Rotary table accurate alignment control method based on multi-modal sensing fusion

Abstract

The invention relates to the technical field of mechanical engineering, and discloses a rotary table accurate alignment control method based on multi-modal sensing fusion, which comprises the following steps: firstly, acquiring multi-modal sensing data of a target rotary table, fusing the multi-modal sensing data into a multi-modal vector, and identifying a real-time pose; calculating a deviation value between a current position and a target position, determining a control stage, and constructing a hierarchical control process; then extracting control positioning data in the process, dividing to obtain an adjustment sequence, and generating a turntable servo motor driving instruction; then, a high-frequency disturbance value during instruction execution is collected, an anti-disturbance control quantity is calculated, and a precision unbalance point is detected; and finally, identifying the precision deviation value of the unbalance point, carrying out cooperative alignment with a preset precision index to obtain a parameter group, and generating an alignment control strategy. According to the invention, the control precision and the adaptive capability of the rotary table in a complex application environment can be improved.

Description

TECHNICAL FIELD

[0001] The application relates to a multi-modal sensor fusion-based precision alignment control method for a rotary table, and belongs to the technical field of mechanical engineering. BACKGROUND

[0002] As a high-precision rotary positioning device, a rotary table is widely used in industrial manufacturing, aerospace and precision measurement fields, and its core function is to realize precise angle positioning of a bearing object.

[0003] At present, the alignment control of a rotary table mainly depends on a single encoder or a photoelectric sensor to obtain position information, and a preset motion curve or a PID control mode is used for positioning adjustment. However, such a method is difficult to cope with complex working conditions under multi-source disturbances, and is prone to overshoot, oscillation or precision reduction in the alignment process due to single sensor data, insufficient dynamic error compensation and weak anti-interference ability. In particular, in the scenarios of load change, external vibration or thermal deformation, stable and high-precision positioning cannot be achieved. Therefore, a multi-modal sensor fusion-based precision alignment control method for a rotary table is needed to improve the control precision and adaptive ability of the rotary table in complex application environments. SUMMARY

[0004] The application provides a multi-modal sensor fusion-based precision alignment control method for a rotary table, which aims to improve the control precision and adaptive ability of the rotary table in complex application environments.

[0005] To achieve the above-mentioned purpose, the application provides a multi-modal sensor fusion-based precision alignment control method for a rotary table, which comprises: acquiring rotary table sensing data of a target rotary table in a multi-modal scene, performing data fusion on the rotary table sensing data to obtain a multi-modal fusion vector, and identifying the real-time pose state corresponding to the target rotary table based on the multi-modal fusion vector; based on the actual pose state, calculating the position deviation between the current position and the target position of the target rotary table, determining the control stage corresponding to the target rotary table according to the position deviation, and constructing the hierarchical control process corresponding to the target rotary table based on the control stage; extracting control positioning data in the hierarchical control process, adjusting and dividing the control positioning data to obtain an adjustment control sequence, and generating a drive instruction corresponding to a rotary table servo motor in the target rotary table according to the adjustment control sequence; collecting a high-frequency disturbance value in the execution process of the drive instruction, calculating an anti-disturbance control amount corresponding to the high-frequency disturbance value, and detecting a precision imbalance point in the operation of the target rotary table based on the anti-disturbance control amount; The precision offset value corresponding to the precision imbalance point is identified, the precision offset value is cooperatively aligned with a preset precision index, an alignment parameter group is obtained, and an alignment control strategy corresponding to the target rotary table is generated based on the alignment parameter group.

[0006] Optionally, the real-time pose state corresponding to the target rotary table is identified based on the multi-mode fusion vector, including: extracting multi-mode pose features in the multi-mode fusion vector; analyzing a sensing pose component associated with the multi-mode pose features; synchronizing the sensing pose component in real time to obtain synchronization component data; Based on the synchronization component data, the current tilt angle and center coordinates of the target rotary table are analyzed; Based on the tilt angle and the center coordinates, the real-time pose state corresponding to the target rotary table is identified.

[0007] Optionally, the real-time synchronization of the sensing pose component is performed to obtain synchronization component data, including: Analyzing the component time sequence corresponding to the sensing pose component; Determine the synchronization time axis corresponding to the component time sequence; Query the data timestamp corresponding to each component in the synchronization time axis; Synchronize and align the data timestamp to obtain a synchronization data set; Extract consistent data in the synchronization data set as synchronization component data.

[0008] Optionally, the hierarchical control flow corresponding to the target rotary table is constructed based on the control stage, including: Query the stage division logic in the control stage; Based on the stage division logic, determine the interlayer association module between the control stages; Generate a topological hierarchical framework corresponding to the interlayer association module; Configure the hierarchical control data corresponding to the hierarchical control framework; Based on the hierarchical control data, a hierarchical control flow corresponding to the target rotary table is constructed.

[0009] Optionally, the topological hierarchical framework corresponding to the hierarchical association module is generated, including: Analyzing the hierarchical mapping relationship in the hierarchical association module; Based on the hierarchical mapping relationship, identify the associated control unit corresponding to the hierarchical association module; Extract the key control nodes in the associated control unit; Reconstruct a topology control chain corresponding to the key control node; Generate a topology hierarchical framework corresponding to the topology control chain.

[0010] Optionally, the adjusting and dividing of the control positioning data to obtain an adjusted control sequence comprises: Parsing a control positioning label in the control positioning data; According to the control positioning label, querying a positioning dominant mode in a preset control mode library; Extracting a dominant performance index corresponding to the positioning dominant mode; Based on the dominant performance index, determining a timing scope corresponding to the control positioning data; Adjusting and dividing the timing scope to obtain an adjusted control sequence.

[0011] Optionally, the determining of the timing scope corresponding to the control positioning data based on the dominant performance index comprises: Extracting a dynamic parameter sequence in the dominant performance index; Coordinately mapping the dynamic parameter sequence and the control positioning data to obtain a mapping data set; Analyzing an effective time span corresponding to the mapping data set; Identifying a key action point in the effective time span; Based on the key action point, determining the timing scope corresponding to the control positioning data.

[0012] Optionally, the calculating of the position deviation amount between the current position and the target position of the target turntable based on the actual pose state comprises: Analyzing a pose state type corresponding to the actual pose state; Based on the pose state type, analyzing a current position and a target position corresponding to the target turntable; Dividing a position axis component corresponding to the current position; Identifying an axis deviation value of the position axis component relative to the target position; Based on the axis deviation value, calculating the position deviation amount between the current position and the target position of the target turntable through the following formula: ; Wherein, The position deviation amount between the current position and the target position of the target turntable is represented by ΔP, The total number of control axes in the target turntable is represented by N, The control axis index corresponding to the target turntable is represented by i, The deviation value on the i th control axis is represented by Δxi. represents the weight coefficient of the control axis.

[0013] Optionally, the collecting the high-frequency disturbance value of the driving instruction in the execution process comprises: identifying an execution time interval corresponding to the driving instruction; extracting driving execution data in the execution time interval; statistically analyzing a high-frequency oscillation index corresponding to a high-frequency component in the driving execution data; determining a disturbance amplitude level corresponding to the high-frequency oscillation index; collecting the high-frequency disturbance value of the driving instruction in the execution process based on the disturbance amplitude level; and calculating the anti-disturbance control amount corresponding to the high-frequency disturbance value through the following formula: ; wherein, represents the anti-disturbance control amount corresponding to the high-frequency disturbance value, represents a proportional coefficient, represents a disturbance amplitude level, represents an execution time interval, represents the high-frequency disturbance value.

[0014] Optionally, the precision offset value and the preset precision index are cooperatively aligned to obtain an alignment parameter group, which comprises: comparing the precision difference between the precision offset value and the preset precision index; based on the precision difference, traversing a set of alignment relationships in a preset alignment rule library; screening a cooperatively matched pair in the set of alignment relationships that meets a precision threshold value; analyzing a cooperative compensation amount matched by the cooperatively matched pair; based on the cooperative compensation amount, cooperatively aligning the precision offset value and the preset precision index to obtain an alignment parameter group.

[0015] Compared with the problems described in the background art, the present application can break through the information limitation of a single sensor by acquiring the turntable sensor data of the target turntable in a multi-modal scene, cover multi-dimensional state information such as position, vibration and temperature in the operation of the turntable, realize comprehensive perception of the working condition of the turntable, avoid state misjudgment caused by the lack of single data dimension, and guarantee the precision and stability of the turntable alignment control from the source. Based on the actual pose state, the present application calculates the position deviation between the current position and the target position of the target turntable, can accurately quantify the specific difference between the current position and the target position, provides clear numerical basis for subsequent control decision, avoids the misalignment of control direction or improper adjustment due to the inability to grasp the deviation degree, reduces the problems such as overshoot and oscillation caused by inaccurate deviation estimation, guarantees the precision of the final turntable alignment control from the key link, further, the present application extracts the control positioning data in the hierarchical control process, can provide accurate and specific information basis for subsequent adjustment and division of the control positioning data, generation of turntable servo motor driving instructions, avoid the lack of clear direction in subsequent control link due to data loss or ambiguity, help to improve the long-term stability and precision of the turntable alignment control, further, the present application collects the high-frequency disturbance value of the driving instruction in the execution process, and calculates the anti-disturbance control amount corresponding to the high-frequency disturbance value, can timely capture the influence of external interference or motor itself fluctuation on the action of the turntable, avoid the expansion of the pose deviation caused by disturbance accumulation, provide data basis for accurate anti-disturbance, further consolidate the reliability of the alignment control, finally, the present application can accurately quantify the degree of deviation of the turntable pose from the target by identifying the precision offset value corresponding to the precision imbalance point, provide clear numerical basis for subsequent correction action; the anti-disturbance or adjustment strategy can be more targeted, avoid resource waste and efficiency loss caused by indiscriminate adjustment; also can trace the root cause of imbalance, provide key reference for optimizing the control logic of the turntable and improving the long-term operation precision, so as to guarantee the stability and precision of the turntable alignment. Therefore, the turntable precise alignment control method based on multi-modal sensor fusion provided by the embodiment of the present application can improve the control precision and self-adaptive ability of the turntable in complex application environment. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 A flowchart of a turntable precise alignment control method based on multi-modal sensor fusion provided by an embodiment of the present application is shown in the figure; Figure 2 A hierarchical diagram of turntable pose control in the turntable precise alignment control method based on multi-modal sensor fusion provided by an embodiment of the present application is shown in the figure; Figure 3 A module diagram of the turntable precise alignment control system based on multi-modal sensor fusion provided by an embodiment of the present application is shown in the figure.

[0017] The objectives, functional characteristics and advantages of the present application will be further illustrated with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0018] It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.

