Non-contact optical positioning and precision mechanical motion cooperative detection method

Abstract

The application discloses a kind of non-contact optical positioning and precision machinery movement collaborative detection method, it is related to tool detection technical field;Method includes first constructs with the transverse of rotating tool rotation plane as x axis, infrared light and so on initial alignment direction as y axis, vertical direction as z axis target three-dimensional coordinate system, then the preset movement operation of z axis motor and detection adjustment motor determines tool tip point and records y axis displacement value, subsequently rotating tool is to the preset detection requirement after repeating movement operation obtains x axis displacement value, finally obtains x, y axis displacement value.This method is by constructing coordinate system to fit tool detection scene, can accurately obtain displacement data, improve the accuracy and efficiency of tool detection.

Description

Technical Field

[0001] This invention belongs to the field of cutting tool inspection technology, specifically relating to a non-contact optical positioning and precision mechanical motion collaborative detection method. Background Technology

[0002] In the fields of machining, tool inspection, and precision manufacturing, the positioning accuracy of the cutting edge of a rotary tool directly determines the dimensional tolerances, surface quality, and machining efficiency of the workpiece, and is a core element in ensuring the stability of precision manufacturing processes. Currently, tool tip positioning inspection methods are mainly divided into two categories: contact inspection and non-contact inspection.

[0003] Contact detection methods obtain position information by directly contacting the detection element with the tool tip. Although the principle is simple and the cost is low, there are significant limitations: First, the contact process can easily cause mechanical wear on the tool tip, which is especially unsuitable for precision tools such as diamond tools and micro-nano scale tools; second, the detection force may cause slight deformation of the tool tip, affecting the positioning accuracy; and third, the detection efficiency is low, making it difficult to adapt to the rapid detection needs of automated production lines. Summary of the Invention

[0004] The purpose of this invention is to solve the problems of insufficient coordination between optical and precision mechanical motion and difficulty in balancing detection accuracy and efficiency in existing non-contact optical detection methods, and to propose a non-contact optical positioning and precision mechanical motion coordinated detection method.

[0005] This invention proposes a non-contact optical positioning and precision mechanical motion coordinated detection method, including a blade control machine, an infrared light emitter, a rotary tool, a grating position sensor, and a carriage. The blade control machine includes a detection and adjustment motor and a Z-axis motor. The grating position sensor is mounted on the Z-axis motor. Both the detection and adjustment motor and the Z-axis motor are connected to the carriage. The detection and adjustment motor drives the carriage to move left and right. The infrared light emitter and the reference axis of the rotary tool are located at the same y-coordinate 0 point. The method includes:

[0006] Step 1: Construct a target three-dimensional coordinate system; the x-axis of the target three-dimensional coordinate system is the transverse coordinate axis of the rotation plane of the rotating tool, the y-axis is the coordinate axis for the initial alignment of the infrared light, the reference axis, and the axis of the rotating tool, and the z-axis is the coordinate axis for the z-axis motor to drive the carriage to move in the vertical direction;

[0007] Step 2: Perform a first preset movement operation on the z-axis motor, and a second preset movement operation on the detection and adjustment motor until the rotating tool senses the grating position sensor, then determine the current coordinate as the final correction position; the rotating tool is on the y-axis;

[0008] Step 3: Record the displacement of the motor to obtain the y-axis displacement value;

[0009] Step 4: Rotate the rotating tool according to the preset rotation angle until the rotating tool meets the preset detection requirements;

[0010] Step 5: Execute step 2 again. This time, the rotating tool is on the x-axis. Record the displacement of the detection and adjustment motor to obtain the x-axis displacement value.

[0011] Optionally, the first preset movement operation includes:

[0012] The z-axis motor moves downwards incrementally in steps Δz, and the downward movement of the z-axis motor is monitored in real time by a grating position sensor until the infrared light is blocked by the rotating tool.

[0013] Optionally, the second preset movement operation includes:

[0014] After the z-axis motor moves downward by Δz each time, the detection and adjustment motor moves Δy in the +y direction from the current coordinate as the origin. If the light from the grating position sensor and the rotating tool is connected after the movement, the origin is updated to the coordinates of the detection and adjustment motor after the movement. Otherwise, the detection and adjustment motor returns to the origin and moves Δy in the -y direction from the origin, updating the origin to the coordinates of the detection and adjustment motor after the movement.

