-engine blades, data acquisition and analysis of the corrected polishing track are crucial. Surface roughness data is collected in real time by a high-precision sensor, and compared with a preset threshold value to preliminarily evaluate the region compliance. For example, the roughness of a certain area is 2.8 μm, which is lower than the threshold value of 3.0 μm, and can be preliminarily determined as qualified, but still needs to be combined with multi-point data to judge the overall uniformity. Based on the state estimator of Kalman filter, multi-source feedback data from roughness sensor, force sensor and position encoder are integrated, and the roughness change trend is obtained by analyzing the first derivative of roughness value with respect to time in the fused data, i.e. the change slope. For example, if the roughness of the middle part of the blade increases from 2.5 μm to 2.9 μm, and is accompanied by uneven pressure, it is identified as a potential problem area, providing a basis for subsequent adjustment. If the comprehensive trend does not meet the quality consistency requirement, the tool path editing and optimization module in Powermill software is started for secondary correction.

[0111] For example, for uneven edge roughness, by calculating the standard deviation of the roughness value of the local area, if the standard deviation is greater than the allowed value (such as 0.15 μm), it is determined that the edge roughness is uneven, and by re-planning the path, increasing the processing time or adjusting the tool angle, the uniformity of the distribution is optimized, and it is verified whether the corrected data meets the standard. The process capability analysis function of Minitab software is used to evaluate the degree of conformity of the corrected quality distribution with the consistency requirements. For example, if the roughness of the blade root is stable at 2.7 μm and the distribution is uniform, based on the calculation result of the process capability index Cpk, if Cpk≥1.33 (corresponding to about 95% of the qualified rate), it is determined that it is a high-quality area; if a point reaches 3.1 μm, it is marked as needing local optimization to achieve accurate positioning. The overall data and historical records are combined to generate a high-quality polishing verification result. For example, if the overall roughness of the blade is distributed between 2.5 and 2.9 μm and there is no abnormality, it is determined that the verification is passed.

[0112] It is worth noting that in the polishing process of the aero-engine blade, for the subsequent processing of the high-quality polishing verification result, multi-angle analysis and optimization are required for the insufficiently processed area. The data processing script based on SQL query statement or Pandas library is used to filter the roughness abnormal area from the verification result, and the measured roughness value of each sampling point is compared with the upper limit value specified in the process specification. If the measured value continuously exceeds the upper limit value, the area to which the point belongs is determined as an abnormal area. For example, when it is detected that the roughness of the edge of the blade is 3.2 μm (the standard value is 3.0 μm), the system will automatically mark the area as insufficiently covered and extract its location information to quickly locate the problem area. According to the roughness deviation degree and the area position, the local details are classified by features. For example, if the deviation of the blade root is 3.1 μm, and the edge area reaches 3.3 μm, the system will classify the edge as a high-priority area. This classification method provides a clear basis for subsequent processing demand assessment, ensuring reasonable allocation of resources.

[0113] In the path planning link, the system sets a priority threshold (such as 80%) to evaluate the processing demand. For example, if the edge area priority evaluation value is 85%, which exceeds the threshold, a supplementary processing instruction is generated, and the processing range is determined as a 5 mm wide strip area on the edge to ensure that critical areas are processed in a timely manner. The trajectory generation tool designs a preliminary processing path based on the coverage completeness data. For example, if the coverage rate of the preliminary path in the edge area is only 70% (the standard is 90%), the system will optimize it by increasing the number of processing times or adjusting the moving speed to increase the coverage rate to 92%, thereby enhancing the comprehensiveness of processing.

[0114] In step S16, according to the polishing adjustment scheme, the running trajectory and pressure parameters of the mechanical arm are updated, and the polishing processing result is obtained by combining real-time response verification and local defect detection, including:

[0115] According to the polishing adjustment scheme, the mechanical arm control instruction is updated and executed to obtain running track data;

[0116] According to the running track data, pressure adjustment is performed to determine a pressure parameter configuration;

[0117] According to the pressure parameter configuration, defect recognition is performed to obtain defect distribution details;

[0118] According to the defect distribution details, a track correction instruction is generated, the mechanical arm control instruction is dynamically updated, a final running track and a final pressure parameter configuration are determined, and a polishing processing result is obtained.