[0019] An embodiment of the present application provides a turntable accurate alignment control method based on multi-modal sensor fusion. An execution subject of the turntable accurate alignment control method based on multi-modal sensor fusion includes but is not limited to at least one of electronic devices such as a server, a terminal and the like which can be configured to execute the method provided by the embodiment of the present application. In other words, the turntable accurate alignment control method based on multi-modal sensor fusion can be executed by software or hardware installed in a terminal device or a server device. The server includes but is not limited to a single server, a server cluster, a cloud server or a cloud server cluster and the like.

[0020] Referring to Figure 1 Fig. 1 is a flowchart of a turntable accurate alignment control method based on multi-modal sensor fusion provided by an embodiment of the present application. In the embodiment, the turntable accurate alignment control method based on multi-modal sensor fusion includes the following steps.

[0021] S1, acquiring turntable sensing data of a target turntable in a multi-modal scene, performing data fusion on the turntable sensing data to obtain a multi-modal fusion vector, and identifying a real-time pose state corresponding to the target turntable based on the multi-modal fusion vector.

[0022] The present application can break through the information limitation of a single sensor, cover multi-dimensional state information such as position, vibration and temperature in the operation of a turntable, realize comprehensive perception of the working condition of the turntable, avoid state misjudgment caused by the lack of a single data dimension, and guarantee the accuracy and stability of turntable alignment control from the source.

[0023] The target rotary table refers to a core rotating device to be positioned and controlled in a precise angle, usually having the function of bearing a specific load and responding to control instructions to complete rotating actions, and is widely used in fields requiring high-precision positioning, and its performance directly determines the work precision of the bearing object, for example, a satellite attitude simulation rotary table in the field of aerospace, with a design load capacity of 80 kg, a rated rotating range of 0°-360°, and an initial positioning accuracy requirement of ±0.005°, which needs to be ensured to be stably maintained at a target angle in satellite attitude calibration operations through multi-modal sensing and control means to meet the high-precision requirements of satellite on-orbit attitude simulation; the multi-modal scene refers to the comprehensive environment faced by the target rotary table during operation, which contains various complex working conditions and interference factors that can individually or cumulatively affect the positioning accuracy of the rotary table, covering temperature fluctuations, external vibrations, electromagnetic interference, load changes and other influencing conditions in different dimensions, for example, a part detection rotary table in an industrial precision machining scene, which is subjected to vibration interference of 10Hz-25Hz and amplitude of 0.03mm generated by surrounding numerical control machines, and electromagnetic radiation from 380V industrial power in the workshop, and needs to capture these composite influences through multi-modal sensing; the rotary table sensing data refers to multi-dimensional data collected by various types of sensors to reflect the real-time running state of the target rotary table, covering key information such as the position, attitude, vibration and temperature of the rotary table, and providing original basis for subsequent data fusion and state recognition, for example, a sensing system configured for a high-precision measurement rotary table can collect real-time angle data output by a high-precision encoder (resolution 262144 lines / revolution), vibration acceleration data output by a MEMS accelerometer (sampling rate 2000Hz, measurement range ±10g), and rotary table shell temperature data output by an infrared temperature sensor (measurement accuracy ±0.2℃), which together constitute a rotary table sensing data set. Optionally, the rotary table sensing data of the target rotary table in the multi-modal scene can be obtained through a multi-modal sensor integration method, such as using a ROS framework to integrate laser radar, IMU and encoder data streams to obtain rotary table sensing data.

[0024] Further, the present application can effectively integrate sensing information in different dimensions, break through the limitations of single sensors in monitoring range, precision or anti-interference ability, avoid one-sidedness of rotary table state recognition caused by missing or biased data of a certain type, realize more comprehensive and stereoscopic description of the rotary table running state, and ensure the smoothness of the entire control process.

[0025] The multi-mode fusion vector refers to a structured numerical set formed after data calibration, feature extraction and collaborative fusion processing of multi-dimensional sensing original data (such as position, vibration, temperature, motor operating parameter, etc.) of the rotary table, which integrates effective information of each sensor, eliminates data redundancy and single sensor error, has unified data format and clear physical meaning, and can concentrate and accurately reflect the real-time running state of the rotary table. For example, the multi-mode fusion vector of a certain rotary table can be expressed as [45.327°, 0.011mm, 29.6℃, 4.9A], which respectively corresponds to the angle, vibration amplitude, shell temperature and motor working current parameter after fusion, and provides unified and high-quality information input for subsequent rotary table pose recognition and control strategy formulation. Optionally, the data fusion on the rotary table sensing data can be realized by a multi-sensor Kalman filtering algorithm, such as using an extended Kalman filter to fuse the pose data of an inertial measurement unit and an encoder, so as to obtain the multi-mode fusion vector.

[0026] Further, based on the multi-mode fusion vector, the real-time pose state corresponding to the target rotary table is identified, which can break through the limitation of single sensing data, integrate multi-dimensional information related to the pose, completely present the current angle, attitude offset and running related characteristics of the rotary table, avoid the pose recognition deviation caused by one-sided information, realize more comprehensive control of the rotary table state, and ensure the overall effect of the rotary table alignment control.

[0027] The real-time pose state refers to a whole state description that can comprehensively reflect the current spatial position and attitude of the rotary table by comprehensively considering the tilt angle, center coordinates and other related parameters (such as the current rotation angle and attitude stability) of the target rotary table, and is the core basis for judging whether adjustment is needed in the rotary table alignment control. For example, the real-time pose state of a certain rotary table can be described as: the current rotation angle is 45.032°, the X-axis tilt angle is 0.08°, the Y-axis tilt angle is 0.05°, the center coordinates are (120.53mm, 80.41mm), and the attitude fluctuation amplitude is ≤0.005° / s. This state clearly indicates whether the current pose of the rotary table meets the positioning accuracy requirement of ±0.01°.

[0028] As an embodiment of the present application, based on the multi-mode fusion vector, the real-time pose state corresponding to the target rotary table is identified, which includes: extracting multi-mode pose features in the multi-mode fusion vector; analyzing the sensing pose components associated with the multi-mode pose features; synchronizing the sensing pose components in real time to obtain synchronous component data; based on the synchronous component data, analyzing the current tilt angle and center coordinates of the target rotary table; based on the tilt angle and the center coordinates, identifying the real-time pose state corresponding to the target rotary table.

[0029] The multi-mode pose feature refers to a key feature set extracted from a multi-mode fusion vector, directly reflecting the key state of the target rotary table pose, covering core information related to the rotary table angle, position, and attitude, and being able to reflect the real-time position and attitude tendency of the rotary table in space, and is the basis for subsequent pose analysis. For example, the multi-mode pose feature of a certain rotary table can include the current absolute angle deviation value of the rotary table (such as ±0.02°), the inclination tendency value of the table surface relative to the reference surface (such as ±0.01° in the X-axis direction), and the offset feature value of the rotation center (such as ±0.015mm), which collectively point to the core state of the rotary table pose and provide key basis for accurate identification. The sensing pose component refers to a specific data component in the multi-mode pose feature that is collected by different types of sensors and is directly related to the rotary table pose. Each component comes from the pose monitoring result of a single sensor and can reflect the local pose information of the rotary table from the perspective of the sensor. For example, the sensing pose component of a certain rotary table can include the current angle of the rotary table collected by a high-precision encoder (such as 45.032°), the inclination data of the table surface collected by a MEMS gyroscope (such as 0.12° in the Y-axis direction), and the offset of the rotation center collected by a laser displacement sensor (such as +0.018mm). These components respectively carry the pose monitoring results of different sensors and need to be further processed to integrate into unified information. The synchronous component data refers to consistent data extracted from a synchronous data set and formed into a unified time, reliable, and contradiction-free pose data set after screening. It has time synchronization and data effectiveness and can directly provide high-quality data support for subsequent analysis of the tilt angle and center coordinates of the rotary table. For example, the synchronous component data of a certain rotary table includes consistent data of "angle 45.031°, inclination 0.119°, offset 0.017mm" at the "1695210800.002ms" node, and all data timestamps are unified with a deviation that meets the requirements, which can be directly used for pose state identification. The tilt angle refers to the inclination degree of the bearing table surface of the target rotary table relative to the preset horizontal reference surface (or specified reference surface) in different axial directions (such as X-axis and Y-axis) in space, quantified by angle value, reflecting whether the table surface is horizontal or in a preset inclined state, and is an important indicator affecting the positioning accuracy of the rotary table. For example, the preset table surface of a certain precision measurement rotary table needs to be kept horizontal (tilt angle 0°), and the actual running analysis obtains a tilt angle of 0.08° in the X-axis direction and 0.05° in the Y-axis direction, which directly reflects the attitude deviation of the table surface. If the angle value exceeds 0.1°, attitude correction needs to be started. The center coordinates refer to the specific coordinate values of the rotation center of the target rotary table in the preset three-dimensional coordinate system (or plane coordinate system), quantifying the position of the rotation center in space and reflecting the spatial positioning of the rotary table rotation axis, which is a key parameter for judging whether the rotary table is in a normal rotation reference state. For example, the preset rotation center of a certain rotary table has coordinates (120.5mm, 80.4mm) in the plane coordinate system, and the current center coordinates obtained by analysis are (120.53mm, 80.41mm), the coordinate deviation is controlled within ±0.02mm, so that the rotary table can rotate around the preset center, and the positioning deviation caused by the center deviation is avoided.

[0030] Further, the extracting the multi-mode pose feature in the multi-mode fusion vector can be realized by a principal component analysis method, such as: using a PCA algorithm to reduce dimension of multi-source sensor fusion data, so as to obtain a multi-mode pose feature; the analyzing the sensor pose component associated with the multi-mode pose feature can be realized by a blind source separation algorithm, such as: using an independent component analysis ICA to decouple the coupled sensor signal components, so as to obtain a sensor pose component; the real-time synchronization of the sensor pose component can be realized by a time stamp alignment method, such as: using a PTP precise time protocol to synchronize multi-source sensor data streams, so as to obtain synchronized component data; the analyzing the current tilt angle of the target rotary table can be realized by a Kalman filter algorithm, such as: using a complementary filter to fuse accelerometer and gyroscope data to calculate the tilt angle, so as to obtain the tilt angle; the analyzing the current center coordinates of the target rotary table can be realized by a least squares estimation method, such as: using a Levenberg-Marquardt algorithm to optimize the spatial coordinates of the visual marker points, so as to obtain the center coordinates; the identifying the real-time pose state corresponding to the target rotary table can be realized by a state machine modeling method, such as: constructing a finite state machine model combined with threshold judgment logic to process the pose data, so as to obtain the real-time pose state.

[0031] As another embodiment of the application, the real-time synchronization of the sensor pose component to obtain synchronized component data comprises: analyzing the component time sequence corresponding to the sensor pose component; determining the synchronization time axis corresponding to the component time sequence; querying the data time stamps corresponding to each component in the synchronization time axis; synchronously aligning the data time stamps to obtain a synchronized data set; and extracting consistent data in the synchronized data set as synchronized component data.

