Position compensation control method and system for machine tool rotary bin

Abstract

The present application provides a position compensation control method and system for a machine tool rotary hopper, the method comprising: determining the standard rotation angle required for station rotation according to the number of stations of the machine tool rotary hopper, and determining the first position information of each station; obtaining image information after the machine tool rotary hopper completes rotation, and determining the second position information of each station through image processing; determining the actual rotation angle of the current ejection station according to the first position information and the second position information; judging whether the rotation angle error of the next ejection station exceeds the rotation error threshold according to the actual rotation angle, the standard rotation angle, and the number of rotations; if the rotation angle error exceeds the rotation error threshold, performing rotation angle compensation on the next rotation of the machine tool rotary hopper according to the rotation angle error, and resetting the number of rotations. Through precise error detection and compensation, the accuracy of the rotation angle is guaranteed, thereby improving the accuracy and efficiency of the entire production process.

Description

Technical Field

[0001] The present application relates to the technical field of machine tool silos, and in particular to a position compensation control method and system for a machine tool rotary silo. Background Art

[0002] In modern manufacturing, machine tool rotary silos, as a crucial component of automated production lines, are widely used in machining, assembly, and other automated operations. These silos are typically used to store and manage the various raw materials required for production, effectively improving production efficiency. However, with the increasing degree of automation in production, the precision requirements for rotary silos are also increasing. Traditional rotary silo control methods rely on fixed rotation angles and simple positioning techniques. These methods often fail to meet the high-precision and high-efficiency production requirements under certain complex working conditions.

[0003] For example, in certain high-precision machining tasks, errors in the station switching of a rotary silo can lead to inaccurate station positions, affecting the handling and processing of raw materials, and even causing quality issues with the product. Especially during the use of a rotary silo, due to the accumulation of rotational errors, the positioning deviation of the actual station will gradually increase as production continues, causing the raw materials to shift position during the robot's grasping process, thereby affecting the accuracy of subsequent operations. Existing technologies are usually only able to perform rough angle control during rotation, and are unable to adjust and compensate for position deviations caused by rotational errors in real time. Furthermore, they are unable to dynamically adjust the rotation angle based on the actual error to ensure accuracy. Summary of the Invention

[0004] The present application provides a position compensation control method and system for a machine tool rotary hopper, which is used to solve the problem in related technologies that the rotation angle cannot be dynamically adjusted according to the actual error.

[0005] In a first aspect, the present application provides a position compensation control method for a machine tool rotary hopper, the position compensation control method for a machine tool rotary hopper comprising:

[0006] Determine the standard rotation angle required for the station rotation according to the number of stations of the machine tool rotary bin, and determine the first position information of each station;

[0007] Obtaining image information of the machine tool rotary hopper after the rotation is completed, and determining second position information of each workstation through image processing;

[0008] Determine the actual rotation angle of the current ejecting station according to the first position information and the second position information;

[0009] Determine whether the rotation angle error of the next ejection station exceeds a rotation error threshold according to the actual rotation angle, the standard rotation angle, and the number of rotations;

[0010] If the rotation angle error exceeds the rotation error threshold, rotation angle compensation is performed on the next rotation of the machine tool rotary silo according to the rotation angle error, and the number of rotations is reset.

[0011] Optionally, in a first implementation of the first aspect of the present application, the step of obtaining image information of the machine tool rotary hopper after the rotation is completed, and determining the second position information of each workstation through image processing includes:

[0012] The binocular vision sensor is used to collect a multi-view image sequence of the rotating silo workstation, and the multi-view image sequence is temporally and spatially registered to generate a three-dimensional point cloud map.

[0013] Based on a preset workstation edge feature template, multi-scale convolution kernel filtering is performed on the three-dimensional point cloud mapping image to extract the first edge contour feature point set of each workstation;

[0014] generating a workstation space coordinate matrix according to a matching degree between the first edge contour feature point set and the workstation edge feature template;

[0015] The workstation space coordinate matrix is ​​input into a coordinate system conversion model, and the second position information of each workstation is output.

[0016] Optionally, in a second implementation of the first aspect of the present application, the step of determining an actual rotation angle of the current ejecting station according to the first position information and the second position information includes:

[0017] Dynamically matching the center point of each workstation according to the first position information and the workstation spatial coordinate matrix to generate a workstation center offset vector group;

[0018] Constructing a coordinate system transformation model based on the workstation center offset vector group, performing posture calculation on the workstation space coordinate matrix, and determining the real-time posture parameters of each workstation in the rotation axis coordinate system;

[0019] Calculating the instantaneous rotation angle deviation value of the current ejection station according to the Euler angle component in the real-time posture parameter and the standard rotation angle;

[0020] The instantaneous rotation angle deviation value is integrated with the historical accumulated error data to determine the actual rotation angle of the current ejection station.

[0021] Optionally, in a third implementation of the first aspect of the present application, the step of determining whether a rotation angle error of the next ejection station exceeds a rotation error threshold based on the actual rotation angle, the standard rotation angle, and the number of rotations includes:

[0022] constructing a dynamic error sequence based on the historical accumulated error data;

[0023] Determine the predicted rotation angle deviation value of the next ejection station by combining the dynamic error sequence with the current instantaneous rotation angle deviation value;

[0024] Performing dynamic weighted correction on the predicted rotation angle deviation value according to the rotation number and a preset error amplification coefficient curve to generate a corrected predicted error value;

[0025] Based on an asymmetric comparison between the corrected prediction error value and the confidence interval boundary of the rotation error threshold, it is determined whether the rotation angle error of the next ejection station exceeds the rotation error threshold.

[0026] Optionally, in a fourth implementation of the first aspect of the present application, the step of performing rotation angle compensation for the next rotation of the machine tool rotary hopper according to the rotation angle error and resetting the number of rotations includes:

[0027] generating a dynamic compensation coefficient according to the rotation angle error;

[0028] Inputting the dynamic compensation coefficient and the standard rotation angle into a servo control model to generate a rotation control instruction;

[0029] Controlling the machine tool rotary bin to perform compensatory rotation according to the rotation control instruction, and collecting the actual offset of the compensated workstation through a laser displacement sensor;

[0030] Based on the actual offset of the workstation and a preset convergence judgment condition, the historical accumulated error data and the number of rotations are reset.