[0119] According to the polishing adjustment scheme, a new mechanical arm control instruction is generated to obtain updated running track data. A preset pressure threshold is determined according to the intersection of the yield limit of the workpiece material to be polished and the safe pressure range of the polishing tool; for example, for an aviation aluminum alloy blade, the yield limit is 250 MPa, and the safe pressure range of the polishing tool is 3-6 N / cm², so the intersection is set to 5 N / cm². According to the running track data, pressure adjustment evaluation is performed, and if the pressure value exceeds the preset pressure threshold, dynamic adjustment is performed, the PID adjustment module of the force controller calculates the compensation amount required to restore the pressure to within the threshold, the compensation amount is converted into a position correction amount of the mechanical arm end in the normal direction of the curved surface, and the real-time path interpolator is used to superimpose this correction amount on the original track to generate a new path that is instantaneously controlled by pressure. The process is executed in a loop until the pressure is stabilized within the threshold range, and a new pressure parameter configuration is determined.

[0120] It is worth noting that in the process of executing the new parameters, the state data of the machined surface is obtained, and defect recognition is performed to obtain defect distribution details; for example, surface topological data is collected by a Keyence laser profiler, and a gray threshold segmentation algorithm is used to identify defect areas. According to the defect distribution details, a custom message publishing node running on a ROS 2 real-time system is used to convert the deviation information into a track correction instruction, and the track correction instruction is sent to the motion planning layer of the mechanical arm controller. The planning layer calls a dynamic motion primitive DMP to instantaneously generate a sub-track that avoids or refines the defect area under the premise of ensuring motion smoothness and continuity, and replaces the corresponding track segment in the original instruction with this sub-track, thereby achieving seamless dynamic updating of the instruction. Finally, a set of final running track and final pressure parameter configuration that can simultaneously ensure track accuracy, appropriate pressure, and defect-free surface are determined, thereby obtaining a lean complex curved surface polishing result.

[0121] In a possible implementation, the related technical subject can be analyzed and applied in detail from multiple perspectives for the implementation of the final polishing path adjustment scheme. The post-processor in the offline programming software RobotStudio generates the robot control instruction, the surface damage critical pressure of the blade material is obtained through a material mechanics experiment, the preset pressure threshold is determined, it is assumed that the preset pressure threshold is 5N per cm2, and the initial trajectory planning shows that the pressure of the blade tip area reaches 6N, which exceeds the threshold. When the pressure exceeds the threshold, the pressure of the polishing tool or the moving speed is dynamically adjusted, and the parameters are optimized to be within 4.8N. The roughness data is collected in real time by using a Keyence LJ-V7000 series laser profiler, and is compared with the roughness qualified range specified in the process file. If the data continuously exceeds the upper limit of the range, it is determined that there is a roughness deviation. If the monitoring finds that there is a roughness deviation at the root of the blade, the rule engine deployed in the MES system analyzes the machining trajectory of the area, generates a correction instruction, and guides the robot to add a local polishing action in the area, and adjusts the tool angle to improve the machining uniformity.

[0122] In summary, the polishing trajectory real-time correction control method disclosed in the present application comprises the following steps: collecting multi-source data of a surface to be polished, preprocessing the multi-source data to obtain a real-time machining data set; extracting deviation distribution characteristics according to the real-time machining data set, determining an initial deviation vector, and generating a temporary machining path according to the initial deviation vector; obtaining key parameters of the temporary machining path and performing matching verification with the surface to be polished, optimizing the curvature of the temporary machining path to obtain an optimized machining path; extracting key path parameters according to the optimized machining path, comparing the key path parameters with preset perception conditions, and determining a polishing trajectory; extracting a polishing verification result according to the polishing trajectory, generating a supplementary machining instruction according to the polishing verification result and executing the supplementary machining instruction, and determining a polishing adjustment scheme; and updating the running trajectory and pressure parameters of the robot according to the polishing adjustment scheme, combining real-time response verification and local defect detection to obtain a polishing machining result. The polishing trajectory real-time correction control method solves the problem that the existing control method lacks a sensing and response mechanism for dynamic changes in the machining process.

[0123] Reference Figure 2 The second embodiment of the present application provides a polishing trajectory real-time correction control system, comprising:

[0124] A data acquisition and processing module is configured to collect multi-source data of a surface to be polished, preprocess the multi-source data to obtain a real-time machining data set, extract deviation distribution characteristics according to the real-time machining data set, determine an initial deviation vector, and generate a temporary machining path according to the initial deviation vector.

[0125] A deviation analysis module is configured to extract deviation distribution characteristics according to the real-time machining data set, determine an initial deviation vector, and generate a temporary machining path according to the initial deviation vector.

[0126] A path optimization module is configured to acquire key parameters of the temporary machining path, perform matching verification with the curved surface to be polished, perform curvature optimization on the temporary machining path, and obtain an optimized machining path.

[0127] An execution correction module is configured to extract key path parameters according to the optimized machining path, compare the key path parameters with preset perception conditions, and determine a polishing track.