[0032] The component time sequence refers to dynamic data sequences of different sensing component changes over time, records the time sequence and corresponding values of the pose information collected by each sensor, reflects the change rule of the sensing component in the time dimension, and is the basis for subsequent synchronization processing. For example, in the component time sequence of a certain rotary table, the angle data time sequence collected by the encoder is "1695210800.001 ms: 45.030°, 1695210800.003 ms: 45.031°", and the tilt data time sequence collected by the gyroscope is "1695210800.002 ms: 0.118°, 1695210800.004 ms: 0.120°", which clearly presents the time and value association of the two types of data. The synchronization time axis refers to a standard time framework constructed for the time reference of each sensing component, which sets fixed time intervals and key time nodes, and is used to align the time dimension of different component time sequences and eliminate the time deviation of data collection by each sensor. For example, for the synchronization requirements of a certain rotary table, the synchronization time axis is constructed with a time interval of 1 ms, and the key time points are marked as "1695210800.001 ms, 1695210800.002 ms, 1695210800.003 ms, …". The data timestamp refers to the time identifier marked for each sensing component data, which uniquely corresponds to the collection time. It is usually presented in the form of "year-month-day hour: minute: second, millisecond" or timestamp value, and is used to accurately locate the collection time of the data. It is the core identifier for judging whether the data is synchronized. For example, the encoder of a certain rotary table collects "45.031°" angle data at 1695210800.002 ms, and the gyroscope collects "0.119°" tilt data at the same timestamp. The consistent timestamps indicate that the data is synchronized. The synchronized data set refers to the data set formed by matching and aligning the original data of each sensing component according to the data timestamp and synchronization time axis, which ensures that the pose data of different sensors corresponds at the same time node and eliminates the data misplacement problem caused by time delay. For example, at the "1695210800.002 ms" node of the synchronization time axis, the synchronized data set integrates the "45.031°" angle data of the encoder, the "0.119°" tilt data of the gyroscope, and the "0.017 mm" offset data of the laser displacement sensor, forming a multi-dimensional pose data set at the same time. The consistent data refers to the data collected by different sensors for the same pose dimension (such as angle or tilt) of the rotary table in the synchronized data set, and the deviation between them is within a pre-set reasonable range (such as angle deviation ≤0.001°, the data is consistent, and the data can truly reflect the actual position state of the turntable, and the interference of abnormal or error data is excluded, for example, in the synchronization data set, the angle data of the encoder and the laser angle sensor at "1695210800.002 ms" is "45.031°" and "45.0315°" respectively, and the deviation is 0.0005°≤0.001°, so the two groups of data belong to consistent data.

[0033] Further, the component time sequence corresponding to the sensing position component can be realized by a time sequence analysis method, such as extracting the frequency domain features of the sensor data by using Fourier transform, so as to obtain the component time sequence; the synchronization time axis corresponding to the component time sequence can be realized by a dynamic time warping algorithm, such as using the DTW algorithm to align the sensor time sequence data with different sampling rates, so as to obtain the synchronization time axis; the data timestamp corresponding to each component in the synchronization time axis can be realized by a timestamp extraction method, such as using the Linux system clock function gettimeofday() to obtain the time mark accurate to microseconds, so as to obtain the data timestamp; the data timestamp can be synchronized and aligned by an interpolation synchronization method, such as using a cubic spline interpolation algorithm to perform time axis resampling on the asynchronously collected data points, so as to obtain a synchronization data set; the consistent data in the synchronization data set can be extracted as synchronization component data by a data consistency verification method, such as using a sliding window variance detection algorithm to remove abnormal data points and retain stable data segments, so as to obtain the synchronization component data.

[0034] S2, based on the actual position state, calculating the position deviation between the current position and the target position of the target turntable, determining the control stage corresponding to the target turntable according to the position deviation, and constructing the hierarchical control process corresponding to the target turntable based on the control stage.

[0035] Based on the actual position state, the present application calculates the position deviation between the current position and the target position of the target turntable, which can accurately quantify the specific difference between the current position and the target position, and provide clear numerical basis for subsequent control decision, avoiding the problem of inaccurate control direction or improper adjustment due to the inability to grasp the deviation degree, reducing the problems of overshoot and oscillation caused by inaccurate deviation estimation, and ensuring the accuracy of the final positioning control of the turntable from the key link.

[0036] The current position refers to a specific parameter set reflecting the current spatial positioning in the real-time pose state of the turntable analyzed based on the multi-mode fusion vector, covering the current rotation angle, rotation center coordinates and position relationship relative to the reference plane of the turntable, which is the actual spatial position of the turntable at a certain moment, directly reflecting the real-time result of pose recognition. For example, the current position of a certain precision detection turntable at 1695210800.5 ms analyzed from the synchronous component data is: rotation angle 45.032°, rotation center coordinates (120.53 mm, 80.41 mm), and vertical distance relative to the reference table 200.05 mm. These parameters together constitute the actual spatial positioning of the turntable at this time. The target position refers to the spatial positioning parameter set expected by the turntable to finally reach according to the task requirements (such as part machining positioning, satellite attitude simulation calibration) of the turntable, including target rotation angle, target rotation center coordinates and target attitude parameters, which is the reference and final target of the turntable position control, directly determining the direction and accuracy standard of the control process. For example, the precision detection turntable needs to complete the center hole detection task of a certain part, and the preset target position is: rotation angle 45.000°, rotation center coordinates (120.50 mm, 80.40 mm), and vertical distance relative to the reference table 200.00 mm. This position is the core basis for judging whether the turntable has completed accurate positioning. The position deviation amount refers to the axis deviation value of all position axis components, which is a quantitative index reflecting the overall deviation degree of the current position of the turntable from the target position calculated by a preset algorithm (such as Euclidean distance formula, weighted summation method), which is the key basis for judging whether the turntable needs to be adjusted and selecting a control strategy. For example, the turntable uses the Euclidean distance method to calculate the position deviation amount. If the X and Y axis units are unified as mm, and the Z axis deviation is converted to linear deviation (assuming a rotation radius of 100 mm, and a linear deviation of about 0.056 mm corresponding to 0.032°), then the deviation amount is √[(0.03)²+(0.01)²+(0.056)²]≈0.065 mm, which directly reflects the overall deviation degree.

[0037] As an embodiment of the present application, the position deviation amount between the current position and the target position of the target turntable is calculated based on the actual pose state, which includes: analyzing the pose state type corresponding to the actual pose state; based on the pose state type, analyzing the current position and the target position corresponding to the target turntable; dividing the position axis component corresponding to the current position; identifying the axis deviation value of the position axis component relative to the target position; based on the axis deviation value, calculating the position deviation amount between the current position and the target position of the target turntable.

[0038] The pose state type refers to the specific category of the real-time pose state of the target turntable according to the task requirements, motion characteristics and pose performance of the target turntable. Different types correspond to different running modes and control priorities of the turntable, directly determine the way of subsequent analysis of the current position and calculation of the deviation amount. For example, the pose state type of a satellite attitude simulation turntable in the aerospace field can be divided into "static positioning type" (such as simulating the satellite geostationary orbit attitude, the turntable angle is fixed at 0°) and "dynamic tracking type" (such as simulating satellite orbit transfer, the turntable rotates at an angular velocity of 30° / s). The pose analysis logic and deviation calculation dimension of the two types are obviously different. The position axis component refers to the decomposition of the current position of the target turntable into independent position parameters in different axial directions according to a preset coordinate system (such as a three-dimensional rectangular coordinate system, a polar coordinate system). Each component corresponds to the spatial positioning information of the turntable in a single axial direction, which is the basis for subsequent calculation of single-axis deviation and integration of overall deviation. For example, a certain industrial precision machining turntable adopts a positioning system of two-dimensional plane coordinate system (X-axis, Y-axis) + rotating axis (Z-axis), and its current position (X: 120.53mm, Y: 80.41mm, Z: 45.032°) can be decomposed into three independent position axis components: X-axis component 120.53mm, Y-axis component 80.41mm, and Z-axis component 45.032°. The axis deviation value refers to the difference between each position axis component and the corresponding axial component of the target position, which quantifies the degree of deviation of the turntable from the target position in a single axial direction. It can be positive or negative (positive indicates exceeding the target value, negative indicates not reaching the target value), and is the core unit of the overall position deviation amount. For example, the target position of the above-mentioned turntable is (X: 120.50mm, Y: 80.40mm, Z: 45.000°), then the X-axis deviation value is 120.53mm-120.50mm=+0.03mm, the Y-axis deviation value is 80.41mm-80.40mm=+0.01mm, and the Z-axis deviation value is 45.032°-45.000°=+0.032°.

[0039] Furthermore, the analysis of the pose state type corresponding to the actual pose state can be achieved through pattern recognition algorithms, such as using a Support Vector Machine (SVM) classifier to classify the pose data to obtain the pose state type; the parsing of the current position corresponding to the target turntable can be achieved through visual SLAM methods, such as using an ORB-SLAM3 system to calculate the spatial coordinates of the turntable in the environment in real time to obtain the current position; the parsing of the target position corresponding to the target turntable can be achieved through path planning algorithms, such as using the A* algorithm to calculate the endpoint coordinates of the turntable's motion trajectory to obtain the target position; the division of the position axis components corresponding to the current position can be achieved through coordinate system projection methods, such as using a three-dimensional coordinate transformation matrix to decompose the global coordinates into the motion axis vectors of the turntable to obtain the position axis components; the identification of the axis deviation value between the position axis component and the target position can be achieved through an error calculation model, such as using the Euclidean distance formula to calculate the difference between the current position and the target position axis by axis to obtain the axis deviation value; the calculation of the position deviation between the current position and the target position of the target turntable can be achieved through the following formula.

[0040] In another embodiment of the present invention, the positional deviation between the current position and the target position of the target turntable is calculated based on the axis deviation value using the following formula: ; in, This indicates the positional deviation (in mm) between the current position and the target position of the target turntable. This indicates the total number of control axes in the target turntable. This indicates the control axis index corresponding to the target turntable. Indicates the first Deviation values ​​on each control axis (unit: mm). Indicates the first Weighting coefficients for each control axis.