[0031] Optionally, in a fifth implementation of the first aspect of the present application, the method further includes:

[0032] Determining a single rotation angle error according to the standard rotation angle and the single rotation angle;

[0033] Determining whether the rotation angle error corresponding to the number of rotations exceeds the rotation error threshold according to the single rotation angle error and the number of rotations;

[0034] If the rotation angle error exceeds the rotation error threshold, rotation angle compensation is performed on the workstation rotation corresponding to the rotation number according to the rotation angle error, and the rotation number is reset.

[0035] Optionally, in a sixth implementation of the first aspect of the present application, the method further includes:

[0036] Acquire image information of the raw material on the ejecting station according to the image information, and extract a second edge contour feature point set of the raw material through image processing;

[0037] Perform feature matching on the edge contour feature point set and a preset standard pose template to generate actual pose parameters of the raw material;

[0038] Calculating a posture deviation value of the raw material based on a difference between the actual posture parameter and the standard posture template;

[0039] If the posture deviation value exceeds a preset posture tolerance threshold, a posture adjustment instruction is generated and the fine-tuning mechanism of the machine tool rotary hopper is driven to correct the posture of the raw material.

[0040] A second aspect of the present application provides a position compensation control device for a machine tool rotary hopper, the position compensation control device for a machine tool rotary hopper comprising:

[0041] A calibration module, used to determine the standard rotation angle required for the station rotation according to the number of stations of the machine tool rotary bin, and determine the first position information of each station;

[0042] an acquisition module, configured to acquire image information of the machine tool rotary hopper after the machine tool completes rotation, and determine second position information of each workstation through image processing;

[0043] a determination module, configured to determine an actual rotation angle of the current ejecting station according to the first position information and the second position information;

[0044] a judgment module, configured to judge whether a rotation angle error of a next ejection station exceeds a rotation error threshold according to the actual rotation angle, the standard rotation angle, and the number of rotations;

[0045] The compensation module is used to perform rotation angle compensation on the next rotation of the machine tool rotary hopper according to the rotation angle error and reset the number of rotations if the rotation angle error exceeds the rotation error threshold.

[0046] A third aspect of an embodiment of the present application provides an electronic device, comprising a memory and a processor, wherein the processor is used to execute a computer program stored on the memory. When the processor executes the computer program, it implements each step of the position compensation control method for the machine tool rotary hopper provided in the first aspect of the embodiment of the present application.

[0047] The fourth aspect of the embodiment of the present application provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the various steps in the position compensation control method of the machine tool rotary hopper provided in the first aspect of the embodiment of the present application are implemented.

[0048] In summary, according to the position compensation control method and system of a machine tool rotary hopper provided by the present application, the standard rotation angle required for station rotation is determined according to the number of stations of the machine tool rotary hopper, and the first position information of each station is determined; the image information of the machine tool rotary hopper after the rotation is completed is obtained, and the second position information of each station is determined through image processing; the actual rotation angle of the current ejecting station is determined according to the first position information and the second position information; whether the rotation angle error of the next ejecting station exceeds the rotation error threshold is judged according to the actual rotation angle, the standard rotation angle and the number of rotations; if the rotation angle error exceeds the rotation error threshold, the next rotation of the machine tool rotary hopper is compensated for the rotation angle according to the rotation angle error, and the number of rotations is reset. Through precise error detection and compensation, the accuracy of the rotation angle is guaranteed, thereby improving the accuracy and efficiency of the entire production process. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A schematic flow chart of a position compensation control method for a machine tool rotary bin provided in an embodiment of the present application;

[0050] Figure 2 A schematic diagram of a program module of a position compensation control device for a machine tool rotary hopper provided in an embodiment of the present application;

[0051] Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0052] In order to make the purpose, features, and advantages of the invention of this application more obvious and easy to understand, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts shall fall within the scope of protection of this application.

[0053] In order to solve the problem in the related art that the rotation angle cannot be dynamically adjusted according to the actual error, the embodiment of the present application provides a position compensation control method for a machine tool rotating bin, such as Figure 1 The flowchart of the position compensation control method of the machine tool rotary bin provided in this embodiment includes the following steps:

[0054] Step 110: Determine the standard rotation angle required for the station rotation according to the number of stations of the machine tool rotary silo, and determine the first position information of each station.

[0055] Specifically, the number of workstations in the rotary silo determines the angular distribution during the rotation process. The standard rotation angle for each workstation is calculated by dividing the entire rotation range of 360 degrees by the total number of workstations. This calculation clearly defines the required rotation angle for each workstation, ensuring that the rotary silo can accurately switch to the predetermined workstation position. On this basis, the initial position information of the workstation is usually determined by obtaining the initial coordinate position of the workstation using a fixed sensor or laser measurement system to ensure the accuracy of the workstation position data. This calibration data provides a reference for subsequent rotation operations, ensuring a one-to-one correspondence between the workstation and the rotation angle.

[0056] Step 120: Obtain image information of the machine tool rotary hopper after the rotation is completed, and determine the second position information of each workstation through image processing.

[0057] Specifically, after the rotary silo completes its rotation, the image information provides data on the actual workstation location. Using binocular vision sensors or other imaging technologies, image data of the rotary silo is acquired. This image data requires processing, such as distortion correction and image registration, to ensure accurate workstation images are captured. Image processing algorithms further identify feature points in the image and, combined with a preset workstation template, determine the exact location of each workstation by comparing the match between the feature points in the image and the template. This processing not only improves the accuracy of workstation positioning but also provides real-time feedback on changes in workstation position, providing a basis for subsequent error judgment and compensation.

