A precision detection method and system for an optical encoder

Abstract

The application discloses a precision detection method and system for an optical encoder, and relates to the technical field of electric digital data processing.The method comprises the following steps: acquiring rotation data and bottom layer parameters under the same working condition, wherein the bottom layer parameters comprise deflection data and spectrum data; acquiring a gap modulation amount according to the deflection data, acquiring a dispersion characteristic according to the spectrum data, and acquiring a steady-state equivalent according to the gap modulation amount and the dispersion characteristic; acquiring a quadrature signal and a frame timestamp according to the rotation data, acquiring a track characteristic according to the quadrature signal, and acquiring a jitter variance according to the frame timestamp; acquiring a compensation parameter according to the steady-state equivalent, the track characteristic and the jitter variance, and associating the compensation parameter to a control node.The application has the advantages of comprehensive coverage, dynamic compensation and unified dimension.

Description

Technical Field

[0001] This invention relates to the field of electronic digital data processing technology, and specifically to a method and system for accuracy detection of photoelectric encoders. Background Technology

[0002] In existing industrial control and precision measurement applications, photoelectric encoders are often installed on motor spindles or moving joints, and the stability of their output accuracy affects the positioning and speed control quality of servo systems. However, existing accuracy evaluation models are usually limited to single static error measurements or mathematical environmental coefficient weighting, failing to reveal the underlying failure mechanisms of photoelectric encoders caused by strong coupling of multiple physical fields under actual operating conditions.

[0003] Specifically, under high-speed variable load conditions, the rotating shaft system will experience dynamic torsion and microscopic deformation. This mechanical deflection will cause nonlinear fluctuations in the axial gap between the code disk surface inside the photoelectric encoder and the photoelectric receiving array, thereby changing the spatial distribution pattern of the interference fringes. At the same time, continuous high-load operation will cause the junction temperature of the internal light-emitting diodes to rise, causing a thermal redshift of the center emission wavelength. This optical dispersion distortion, combined with the mechanical gap fluctuations, will destroy the pitch and light intensity contrast of the steady-state moiré fringes. In addition, high-frequency mechanical oscillations will not only cause distortion at the optical end, but will also be transmitted to the electrical signal layer, causing phase shift and amplitude attenuation of the two analog orthogonal signals output by the photoelectric receiving array. This will cause the ideal circular vector synthesis trajectory in the phase plane to deteriorate into an ellipse, destroying the spatial orthogonality of the signal. Furthermore, this underlying electromechanical disturbance will also be transmitted to the digital communication protocol layer, causing time lag in the internal processor when performing zero-crossing calculations and state machine switching. This will ultimately manifest as nonlinear jitter in the data frame arrival timestamp on the communication bus.

[0004] Therefore, existing accuracy detection and compensation mechanisms often treat mechanical deformation, optical temperature drift, electrical phase difference and protocol communication delay separately, lacking unified mapping and collaborative decoupling capabilities for multi-physics nonlinear coupling errors. As a result, when faced with complex working conditions of strong vibration, large temperature rise and variable load superposition, they cannot provide servo drive systems with real-time and dimensionally consistent dynamic angular displacement feedforward compensation parameters. Summary of the Invention

[0005] In view of the technical problems described in the background art, the present invention provides a method and system for accuracy detection of photoelectric encoders.

[0006] A method for accuracy detection of an optical encoder includes: acquiring rotational data and underlying parameters under the same operating condition, the underlying parameters including deflection data and spectral data; obtaining gap modulation based on the deflection data and dispersion characteristics based on the spectral data, and obtaining a steady-state equivalent based on the gap modulation and dispersion characteristics; acquiring orthogonal signals and frame timestamps based on the rotational data, obtaining trajectory characteristics based on the orthogonal signals, and obtaining jitter variance based on the frame timestamps; obtaining compensation parameters based on the steady-state equivalent, trajectory characteristics, and jitter variance, and associating the compensation parameters with a control node.

[0007] Optionally, the rotational data and underlying parameters under the same operating condition are obtained, including: obtaining continuous frames based on the underlying raw data and using the continuous frames as rotational data; obtaining stator current, obtaining electromagnetic torque based on torque constant and stator current, obtaining deformation based on elastic modulus, geometric dimensions and electromagnetic torque, and using the deformation as deflection data; obtaining voltage drop data, obtaining temperature rise data based on voltage drop data, and using the temperature rise data as spectral data.