[0015] Optional preset rotation angles include:

[0016] The preset rotation angle is 90 degrees for rotating the cutting tool.

[0017] Optional, preset detection requirements include:

[0018] The preset testing requirement is that the light can be conducted when the motor moves in both directions.

[0019] Optionally, step 1 may be followed by:

[0020] The rotary tool is rotated according to the target three-dimensional coordinate system to obtain full-angle offset data;

[0021] The full-angle offset data is divided into sub-data segment sets according to the rotation angle interval;

[0022] Frequency domain transformation is performed on each group of sub-data segments in the sub-data segment set to obtain the sub-data segment spectrum distribution set;

[0023] The fundamental wave feature set of each sub-data segment is obtained by identifying the amplitude and phase characteristics of the fundamental wave component in the spectral distribution of each group of sub-data segments in the spectral distribution set.

[0024] By fusing and analyzing the fundamental component characteristics of multiple sub-data segments, calibration deviation values ​​with spatial globality are obtained.

[0025] Optionally, step 3 may be followed by:

[0026] The time series and spatial distribution features of the calibration deviation values ​​are extracted to obtain temporal and spatial features;

[0027] Based on the temporal and spatial characteristics, a deviation correlation matrix is ​​obtained through relationship analysis.

[0028] The deviation correction amount is obtained by performing matrix operations on the y-axis displacement value and the deviation correlation matrix;

[0029] The correction value is obtained by iteratively calibrating the y-axis displacement value based on the deviation correction amount;

[0030] The residual deviation value is obtained by calculating the residual deviation between the correction value and the calibration deviation value;

[0031] The residual deviation is quantified to obtain the compensation value;

[0032] The correction value is adjusted based on the compensation value to obtain the precise deviation value of the y-axis.

[0033] Optionally, step 5 may be followed by:

[0034] The static deviation component and the dynamic deviation component are obtained by analyzing the calibration deviation value;

[0035] The static displacement value is obtained by performing basic compensation correction on the calibration deviation value based on the static deviation component.

[0036] The dynamic displacement value is obtained by dynamically compensating and correcting the calibration deviation value based on the rotation angle and the dynamic deviation component.

[0037] The static displacement value and the dynamic displacement value are fused to obtain the accurate x-axis deviation value.

[0038] The beneficial effects of this invention are as follows: This invention proposes a non-contact optical positioning and precision mechanical motion collaborative detection method. First, a target three-dimensional coordinate system is constructed with the horizontal axis of the rotating tool's rotation plane as the x-axis, the initial alignment direction (such as infrared light) as the y-axis, and the vertical direction as the z-axis. Then, the tool tip point is determined and the y-axis displacement value is recorded through preset movement operations of the z-axis motor and the detection adjustment motor. Subsequently, after rotating the tool to the preset detection requirement, the movement operation is repeated to obtain the x-axis displacement value, ultimately yielding the x and y-axis displacement values. This method, by constructing a coordinate system that fits the tool detection scenario, can accurately acquire displacement data, improving the accuracy and efficiency of tool detection. Attached Figure Description

[0039] The present invention will now be further described with reference to the accompanying drawings.

[0040] Figure 1 The flowchart illustrates a non-contact optical positioning and precision mechanical motion co-detection method provided in this embodiment of the invention. Detailed Implementation

[0041] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0042] Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0043] This invention provides a non-contact optical positioning and precision mechanical motion coordinated detection method, including a blade control machine, an infrared light emitter, a rotary tool, a grating position sensor, and a carriage. The blade control machine includes a detection and adjustment motor and a Z-axis motor. The grating position sensor is mounted on the Z-axis motor. Both the detection and adjustment motor and the Z-axis motor are connected to the carriage. The detection and adjustment motor drives the carriage to move left and right. The infrared light emitter and the reference axis of the rotary tool are located at the same y-coordinate 0 point. The method is characterized by:

[0044] Step 1: Construct the target's three-dimensional coordinate system;

[0045] Step 2: Perform the first preset movement operation on the z-axis motor, and the second preset movement operation on the detection and adjustment motor until the rotating tool senses the grating position sensor, then determine the current coordinate as the final correction position;

[0046] Step 3: Record the displacement of the motor to obtain the y-axis displacement value;

[0047] Step 4: Rotate the rotating tool according to the preset rotation angle until the rotating tool meets the preset detection requirements;

[0048] Step 5: Execute step 2 again. This time, the rotating tool is on the x-axis. Record the displacement of the detection and adjustment motor to obtain the x-axis displacement value.