[0058] In an optional implementation of the present embodiment, the steps of obtaining image information of the machine tool rotary hopper after the rotation is completed, and determining the second position information of each workstation through image processing include: collecting a multi-perspective image sequence of the rotary hopper workstation through a binocular vision sensor, performing spatiotemporal registration on the multi-perspective image sequence, and generating a three-dimensional point cloud mapping image; performing multi-scale convolution kernel filtering on the three-dimensional point cloud mapping image based on a preset workstation edge feature template, and extracting a first edge contour feature point set of each workstation; generating a workstation space coordinate matrix based on the matching degree between the first edge contour feature point set and the workstation edge feature template; inputting the workstation space coordinate matrix into a coordinate system conversion model, and outputting the second position information of each workstation.

[0059] Specifically, a binocular vision system consists of two cameras spaced a certain distance apart. It simultaneously captures images within the fields of view of both cameras and uses the parallax between the images to reconstruct the object's three-dimensional structure. The resulting multi-view image sequences provide visual data from multiple angles for each workstation in the rotary silo. These images can capture the appearance of each workstation at different rotation angles. The variations in the image of each workstation under different viewpoints provide rich spatial information, helping to accurately construct a three-dimensional model of the rotary silo. Spatiotemporal registration involves uniformly aligning images from different viewpoints under different temporal and spatial conditions, eliminating temporal and spatial deviations and ensuring that the images are effectively aligned in the same coordinate system. Spatiotemporal registration ensures that image data captured from multiple viewpoints can be accurately fused into a high-quality image sequence, generating a 3D point cloud map. Edge features are prominent indicators of an object's surface contour and can typically be extracted using image processing algorithms. In a 3D point cloud map, workstation edge features typically appear as areas with significant variations in point cloud density. Therefore, multi-scale convolution kernel filtering can effectively extract this edge information from the point cloud data. Convolution kernel filtering uses convolution kernels of different sizes to convolve an image, extracting features at different scales. The significance of multi-scale convolution kernel filtering lies in the fact that edge information of workstations may vary in scale due to different angles in the rotating silo or different camera angles. Therefore, multi-scale filtering can effectively identify and extract edge features at multiple levels, avoiding the loss of important edge information due to scale differences. Precise positioning of the workstations can be achieved by matching the extracted first edge contour feature point set against a preset workstation edge feature template. A workstation edge feature template is a pre-defined template containing workstation feature points and edge shapes based on the morphology and structure of the rotating silo workstations. These templates can be obtained through actual calibration or design models and reflect the basic contour features of the workstations. The matching process typically relies on standard image matching algorithms, such as those based on SIFT (Scale-Invariant Feature Transform) or SURF (Speeded Up Robust Features). The similarity between feature point sets is calculated to determine the degree of match between the first edge contour feature point set and the template. If the degree of match is high, it means that the outline of the current workstation matches the template, and the position of the workstation is confirmed. Finally, the workstation space coordinate matrix is ​​input into the coordinate system conversion model to achieve the conversion from one coordinate system to another, and output the second position information of each workstation. The core of the coordinate system conversion model is to convert the workstation data in three-dimensional space from the image coordinate system or camera coordinate system to the world coordinate system or the preset coordinate system of the control system, so as to ensure that the actual position of the workstation can be connected to the control system. Through this conversion, the spatial coordinate information of the workstation can not only be represented in the visual data, but also can be actually applied in the control system.

[0060] Step 130: Determine the actual rotation angle of the current ejecting station according to the first position information and the second position information.

[0061] Specifically, by comparing the first and second position information, the magnitude of the rotational error can be determined. The difference between the actual station position and the standard station position reflects the actual rotation angle deviation of the current station, which in turn determines the actual rotation angle of the ejection station. This process is calculated using a mathematical model. Based on the initial coordinates of each station in the first position information and the image position obtained from the second position information, a dynamic offset vector is generated to calculate the rotational angle error of the station. This calculation step ensures the positioning accuracy of the rotary silo and eliminates the influence of rotational error, providing an important basis for error compensation in the next step.

[0062] In an optional implementation of the present embodiment, the step of determining the actual rotation angle of the current material ejecting station based on the first position information and the second position information includes: dynamically matching the center point of each station based on the first position information and the station space coordinate matrix to generate a station center offset vector group; constructing a coordinate system transformation model based on the station center offset vector group, performing posture solution on the station space coordinate matrix, and determining the real-time posture parameters of each station in the rotation axis coordinate system; calculating the instantaneous rotation angle deviation value of the current material ejecting station based on the Euler angle component in the real-time posture parameter and the standard rotation angle; and fusing the instantaneous rotation angle deviation value with the historical accumulated error data to determine the actual rotation angle of the current material ejecting station.

[0063] Specifically, in a rotary silo, the initial coordinates of each workstation are the ideal positions of the workstations, pre-determined using precision measuring instruments. These coordinates provide a benchmark, ensuring the known positions of the workstations when not rotating. In contrast, the workstation spatial coordinate matrix is ​​dynamically generated during the operation of the rotary silo using technologies such as vision sensors or laser scanning. It reflects the actual spatial positions of the workstations during rotation. By comparing these two, deviations in workstation positions can be detected, and the center point offset of each workstation can be calculated. This offset vector group records the actual offset of each workstation relative to its initial coordinates, ensuring accurate tracking of workstation position changes. For example, if a workstation shifts due to mechanical control or environmental factors, the matching process captures this deviation and generates a vector representing the specific offset of the workstation relative to its initial position, allowing precise adjustment of the workstation position. Based on these offset vector groups, a coordinate transformation model is constructed to convert the workstation spatial coordinate matrix into the rotary axis coordinate system. The function of the coordinate transformation model is to transform the workstations from the coordinate system used by the vision system to the rotary axis coordinate system used for actual control of the rotary silo. The transformed coordinate data can be used for pose calculation to determine the real-time pose parameters of each workstation in the rotation axis coordinate system. Pose calculation combines the position of each workstation with its rotation angle using a mathematical model to determine the workstation's pose information. Pose not only includes the position's translation vector but also its rotation angle, which is typically expressed using Euler angles. Euler angles describe the rotation of an object in three-dimensional space using three rotation angles (about the X, Y, and Z axes). Through pose calculation, the system accurately calculates the pose parameters of each workstation in the rotation axis coordinate system based on the actual motion trajectory of the rotating silo. The instantaneous rotation angle deviation of the current ejection station is calculated based on the Euler angle components in the real-time pose parameters and the standard rotation angle. The standard rotation angle is a target angle preset based on production requirements or system design, representing the ideal position that the workstation should achieve. The Euler angle components in the real-time pose parameters allow the system to determine the actual rotation angle of the workstation. The instantaneous rotation angle deviation is the difference between the current workstation's actual rotation angle and the standard rotation angle. This deviation can be calculated using a simple subtraction method and represents the error in the current rotation process. Finally, the instantaneous rotation angle deviation value is combined with historical cumulative error data to generate the actual rotation angle of the current ejection station. Historical cumulative error data refers to the long-term accumulated errors caused by mechanical errors, sensor accuracy, and other factors during the operation of the rotary silo. These errors can gradually increase with the number of rotations. Therefore, relying solely on the instantaneous deviation value is insufficient to reflect the actual system error. By combining historical cumulative errors with the current instantaneous error, a more accurate rotation angle can be obtained, thereby correcting any accumulated deviations in the system.