[0041] In detail, the position deviation can represent the overall degree of deviation of the current position of the target turntable from the target position. It is a quantitative value obtained by weighting, averaging, and then taking the square root of the deviations of each control axis. It can comprehensively reflect the position accuracy under multi-axis coordination. For example, for a three-control-axis turntable (n=3, including the X-axis, Y-axis, and rotation axis), the X-axis deviation... Weight (X-axis positioning accuracy requirements are higher), Y-axis deviation Weight After the rotation axis deviation is converted into a linear deviation (Corresponding angular deviation 0.028°, deviation per degree line 2mm), weight Substituting into the formula, we get , the control axis can represent the independent motion axis of the target rotary table to achieve the pose adjustment, which is the core dimension of the rotary table space pose, each control axis corresponds to a motion or positioning direction (such as X axis, Y axis, rotation axis, etc.), n is the total number of control axes, for example, the above three control axis rotary table, the X axis controls the horizontal transverse displacement, the Y axis controls the horizontal longitudinal displacement, and the rotation axis controls the angle of the table surface, and the three axes jointly determine the rotary table space position, so n = 3; the deviation value can represent the difference between the current position and the target position of the rotary table on the i-th control axis, which quantifies the deviation degree of the single-axis direction; if it is a rotation axis, it needs to be converted into a unit consistent with the straight line axis (such as mm), for example, the above rotary table X axis current position 120.53mm, target position 120.50mm, so ; the target angle of the rotation axis is 45.000°, and the current angle is 45.028°, because the rotation radius is 100mm, the angle deviation 0.028° corresponds to the arc length (linear deviation) 2π×100× ≈0.0489mm, approximately , so as to unify it as a linear deviation unit; the weight coefficient can represent the importance of the i-th control axis in the overall pose accuracy of the rotary table, which is set according to the positioning accuracy requirement of the axis, the dependence of the work on the axis, etc. The greater the weight, the stronger the influence of the axis deviation on the overall position deviation, for example, the above rotary table X axis is responsible for the transverse positioning of the key hole of the part, and the accuracy requirement is high, so ; the positioning requirement of the Y axis is secondary, so ; the rotation axis mainly assists the angle alignment, and has a relatively weak direct influence on the final position deviation, so , the importance difference of each axis is reflected through the weight.

[0042] The present application determines the corresponding control stage of the target rotary table according to the position deviation, can accurately match the adaptability of different deviation degrees and control strategies, avoid the waste of efficiency or insufficient accuracy caused by unified control mode; can clearly divide the stages of coarse adjustment and fine adjustment, make the control target of each stage more focused, and ensure that the rotary table completes the pose adjustment quickly and accurately.

[0043] The control stage refers to a work stage corresponding to different control strategies and accuracy requirements divided for the target turntable according to the size of the position deviation. Different stages match targeted control parameters (such as driving force, adjustment step, etc.), so as to efficiently and accurately complete the pose adjustment and avoid the inefficiency of a single control strategy in the full deviation range. For example, when the position deviation of a certain turntable is greater than 0.5 mm, it is in the “coarse adjustment stage” and adopts a fast adjustment strategy with large driving force and large step size; when the deviation is between 0.05 mm and 0.5 mm, it is in the “transition adjustment stage” and the driving force is reduced and the step size is reduced; when the deviation is less than 0.05 mm, it is in the “fine adjustment stage” and achieves micron-level accurate positioning with very small driving force and step size. Optionally, the determination of the control stage corresponding to the target turntable can be realized by a finite state machine modeling method, such as defining the state transition conditions of turntable start, run, stop, etc. in the FSM framework, so as to obtain the control stage.

[0044] Further, based on the control stage, the layered control process corresponding to the target turntable is constructed, which can accurately match different deviation degrees and adaptive control strategies, avoid the inefficiency of a single strategy in the full deviation range, optimize the dynamic allocation of control resources, ensure the convergence efficiency in large deviation and the positioning accuracy in small deviation, and thus improve the accuracy and efficiency of the turntable pose adjustment as a whole.

[0045] The layered control process refers to a turntable control process formed by a topological layered framework and layered control data, which is executed in stages and cooperates in each stage to realize accurate adjustment from the current pose to the target pose. For example, the layered control process of the turntable is as follows: when the position deviation is greater than 0.5 mm, it enters the coarse adjustment stage and is adjusted quickly with a driving force of 10 N and a step size of 0.2 mm; when the deviation is reduced to 0.5 mm, it is switched to the transition adjustment stage by an interlayer association module and continues to be adjusted with a driving force of 7 N and a step size of 0.05 mm; when the deviation is less than 0.05 mm, it is switched to the fine adjustment stage and completes micron-level positioning with a driving force of 3 N and a step size of 0.005 mm, and the control is completed in an orderly manner.

[0046] As an embodiment of the present application, based on the control stage, the layered control process corresponding to the target turntable is constructed, which includes: querying the stage division logic in the control stage; determining the interlayer association module between the control stages based on the stage division logic; generating the topological layered framework corresponding to the interlayer association module; configuring the layered control data corresponding to the layered control framework; and constructing the layered control process corresponding to the target turntable based on the layered control data.

[0047] The stage division logic refers to a rule system for classifying and defining control stages according to core indicators such as position deviation, and clearly defines the trigger conditions, deviation value range, and core control targets of each control stage, which is the basis for hierarchical control. For example, a turntable divides the stages into "coarse adjustment (deviation > 0.5 mm, target fast convergence deviation), transition adjustment (0.05-0.5 mm, target smooth transition adjustment), and fine adjustment (deviation < 0.05 mm, target micron-level accurate positioning)" with 0.5 mm and 0.05 mm as the position deviation thresholds. This series of classification rules is the stage division logic. The inter-layer association module refers to a functional module responsible for stage switching, data transfer, and control strategy connection between different control stages, ensuring smooth transition and collaborative operation of each stage control process, and avoiding control faults between stages. For example, the inter-layer association module from "coarse adjustment" to "transition adjustment" needs to trigger the parameter switching logic of "30% driving force attenuation and adjustment step size reduction to 1 / 2 of the original" when the position deviation decreases to 0.5 mm, and transfer the real-time data of the current pose to ensure the continuity of the control after the transition. The topology hierarchical framework refers to the overall control framework that integrates all control stages, associated control units, and corresponding topology control chains, and presents a visual hierarchical structure. It directly displays the connection relationship and data interaction path of the position, topology control chain of each control layer (coarse adjustment, transition adjustment, and fine adjustment). For example, the topology hierarchical framework of a turntable presents "coarse adjustment layer → transition adjustment layer → fine adjustment layer" in horizontal layering, and connects each layer through two topology control chains (coarse adjustment-transition adjustment chain and transition adjustment-fine adjustment chain) in vertical layering. It also labels the key control nodes of each chain (such as the 0.5 mm threshold node and the 7N parameter node), forming a "layer-chain-node" three-in-one structured framework that provides an intuitive basis for hierarchical control process execution. The hierarchical control data refers to the parameter set configured for each control stage and inter-layer association module in the hierarchical control framework, which supports the operation of control logic and covers control parameters (such as driving force and adjustment step size) of each stage, stage switching thresholds, data interaction formats, etc. For example, for the hierarchical control of a turntable, the coarse adjustment layer is configured with "driving force 10N, adjustment step size 0.2 mm, and switching threshold 0.5 mm", the transition adjustment layer is configured with "driving force 7N, adjustment step size 0.05 mm, and switching threshold 0.05 mm", and the fine adjustment layer is configured with "driving force 3N, adjustment step size 0.005 mm". These parameters collectively constitute the hierarchical control data.

[0048] Further, the query stage division logic in the control stage can be implemented by a state transition analysis method, such as: using a Petri net modeling tool to analyze the conditional trigger relationship between stages, thereby obtaining the stage division logic; the determination of the interlayer association module between the control stages can be implemented by an interface mapping method, such as: using a SysML system modeling language to define the data exchange interface specification between levels, thereby obtaining the interlayer association module; the generation of the topological layered framework corresponding to the interlayer association module can be implemented by a hierarchical modeling algorithm, such as: using a directed acyclic graph (DAG) to construct the dependency relationship topological structure of the control layer, thereby obtaining the topological layered framework; the configuration of the layered control data corresponding to the layered control framework can be implemented by a parameterized configuration method, such as: using an XML configuration file to define the control parameters and constraint conditions of each level, thereby obtaining the layered control data; the construction of the layered control process corresponding to the target turret can be implemented by a workflow engine, such as: using an Activiti process engine to arrange the execution order of multi-level control instructions, thereby obtaining the layered control process.

[0049] As another embodiment of the application, the generation of the topological layered framework corresponding to the hierarchical association module comprises: analyzing the level mapping relationship in the hierarchical association module; based on the level mapping relationship, identifying the associated control unit corresponding to the hierarchical association module; extracting the key control node in the associated control unit; reconstructing the topological control chain corresponding to the key control node; and generating the topological layered framework corresponding to the topological control chain.

[0050] The hierarchical mapping relationship refers to the corresponding connection rules between the interlayer association module and different control stages, and clearly indicates the upper and lower control stages, data interaction direction and trigger condition that a certain association module needs to connect, which is the logical basis for constructing the topology structure. For example, in the turret control, the "coarse adjustment-transition adjustment" association module (M1) maps the "coarse adjustment stage (deviation > 0.5 mm)" as the upper stage, and the "transition adjustment stage (0.05 mm-0.5 mm)" as the lower stage, and the trigger condition is that the position deviation is reduced to 0.5 mm; the "transition adjustment-fine adjustment" association module (M2) maps the "transition adjustment stage" as the upper stage, and the "fine adjustment stage (deviation < 0.05 mm)" as the lower stage, and the trigger condition is that the deviation is reduced to 0.05 mm. These two sets of corresponding rules are the hierarchical mapping relationship. The associated control unit refers to a specific functional unit that realizes the connection of a specific upper and lower control stage according to the hierarchical mapping relationship, and integrates the core logic of deviation monitoring, parameter conversion, instruction transmission, etc., and is the execution carrier of stage switching. For example, the associated control unit corresponding to the "coarse adjustment-transition adjustment" mapping relationship includes a deviation real-time monitoring chip (used to capture whether the deviation is reduced to 0.5 mm) with a sampling rate of 1 kHz, a parameter conversion module (converts the 10N driving force of coarse adjustment into the 7N driving force of transition adjustment), and an instruction synchronization module (sends a switching instruction to the turret servo system). When the trigger condition is met, the unit can automatically complete the control connection of the two stages. The key control node refers to the core component in the associated control unit that determines the stage switching, data processing or instruction output, and is the core support point of the topology control chain, which directly affects the accuracy and timeliness of control connection. For example, in the above-mentioned "coarse adjustment-transition adjustment" associated control unit, the key control nodes include: a deviation threshold comparison node (preset 0.5 mm threshold to judge whether the deviation meets the standard), a control parameter storage node (pre-stores the 7N driving force and 0.05 mm adjustment step parameter of the transition adjustment stage), and an instruction sending node (sends a parameter switching instruction to the servo motor after the deviation meets the standard). The absence or failure of these nodes will cause the stage switching to fail. The topology control chain refers to the chain structure formed by connecting the key control nodes corresponding to the same association module in the logical order of "data input-processing-output", which clearly presents the transmission path of the control signal and data between the nodes. For example, the topology control chain of the "coarse adjustment-transition adjustment" association module is: coarse adjustment layer data output node→deviation monitoring node (sampling deviation value)→threshold comparison node (compared with 0.5 mm)→parameter calling node (calls the transition adjustment parameter)→instruction generation node (generates the switching instruction)→transition adjustment layer data receiving node. Each node transmits information in order to ensure that the stage switching process is executed in order.