[0008] Optionally, the gap modulation amount is obtained based on the deflection data, and the dispersion characteristics are obtained based on the spectral data. The steady-state equivalent is obtained based on the gap modulation amount and the dispersion characteristics, including: obtaining the gap displacement based on the deflection data and the flange distance, and using the gap displacement as the gap modulation amount; obtaining the offset based on the temperature rise data and the temperature drift coefficient, and using the offset as the dispersion characteristics; and adding the product of the gap modulation amount and the sensitivity constant, and the ratio of the sum of the center wavelength and the offset to the grating constant, to obtain the steady-state equivalent.

[0009] Optionally, obtaining orthogonal signals and frame timestamps based on rotation data includes: obtaining two signals with a preset phase difference angle based on rotation data, using them as a first signal and a second signal respectively, and using the first signal and the second signal as orthogonal signals; obtaining the arrival time of adjacent data frames based on rotation data, and using the arrival time as a frame timestamp.

[0010] Optionally, the trajectory features are obtained based on the orthogonal signals, including: using the first signal as the abscissa and the second signal as the ordinate, fitting a feature ellipse in the phase plane; obtaining the major axis length and minor axis length based on the feature ellipse, calculating the ratio of the square of the minor axis length to the square of the major axis length, and taking the square root of the result after subtracting the ratio to obtain the trajectory features.

[0011] Optionally, the jitter variance is obtained based on the frame timestamp, including: calculating the time difference between adjacent frame timestamps and using the time difference as the actual period; calculating the mean of the squares of the difference between the actual period and the nominal period and using the mean of the squared differences as the jitter variance.

[0012] Optionally, compensation parameters are obtained based on steady-state equivalent, trajectory characteristics, and jitter variance, including: obtaining scaling constant, electrical coefficient, and protocol coefficient; multiplying the steady-state equivalent by the scaling constant to obtain a first product; multiplying the trajectory characteristics by the electrical coefficient to obtain a second product; multiplying the jitter variance by the protocol coefficient to obtain a third product; and adding the first, second, and third products to obtain the compensation parameters.

[0013] A precision detection system for photoelectric encoders is also provided. The system includes: an acquisition module configured to acquire rotational data and underlying parameters under the same working condition, the underlying parameters including deflection data and spectral data; a mapping module configured to acquire gap modulation based on deflection data, and dispersion characteristics based on spectral data, and steady-state equivalent based on gap modulation and dispersion characteristics; an extraction module configured to acquire orthogonal signals and frame timestamps based on rotational data, acquire trajectory characteristics based on orthogonal signals, and acquire jitter variance based on frame timestamps; and an execution module configured to acquire compensation parameters based on steady-state equivalent, trajectory characteristics, and jitter variance, and associate the compensation parameters with control nodes.

[0014] Optionally, the acquisition module is also configured to: acquire continuous frames based on the underlying raw data and use the continuous frames as rotation data; acquire stator current, acquire electromagnetic torque based on torque constant and stator current, acquire deformation based on elastic modulus, geometric dimensions and electromagnetic torque, and use the deformation as deflection data; acquire voltage drop data, acquire temperature rise data based on voltage drop data, and use the temperature rise data as spectral data.

[0015] Optionally, the mapping module is also configured to: obtain the gap displacement based on the deflection data and flange distance, and use the gap displacement as the gap modulation amount; obtain the offset based on the temperature rise data and temperature drift coefficient, and use the offset as the dispersion feature; and add the product of the gap modulation amount and the sensitivity constant, and the ratio of the sum of the center wavelength and the offset to the grating constant to obtain the steady-state equivalent.