[0064] Step 140: Determine whether the rotation angle error of the next ejection station exceeds a rotation error threshold based on the actual rotation angle, the standard rotation angle, and the number of rotations.

[0065] Specifically, in this step, the error in the current rotation process is calculated by comparing the actual rotation angle with the standard rotation angle. At the same time, considering the cumulative effect of the number of rotations, the rotation error will gradually increase with each rotation. Therefore, it is necessary to determine whether the rotation angle error of the next ejection station exceeds the preset rotation error threshold. This step relies on the setting of the rotation error threshold and compares the historical cumulative error with the current rotation angle. Through this judgment, the system can determine whether error compensation is necessary to ensure the accuracy of the station rotation and the stability of the production process.

[0066] In an optional implementation of the present embodiment, the step of judging whether the rotation angle error of the next material ejecting station exceeds the rotation error threshold based on the actual rotation angle, the standard rotation angle and the number of rotations includes: constructing a dynamic error sequence based on historical accumulated error data; determining the predicted rotation angle deviation value of the next material ejecting station by combining the dynamic error sequence with the current instantaneous rotation angle deviation value; performing dynamic weighted correction on the predicted rotation angle deviation value based on the number of rotations and a preset error amplification coefficient curve to generate a corrected predicted error value; and performing an asymmetric comparison based on the confidence interval boundary of the corrected predicted error value and the rotation error threshold to judge whether the rotation angle error of the next material ejecting station exceeds the rotation error threshold.

[0067] Specifically, in the rotary silo control system, in order to determine whether the rotation angle error of the next ejection station exceeds the set threshold, it is necessary to conduct a systematic analysis based on the actual rotation angle, standard rotation angle and number of rotations before each station switch. First, the system records the cumulative error values ​​in the historical station switches to construct a dynamic error sequence. The sequence is arranged in chronological order and reflects the actual error generated by each rotation. The construction basis is the difference between the standard rotation angle and the actual rotation angle of the corresponding station. After filtering, the difference retains a stable and effective error trend to form a dynamic error curve. The existence of this error sequence is very critical for identifying error growth patterns and error periodicity, and is especially suitable for non-integer indexing angle silo systems. For example, in a 7-station system, the theoretical rotation angle of each rotation is 51.42857 degrees, but the actual control signal can only be issued 51.42 degrees, resulting in an error of 0.00857 degrees each time. If this error is not processed, it will reach a cumulative deviation of 0.06 degrees after seven rotations, causing the silo station to be misaligned. The system fuses the constructed dynamic error sequence with the instantaneous rotation angle deviation from the current rotation operation to determine the predicted rotation angle deviation for the next ejection station. This fusion process does not directly average the error, but instead considers the potential for instantaneous errors to predict future rotation results. By calculating the recent error growth rate and combining it with the changing trend of the instantaneous deviation, the system predicts the total deviation likely to occur after the next station rotation. Furthermore, to account for the cumulative effect of the number of rotations on the error, the control system inputs the predicted rotation angle deviation into an error amplification factor curve for dynamic weighted correction. The error amplification factor curve is a response curve generated by a nonlinear mapping of the number of rotations. The curve can be designed as an exponential, logistic, or power function, and is used to control the weight of historical errors at different stages of the rotation. If the number of rotations is small, the amplification factor is low, and the system is more tolerant of early deviations. However, as the number of rotations increases, the weight increases rapidly, causing historical errors to accumulate more quickly and triggering corrections. This mechanism implements a dynamic error amplification strategy that enables the control system to adaptively enhance its perception and response to long-term deviations. Finally, the system performs an asymmetric comparison of the corrected predicted error value against the set rotational error threshold to determine whether the confidence interval bounds of the error tolerance are exceeded. This comparison does not employ a symmetrical threshold. Instead, different trigger limits are set for positive and negative errors, allowing for adjustment of error sensitivity based on the silo's rotational direction and structural eccentricity. For example, if the system rotates clockwise and the structure is right-leaning, a lower threshold (e.g., 0.04 degrees) can be set for the clockwise direction, while a higher threshold (e.g., 0.06 degrees) can be set for the counterclockwise direction to prevent uncontrolled right-side displacement accumulation. This asymmetric judgment mechanism ensures a more refined and secure error compensation strategy.When the system detects that the corrected prediction error has exceeded the corresponding directional threshold, it immediately issues a compensation signal to adjust the target angle for the next ejection station. It also records the compensation behavior to support subsequent self-updates of the control strategy, thus forming a closed-loop rotation control and error correction mechanism. This approach enables the system to achieve long-term, stable control of the accuracy of the rotating silo station, effectively preventing problems such as grasping failure and material misalignment caused by accumulated rotational errors.

[0068] Step 150: If the rotation angle error exceeds the rotation error threshold, perform rotation angle compensation on the next rotation of the machine tool rotary bin according to the rotation angle error, and reset the number of rotations.