[0049] Among them, the x-axis of the target three-dimensional coordinate system is the horizontal coordinate axis of the rotating tool's rotation plane, the y-axis is the coordinate axis for the initial alignment of the infrared light, the reference axis, and the rotating tool's axis, and the z-axis is the coordinate axis for the z-axis motor to drive the carriage to move in the vertical direction;

[0050] The rotating tool is on the y-axis.

[0051] Based on the non-contact optical positioning and precision mechanical motion collaborative detection method provided in this invention, a target three-dimensional coordinate system is first constructed with the horizontal direction of the rotating tool's rotation plane as the x-axis, the initial alignment direction of infrared light as the y-axis, and the vertical direction as the z-axis. Then, the tool tip point is determined and the y-axis displacement value is recorded through preset movement operations of the z-axis motor and the detection adjustment motor. Subsequently, after rotating the tool to the preset detection requirement, the movement operation is repeated to obtain the x-axis displacement value, ultimately yielding the x and y-axis displacement values. This method, by constructing a coordinate system that fits the tool detection scenario, can accurately acquire displacement data, improving the accuracy and efficiency of tool detection.

[0052] In one implementation, it is assumed that the axis of the rotating tool deviates from the reference axis, with a basic offset of y in the y-direction and x in the x-direction. The detection infrared light is at the same Y-coordinate position as the axis of the rotating tool deviating from the reference axis. The detection infrared light is a beam of light with a tiny diameter. A grating position sensor is installed on the z-axis motor of the trolley to monitor the z-axis feed in real time. The z-axis motor of the trolley in the working state is defined as z-coordinate 0 point, and the coordinate value of the spring reset position point is +z. The first step is that the trolley moves downwards in the z-direction by a very small displacement, defined as Δz, until the light is blocked. The y-axis drive motor moves a very small distance in the +y direction and a very small distance in the -y direction, defined as Δy, and each is run once. If the light is connected at one of these times, the trolley moves downwards in the z-direction by a very small displacement, and at the same time, the y-axis drive motor moves a very small distance towards the side where the light is connected. The trolley continues to move downwards in the z-direction by a very small displacement. The displacement cycle continues until the y-axis drive motor moves a very small distance in the +y direction and then a very small distance in the -y direction, each cycle repeating once. The light beam is activated in both cycles, at which point the outermost point of the rotating blade and the detection infrared beam can be identified, which is the blade tip. The first step is to record the displacement y1 driven by the y-axis drive motor at this point. This value is calculated from the total number of rotations of the y-axis motor and is defined as the distance from the blade tip to the detection infrared beam, a vector value of the y-coordinate. This value represents the blade's deviation in the y-direction. The third step is to actively rotate the blade 90°. Specifically, the left hand holds the reference axis of the rotating blade, the thumb points in the z+ direction, and the four fingers indicate the rotation direction. The first step continues. The fourth step is to record the displacement y2 driven by the y-axis drive motor at this point. This value is calculated from the total number of rotations of the y-axis motor and is defined as the distance from the blade tip to the detection infrared beam, a vector value of the y-coordinate. In reality, this value represents the blade's deviation in the x-direction before rotation.

[0053] In one implementation, the first preset movement operation includes:

[0054] The first preset movement operation is that the z-axis motor moves downwards gradually with a displacement step size Δz, and the downward movement of the z-axis motor is monitored in real time by a grating position sensor until the infrared light is blocked by the rotating tool.

[0055] In one implementation, the values ​​of Δy and Δz range from 0.001mm to 0.01mm.