[0051] Furthermore, the parsing of the hierarchical mapping relationships in the hierarchical association modules can be achieved through graph theory analysis methods, such as using adjacency matrix modeling tools to parse the bidirectional dependencies between levels, thereby obtaining the hierarchical mapping relationships; the identification of the associated control units corresponding to the hierarchical association modules can be achieved through modular decomposition methods, such as using UML component graph tools to identify control unit clusters with independent functions, thereby obtaining the associated control units; the extraction of key control nodes in the associated control units can be achieved through centrality measurement algorithms, such as using the PageRank algorithm to calculate the influence weight of nodes in the control network, thereby obtaining the key control nodes; the reconstruction of the topology control chain corresponding to the key control nodes can be achieved through path optimization algorithms, such as using the Dijkstra algorithm to reconstruct the shortest control path between key nodes, thereby obtaining the topology control chain; the generation of the topology hierarchical framework corresponding to the topology control chain can be achieved through hierarchical topology sorting methods, such as using the Kahn algorithm to generate a hierarchical topology structure based on control chain dependencies, thereby obtaining the topology hierarchical framework.

[0052] Specifically, for a more intuitive understanding of the execution logic and data flow relationships corresponding to the target turntable pose control in this solution, please refer to [reference needed]. Figure 2 The layered diagram of turntable pose control, Figure 2 As the core process framework of the turntable pose control system, it clearly presents the complete link from actual pose input to hierarchical control process output: The input layer focuses on the actual pose state of the turntable, which is the basis for subsequent deviation calculation and control strategy generation; the processing layer completes position deviation quantification, control stage matching, topology hierarchical framework and process generation in sequence through the hierarchical logic of "data preparation and deviation calculation → control stage analysis → process construction and output", transforming pose data into an executable hierarchical control strategy; the output layer takes the "hierarchical control process" as the final result. It should be noted that the connection between each link in the flowchart is essentially an abstract refinement of the turntable pose control logic. In actual scenarios, the complexity of pose analysis (such as the real-time requirements of multi-mode fusion vector analysis) and the diversity of stage adaptation (different deviation amounts correspond to differentiated control parameter rules) are far greater than what is shown in the diagram. This architecture is only a concise display of the core logic to provide an intuitive reference for understanding the systematic thinking of turntable pose control.

[0053] S3. Extract the control positioning data from the hierarchical control process, adjust and divide the control positioning data to obtain an adjustment control sequence, and generate the drive command corresponding to the turntable servo motor in the target turntable based on the adjustment control sequence.

[0054] The present application can provide accurate and specific information basis for subsequent adjustment and division of control positioning data and generation of turntable servo motor driving instructions by extracting control positioning data in the hierarchical control flow, avoid lack of clear direction of subsequent control links due to data loss or ambiguity, and help improve the long-term stability and accuracy of turntable positioning control.

[0055] The control positioning data refers to a set of key data directly related to positioning control extracted from each stage of the hierarchical control flow of the turntable, covering control parameters (such as driving force, adjustment step), real-time positioning results (such as current position coordinates), and deviation change data of each control stage, which is the core basis for subsequent division and adjustment of control sequence and generation of servo motor driving instructions. For example, in the hierarchical control of a certain turntable, the coarse adjustment stage extracts "driving force 10N, adjustment step 0.2mm, current X-axis position 120.4mm, deviation 0.4mm", the transition adjustment stage extracts "driving force 7N, step 0.05mm, X-axis position 120.52mm, deviation 0.02mm", and the fine adjustment stage extracts "driving force 3N, step 0.005mm, X-axis position 120.50mm, deviation 0.001mm". These data collectively constitute the control positioning data. Optionally, the extraction of the control positioning data in the hierarchical control flow can be realized by data mining methods, such as using association rule mining algorithm to extract position parameters of each control node in the flow to obtain the control positioning data.

[0056] Further, the present application can convert scattered positioning data into an ordered and structured control step set by adjusting and dividing the control positioning data to obtain an adjustment control sequence, providing a clear and executable basis for subsequent generation of servo motor driving instructions, avoiding disorder or deviation in driving instruction generation due to chaotic data, and improving the refinement level of turntable positioning.

[0057] The adjustment control sequence refers to an ordered control step set formed by time segmentation and action decomposition of the time sequence scope according to the dominant performance index, each step containing specific adjustment parameters (such as driving force, step), execution time point, and target deviation, which can directly guide the action of the servo motor. For example, the adjustment control sequence of the above-mentioned transition adjustment stage is: "1695210800.0s: driving force 7N, step 0.05mm, target deviation 0.04mm; 1695210802.0s: driving force 6.5N, step 0.03mm, target deviation 0.03mm; 1695210805.0s: driving force 6N, step 0.02mm, target deviation 0.02mm", the steps are ordered and meet the performance index requirements.

[0058] As an embodiment of the present application, the adjusting and dividing of the control positioning data to obtain an adjusted control sequence comprises: parsing a control positioning label in the control positioning data; querying a positioning dominant mode in a preset control mode library according to the control positioning label; extracting a dominant performance index corresponding to the positioning dominant mode; determining a time sequence scope corresponding to the control positioning data based on the dominant performance index; and adjusting and dividing the time sequence scope to obtain an adjusted control sequence.

[0059] The control positioning label refers to a classification label added to control positioning data to identify its core attributes, covering key information such as the control stage to which the data belongs, the corresponding control axis, and the positioning accuracy requirement, and is the core basis for matching the preset control mode. For example, the control positioning data label of a certain rotary table can be marked as "stage: coarse adjustment, axis type: X axis, accuracy level: fast convergence level". Through this label, the corresponding control scenario of the data can be quickly identified, avoiding directional deviation during subsequent mode matching, and ensuring that the appropriate control logic is called from the mode library. The preset control mode library refers to a database that is pre-constructed and stores multiple rotary table positioning control modes. Each mode corresponds to a specific control scenario (such as coarse adjustment or fine adjustment) and includes the control strategy, parameter range, and execution logic for that scenario, providing an adaptive solution for positioning data with different labels. For example, the mode library stores "coarse adjustment-fast convergence mode" (drive force 8-12 N, step size 0.1-0.3 mm, sampling rate 500 Hz) and "fine adjustment-high precision stability mode" (drive force 2-5 N, step size 0.001-0.01 mm, sampling rate 2000 Hz), which can be directly retrieved according to the positioning label. The positioning dominant mode refers to the core control mode that is most suitable for the current positioning scenario, matched from the preset control mode library based on the control positioning label. It determines the core strategy and parameter direction of the positioning adjustment at this stage and serves as the basis for subsequent adjustment sequence division. For example, when the control positioning label is "stage: transition adjustment, axis type: Y axis, accuracy level: smooth connection level", the "transition adjustment-smooth connection mode" (drive force 6-8 N, step size 0.03-0.08 mm, switching threshold 0.05 mm) is matched from the mode library, which is the current positioning dominant mode and guides the subsequent adjustment division. The dominant performance indicator is a key performance parameter in the positioning dominant mode that determines the adjustment effect, covering adjustment speed, accuracy threshold, stability requirement, etc. It directly constrains the division of the time sequence scope and the design of the adjustment steps, ensuring that the adjustment process meets the core objectives of the mode. For example, the dominant performance indicators of the "transition adjustment-smooth connection mode" are "convergence time ≤ 5 s, deviation fluctuation ≤ 0.02 mm, step size adjustment gradient ≤ 0.02 mm / step", which clearly define the speed and accuracy standards that the adjustment needs to meet, avoiding deviation of the divided sequence from the mode requirements. The time sequence scope refers to a specific time sub-interval within the effective time span, divided based on key action points. Each sub-interval corresponds to a continuous adjustment action and includes dynamic parameters and positioning data within that interval, clearly defining the effective time range of a certain group of adjustment strategies. For example, in the above-mentioned transition adjustment stage, three time sequence scopes are divided based on key action points: "1695210800.0s-1695210802.0s (step size 0.05 mm takes effect), 1695210802.0s-1695210805.0s (step size 0.03 mm to 0.02 mm takes effect)". The parameters within each scope are stable, providing a clear time boundary for adjustment division.

[0060] Further, the parsing of the control positioning label in the control positioning data can be realized by a semantic analysis method, such as: using a natural language processing technology NLP to extract key identifiers in the data, thereby obtaining the control positioning label; the query of the positioning dominant mode in the preset control mode library can be realized by a mode matching algorithm, such as: using a K nearest neighbor classifier KNN to match the most similar positioning control mode, thereby obtaining the positioning dominant mode; the extraction of the dominant performance index corresponding to the positioning dominant mode can be realized by a feature extraction method, such as: using a principal component analysis PCA dimension reduction technology to obtain a mode core performance parameter, thereby obtaining the dominant performance index; the determination of the time sequence scope corresponding to the control positioning data can be realized by a time window analysis method, such as: using a sliding window algorithm to divide a data effective action time interval, thereby obtaining the time sequence scope; the adjustment and division of the time sequence scope can be realized by a dynamic segmentation method, such as: using an adaptive threshold algorithm to divide a time period according to a control requirement, thereby obtaining an adjustment control sequence.

[0061] As another embodiment of the application, the determination of the time sequence scope corresponding to the control positioning data based on the dominant performance index comprises: extracting a dynamic parameter sequence in the dominant performance index; performing a cooperative mapping of the dynamic parameter sequence and the control positioning data to obtain a mapping data set; analyzing an effective time span corresponding to the mapping data set; identifying a key action point in the effective time span; and determining the time sequence scope corresponding to the control positioning data based on the key action point.

[0062] The dynamic parameter sequence refers to a set of key parameters in the dominant performance index that dynamically change over time, records the parameter values and change trends in chronological order, reflects the dynamic characteristics of performance requirements in the positioning adjustment process, and is the core link of the correlation control positioning data. For example, the dynamic parameter sequence of the "transition adjustment-stable connection mode" is "0s: step 0.05mm, driving force 7N; 2s: step 0.03mm, driving force 6.5N; 5s: step 0.02mm, driving force 6N", which clearly presents the parameter changes at different times and provides a time dimension reference benchmark for matching control positioning data. The mapping data set refers to a data set formed by correlating and matching the dynamic parameter sequence and the control positioning data in the time dimension, so that the dynamic parameters (such as step and driving force) at each time point correspond to the positioning data (such as current position and deviation value) at the corresponding time point one by one, eliminating the time dislocation between the data. For example, the mapping data set of the transition adjustment phase of a certain turntable is "0s: step 0.05mm + position 120.4mm + deviation 0.4mm; 2s: step 0.03mm + position 120.48mm + deviation 0.02mm; 5s: step 0.02mm + position 120.50mm + deviation 0.001mm", which realizes the time alignment of parameters and positioning data. The effective time span refers to the time interval in the mapping data set where the dynamic parameter sequence and the control positioning data are both effective and matched with each other, that is, the time range from the start of the adjustment action to the end time when the dominant performance index requirements are met. Data beyond this range is considered invalid. For example, in the mapping data set of the above transition adjustment phase, the dynamic parameter is effective from 0s, and the deviation decreases to 0.001mm (meeting the index of ≤0.02mm) at 5s, so the effective time span is "1695210800.0s-1695210805.0s", ensuring that subsequent analysis is based only on data within the effective time. The key action point refers to the time node in the effective time span where the dynamic parameter changes significantly or the positioning data reaches an important threshold. These nodes mark the turning point of the adjustment strategy (such as parameter switching) or the achievement of the stage goal, and are the core basis for dividing the time sequence scope. For example, in the effective time span of the above transition adjustment phase, the key action points include "0s (adjustment start, initial parameter effective), 2s (step from 0.05mm to 0.03mm), 5s (deviation meets the standard, adjustment ends)", each node corresponds to an important turning point in the adjustment process, and determines the segmentation logic of the time sequence scope.