[0069] Specifically, in this step, if the error judgment result exceeds the set threshold, the system will automatically activate the compensation mechanism. Based on the size of the rotation angle error, the corresponding dynamic compensation coefficient is generated and input into the control system for compensation. During the compensation process, the rotation angle is adjusted to ensure that the rotary hopper can accurately reach the target workstation position. In addition, to avoid error accumulation, the system will reset the number of rotations after each compensation to prevent the continued accumulation of errors and the degradation of system performance. Through this operation, the system can maintain high-precision workstation switching and an efficient production process without stopping production.

[0070] In an optional implementation of this embodiment, the steps of performing rotation angle compensation for the next rotation of the machine tool rotary hopper according to the rotation angle error and resetting the number of rotations include: generating a dynamic compensation coefficient according to the rotation angle error; inputting the dynamic compensation coefficient and the standard rotation angle into the servo control model to generate a rotation control instruction; controlling the machine tool rotary hopper to perform compensatory rotation according to the rotation control instruction, and collecting the actual offset of the work station after compensation through a laser displacement sensor; resetting the historical accumulated error data and the number of rotations based on the actual offset of the work station and the preset convergence judgment condition.

[0071] Specifically, first, a dynamic compensation coefficient is generated based on the rotation angle error, and the error value needs to be mapped through a nonlinear mapping curve to achieve flexible response to errors of different amplitudes. The mapping curve can be in the form of The function of represents the rotation angle error, and is an adjustable gain constant, which makes the compensation coefficient close to 1 when the error is small, and automatically increases the compensation strength when the error is large. =0.02 and set β =0.1, γ= 10, the compensation coefficient C ≈ 1.23, allowing for appropriate angle correction. Next, the dynamic compensation coefficient and the target rotation angle are input into the servo control model. This model integrates feedforward torque compensation and closed-loop feedback control, combined with a proportional-integral-derivative (PID) algorithm, to output a pulse density or analog current command. The feedforward phase uses the compensation coefficient to modify the target angle curve, generating a pre-compensated angle signal. The feedback phase then performs error correction based on the measured value, resulting in a highly accurate rotation control command. This control command balances speed and torque requirements, ensuring the rotating silo moves smoothly along the corrected angle trajectory. This command then drives the rotating silo to perform the compensation action, while a laser displacement sensor simultaneously collects the actual offset of the compensated workstation. The laser displacement sensor utilizes the principle of coherent interferometry to measure displacement in real time, ranging from millimeters to nanometers. Its output signal is de-noised using a Kalman filter, resulting in displacement data with a high signal-to-noise ratio. If the measured offset after a compensation is 0.015 mm, this value is fed back to the control system to assess the effectiveness of the compensation. Finally, the actual offset of the collected workstation is compared with the preset convergence judgment criteria. When the offset values ​​measured multiple times fall within the threshold range (for example, less than 0.02 mm), the system is determined to have entered a convergence state, and the historical accumulated error data and the number of rotations are reset to zero to prevent the superposition interference of past errors on subsequent compensation. The convergence judgment criteria can be designed as a multi-level threshold or adaptive window. For example, a reset is performed only when the accuracy requirements are met for at least three of the last five compensations. This mechanism ensures that in long-term operation, each time the system starts compensation, it is based on a clear and accurate initial state, maintaining high accuracy and stability.

[0072] In an optional implementation of this embodiment, the single rotation angle error is determined based on the standard rotation angle and the single rotation angle; whether the rotation angle error corresponding to the number of rotations exceeds the rotation error threshold is judged based on the single rotation angle error and the number of rotations; if the rotation angle error exceeds the rotation error threshold, the workstation rotation corresponding to the number of rotations is compensated for the rotation angle based on the rotation angle error, and the number of rotations is reset.

[0073] Specifically, within the silo control program, the angular error of each rotation is determined by comparing the actual single rotation angle issued each time with the pre-calculated standard rotation angle. The so-called standard rotation angle is a theoretical value obtained by dividing 360 degrees by the total number of workstations. If the number of workstations is 7, the theoretical value is 360 / 7≈51.42857 degrees. However, the servo command can only issue 51.42 degrees, so the decimal portion of approximately 0.00857 degrees is lost each time. This value is the single rotation angle error. This error is then combined with the historical cumulative error, and its increment is calculated based on the number of rotations. When the controller detects that the current cumulative error reaches or exceeds the pre-set rotation error threshold, it means that the error has exceeded the acceptable range. After determining whether the error exceeds the limit, if the next station is the first, the target angle automatically returns to the initial angle without additional compensation. Otherwise, the cumulative error E = n × e is calculated based on the number of rotations n and the single error e. E is then compared with the rotation error threshold. If E ≥ the threshold, the target angle is adjusted from n × the standard angle to n × the standard angle + E. This compensated target angle now incorporates all remaining errors from the previous n rotations, allowing full correction of the error during the next switch. After issuing the compensation rotation command, the system immediately resets the number of rotations to zero to prevent double-counting of compensated errors. For example, when the number of stations is seven, the difference of 0.00857 degrees between the theoretical angle of 51.42857 degrees and the actual angle of 51.42 degrees would accumulate to 0.04285 degrees after five consecutive switches. If the error threshold is set to 0.04 degrees, before the fifth rotation, the system automatically corrects the target angle from 5 × 51.42 degrees (257.10 degrees) to 257.10 degrees + 0.04285 degrees ≈ 257.14285 degrees, and then sends it to the servo to ensure accurate positioning of the rotating silo. Through this dynamic compensation mechanism based on single error, number of rotations, and threshold, the accumulated error is promptly corrected each time the conditions are met, ensuring high-precision alignment of the silo after multiple consecutive rotations.

[0074] In an optional implementation of this embodiment, image information of the raw material on the ejection station is obtained based on image information, and a second edge contour feature point set of the raw material is extracted through image processing; feature matching is performed between the edge contour feature point set and a preset standard posture template to generate actual posture parameters of the raw material; based on the difference between the actual posture parameters and the standard posture template, the posture deviation value of the raw material is calculated; if the posture deviation value exceeds a preset posture tolerance threshold, a posture adjustment instruction is generated and the fine-tuning mechanism of the machine tool's rotating hopper is driven to correct the raw material posture.