[0128] A supplementary machining module is configured to extract a polishing verification result according to the polishing track, generate a supplementary machining instruction according to the polishing verification result, and perform the supplementary machining instruction to determine a polishing adjustment scheme.

[0129] A dynamic updating module is configured to update a running track and a pressure parameter of a mechanical arm according to the polishing adjustment scheme, combine real-time response verification and local defect detection, and obtain a polishing machining result.

[0130] It should be noted that the polishing track real-time correction control system provided by the embodiments of the present application is used to execute all process steps of the polishing track real-time correction control method of the above-mentioned embodiments, and the working principles and beneficial effects of the two are one-to-one correspondence, thus not being repeated.

[0131] The embodiments of the present application also provide an electronic device. The electronic device includes a processor, a memory, and a computer program, such as a polishing track real-time correction control program, stored in the memory and executable on the processor. The processor executes the computer program to implement the steps in the above-mentioned various polishing track real-time correction control method embodiments, such as the step S11 shown in the figure. Figure 1 Alternatively, the processor executes the computer program to implement the functions of each module / unit in the above-mentioned various device embodiments.

[0132] For example, the computer program can be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present application. The one or more modules / units can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program in the electronic device.

[0133] The electronic device can be a computing device such as a desktop computer, a notebook computer, a palm computer, a smart tablet, etc. The electronic device can include, but is not limited to, a processor, a memory. Those skilled in the art can understand that the above components are only examples of the electronic device and do not constitute a limitation on the electronic device, and the electronic device can include more or fewer components than the above, or combine certain components, or different components, for example, the electronic device can also include an input / output device, a network access device, a bus, etc.

[0134] The processor can be a central processing unit (CPU), and can also be other general-purpose processors, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The processor is the control center of the electronic device, and connects various parts of the electronic device through various interfaces and lines.

[0135] The memory can be used to store the computer programs and / or modules, and the processor realizes various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required by a function (such as a sound playing function, an image playing function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, for example, a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state memory device.

[0136] The modules / units integrated in the electronic device, if realized in the form of software function units and sold or used as independent products, can be stored in a computer readable storage medium. Based on such understanding, all or part of the processes in the above-mentioned embodiment methods can also be completed by a computer program instructing related hardware, and the computer program can be stored in a computer readable storage medium. The computer program can implement the steps of each method embodiment when executed by a processor. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or some intermediate forms, etc. The computer readable medium can include any entity or device, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. that can carry the computer program code. It should be noted that the contents included in the computer readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction, for example, in some jurisdictions, according to legislation and patent practice, the computer readable medium does not include electrical carrier signals and telecommunication signals.

[0137] It should be noted that the above-described device embodiments are only schematic, and the units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or distributed on multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the embodiment. In addition, the connection relationship between the modules in the device embodiment provided by the present application indicates that there is a communication connection between them, which can be realized as one or more communication buses or signal lines. Those skilled in the art can understand and implement it without creative labor.

[0138] The above-described specific embodiments further illustrate the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above-described specific embodiments are only for the specific embodiments of the present application and are not used to limit the protection scope of the present application. It is particularly pointed out that any modification, equivalent replacement, improvement, etc. made by those skilled in the art within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A method for real-time correction and control of polishing trajectory, characterized in that, include: Collect multi-source data of the surface to be polished, and preprocess the multi-source data to obtain a real-time processing dataset; Based on the real-time processing dataset, the deviation distribution features are extracted, the initial deviation vector is determined, and a temporary processing path is generated based on the initial deviation vector. The key parameters of the temporary processing path are obtained and their matching with the surface to be polished is verified. The curvature of the temporary processing path is optimized to obtain the optimized processing path. Based on the optimized processing path, key path parameters are extracted, and the key path parameters are compared with preset sensing conditions to determine the polishing trajectory. Based on the polishing trajectory, the polishing verification result is extracted, and based on the polishing verification result, a supplementary processing instruction is generated and executed to determine the polishing adjustment scheme; Based on the polishing adjustment scheme, the robotic arm's running trajectory and pressure parameters are updated, and the polishing results are obtained by combining real-time response verification and local defect detection.

2. The real-time correction and control method for polishing trajectory according to claim 1, characterized in that, The process involves collecting multi-source data from the surface to be polished, preprocessing the multi-source data to obtain a real-time processing dataset, including: Force feedback and surface roughness data of the surface to be polished are acquired by a sensor array to obtain multi-source data. The multi-source data is then denoised to obtain a denoised multi-source dataset. The denoised multi-source dataset is standardized to obtain a fused dataset. The fused dataset is then filtered using a preset standard threshold to obtain a processed dataset. Real-time features are extracted from the processing dataset to determine the stability of the processing environment, thus obtaining the real-time processing dataset.