[0063] Further, the extraction of the dynamic parameter sequence in the dominant performance index can be realized by a time series decomposition method, such as: using an STL seasonal decomposition algorithm to extract the trend component and the periodic component in the index, so as to obtain the dynamic parameter sequence; the collaborative mapping of the dynamic parameter sequence and the control positioning data can be realized by a data association analysis method, such as: using a dynamic time warping (DTW) algorithm to align the time series features of the two types of data, so as to obtain a mapping data set; the analysis of the effective time span corresponding to the mapping data set can be realized by a sliding window statistical method, such as: using an autocorrelation function (ACF) to analyze the significance interval of the data in the time dimension, so as to obtain the effective time span; the identification of the key action points in the effective time span can be realized by an extreme value detection algorithm, such as: using a peak detection algorithm to identify the local maximum value and the mutation point in the time series, so as to obtain the key action points; the determination of the time sequence scope corresponding to the control positioning data can be realized by a time domain division method, such as: using a density-based spatial clustering of applications with noise (DBSCAN) algorithm to perform clustering analysis on the time stamp, so as to obtain the time sequence scope.

[0064] According to the adjustment control sequence, the drive instruction corresponding to the servo motor of the target turntable is generated, the ordered adjustment steps can be converted into accurate signals that can be directly executed by the motor, the motor action deviation caused by ambiguous or disordered instructions is avoided, an explicit execution basis is provided for the turntable pose adjustment, and the precision basis of the alignment control is further consolidated.

[0065] The turntable servo motor refers to a driving execution component specially designed for target turntable pose adjustment, having high-precision closed-loop control capability, and being capable of accurately adjusting the rotation speed, torque and rotation angle according to the input control signal, adapting to the power requirements of different stages such as turntable coarse adjustment and fine adjustment, and being the core power source for realizing the movement of the turntable from the current pose to the target pose. For example, the servo motor configured for a certain turntable is of the MS-200 model, has a rated power of 2.0kW, a rated rotation speed of 3600r / min, and a positioning accuracy of ±0.0005°. In the coarse adjustment stage, it can output a large torque of 12N·m to realize rapid rotation, and in the fine adjustment stage, it can maintain a small torque of 0.5N·m to realize micron-level angle control, thereby ensuring the accuracy and efficiency of the turntable pose adjustment. The driving instruction refers to a digital control signal generated according to the adjustment control sequence and recognizable by the turntable servo motor, which contains the motor action type (such as position adjustment and speed adjustment), specific parameters (rotation speed, torque, target angle) and execution time, and can directly guide the motor to complete the pose adjustment action according to the preset logic. For example, for the above-mentioned MS-200 motor, a certain driving instruction is “action type: position adjustment; target rotation speed: 1800r / min; output torque: 5N·m; execution time: 1.5s; target angle: 45.000°”. After receiving the instruction, the motor will rotate at a rotation speed of 1800r / min and a torque of 5N·m within 1.5 seconds, and finally stop at the target angle of 45.000°, matching the positioning requirement in the fine adjustment stage of the turntable. Alternatively, the generation of the driving instruction corresponding to the turntable servo motor in the target turntable can be realized by a PID control algorithm, such as calculating the motor rotation speed and rotation control amount by using a position-type PID controller, thereby obtaining the driving instruction.

[0066] S4, collect the high-frequency disturbance value of the driving instruction in the execution process, calculate the anti-disturbance control amount corresponding to the high-frequency disturbance value, and detect the precision imbalance point in the operation of the target turntable based on the anti-disturbance control amount.

[0067] The present application can timely capture the influence of external interference or motor itself fluctuation on the turntable action by collecting the high-frequency disturbance value of the driving instruction in the execution process and calculating the anti-disturbance control amount corresponding to the high-frequency disturbance value, thereby avoiding the expansion of pose deviation caused by disturbance accumulation, providing a data basis for precise anti-disturbance, and further improving the reliability of the position control.

[0068] The high-frequency disturbance value refers to a specific value quantifying the high-frequency interference size collected from the high-frequency component of the driving execution data based on the disturbance amplitude level. Low-level disturbance is collected at normal precision, and high-level disturbance is collected at higher precision, ensuring that the data can truly reflect the actual impact of disturbances of different intensities. For example, the above-mentioned "medium level" rotational speed disturbance is collected at a precision of 0.01 r / min, and the high-frequency disturbance value is "± 0.82 r / min". The "low level" torque disturbance is collected at a precision of 0.05 N·m, and the high-frequency disturbance value is "± 0.38 N·m". These values are directly used in subsequent anti-disturbance control quantity calculation.

[0069] As an embodiment of the present application, the collection of the high-frequency disturbance value during the execution of the driving instruction includes: identifying the execution time interval corresponding to the driving instruction; extracting driving execution data within the execution time interval; counting the high-frequency oscillation index corresponding to the high-frequency component in the driving execution data; determining the disturbance amplitude level corresponding to the high-frequency oscillation index; and collecting the high-frequency disturbance value during the execution of the driving instruction based on the disturbance amplitude level.

[0070] The execution time interval refers to the complete time range from the start of the driving instruction to the completion of the instruction action by the servo motor, clearly defines the time boundary of the disturbance value collection, ensures that only the data during the instruction execution process is analyzed, and avoids irrelevant time data interference. For example, the "position adjustment" driving instruction of a certain rotary table servo motor has a preset execution time of 1.5 s, and the actual instruction receiving time is 10:00:00.000 and the target angle adjustment is completed at 10:00:01.500. The execution time interval of the instruction is "10:00:00.000-10:00:01.500", and all disturbance collection operations need to be completed within this interval. The driving execution data refers to the key parameter data reflecting the state of the rotary table servo motor executing the driving instruction, which is collected in real time within the execution time interval, including motor speed, output torque, stator current, actual rotation angle, etc., and is the basic data source for extracting high-frequency disturbance features. For example, within the above-mentioned 1.5 s execution interval, the collected driving execution data includes: speed (1798 r / min-1802 r / min fluctuation), torque (4.8 N·m-5.2 N·m change), and actual angle (44.995°-45.005° adjustment). These data completely record the dynamic process of the motor executing the instruction. The high-frequency oscillation index refers to the index of quantifying the oscillation strength of the high-frequency component (usually referring to the frequency > 100 Hz fluctuation) extracted from the driving execution data through signal processing (such as Fourier transform). It can be calculated by the amplitude standard deviation or peak difference of the high-frequency component, and directly reflects the intensity of the high-frequency disturbance. For example, the above-mentioned speed data is processed to separate the 120 Hz high-frequency fluctuation component, whose amplitude changes between 1798 r / min and 1802 r / min, and the standard deviation is calculated as 0.8 r / min. The "0.8 r / min" is the high-frequency oscillation index corresponding to the speed, which reflects the strength of the high-frequency disturbance of the speed. The disturbance amplitude level refers to the high-frequency disturbance intensity level divided according to the numerical range of the high-frequency oscillation index. It is usually divided into low, medium, and high three levels, and different levels correspond to different collection accuracy and subsequent anti-disturbance strategies, providing a hierarchical basis for accurate collection of high-frequency disturbance values. For example, the preset high-frequency oscillation index <0.5 is "low level", 0.5-1.0 is "medium level", and >1.0 is "high level". The above-mentioned speed high-frequency oscillation index 0.8 r / min belongs to "medium level", and the torque high-frequency oscillation index 0.4 N·m belongs to "low level", which correspond to different disturbance collection accuracy requirements.

[0071] Furthermore, identifying the execution time interval corresponding to the drive command can be achieved through timestamp parsing methods, such as using a command cycle analysis tool to extract the effective time period of the PWM waveform to obtain the execution time interval; extracting the drive execution data within the execution time interval can be achieved through a data acquisition card, such as using the NIDAQmx tool to record the current and voltage waveforms of the motor driver in real time to obtain the drive execution data; statistically analyzing the high-frequency oscillation index corresponding to the high-frequency components in the drive execution data can be achieved through spectrum analysis methods, such as using Fast Fourier Transform (FFT) to calculate the energy integral value of the frequency band above 5kHz to obtain the high-frequency oscillation index; determining the disturbance amplitude level corresponding to the high-frequency oscillation index can be achieved through threshold grading methods, such as using the ISO 10816 vibration standard to divide the danger level range of the oscillation index to obtain the disturbance amplitude level; collecting the high-frequency disturbance value of the drive command during execution can be achieved through bandpass filtering methods, such as using a digital filter to extract the effective value of the vibration signal in the 2kHz-10kHz frequency band to obtain the high-frequency disturbance value.

[0072] In another embodiment of the present invention, the anti-interference control quantity corresponding to the high-frequency disturbance value is calculated using the following formula: ; in, This represents the disturbance rejection control quantity (unit: Nm) corresponding to the high-frequency disturbance value. Represents the proportionality coefficient. Indicates the level of disturbance amplitude. Indicates the execution time interval (unit: seconds). The value of the high-frequency disturbance is expressed in Nm.

[0073] In detail, the disturbance rejection control quantity can be expressed as the additional control quantity (unit: Nm) needed to counteract the influence of high-frequency disturbances on the turntable servo motor. This quantity guides the motor to output a reverse force for stable operation. For example, when a turntable servo motor is subjected to a high-frequency disturbance, the calculated... =1.2Nm, the motor will output a reverse torque of 1.2Nm to counteract the disturbance; the proportional coefficient can represent an adjustment coefficient set according to the mechanical characteristics of the turntable (such as inertia, stiffness), used to adapt to the disturbance rejection intensity and avoid excessive or insufficient disturbance rejection. For example, if the turntable has a small inertia and a fast response, K=0.5 is set to match the disturbance rejection control quantity with the response capability of the mechanical system; the disturbance amplitude level can represent the classification of the high-frequency disturbance intensity (such as low level 1, medium level 2, high level 3). The higher the level, the more urgent the disturbance rejection requirement. The formula will amplify or reduce the disturbance rejection control quantity through this value. For example, if the disturbance is at a high level at a certain moment, =3, which makes the calculated disturbance rejection control quantity more significant; the execution time interval can represent the execution duration of the drive instruction (unit: s), which determines the time range of the integral in the formula, used to statistically analyze the cumulative effect of the disturbance within that time period. For example, if the drive instruction execution time is T=1s, the integral extends from t=0 to t=1, covering the disturbance during the entire instruction execution process; the high-frequency disturbance value can represent the high-frequency interference value (unit: Nm) that changes with time, reflecting the magnitude of the disturbance at different times. By integrating its square, the energy accumulation of the disturbance can be reflected. For example, within T=2s, =0.3sin(100πt) (simulating 100Hz high-frequency oscillation), substituting into the formula, the energy contribution of the disturbance can be calculated, and then the disturbance rejection control quantity can be obtained.