[0016] The beneficial effects of this invention are reflected in: In the entire accuracy detection method for photoelectric encoders, firstly, by acquiring the underlying spectral parameters including the dynamic deflection of the shaft system and the temperature rise of the light-emitting components, the physical boundary conditions under variable load and thermal accumulation conditions are obtained, providing a basic physical benchmark for subsequent error tracing. Secondly, by equivalently mapping the axial clearance displacement at the mechanical level with the emission wavelength shift at the optical level, the dimensionless steady-state spatial moiré fringe distortion equivalent is derived, quantifying the degree of microscopic distortion of the interference fringes under the machine-optical coupling effect, avoiding the singleness of existing static geometric error measurements. Based on this, the method delves deeper into the electrical layer and digital coordination... The protocol layer quantifies the degree of orthogonality degradation caused by high-frequency vibration by fitting an orthogonal composite feature ellipse in a two-dimensional phase plane and extracting the ratio of its minor and major axes. Simultaneously, it extracts the jitter variance reflecting clock modulation effects by calculating the mean of the squared difference between the actual communication cycle and the nominal cycle, thus measuring dynamic response hysteresis. Finally, relying on preset spatial scaling constants, electrical coefficients, and protocol coefficients, the dimensionless steady-state equivalent, trajectory characteristics, and time-dimensional jitter variance are reduced in dimension and mapped to feedforward compensation parameters with angular displacement dimensions, which are then associated with the position loop control node of the servo driver. This closed-loop decoupling and multi-dimensional scale mapping mechanism not only achieves dimensional unification across physical fields in the underlying mathematical expression but also adaptively cancels position feedback deviations caused by composite disturbances in the servo electromagnetic circuit, improving the dynamic measurement accuracy of the photoelectric encoder under complex operating conditions. Attached Figure Description

[0017] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0018] Figure 1 This is a schematic diagram of the steps in the accuracy detection method for an optical encoder according to the present invention; Figure 2 This is a schematic diagram of a portion of step S1 in the accuracy detection method for photoelectric encoders of the present invention; Figure 3 This is a schematic diagram of a portion of step S2 in the accuracy detection method for photoelectric encoders of the present invention; Figure 4 This is a schematic diagram of a portion of step S3 in the accuracy detection method for photoelectric encoders of the present invention; Figure 5 This is a schematic diagram of a portion of step S4 in the accuracy detection method for photoelectric encoders of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0020] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0021] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0022] This invention provides a method for accuracy detection of an optical encoder, such as... Figure 1 As shown, in one embodiment, the method specifically includes: S1. Obtain rotational data and underlying parameters under the same working condition. The underlying parameters include deflection data and spectral data. S2. Obtain the gap modulation amount based on the deflection data, and obtain the dispersion characteristics based on the spectral data. Obtain the steady-state equivalent based on the gap modulation amount and dispersion characteristics. S3. Obtain orthogonal signals and frame timestamps based on rotation data, obtain trajectory features based on orthogonal signals, and obtain jitter variance based on frame timestamps; S4. Obtain compensation parameters based on steady-state equivalent, trajectory characteristics and jitter variance, and associate the compensation parameters with the control node.

[0023] In this embodiment, it should be noted that in S1, the actual underlying physical state parameters of the device under complex working conditions are obtained from the source of opto-mechatronic coupling, solving the problem that existing detection methods rely only on surface environmental data or static parameters and ignore the internal dynamic load and thermal accumulation effects. In specific applications, when the CNC milling machine spindle is running at a high-speed variable load of 15,000 revolutions per minute, it first continuously receives undecoded continuous frame sequences as rotation data through the communication interface. At the same time, the real-time stator current of the spindle servo motor is collected, for example, if the stator current is measured to be 18.5 amperes, combined with the fixed motor torque constant and rotor elastic modulus and other mechanical parameters, the dynamic deformation of the rotating shaft system under the current cutting stress is calculated through the physical force model and used as deflection data. In addition, the forward voltage drop reduction value of the light-emitting diode inside the photoelectric encoder is monitored synchronously, and the real-time junction temperature rise data is calculated accordingly, for example, if the temperature rise is 20 Kelvin, and this is used as spectral data input into the underlying model. This data acquisition logic is no longer limited to monitoring the external macro environment, but goes deep into the junction temperature change of the light-emitting components and the micro-torsion of the main shaft. It objectively records the physical boundary conditions of the equipment when it is running at high speed, providing basic data support for the subsequent in-depth tracing and coupling analysis of cross-physical field errors, and enabling the accuracy assessment to have a real underlying physical benchmark.