3. The real-time correction and control method for polishing trajectory according to claim 1, characterized in that, The step of extracting deviation distribution features based on the real-time processing dataset, determining an initial deviation vector, and generating a temporary processing path based on the initial deviation vector includes: The real-time processing dataset is matched point by point with the preset trajectory data to determine the initial set of deviations; Spatial distribution processing is performed on the deviation values ​​of the preliminary deviation set to remove outliers in the distribution characteristics, thereby obtaining the deviation distribution characteristics; The deviation distribution characteristics are comprehensively processed to determine the initial deviation vector. The initial deviation vector is then matched one by one with a preset threshold range to obtain the deviation exceeding the limit points. The angle and velocity parameters of the deviation exceeding the limit points are extracted to obtain parameter adjustment data. Abnormal fluctuations in the parameter adjustment data are corrected, and key control parameters are extracted to obtain the adjustment range. A temporary processing path is obtained by performing dynamic simulation based on the adjustment range.

4. The real-time correction and control method for polishing trajectory according to claim 1, characterized in that, The process of obtaining key parameters of the temporary processing path, verifying their compatibility with the surface to be polished, and optimizing the curvature of the temporary processing path to obtain an optimized processing path includes: Extract key parameters from the temporary processing path, compare the areas where the key parameters do not match the preset parameter range, mark and obtain the mismatched areas; The polishing pressure distribution in the mismatched area is detected, the polishing pressure distribution is calibrated, and the curvature change of the surface to be polished is analyzed to obtain curvature change data. Based on the curvature change data, the temporary processing path is dynamically modified to correct path deviations and obtain an optimized processing path.

5. The real-time correction and control method for polishing trajectory according to claim 1, characterized in that, The step of extracting key path parameters based on the optimized processing path and comparing the key path parameters with preset sensing conditions to determine the polishing trajectory includes: Based on the optimized processing path, key path parameters are extracted to obtain control point data; Based on the control point data, the feedback distribution and response matching degree are determined, and the response data is obtained; Based on the response data and the characteristics of the surface to be polished, parameter set data is obtained; The parameter set data is compared with preset sensing conditions to calibrate the operating parameters of the robotic arm and determine the polishing trajectory.

6. The real-time correction and control method for polishing trajectory according to claim 1, characterized in that, The steps include extracting polishing verification results based on the polishing trajectory, generating and executing supplementary processing instructions based on the polishing verification results, and determining a polishing adjustment plan, including: Based on the polishing trajectory, the surface roughness is obtained and compared with a preset roughness threshold to obtain the initial distribution result; Based on the initial distribution results, quality distribution information is obtained and trend analysis is performed to determine the polishing verification results; Based on the polishing verification results, the surface to be polished is classified to obtain classification region information; Based on the classified area information, supplementary processing instructions are generated and executed to obtain a polishing adjustment scheme.

7. The real-time correction and control method for polishing trajectory according to claim 1, characterized in that, The process involves updating the robotic arm's trajectory and pressure parameters according to the polishing adjustment scheme, and combining real-time response verification and local defect detection to obtain the polishing result, including: Based on the polishing adjustment scheme, update and execute the robotic arm control commands to obtain the running trajectory data; Based on the aforementioned operating trajectory data, pressure adjustment is performed to determine the pressure parameter configuration; Based on the pressure parameter configuration, defect identification is performed to obtain detailed defect distribution information; Based on the defect distribution details, a trajectory correction instruction is generated, and the robotic arm control instruction is dynamically updated to determine the final running trajectory and final pressure parameter configuration, thereby obtaining the polishing result.

8. A real-time polishing trajectory correction control system, characterized in that, include: The data acquisition and processing module is used to acquire multi-source data of the surface to be polished, preprocess the multi-source data, and obtain a real-time processing dataset. The deviation analysis module is used to extract deviation distribution features based on the real-time processing dataset, determine the initial deviation vector, and generate a temporary processing path based on the initial deviation vector. The path optimization module is used to obtain the key parameters of the temporary processing path, verify the matching with the surface to be polished, optimize the curvature of the temporary processing path, and obtain the optimized processing path. The execution correction module is used to extract key path parameters based on the optimized processing path, compare the key path parameters with preset sensing conditions, and determine the polishing trajectory. The supplementary processing module is used to extract the polishing verification results based on the polishing trajectory, generate and execute supplementary processing instructions based on the polishing verification results, and determine the polishing adjustment plan. The dynamic update module is used to update the robotic arm's running trajectory and pressure parameters according to the polishing adjustment scheme, and obtain the polishing result by combining real-time response verification and local defect detection.

Images