[0074] Based on the aforementioned anti-disturbance control quantity, this invention detects the accuracy imbalance point during the operation of the target turntable, which can accurately locate the root cause of the pose deviation caused by the disturbance and avoid aimless accuracy investigation; it can capture the potential trend of accuracy decline in advance and intervene before the imbalance intensifies to prevent serious deviations; it also provides a basis for subsequent optimization of anti-disturbance strategies and continuous improvement of the long-term operating accuracy of the turntable, effectively ensuring the stability of alignment control.

[0075] The precision imbalance point refers to a specific position or moment during the operation of the target turntable where the actual pose accuracy exceeds a preset allowable threshold due to factors such as high-frequency disturbances, mechanical wear, or control delays. This is a key indicator of decreased turntable positioning stability, and the concentrated outbreak point of the positioning deviation needs to be deduced through the anti-disturbance control quantity. For example, a precision turntable has a preset allowable X-axis positioning deviation of ±0.002mm. When executing drive commands, monitoring through the anti-disturbance control quantity reveals that when the turntable moves to the X-axis position of 120.50mm, the actual deviation remains stable at 0.005mm (exceeding the threshold), and the anti-disturbance control quantity needs to be increased to 1.8Nm to barely maintain this position. This 120.50mm position is a typical precision imbalance point, reflecting a significant disturbance or mechanical coordination problem. Optionally, the detection of the precision imbalance point during the operation of the target turntable can be achieved through laser interferometry, such as using a Renishaw XL-80 laser interferometer to monitor the turntable positioning deviation and repeatability accuracy, thereby obtaining the precision imbalance point.

[0076] S5. Identify the precision offset value corresponding to the precision imbalance point, coordinate the precision offset value with the preset precision index to obtain the alignment parameter group, and generate the alignment control strategy corresponding to the target turntable based on the alignment parameter group.

[0077] The application can accurately quantify the degree of the target pose deviation of the rotary table by identifying the precision offset value corresponding to the precision imbalance point, provide clear numerical basis for subsequent correction actions, make the anti-interference or adjustment strategy more targeted, avoid resource waste and efficiency loss caused by indiscriminate adjustment, trace the root cause of imbalance, provide key reference for optimizing the rotary table control logic and improving long-term operation precision, and thus guarantee the stability and accuracy of the rotary table alignment.

[0078] The precision offset value refers to the deviation value between the actual pose (such as position, angle, etc.) and the preset target pose of the target rotary table at the precision imbalance point, can quantify the pose deviation degree of the imbalance point, provide specific basis for subsequent deviation correction, for example, the X-axis target position of a rotary table is 120.500 mm, the actual position at the precision imbalance point is 120.505 mm, the X-axis precision offset value is 0.005 mm, the Y-axis target position is 80.200 mm, and the actual position is 80.197 mm, the Y-axis precision offset value is -0.003 mm, which directly reflects the deviation of each axis at the imbalance point, and the identification of the precision offset value corresponding to the precision imbalance point can be realized by a deviation quantification analysis method, such as using the least square method to fit the coordinate deviation between the theoretical trajectory and the actual trajectory, so as to obtain the precision offset value.

[0079] Further, the precision offset value and the preset precision index are cooperatively aligned to obtain an alignment parameter group, the quantitative correlation between the actual pose deviation and the standard requirement can be established, it is clear to define whether the offset exceeds the index range and the specific exceeding degree, provide clear reference basis for subsequent processing, and help the long-term stable rotary table to run in a high-precision state.

[0080] The alignment parameter group refers to a parameter set formed by integrating the precision offset value, the precision difference value, the cooperative compensation amount, the preset precision index and the target correction value, contains all key data required for the alignment of the rotary table, can directly guide the subsequent precision correction action, for example, the alignment parameter group of the X-axis is: "precision offset value: 0.005 mm; preset precision index: ±0.002 mm; precision difference value: 0.003 mm; cooperative compensation amount: 0.0039 mm; target corrected position: 120.500 mm-0.0039 mm=120.4961 mm", which completely presents the alignment logic and the correction target, and provides a clear basis for the adjustment of the rotary table.

[0081] As an embodiment of the present application, the precision offset value is cooperatively aligned with the preset precision index to obtain an alignment parameter group, comprising: comparing the precision difference between the precision offset value and the preset precision index; based on the precision difference, traversing the alignment relationship set in the preset alignment rule library; screening the cooperative matching pairs in the alignment relationship set that satisfy the precision threshold; analyzing the cooperative compensation amount matched by the cooperative matching pairs; based on the cooperative compensation amount, cooperatively aligning the precision offset value with the preset precision index to obtain an alignment parameter group.

[0082] The precision difference value is an absolute difference value or a relative difference value between the precision offset value and the preset precision index, is used for quantifying the degree of actual pose of the turntable deviating from the standard requirement, and is a core quantitative basis for judging whether compensation is needed and a compensation amplitude. For example, a preset precision index of an X-axis of a certain turntable is ±0.002 mm (that is, an allowed deviation range), and a precision offset value measured at a precision imbalance point is 0.005 mm. At this time, the precision difference value is 0.005 mm-0.002 mm=0.003 mm, clearly reflecting the specific value of the actual offset exceeding the standard, and providing a basis for subsequent matching compensation rules. The preset alignment rule library is a database that stores corresponding rules of the precision difference value and the compensation strategy, covers collaborative matching logic, a compensation mode and a parameter calculation method corresponding to different precision difference value intervals, provides a standardized basis for collaborative alignment of the precision offset value and the preset index, and stores, for example, "precision difference value 0.001-0.003 mm: linear compensation, compensation coefficient 1.2; precision difference value 0.003-0.005 mm: nonlinear compensation, compensation coefficient 1.5; difference value >0.005 mm: trigger an emergency calibration process" in the rule library. The alignment relationship set is a specific rule set containing "precision difference value range-compensation type-compensation coefficient" classified according to the precision difference value interval in the preset alignment rule library, is a direct object of traversal matching, and each relationship corresponds to a group of specific collaborative alignment logic. For example, the alignment relationship set for a "mild offset" scenario is "difference value 0.001-0.002 mm: linear compensation, coefficient 1.0; difference value 0.002-0.003 mm: linear compensation, coefficient 1.2", and the relationship set for a "moderate offset" scenario is "difference value 0.003-0.004 mm: nonlinear compensation, coefficient 1.3; difference value 0.004-0.005 mm: nonlinear compensation, coefficient 1.5", facilitating accurate matching according to the difference value. The collaborative matching pair is a "precision difference value range-compensation rule" pair selected from the alignment relationship set and completely matched with the current precision difference value, meets a preset precision threshold condition (that is, the current difference value falls within a certain interval of the relationship set), and is a direct basis for determining a collaborative compensation amount. For example, the current calculated precision difference value of the X-axis is 0.003 mm, and after traversing the alignment relationship set, it is found that the current difference value falls within the relationship interval "difference value 0.003-0.004 mm: nonlinear compensation, coefficient 1.3". The combination of "0.003-0.004 mm (difference value range)-nonlinear compensation+1.3 (compensation rule)" is the collaborative matching pair. The collaborative compensation amount is a specific value used for correcting the precision offset value and calculated based on the compensation rule (such as the compensation type and the coefficient) in the collaborative matching pair, can make the actual pose of the turntable approach the preset precision index, and is a core execution parameter of collaborative alignment. For example, the above collaborative matching pair is "nonlinear compensation, coefficient 1.3", and the current precision difference value is 0.003 mm.003mm, the cooperative compensation amount is calculated according to the rule: 0.003mm*1.3=0.0039mm, and the value is the compensation amount required for adjusting the target position of the X-axis, which is used to offset the precision deviation of 0.005mm.

[0083] Further, the precision difference between the precision deviation value and the preset precision index can be realized by a difference calculation method, such as: using an absolute value difference algorithm to calculate the numerical difference between the measured deviation value and the standard index, so as to obtain the precision difference; the traversal of the alignment relationship set in the preset alignment rule library can be realized by a graph database query method, such as: using the Cypher language of the Neo4j graph database to traverse all node relationship paths, so as to obtain the alignment relationship set; the screening of the cooperative matching pairs in the alignment relationship set that meet the precision threshold can be realized by a filtering algorithm, such as: using a Bloom filter to quickly screen matching combinations that meet the precision tolerance range, so as to obtain the cooperative matching pairs; the analysis of the cooperative compensation amount matched by the cooperative matching pairs can be realized by a regression analysis method, such as: using a multiple linear regression model to calculate the compensation coefficient matrix of each matching pair, so as to obtain the cooperative compensation amount; the cooperative alignment of the precision deviation value and the preset precision index can be realized by a data registration method, such as: using an ICP iterative closest point algorithm to realize the spatial alignment of the deviation data and the standard data, so as to obtain the alignment parameter set.

[0084] Based on the alignment parameter set, the alignment control strategy corresponding to the target turntable is generated, which can make the strategy deeply fit the actual precision deviation and compensation demand of the turntable, avoid the control blindness caused by the lack of quantitative data, provide a clear direction for accurate pose correction, provide data support for subsequent iterative optimization control logic and adaptation to alignment requirements under different working conditions, and help long-term stable turntable high-precision operation state.

[0085] The alignment control strategy refers to a systematic control scheme for correcting the accuracy deviation of the rotary table and returning it to the preset accuracy index based on core data such as accuracy offset values and collaborative compensation amounts in the alignment parameter group, and contains control targets, execution steps, parameter configurations, and monitoring logic, which can directly guide the rotary table to complete position correction. For example, the X-axis alignment parameter group of a certain rotary table is "accuracy offset value 0.005 mm, collaborative compensation amount 0.0039 mm, target correction position 120.4961 mm", and the corresponding alignment control strategy is: first, the servo motor executes a compensation movement of 0.0039 mm at a torque of 5 N·m; second, the real-time deviation is monitored at a frequency of 1 kHz; third, when the deviation is less than or equal to 0.002 mm, the torque is switched to 3 N·m to maintain stability, ensuring that the rotary table accurately returns to the target position. Optionally, the generation of the alignment control strategy corresponding to the target rotary table can be achieved through a fuzzy logic control method, such as using a Mamdani-type fuzzy inference system to generate a multi-level adjustment strategy based on the alignment parameter group, thereby obtaining the alignment control strategy.