[0075] Specifically, in a machine tool's rotary silo system, to accurately perceive and dynamically correct the posture of the material at the ejection station, an industrial camera must first capture images of the current station. The image capture must cover the entire material area to ensure a complete data foundation for subsequent processing. Once the material image is acquired, the image processing process detects the material's edges. A common method is the Canny edge detection algorithm, which uses image gradients to identify areas of grayscale abrupt changes and extracts the material's contour with pixel-level accuracy. To enhance boundary accuracy and avoid false edges caused by illumination or reflection, morphological filtering can be combined to further refine the boundary structure, and sub-pixel interpolation can be used to enhance the geometric accuracy of the contour. After processing, the system extracts several geometrically significant feature points from the edge contour, such as corner points, curvature change points, and symmetry centers. These points constitute the material's second edge contour feature point set, which serves as the key input for subsequent pose analysis. After extracting the feature point set, the control system performs feature matching against a pre-set standard pose template. The standard pose template is an idealized edge morphology model derived through visual modeling of the material in its standard placement. It contains the theoretical outline distribution of the material and its position and orientation in a standard coordinate system. During the matching process, the system normalizes, rotates, and translates the feature points in the material image to establish a spatial mapping relationship with the template. This feature matching process utilizes image registration techniques, typically including SIFT or ORB feature descriptor comparison, RANSAC robust estimation to eliminate outlier matches, and transformation matrix optimization. These methods calculate the offset and rotation angle of the material relative to the template, thereby generating the actual pose parameters of the material. The pose parameters, composed of a translation vector and a rotation angle, describe the pose change of the material in three-dimensional space relative to the standard state. Once the actual pose parameters are obtained, the system compares them with the standard pose template to calculate the pose deviation of the material. Deviation values ​​encompass two aspects: positional deviation, which refers to the spatial displacement of the workpiece's center point relative to the template's reference point, typically measured in millimeters; and posture deviation, which refers to the rotational change in the workpiece about its axis, typically expressed in degrees. During the analysis process, the control system compares the three-axis translational deviations and three-axis rotational angles against pre-set posture tolerances. These are combined into an overall deviation metric, which is used to comprehensively assess the stability and acceptability of the workpiece's posture. The goal of this comparison is to determine whether the workpiece meets the posture accuracy requirements for gripping, processing, or inspection. If the system detects that any posture deviation exceeds the pre-set tolerance threshold, it deems the workpiece's posture abnormal and prevents it from proceeding directly to the next step. The system then automatically generates a posture adjustment command and transmits it to the controller. This command contains the desired correction direction, displacement, and rotation angle, serving as the target input for the actuator.Upon receiving the command, the rotary silo's fine-tuning mechanism begins to operate. This fine-tuning mechanism, which can consist of a two-dimensional slide, an electric rotary table, or a piezoelectric actuator, is capable of performing micron-level position corrections and angular posture adjustments on the raw material. For example, if the raw material deflects due to inertia, the fine-tuning mechanism will perform a small reverse rotation in conjunction with translational motion to reposition the material at the reference point and reference angle defined by the standard posture template. Through this process, the rotary silo system achieves high-precision detection and online correction of the raw material's posture, thereby ensuring the stability of automatic loading and unloading or subsequent processing processes and product consistency.

[0076] According to a position compensation control method for a machine tool rotary hopper provided by the present application, the standard rotation angle required for station rotation is determined according to the number of stations of the machine tool rotary hopper, and the first position information of each station is determined; image information of the machine tool rotary hopper after the rotation is completed is obtained, and the second position information of each station is determined through image processing; the actual rotation angle of the current ejecting station is determined according to the first position information and the second position information; whether the rotation angle error of the next ejecting station exceeds the rotation error threshold is judged according to the actual rotation angle, the standard rotation angle and the number of rotations; if the rotation angle error exceeds the rotation error threshold, the rotation angle compensation is performed on the next rotation of the machine tool rotary hopper according to the rotation angle error, and the number of rotations is reset. Through precise error detection and compensation, the accuracy of the rotation angle is guaranteed, thereby improving the accuracy and efficiency of the entire production process.

[0077] Figure 2 The present invention provides a position compensation control device for a machine tool rotary silo, which can be used to implement the position compensation control method for a machine tool rotary silo in the aforementioned embodiment. Figure 2 As shown, the position compensation control device of the machine tool rotary hopper mainly includes:

[0078] The calibration module 10 is used to determine the standard rotation angle required for the station rotation according to the number of stations of the machine tool rotary bin, and determine the first position information of each station;

[0079] An acquisition module 20 is used to acquire image information of the machine tool rotary bin after the rotation is completed, and determine the second position information of each station through image processing;

[0080] A determination module 30 is used to determine an actual rotation angle of the current ejection station according to the first position information and the second position information;

[0081] A judgment module 40 is used to judge whether the rotation angle error of the next ejection station exceeds the rotation error threshold according to the actual rotation angle, the standard rotation angle and the number of rotations;

[0082] The compensation module 50 is configured to perform rotation angle compensation on the next rotation of the machine tool rotary bin according to the rotation angle error and reset the number of rotations if the rotation angle error exceeds a rotation error threshold.

[0083] In an optional implementation of this embodiment, the acquisition module is specifically used to: collect a multi-perspective image sequence of the rotating silo workstation through a binocular vision sensor, perform spatiotemporal alignment on the multi-perspective image sequence, and generate a three-dimensional point cloud mapping image; based on a preset workstation edge feature template, perform multi-scale convolution kernel filtering on the three-dimensional point cloud mapping image to extract the first edge contour feature point set of each workstation; generate a workstation space coordinate matrix based on the matching degree of the first edge contour feature point set and the workstation edge feature template; input the workstation space coordinate matrix into the coordinate system conversion model, and output the second position information of each workstation.