[0024] In S2, the acquired underlying physical parameters are transformed into optical distortion indices, resolving the problem of mechanical deformation and optical temperature drift being treated separately, thus preventing a comprehensive quantification of interference fringe degradation. Based on previously acquired deflection data and the encoder's preset flange distance, a geometric mapping is performed to calculate the axial gap displacement between the code disk surface and the photoelectric receiving array, for example, establishing a gap modulation amount of 2.5 micrometers. Simultaneously, combining a junction temperature rise of 20 Kelvin and a calibrated temperature drift coefficient of 0.2 nanometers per Kelvin, the numerical offset of the nominal 850-nanometer center emission wavelength of the LED is calculated as a dispersion characteristic. Logically, the gap modulation amount is multiplied by a preset optical sensitivity constant of 0.02 per micrometer to obtain the attenuation term of contrast due to mechanical disturbance; then, the sum of the center wavelength and the offset is divided by a grating constant of 20,000 nanometers to obtain the effect of wavelength drift on the pitch; finally, the two terms are added together to obtain a steady-state spatial moiré fringe distortion equivalent of 0.0927. This approach transforms the invisible mechanical deflection and thermal redshift into dimensionless unified parameters, objectively quantifying the microscopic distortion state of interference fringes under machine-optical coupling. It overcomes the limitations of existing single static geometric error measurements and enables the measurement of the degree of variation in optical measurement references.

[0025] In S3, the main focus is on feature extraction for high-frequency dynamic distortion generated under high-speed operation, solving the problem that existing evaluation mechanisms cannot measure the damage to the orthogonality of electrical signals and the latency of digital protocols caused by mechanical vibration. Under a vibration environment of 15,000 revolutions per minute, two analog orthogonal signals with a 90-degree phase difference are simultaneously acquired, and a feature ellipse is fitted in a two-dimensional phase plane, for example, with a major axis length of 150 millivolts and a minor axis length of 120 millivolts. By calculating the square of the ratio of the minor axis to the major axis, subtracting this ratio from one, and then taking the square root, a trajectory feature with a value of 0.6 is obtained. Simultaneously, the timestamps of adjacent frames arriving in the rotation data are extracted to calculate the actual communication period, and the mean of the squared differences of the deviation from the nominal 50 microsecond period is calculated, yielding a jitter variance of 120 square nanoseconds. This calculation logic quantifies the degree of degradation of the orthogonality of the analog signal from the eccentricity dimension of the geometric shape, and extracts the time dispersion caused by the internal processor clock being disturbed by oscillations from the statistical variance dimension. This extends the coverage of accuracy detection from the analog front end to the digital communication layer, enabling cross-level extraction of dynamic response hysteresis and spatial orthogonality degradation, and providing multi-dimensional feature data for constructing a compensation model for high-frequency electromechanical response.

[0026] In S4, multiphysics decoupling and final compensation are implemented, resolving the issue of inconsistent error dimensions across different dimensions preventing comprehensive real-time correction at the servo's underlying loop. Pre-calibrated transformation parameters for each dimension are acquired, such as a spatial scaling constant of 20 arcseconds, an electrical coefficient of 5 arcseconds, and a protocol coefficient of 0.05 arcseconds per square nanosecond. Logically, the steady-state equivalent of 0.0927 is multiplied by the scaling constant to obtain 1.854 arcseconds, the trajectory characteristic of 0.6 is multiplied by the electrical coefficient to obtain triangular seconds, and the jitter variance of 120 square nanoseconds is multiplied by the protocol coefficient to obtain 6 arcseconds. Finally, these three products are summed to output a dynamic angular displacement feedforward compensation parameter totaling 10.854 arcseconds. This compensation parameter with a unified angular dimension is then written into the feedforward control register of the servo driver. This dimension reduction mapping and decoupled computation logic, without changing the underlying hardware architecture, maps all disturbances in different physical units, such as optical distortion, electrical phase shift, and protocol delay, to standard angular displacement units through specific conversion coefficients. This achieves cross-physical dimension unification in mathematical expression and adaptively cancels position feedback deviations caused by composite physical disturbances in the motor position loop, thereby improving the measurement fidelity of the encoder under complex working conditions.