[0086] Compared with the problems described in the background art, the present application can break through the information limitation of a single sensor by acquiring the turntable sensor data of the target turntable in a multi-modal scene, cover multi-dimensional state information such as position, vibration and temperature in the operation of the turntable, realize comprehensive perception of the working condition of the turntable, avoid state misjudgment caused by the lack of single data dimension, and ensure the precision and stability of the turntable alignment control from the source. Based on the actual pose state, the present application calculates the position deviation between the current position and the target position of the target turntable, can accurately quantify the specific difference between the current position and the target position, provides clear numerical basis for subsequent control decision, avoids the problem of overshoot, oscillation and other problems caused by inaccurate deviation estimation, ensures the precision of the final turntable alignment control from the key link, further, the present application extracts the control positioning data in the hierarchical control process, can provide accurate and specific information basis for subsequent adjustment and division of control positioning data, generation of turntable servo motor driving instruction, avoid the lack of clear direction of subsequent control link caused by data loss or ambiguity, help to improve the long-term stability and precision of turntable alignment control, further, the present application collects the high-frequency disturbance value of the driving instruction in the execution process, and calculates the anti-disturbance control amount corresponding to the high-frequency disturbance value, can capture the influence of external interference or motor itself fluctuation on the action of the turntable in time, avoid the expansion of the pose deviation caused by disturbance accumulation, provide data basis for accurate anti-disturbance, further consolidate the reliability of the alignment control, finally, the present application can accurately quantify the degree of deviation of the turntable pose from the target by identifying the precision offset value corresponding to the precision imbalance point, provide clear numerical basis for subsequent correction action; the anti-disturbance or adjustment strategy can be more targeted, avoid resource waste and efficiency loss caused by indiscriminate adjustment; it can also trace the root cause of imbalance, provide key reference for optimizing turntable control logic and improving long-term operation precision, so as to ensure the stability and precision of turntable alignment. Therefore, the turntable precise alignment control method based on multi-modal sensor fusion provided by the embodiment of the present application can improve the control precision and self-adaptive ability of the turntable in complex application environment.

[0087] As Figure 3 shown, it is a functional module diagram of a turntable precise alignment control system based on multi-modal sensor fusion.

[0088] The multi-modal sensor fusion-based turntable precise alignment control system 200 can be installed in an electronic device. According to the functions implemented, the multi-modal sensor fusion-based turntable precise alignment control system can include a state recognition module 201, a flow construction module 202, an instruction generation module 203, an imbalance point detection module 204, and a strategy generation module 205. The modules in the present application can also be referred to as units, which refer to a series of computer program segments that can be executed by an electronic device processor and can complete a fixed function, which are stored in the memory of the electronic device.

[0089] In the embodiments of the present application, the functions of each module / unit are as follows: The state recognition module 201 is configured to obtain turntable sensor data of a target turntable in a multi-modal scene, perform data fusion on the turntable sensor data to obtain a multi-modal fusion vector, and recognize a real-time pose state corresponding to the target turntable based on the multi-modal fusion vector. The flow construction module 202 is configured to calculate a position deviation between a current position and a target position of the target turntable based on the actual pose state, determine a control stage corresponding to the target turntable according to the position deviation, and construct a hierarchical control flow corresponding to the target turntable based on the control stage. The instruction generation module 203 is configured to extract control positioning data in the hierarchical control flow, adjust and divide the control positioning data to obtain an adjusted control sequence, and generate a driving instruction corresponding to a turntable servo motor in the target turntable according to the adjusted control sequence. The imbalance point detection module 204 is configured to collect a high-frequency disturbance value of the driving instruction during execution, calculate an anti-disturbance control amount corresponding to the high-frequency disturbance value, and detect an accuracy imbalance point in the operation of the target turntable based on the anti-disturbance control amount. The strategy generation module 205 is configured to identify an accuracy offset value corresponding to the accuracy imbalance point, cooperatively align the accuracy offset value with a preset accuracy indicator to obtain an alignment parameter group, and generate an alignment control strategy corresponding to the target turntable based on the alignment parameter group.

[0090] In detail, the modules in the multi-modal sensor fusion-based turntable precise alignment control system 200 in the embodiments of the present application use the same technical means as the multi-modal sensor fusion-based turntable precise alignment control method in the above Figure 1 , and can produce the same technical effects, which will not be described here.

[0091] It is apparent for a person skilled in the art that the present application is not limited to the details of the above exemplary embodiments, but can be implemented in other concrete forms without departing from the spirit or essential characteristics of the present application.

[0092] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not limiting. In the above embodiments, each embodiment can be combined with or independent of the other embodiments, and deleting any one of them does not affect the technical implementation of the other embodiments. Although the present application has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the present application.

Claims

1. A method for precise alignment control of a rotary table based on multi-modal sensor fusion, characterized in that, The method comprises: acquiring the turret sensor data of the target turret in a multi-modal scene, performing data fusion on the turret sensor data to obtain a multi-modal fusion vector, and identifying the real-time pose state corresponding to the target turret based on the multi-modal fusion vector; based on the real-time pose state, calculating the position deviation between the current position and the target position of the target turret, determining the control stage corresponding to the target turret according to the position deviation, and constructing the hierarchical control process corresponding to the target turret based on the control stage; extracting control positioning data in the hierarchical control process, adjusting and dividing the control positioning data to obtain an adjustment control sequence, and generating the driving instructions corresponding to the turret servo motor in the target turret according to the adjustment control sequence; collecting the high-frequency disturbance value in the execution process of the driving instructions, calculating the anti-disturbance control amount corresponding to the high-frequency disturbance value, and detecting the precision imbalance point in the operation of the target turret based on the anti-disturbance control amount; identifying the precision offset value corresponding to the precision imbalance point, cooperatively aligning the precision offset value with the preset precision index to obtain an alignment parameter group, and generating the alignment control strategy corresponding to the target turret based on the alignment parameter group.

2. The multi-modal sensor fusion-based precision alignment control method of a turntable according to claim 1, wherein, The method comprises: extracting multi-modal pose features in the multi-modal fusion vector; analyzing the sensing pose components associated with the multi-modal pose features; synchronously obtaining synchronization component data by synchronizing the sensing pose components in real time; based on the synchronization component data, analyzing the current tilt angle and center coordinates of the target turret; based on the tilt angle and the center coordinates, identifying the real-time pose state corresponding to the target turret.

3. The multi-modal sensor fusion-based precision alignment control method of a turntable according to claim 2, characterized in that, The method comprises: analyzing the component time sequence corresponding to the sensing pose components; determining the synchronization time axis corresponding to the component time sequence; querying the data time stamp corresponding to each component in the synchronization time axis; synchronously aligning the data time stamp to obtain a synchronization data set; extracting consistent data in the synchronization data set as synchronization component data.

4. The multi-modal sensor fusion-based precision alignment control method of a turntable according to claim 1, wherein, The method comprises: querying the stage division logic in the control stage; based on the stage division logic, determining the inter-layer association module between the control stages; generating a topological hierarchical framework corresponding to the inter-layer association module; configuring the hierarchical control data corresponding to the hierarchical control framework; based on the hierarchical control data, constructing the hierarchical control process corresponding to the target turret.

5. The multi-modal sensor fusion-based precision alignment control method of a turntable according to claim 4, characterized in that, The method comprises: analyzing the hierarchical mapping relationship in the inter-layer association module; based on the hierarchical mapping relationship, identifying the associated control unit corresponding to the inter-layer association module; extracting the key control nodes in the associated control unit; reconstructing the topological control chain corresponding to the key control nodes; generating a topological hierarchical framework corresponding to the topological control chain.

6. The multi-modal sensor fusion-based precision alignment control method of a turntable according to claim 1, wherein, The adjusting and dividing of the control positioning data comprises: parsing a control positioning label in the control positioning data; querying a positioning dominant mode in a preset control mode library according to the control positioning label; extracting a dominant performance index corresponding to the positioning dominant mode; determining a time sequence scope corresponding to the control positioning data based on the dominant performance index; adjusting and dividing the time sequence scope to obtain an adjusted control sequence.

7. The multi-modal sensor fusion-based precision alignment control method of a turntable according to claim 6, characterized in that, The determining of the time sequence scope corresponding to the control positioning data based on the dominant performance index comprises: extracting a dynamic parameter sequence in the dominant performance index; cooperatively mapping the dynamic parameter sequence and the control positioning data to obtain a mapping data set; analyzing an effective time span corresponding to the mapping data set; identifying a key action point in the effective time span; determining the time sequence scope corresponding to the control positioning data based on the key action point.

8. The multi-modal sensor fusion-based precision alignment control method of a turntable according to claim 1, wherein, The calculating of the position deviation amount between the current position and the target position of the target turntable based on the real-time pose state comprises: analyzing a pose state type corresponding to the real-time pose state; parsing a current position and a target position corresponding to the target turntable based on the pose state type; dividing a position axis component corresponding to the current position; identifying an axis deviation value of the position axis component relative to the target position; calculating the position deviation amount between the current position and the target position of the target turntable based on the axis deviation value through the following formula: ； wherein, represents a position deviation amount between a current position and a target position of the target turntable, represents a total number of control axes in the target turntable, represents a control axis index corresponding to the target turntable, represents a deviation value on a first control axis, represents a weight coefficient of a first control axis.

9. The multi-modal sensor fusion-based precision alignment control method of a turntable according to claim 1, wherein, The collecting of the high-frequency disturbance value of the driving instruction in the execution process comprises: identifying an execution time interval corresponding to the driving instruction; extracting driving execution data in the execution time interval; counting a high-frequency oscillation index corresponding to a high-frequency component in the driving execution data; determining a disturbance amplitude level corresponding to the high-frequency oscillation index; collecting the high-frequency disturbance value of the driving instruction in the execution process based on the disturbance amplitude level; calculating an anti-disturbance control amount corresponding to the high-frequency disturbance value through the following formula: ； wherein, represents the disturbance value corresponding to the high-frequency disturbance value, represents a proportional coefficient, represents a disturbance amplitude level, represents an execution time interval, represents the high-frequency disturbance value.

10. The multi-modal sensor fusion-based precision alignment control method of a turntable according to claim 1, wherein, The cooperatively aligning of the precision offset value and the preset precision index to obtain an alignment parameter group comprises: comparing a precision difference value between the precision offset value and the preset precision index; traversing an alignment relationship set in a preset alignment rule library based on the precision difference value; screening a cooperatively matched pair in the alignment relationship set that satisfies a precision threshold value; parsing a cooperatively compensating amount matched by the cooperatively matched pair; cooperatively aligning the precision offset value and the preset precision index based on the cooperatively compensating amount to obtain the alignment parameter group.

Images