[0084] In an optional implementation of this embodiment, the determination module is specifically used to: dynamically match the center point of each workstation according to the first position information and the workstation space coordinate matrix to generate a workstation center offset vector group; construct a coordinate system transformation model based on the workstation center offset vector group, perform posture solution on the workstation space coordinate matrix, and determine the real-time posture parameters of each workstation in the rotation axis coordinate system; calculate the instantaneous rotation angle deviation value of the current material ejection workstation according to the Euler angle component in the real-time posture parameter and the standard rotation angle; fuse the instantaneous rotation angle deviation value with the historical accumulated error data to determine the actual rotation angle of the current material ejection workstation.

[0085] In an optional implementation of this embodiment, the judgment module is specifically used to: construct a dynamic error sequence based on historical accumulated error data; determine the predicted rotation angle deviation value of the next material ejection station by combining the dynamic error sequence with the current instantaneous rotation angle deviation value; perform dynamic weighted correction on the predicted rotation angle deviation value based on the number of rotations and a preset error amplification coefficient curve to generate a corrected predicted error value; perform an asymmetric comparison based on the confidence interval boundary of the corrected predicted error value and the rotation error threshold to determine whether the rotation angle error of the next material ejection station exceeds the rotation error threshold.

[0086] In an optional implementation of this embodiment, the compensation module is specifically used to: generate a dynamic compensation coefficient based on the rotation angle error; input the dynamic compensation coefficient and the standard rotation angle into the servo control model to generate a rotation control instruction; control the machine tool rotating hopper to perform compensatory rotation according to the rotation control instruction, and collect the actual offset of the work station after compensation through the laser displacement sensor; reset the historical accumulated error data and the number of rotations based on the actual offset of the work station and the preset convergence judgment condition.

[0087] In an optional implementation of this embodiment, the compensation module is also used to: determine the single rotation angle error based on the standard rotation angle and the single rotation angle; judge whether the rotation angle error corresponding to the number of rotations exceeds the rotation error threshold based on the single rotation angle error and the number of rotations; if the rotation angle error exceeds the rotation error threshold, perform rotation angle compensation on the workstation rotation corresponding to the number of rotations based on the rotation angle error, and reset the number of rotations.

[0088] In an optional implementation of this embodiment, the position compensation control device further includes a correction module. The correction module is configured to: obtain image information of the raw material on the ejection station based on image information, extract a second edge contour feature point set of the raw material through image processing; perform feature matching between the edge contour feature point set and a preset standard pose template to generate actual pose parameters of the raw material; calculate a pose deviation value of the raw material based on the difference between the actual pose parameters and the standard pose template; and generate a pose adjustment instruction if the pose deviation value exceeds a preset pose tolerance threshold and drive the fine-tuning mechanism of the machine tool's rotary hopper to correct the raw material pose.

[0089] According to the position compensation control device for a machine tool rotary hopper provided by the present application, the standard rotation angle required for station rotation is determined according to the number of stations of the machine tool rotary hopper, and the first position information of each station is determined; image information of the machine tool rotary hopper after the rotation is completed is obtained, and the second position information of each station is determined through image processing; the actual rotation angle of the current ejecting station is determined according to the first position information and the second position information; whether the rotation angle error of the next ejecting station exceeds the rotation error threshold is judged according to the actual rotation angle, the standard rotation angle and the number of rotations; if the rotation angle error exceeds the rotation error threshold, the rotation angle compensation is performed on the next rotation of the machine tool rotary hopper according to the rotation angle error, and the number of rotations is reset. Through precise error detection and compensation, the accuracy of the rotation angle is guaranteed, thereby improving the accuracy and efficiency of the entire production process.

[0090] According to the application plan Figure 3 An electronic device provided in an embodiment of the present application can be used to implement the position compensation control method for the machine tool rotary hopper in the aforementioned embodiment, mainly comprising:

[0091] Memory 301, processor 302, and computer program 303 stored in memory 301 and executable on processor 302. Memory 301 and processor 302 are connected via communication. When processor 302 executes computer program 303, the position compensation control method for a machine tool rotary hopper described in the aforementioned embodiment is implemented. The number of processors may be one or more.

[0092] The memory 301 can be a high-speed random access memory (RAM) memory or a non-volatile memory such as a disk drive. The memory 301 is used to store executable program code. The processor 302 is coupled to the memory 301 .

[0093] Furthermore, the embodiment of the present application also provides a computer-readable storage medium, which can be provided in the electronic device in the above embodiments. The computer-readable storage medium can be the above Figure 3 Memory in the illustrated embodiment.

[0094] The computer-readable storage medium stores a computer program that, when executed by a processor, implements the position compensation control method for a machine tool rotary hopper in the aforementioned embodiment. Furthermore, the computer-readable storage medium may be a USB flash drive, a mobile hard drive, a read-only memory (ROM), RAM, a magnetic disk, or an optical disk, among other media capable of storing program code.

[0095] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0096] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program code.