[0027] In summary, the entire accuracy detection method for photoelectric encoders firstly achieves the acquisition of physical boundary conditions under varying loads and thermal accumulation by obtaining the underlying spectral parameters, including the dynamic deflection of the shaft system and the temperature rise of the light-emitting components. This provides a fundamental physical benchmark for subsequent error tracing. Secondly, by equivalently mapping the axial clearance displacement at the mechanical level with the emission wavelength shift at the optical level, a dimensionless steady-state spatial moiré fringe distortion equivalent is derived. This quantifies the degree of microscopic distortion of the interference fringes under the mechanical-optical coupling effect, avoiding the limitations of existing static geometric error measurements. Based on this, the method delves deeper into the electrical and digital layers. At the protocol layer, the degree of orthogonality degradation caused by high-frequency vibration is quantified by fitting an orthogonal composite feature ellipse in a two-dimensional phase plane and extracting the ratio of its minor and major axes. Simultaneously, jitter variance reflecting clock modulation effects is extracted by calculating the mean of the squared difference between the actual communication cycle and the nominal cycle, thus measuring dynamic response hysteresis. Finally, relying on preset spatial scaling constants, electrical coefficients, and protocol coefficients, the dimensionless steady-state equivalent, trajectory characteristics, and time-dimensional jitter variance are reduced in dimension and mapped to feedforward compensation parameters with angular displacement dimensions, which are then associated with the position loop control node of the servo driver. This closed-loop decoupling and multi-dimensional scale mapping mechanism not only achieves dimensional unification across physical fields in the underlying mathematical expression but also adaptively cancels position feedback deviations caused by composite disturbances in the servo electromagnetic circuit, improving the dynamic measurement accuracy of the photoelectric encoder under complex operating conditions.

[0028] like Figure 2 As shown, in one specific embodiment, S1 includes: S11, obtaining continuous frames based on the underlying raw data, and using the continuous frames as rotation data. Here, the underlying raw data refers to the undecoded bit stream transmitted by the photoelectric encoder through the physical interface, and the continuous frame refers to the complete data packet sequence output in chronological order within a continuous rotation cycle.

[0029] S12. Obtain the stator current, and obtain the electromagnetic torque based on the torque constant and stator current. Obtain the deformation based on the elastic modulus, geometric dimensions, and electromagnetic torque, and use the deformation as deflection data. Here, the torque constant, elastic modulus, and geometric dimensions are all inherent mechanical parameters of the motor and rotor shafts; the deflection data is used to characterize the micron-level physical torsion and bending generated by the rotating shaft system under load.

[0030] The method for determining the torque constant is primarily based on historical dynamometer test data from a limited number of tests conducted on the servo motor under load. In practice, the servo motor under test is mounted alongside a high-precision magnetic particle dynamometer. The motor is instructed to operate with a step-given stator current, for example, from five amperes to twenty-five amperes, in two-ampere increments, acquiring historical data for eleven test nodes. The actual output mechanical torque fed back by the dynamometer at each steady-state current node is recorded synchronously. Subsequently, the input stator current is used as the independent variable, and the measured mechanical torque as the dependent variable. A linear regression algorithm is used to fit their corresponding relationship. The proportionality coefficient in this fitted relationship is extracted, and after removing bearing friction loss and core hysteresis loss terms during data processing, it is finally calibrated as the torque constant. This method provides objective parameter support for inverting mechanical shaft deflection from underlying electrical signals.

[0031] S13. Obtain voltage drop data, and obtain temperature rise data based on the voltage drop data. Use the temperature rise data as spectral data. Here, voltage drop data refers to the decrease in forward voltage of the light-emitting diode inside the photoelectric encoder; spectral data is used as a reference to characterize the wavelength drift of the light-emitting components caused by temperature.

[0032] In this embodiment, it should be noted that in S11, the underlying data stream foundation for dynamically acquiring data is constructed, solving the problems of insufficient data sampling rate and lag in dynamic tracking under high-speed conditions in existing detection mechanisms. Specifically, when the spindle of the CNC milling machine under test is running at a high speed range of 15,000 revolutions per minute, the undecoded bit stream transmitted from the physical layer is continuously received through the high-speed synchronous serial interface configured on the photoelectric encoder. This data acquisition method bypasses the filtering and interpolation algorithms of the upper-layer control, directly acquiring the complete data packet sequence output in chronological order within the encoder's continuous rotation cycle, defining it as rotation data. By capturing the most original continuous frames, the slight nonlinear fluctuation characteristics caused by the high-frequency rotation of the spindle can be preserved. This step provides an undistorted input source for subsequent error coupling analysis, ensuring the data fidelity of mechanical, optical, and electrical condition assessments, and enabling accuracy detection to be built on a data foundation that truly reflects the real-time operating conditions of the equipment.