[0056] In one implementation, the second preset movement operation includes:

[0057] After the z-axis motor moves downward by Δz each time, the detection and adjustment motor moves Δy in the +y direction from the current coordinate as the origin. If the light from the grating position sensor and the rotating tool is connected after the movement, the origin is updated to the coordinates of the detection and adjustment motor after the movement. Otherwise, the detection and adjustment motor returns to the origin and moves Δy in the -y direction from the origin, updating the origin to the coordinates of the detection and adjustment motor after the movement.

[0058] In one implementation, the preset rotation angle includes:

[0059] The preset rotation angle is 90 degrees for rotating the cutting tool.

[0060] In one implementation, the preset detection requirements include:

[0061] The preset testing requirement is that the light can be conducted when the motor moves in both directions.

[0062] In one embodiment, step 1 is followed by:

[0063] The rotary tool is rotated according to the target's three-dimensional coordinate system to obtain full-angle offset data;

[0064] The full-angle offset data is divided into sub-data segments according to the rotation angle range;

[0065] Perform frequency domain transformation on each group of sub-data segments in the sub-data segment set to obtain the sub-data segment spectrum distribution set;

[0066] The fundamental wave feature set of each sub-data segment is obtained by identifying the amplitude and phase characteristics of the fundamental wave component in the spectral distribution of each group of sub-data segments in the spectral distribution set;

[0067] By fusing and analyzing the fundamental component characteristics of multiple sub-data segments, calibration deviation values ​​with spatial globality are obtained.

[0068] In one implementation, a dedicated pre-filtering and transform parameter dynamic adjustment dual module is first matched for each sub-data segment. The pre-filtering stage does not use a uniform cutoff frequency, but generates a variable bandwidth filter that varies with local data characteristics by analyzing the angular span and offset fluctuation density of the sub-data segment. Then, at the transform parameter level, the Fourier transform with a fixed number of sampling points is abandoned, and a dynamic balance mechanism between angular resolution and spectral accuracy is introduced. The order of the Fourier transform is automatically adjusted according to the angular interval length of the sub-data segment. For example, a 1024th order transform is used for a 10° short interval sub-segment to improve local accuracy, while a 256th order transform is used for a 60° long interval sub-segment to balance computational efficiency. At the same time, a transition correction layer based on wavelet packet decomposition is embedded to adaptively compensate for the spectral leakage caused by the Fourier transform when sampling in non-integer periods. The compensation coefficient can be dynamically calculated by the continuity difference between the first and last data of the sub-segment. Finally, after processing by this framework, the output spectral distribution of each sub-data segment not only contains traditional frequency-amplitude information, but also adds local confidence labels, forming a sub-data segment spectral distribution set that combines accuracy, adaptability, and interpretability.

[0069] In one implementation, when identifying the fundamental wave feature of the spectral distribution of each group of sub-data segments, the limitations of traditional single-threshold screening of the fundamental wave are overcome, and a multi-dimensional feature-anchored dynamic confidence iteration two-layer identification framework is constructed. First, in the feature anchoring layer, the identification logic that relies solely on frequency peaks is abandoned. Instead, three correlation features related to the theoretical value of the tool rotation fundamental frequency in the spectrum are extracted simultaneously. These include candidate peaks on the frequency axis whose deviation from the fundamental frequency is within a preset threshold, the energy concentration within 5% of the frequency bandwidth to the left and right of the peak, and the coupling degree between the phase value corresponding to the peak and the starting angle of the sub-data segment. Next, in the dynamic confidence iteration layer, initial weights are assigned to the three features and the comprehensive confidence is calculated. If the initial confidence is lower than the threshold, an adaptive iteration mechanism is activated: for the spectrum with low energy concentration, the bandwidth is automatically expanded to 10%, the energy proportion is recalculated, and its weight is increased; for the spectrum with abnormal phase coupling, the phase trend of adjacent sub-data segments is introduced for correction, and the weights of the correlation features are adjusted synchronously. Finally, when the overall confidence level stably exceeds the threshold, the peak value is locked as the fundamental component, and its amplitude, which has been corrected by energy concentration, and its phase, which has been iteratively calibrated by coupling degree, are output simultaneously.