[0097] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for position compensation control of a machine tool rotary bin, characterized in that: include: Determine the standard rotation angle required for the station rotation according to the number of stations of the machine tool rotary bin, and determine the first position information of each station; Obtaining image information of the machine tool rotary silo after the rotation is completed, and determining second position information of each workstation through image processing, including: collecting a multi-view image sequence of the rotary silo workstation through a binocular vision sensor, performing spatiotemporal registration on the multi-view image sequence, and generating a three-dimensional point cloud mapping image; performing multi-scale convolution kernel filtering on the three-dimensional point cloud mapping image based on a preset workstation edge feature template to extract a first edge contour feature point set of each workstation; and generating a workstation spatial coordinate matrix based on a matching degree between the first edge contour feature point set and the workstation edge feature template; Determining the actual rotation angle of the current ejecting station according to the first position information and the second position information, including: dynamically matching the center point of each station according to the first position information and the station space coordinate matrix to generate a station center offset vector group; constructing a coordinate system transformation model based on the station center offset vector group, performing posture calculation on the station space coordinate matrix, and determining the real-time posture parameters of each station in the rotation axis coordinate system; calculating the instantaneous rotation angle deviation value of the current ejecting station according to the Euler angle component in the real-time posture parameter and the standard rotation angle; fusing the instantaneous rotation angle deviation value with the historical accumulated error data to determine the actual rotation angle of the current ejecting station; Judging whether the rotation angle error of the next ejecting station exceeds the rotation error threshold value according to the actual rotation angle, the standard rotation angle and the number of rotations, including: constructing a dynamic error sequence according to the historical accumulated error data; determining the predicted rotation angle deviation value of the next ejecting station by combining the dynamic error sequence with the current instantaneous rotation angle deviation value; performing dynamic weighted correction on the predicted rotation angle deviation value according to the number of rotations and a preset error amplification coefficient curve to generate a corrected predicted error value; and judging whether the rotation angle error of the next ejecting station exceeds the rotation error threshold value based on an asymmetric comparison between the corrected predicted error value and the confidence interval boundary of the rotation error threshold value; If the rotation angle error exceeds the rotation error threshold, the next rotation of the machine tool rotary hopper is compensated for the rotation angle according to the rotation angle error, and the number of rotations is reset, including: generating a dynamic compensation coefficient according to the rotation angle error; inputting the dynamic compensation coefficient and the standard rotation angle into a servo control model to generate a rotation control instruction; controlling the machine tool rotary hopper to perform compensatory rotation according to the rotation control instruction, and collecting the actual offset of the workstation after compensation through a laser displacement sensor; resetting the historical accumulated error data and the number of rotations based on the actual offset of the workstation and a preset convergence judgment condition.

2. The position compensation control method for a machine tool rotary hopper according to claim 1, characterized in that: The workstation space coordinate matrix is ​​input into a coordinate system conversion model, and the second position information of each workstation is output.

3. The position compensation control method for a machine tool rotary hopper according to claim 1, characterized in that: The method further comprises: Determining a single rotation angle error according to the standard rotation angle and the single rotation angle; Determining whether the rotation angle error corresponding to the number of rotations exceeds the rotation error threshold according to the single rotation angle error and the number of rotations; If the rotation angle error exceeds the rotation error threshold, rotation angle compensation is performed on the workstation rotation corresponding to the number of rotations according to the rotation angle error, and the number of rotations is reset.

4. The position compensation control method for a machine tool rotary bin according to claim 1, characterized in that: The method further comprises: Acquire image information of the raw material on the ejecting station according to the image information, and extract a second edge contour feature point set of the raw material through image processing; Perform feature matching on the edge contour feature point set and a preset standard pose template to generate actual pose parameters of the raw material; Calculating a posture deviation value of the raw material based on a difference between the actual posture parameter and the standard posture template; If the posture deviation value exceeds a preset posture tolerance threshold, a posture adjustment instruction is generated and the fine-tuning mechanism of the machine tool rotary hopper is driven to correct the posture of the raw material.

5. A position compensation control device for a machine tool rotary bin, characterized in that: The position compensation control device of the machine tool rotary bin includes: A calibration module, used to determine the standard rotation angle required for the station rotation according to the number of stations of the machine tool rotary bin, and determine the first position information of each station; An acquisition module is configured to acquire image information of the machine tool rotary silo after the machine tool completes rotation, and determine second position information of each workstation through image processing, including: collecting a multi-view image sequence of the rotary silo workstation through a binocular vision sensor, performing spatiotemporal registration on the multi-view image sequence, and generating a three-dimensional point cloud mapping image; performing multi-scale convolution kernel filtering on the three-dimensional point cloud mapping image based on a preset workstation edge feature template to extract a first edge contour feature point set of each workstation; and generating a workstation spatial coordinate matrix based on a matching degree between the first edge contour feature point set and the workstation edge feature template; A determination module, for determining the actual rotation angle of the current ejecting station according to the first position information and the second position information, comprising: dynamically matching the center point of each station according to the first position information and the station space coordinate matrix to generate a station center offset vector group; constructing a coordinate system transformation model based on the station center offset vector group, performing posture calculation on the station space coordinate matrix, and determining the real-time posture parameters of each station in the rotation axis coordinate system; calculating the instantaneous rotation angle deviation value of the current ejecting station according to the Euler angle component in the real-time posture parameter and the standard rotation angle; fusing the instantaneous rotation angle deviation value with the historical accumulated error data to determine the actual rotation angle of the current ejecting station; a judgment module, for judging whether a rotation angle error of a next ejecting station exceeds a rotation error threshold value based on the actual rotation angle, the standard rotation angle, and the number of rotations, comprising: constructing a dynamic error sequence based on the historical accumulated error data; determining a predicted rotation angle deviation value of the next ejecting station by combining the dynamic error sequence with a current instantaneous rotation angle deviation value; performing a dynamic weighted correction on the predicted rotation angle deviation value based on the number of rotations and a preset error amplification coefficient curve to generate a corrected predicted error value; and judging whether the rotation angle error of the next ejecting station exceeds the rotation error threshold value based on an asymmetric comparison between the corrected predicted error value and a confidence interval boundary of the rotation error threshold value; A compensation module is used to perform rotation angle compensation on the next rotation of the machine tool rotary hopper according to the rotation angle error and reset the number of rotations if the rotation angle error exceeds the rotation error threshold, including: generating a dynamic compensation coefficient according to the rotation angle error; inputting the dynamic compensation coefficient and the standard rotation angle into a servo control model to generate a rotation control instruction; controlling the machine tool rotary hopper to perform compensatory rotation according to the rotation control instruction, and collecting the actual offset of the workstation after compensation through a laser displacement sensor; and resetting the historical accumulated error data and the number of rotations based on the actual offset of the workstation and a preset convergence judgment condition.

6. An electronic device, characterized in that: Comprising a memory and a processor, wherein: The processor is configured to execute a computer program stored in the memory; When the processor executes the computer program, the steps of the position compensation control method of the machine tool rotary hopper according to any one of claims 1 to 4 are implemented.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the position compensation control method of the machine tool rotary hopper according to any one of claims 1 to 4 are implemented.

Images