[0033] In S12, the internal mechanical micro-deformation, which is difficult to measure directly, is transformed into a calculable parameter, solving the problem that the shaft torsion of the encoder cannot be quantified in a closed installation state. Under high-speed variable load conditions of 15,000 revolutions per minute, directly installing a deformation sensor inside the confined encoder is physically space-constrained. Therefore, the real-time stator current of the spindle servo motor is acquired through a current transformer on the servo drive side, and the stator current is measured to be 18.5 amperes. Subsequently, this current is converted into real-time electromagnetic torque using a pre-preserved motor torque constant, and based on the rotor shaft's elastic modulus and geometry, the torsional deformation formula from mechanics of materials (e.g., ...) is applied. ,in For electromagnetic torque, The axis length is For shear modulus, The torsional and bending deformations of the rotating axis system under the current stress are calculated by inversion using the polar moment of inertia, forming deflection data. This indirect calculation logic uses the principles of electromechanical stiffness mechanics to penetrate physical obstruction, realizing the measurement of mechanical disturbances under variable load conditions, and providing a prerequisite for calculating optical clearance.

[0034] In S13, the true temperature rise of the light-emitting components is obtained from a thermodynamic perspective, overcoming the technical limitation that external ambient temperature measurements cannot represent the actual operating temperature of the core light-emitting element. Under continuous high-load operation, the thermal accumulation effect of the LEDs inside the encoder can cause spectral variations. Instead of the existing approach of placing temperature sensors on the encoder housing, the current method directly monitors the reduction in the forward voltage drop of the LEDs, for example, detecting a reduction of 0.12 volts. Utilizing the physical characteristic that the forward voltage drop of a semiconductor PN junction is linearly negatively correlated with temperature, the real-time junction temperature rise is calculated from this voltage drop change, resulting in a core temperature rise of 20 Kelvin, which is then used as the spectral data input. This logic delves into the semiconductor physics of optoelectronic devices, eliminating measurement errors caused by thermal conduction delays and establishing a fundamental physical measurement benchmark for quantifying optical dispersion phenomena such as thermally induced redshift.

[0035] like Figure 3 As shown, in one specific embodiment, S2 includes: S21, obtaining the gap displacement based on the deflection data and the flange distance, and using the gap displacement as the gap modulation amount. Here, the flange distance is the fixed installation distance of the photoelectric encoder code disk; the gap modulation amount is used to characterize the spatial distance variation along the axial direction between the code disk surface and the photoelectric receiving array.

[0036] S22. Obtain the offset based on the temperature rise data and temperature drift coefficient, and use the offset as a dispersion characteristic. Here, the temperature drift coefficient is the calibration coefficient for the wavelength of the light-emitting diode as a function of temperature; the dispersion characteristic is used to characterize the numerical shift in the central emission wavelength.

[0037] The method for determining the temperature drift coefficient is primarily based on historical spectral data extracted from a limited number of high- and low-temperature alternating experiments. Specifically, LEDs from the same batch inside the photoelectric encoder are placed in a high- and low-temperature test chamber. Within an operating temperature range of 20 to 80 degrees Celsius, thirteen steady-state test nodes are set in increments of five degrees Celsius. After thermal equilibrium is reached at each node, the actual emission wavelength of the diodes is collected using a high-precision spectrometer. For example, the historical process of the center wavelength drifting almost linearly from 850 nm to 862 nm with increasing temperature is recorded. Subsequently, the measured steady-state junction temperature is used as the independent variable, and the corresponding measured center wavelength is used as the dependent variable. A linear fit is performed using the least squares method. The slope of the fitted line is then extracted to establish the temperature drift coefficient. This data processing eliminates the uncertainty of a single measurement and establishes a conversion benchmark for the physical dimensions of thermal redshift.