[0070] In one implementation, full-angle offset data is acquired by rotating the rotary tool. This data is then divided into sub-data segments according to the rotation angle interval. Frequency domain transformation is performed on each sub-data segment to obtain a spectral distribution set. The amplitude and phase characteristics of the fundamental wave component of each spectral distribution are identified to form a fundamental wave feature set. Multiple sets of fundamental wave features are then fused and analyzed to obtain a calibration deviation value with spatial globality. The advantage of this approach is that it overcomes the limitations of previous methods that only detected specific positions and single 90-degree rotation angles and could only acquire local displacement data. By acquiring full-angle offset data and dividing the intervals, the detection range is expanded from local to full-angle space. Frequency domain transformation and fundamental wave feature identification eliminate random errors and uncover the essential laws of offset data, improving the scientific nature of deviation calibration. At the same time, the constructed spatial globality calibration deviation value provides a comprehensive and reliable benchmark for subsequent accurate correction of x and y axis displacement values, avoiding the problem of insufficient detection accuracy caused by not considering the global systematic deviation.

[0071] In one embodiment, step 3 is followed by:

[0072] Temporal and spatial features are obtained by extracting time series and spatial distribution features from the calibration deviation values;

[0073] The deviation correlation matrix is ​​obtained by performing relationship analysis based on temporal and spatial characteristics;

[0074] The deviation correction amount is obtained by performing matrix operations on the y-axis displacement value and the deviation correlation matrix.

[0075] The correction value is obtained by iteratively calibrating the y-axis displacement value based on the deviation correction amount;

[0076] The residual deviation value is obtained by calculating the residual deviation between the correction value and the calibration deviation value;

[0077] The residual deviation is quantified to obtain the compensation value;

[0078] The precise deviation value of the y-axis is obtained by adjusting the correction value based on the compensation value.

[0079] In one implementation, firstly, time series features are extracted by introducing a multi-scale wavelet packet deep learning fusion module. After decomposing the calibration deviation time series into subsequences of different scales, the subsequences are input into a customized spatiotemporal-aware convolutional neural network. This network embeds a spatial angle encoding layer, enabling the time series analysis to associate corresponding spatial angle information and extract multi-dimensional time features containing time-domain trend scale-specific spatial correlation. Secondly, for spatial distribution feature extraction, an angle grid feature topology mapping mechanism is designed to divide the full-angle space into triangular grid cells with topological correlation. For the calibration deviation value within each grid cell, a local feature manifold is constructed using a manifold learning algorithm, and the topological similarity between manifolds is calculated as a spatial correlation feature. At the same time, a spatial phase field strength map is constructed by combining the angular distribution of the fundamental phase.

[0080] In one implementation, the time series is first divided into sliding window blocks and core features such as trend inflection points and periodic correlations are extracted. The spatial elements are then quantified by topological distance and divided into regions by clustering. An attention mechanism is then introduced to assign differentiated weights to features in different time stages and spatial regions. The dual-dimensional features are embedded into a high-dimensional correlation space through an improved kernel function mapping. Finally, the correlation matrix that can accurately represent the coupling relationship between time and spatial deviations is generated through feature interaction and iterative calculation.

[0081] In one implementation, the systematic error, random error and coupled error components in the residual deviation are first decomposed by variational mode decomposition technology. Then, a Bayesian inference model is introduced to dynamically update the confidence weight of each error component to highlight the contribution of key deviations. By combining error propagation path tracing, a nonlinear mapping relationship between deviation and compensation is established. The optimal compensation parameters are solved by an improved adaptive particle swarm optimization algorithm, and finally, an accurate compensation value that can reverse the multi-source coupled deviation is generated.

[0082] In one implementation, the time series and spatial distribution features of the calibration deviation values ​​are first extracted and their relationship is analyzed to obtain a deviation correlation matrix. Then, the deviation correction amount of the y-axis displacement value is obtained through matrix operations. The y-axis displacement value is iteratively calibrated to obtain the correction value. The residual deviation between the correction value and the calibration deviation value is calculated and quantified as a compensation value. Finally, the precise deviation value of the y-axis is obtained. Its advantage is that it makes up for the shortcomings of the previous method, which only obtained the basic y-axis displacement value and the global calibration deviation value, did not establish the spatiotemporal correlation of the deviation, and did not perform precise correction of the y-axis displacement. Through spatiotemporal feature extraction and deviation correlation matrix construction, the variation law of the deviation in the time dimension and the distribution correlation in the spatial dimension are clearly sorted out, making the deviation correction more targeted. With the dual optimization of iterative calibration and residual deviation compensation, systematic and random errors in the y-axis displacement data are effectively eliminated, significantly improving the accuracy of y-axis displacement detection and providing more reliable parameter support for subsequent mechanical motion control based on precise y-axis data.

[0083] In one embodiment, step 5 is followed by:

[0084] The static and dynamic deviation components are obtained by analyzing the calibration deviation values.

[0085] The static displacement value is obtained by performing basic compensation correction on the calibration deviation value based on the static deviation component.

[0086] The dynamic displacement value is obtained by dynamically compensating and correcting the calibration deviation value based on the rotation angle and dynamic deviation component.

[0087] The static and dynamic displacement values ​​are fused to obtain the precise x-axis deviation value.

[0088] In one implementation, when analyzing the calibration deviation value, a time-amplitude dual-threshold filtering model is constructed to separate the constant component of the deviation value that does not change with time into a static deviation component, while extracting the time-varying component that fluctuates dynamically with the operating conditions into a dynamic deviation component. Specifically, this can be expressed as follows: ,in To calibrate the deviation value, For static deviation components, The dynamic deviation component varies with time t, thus decoupling the dynamic and static characteristics of the deviation. When performing basic compensation correction based on the static deviation component, a static compensation coefficient k1 is introduced, and a linear correction formula is used. The static displacement value is calculated, where S1 is the displacement value after static compensation. By offsetting the constant static deviation, the basic accuracy for subsequent dynamic compensation is established. k1 is determined through calibration experiments, reflecting the weighting of the static deviation on the displacement. When performing dynamic compensation correction based on the rotation angle and dynamic deviation components, the rotation angle is considered. To modulate the dynamic deviation, a dynamic compensation model is constructed. ,in An orthogonal derivative term of the dynamic deviation component is used to reflect the lateral dynamic effect caused by rotation. This model couples the rotation angle with the spatiotemporal characteristics of the dynamic deviation, outputting a dynamic displacement value S2 to achieve precise cancellation of time-varying deviations. When fusing static and dynamic displacement values, a spatiotemporal weighting factor is introduced. A weighted fusion formula is adopted. ,in, Dynamically adjusted over time, with static periods dominating. Approaching 1, the dynamic dominant period approaches 0. To provide a precise x-axis deviation value, weights are dynamically allocated to ensure that the fusion result can balance static basic accuracy and dynamic real-time correction effect under different working conditions, ultimately outputting a high-precision x-axis deviation value.

[0089] In one implementation, the calibration deviation value is first analyzed to obtain the static deviation component and the dynamic deviation component. Then, the calibration deviation value is corrected by basic compensation based on the static deviation component to obtain the static displacement value. The dynamic displacement value is obtained by combining the rotation angle and the dynamic deviation component. Finally, the static and dynamic displacement values ​​are merged to obtain the accurate x-axis deviation value. The advantage of this approach is that it makes up for the shortcomings of only completing the accurate correction of the y-axis without specifically compensating for the x-axis displacement data in combination with the characteristics of the deviation component. By separating the static and dynamic deviation components, the influence of static factors such as inherent equipment errors on x-axis detection is eliminated, and the dynamic error caused by the change of rotation angle is accurately adapted. This avoids the problem that a single compensation method cannot take into account both static and dynamic deviations. It makes the x-axis displacement detection accuracy match the y-axis high-precision level, further improving the accuracy of the entire collaborative detection scheme and providing more comprehensive and reliable dual-axis accurate data support for the subsequent precision motion control of the rotating tool.

[0090] The foregoing has described one embodiment of the present invention in detail, but this content is merely a preferred embodiment and should not be considered as limiting the scope of the present invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the scope of the claims of this invention.

Claims

1. A non-contact optical positioning and precision mechanical motion coordinated detection method, comprising a blade control machine, an infrared light emitter, a rotary cutting tool, a grating position sensor, and a carriage, wherein the blade control machine includes a detection and adjustment motor and a Z-axis motor, the grating position sensor is mounted on the Z-axis motor, both the detection and adjustment motor and the Z-axis motor are connected to the carriage, the detection and adjustment motor drives the carriage to move left and right, and the infrared light emitter and the reference axis of the rotary cutting tool are located at the same y-coordinate 0 point, characterized in that... The method includes: Step 1: Construct a target three-dimensional coordinate system; the x-axis of the target three-dimensional coordinate system is the transverse coordinate axis of the rotation plane of the rotating tool, the y-axis is the coordinate axis for the initial alignment of the infrared light, the reference axis, and the axis of the rotating tool, and the z-axis is the coordinate axis for the z-axis motor to drive the carriage to move in the vertical direction; Step 2: Perform a first preset movement operation on the z-axis motor, and a second preset movement operation on the detection and adjustment motor until the rotating tool senses the grating position sensor, then determine the current coordinate as the final correction position; the rotating tool is on the y-axis; Step 3: Record the displacement of the motor to obtain the y-axis displacement value; Step 4: Rotate the rotating tool according to the preset rotation angle until the rotating tool meets the preset detection requirements; Step 5: Execute step 2 again. This time, the rotating tool is on the x-axis. Record the displacement of the detection and adjustment motor to obtain the x-axis displacement value. Step 1 is followed by: The rotary tool is rotated according to the target three-dimensional coordinate system to obtain full-angle offset data; The full-angle offset data is divided into sub-data segment sets according to the rotation angle interval; Frequency domain transformation is performed on each group of sub-data segments in the sub-data segment set to obtain the sub-data segment spectrum distribution set; The fundamental wave feature set of each sub-data segment is obtained by identifying the amplitude and phase characteristics of the fundamental wave component in the spectral distribution of each group of sub-data segments in the spectral distribution set. By fusing and analyzing the fundamental component characteristics of multiple sub-data segments, a calibration deviation value with spatial globality is obtained. Step 3 is followed by: The time series and spatial distribution features of the calibration deviation values ​​are extracted to obtain temporal and spatial features; Based on the temporal and spatial characteristics, a deviation correlation matrix is ​​obtained through relationship analysis. The deviation correction amount is obtained by performing matrix operations on the y-axis displacement value and the deviation correlation matrix; The correction value is obtained by iteratively calibrating the y-axis displacement value based on the deviation correction amount; The residual deviation value is obtained by calculating the residual deviation between the correction value and the calibration deviation value; The residual deviation is quantified to obtain the compensation value; The correction value is adjusted based on the compensation value to obtain the precise deviation value of the y-axis; Step 5 is followed by: The static deviation component and the dynamic deviation component are obtained by analyzing the calibration deviation value; The static displacement value is obtained by performing basic compensation correction on the calibration deviation value based on the static deviation component. The dynamic displacement value is obtained by dynamically compensating and correcting the calibration deviation value based on the rotation angle and the dynamic deviation component. The static displacement value and the dynamic displacement value are fused to obtain the accurate x-axis deviation value.

2. The method for co-detection of non-contact optical positioning and precision mechanical motion according to claim 1, characterized in that, The first preset movement operation includes: The z-axis motor moves downwards incrementally in steps Δz, and the downward movement of the z-axis motor is monitored in real time by a grating position sensor until the infrared light is blocked by the rotating tool.

3. The method for co-detection of non-contact optical positioning and precision mechanical motion according to claim 2, characterized in that, The second preset movement operation includes: After the z-axis motor moves downward by Δz each time, the detection and adjustment motor moves Δy in the +y direction from the current coordinate as the origin. If the light from the grating position sensor and the rotating tool is connected after the movement, the origin is updated to the coordinates of the detection and adjustment motor after the movement. Otherwise, the detection and adjustment motor returns to the origin and moves Δy in the -y direction from the origin, updating the origin to the coordinates of the detection and adjustment motor after the movement.

4. The method for coordinated detection of non-contact optical positioning and precision mechanical motion according to claim 1, characterized in that, Preset rotation angles include: The preset rotation angle is 90 degrees for rotating the cutting tool.

5. The method for co-detection of non-contact optical positioning and precision mechanical motion according to claim 1, characterized in that, Preset testing requirements include: The preset testing requirement is that the light can be conducted when the motor moves in both directions.

Images