[0038] S23. Add the product of the gap modulation amount and the sensitivity constant, and the ratio of the sum of the center wavelength and the offset to the grating constant, to obtain the steady-state equivalent. The steady-state equivalent is used to quantify the physical distortion error of the interference fringes under steady-state conditions. Its specific calculation process is as follows:

[0039] in, It is a steady-state equivalent and is dimensionless. The sensitivity constant refers to the influence coefficient of the pre-calibrated optical gap on the contrast of the interference fringes, and its unit is 1000 ppm. ; This is the gap modulation amount, in units of ; The center wavelength refers to the nominal emission wavelength of the light-emitting diode, and the unit is 100 nm. ; This is the temperature drift coefficient, in units of... ; Temperature rise data, unit: ; The grating constant refers to the spacing between adjacent light-transmitting slits on the code disk of a photoelectric encoder, measured in units of... .

[0040] Among them, the sensitivity constant The method for determining the value is primarily based on historical data from a limited number of bench calibration experiments, obtained through curve fitting. Specifically, under a constant temperature and humidity experimental environment, the axial gap of the photoelectric encoder is modulated in steps using a piezoelectric ceramic micro-displacement stage, while simultaneously recording the amplitude contrast of the orthogonal signals output by the photoelectric receiving array at each gap node using a data acquisition card. For example, within the gap operation range of 2.0 micrometers to 3.0 micrometers, 11 sets of measured historical data are acquired in steps of 0.1 micrometers, recording the physical process of the interference fringe contrast linearly decreasing from 0.85 to 0.83. Subsequently, the displacement change is used as the independent variable, and the corresponding contrast decay rate is used as the dependent variable, and linear regression analysis is performed using the least squares method. The absolute value of the slope of the fitted line is extracted and established as the sensitivity constant.

[0041] In this embodiment, it should be noted that in S21, the extracted mechanical deflection is transformed into a key boundary condition for optical measurement, solving the mapping problem between macroscopic mechanical deformation and microscopic optical interference state. The signal quality of the photoelectric encoder is highly dependent on the spatial distance between the code disk and the photoelectric receiving array. Using the deflection data obtained in S12, combined with the fixed geometric parameter of the preset flange distance of the photoelectric encoder, the positional displacement of the internal code disk surface along the rotation axis when the main shaft is subjected to torsion is calculated through a spatial geometric transformation model. In the application scenario, the axial gap displacement is established as 2.5 micrometers, which is defined as the gap modulation amount. This calculation logic projects the external macroscopic mechanical cutting force onto the micrometer-level optical reading gap inside the encoder, intuitively reflecting how mechanical deformation changes the diffraction boundary of the grating slit, thus providing a specific spatial physical scale for evaluating the attenuation of interference fringe contrast.

[0042] In step S22, the thermal temperature rise is converted into a physical shift in the optical wavelength, solving the problem that the change in interference fringe pitch caused by LED heating cannot be detected early. Combining the 20 Kelvin junction temperature rise data obtained in step S13 with the wavelength temperature drift coefficient of 0.2 nanometers per Kelvin calibrated at the LED's manufacturing stage, a numerical shift in the center emission wavelength is extracted using linear multiplication logic. This shift represents the spectral redshift phenomenon caused by the LED at high temperatures and is used as a dispersion characteristic. This conversion logic quantifies the degree to which the thermal accumulation effect alters the physical properties of the light source. This change causes a slight expansion of the angle of the diffracted beam passing through the code disk, leading to a drift in the interference fringe pitch projected onto the receiving array. This step bridges the thermodynamic temperature rise data to optical physical quantities, providing another dimension of parameter for comprehensively evaluating the degradation of the optical reference.

[0043] In S23, dimensionless fusion of mechanical and thermal perturbations at the optical level was achieved, solving the technical bottleneck of the inability to uniformly evaluate errors in the field of mechatronics. Specifically, the calculation expression... A low-level coupled model of optomechanical-thermal multiphysics under steady-state conditions was constructed, and its computation process is designed with rigorous physical mapping logic.

[0044] The first term in the expression The problem addressed is the quantification of the conversion of mechanical disturbances into optical features, among which... It represents the gap modulation amount, which is essentially the change in the actual physical distance between the code disk and the receiving array caused by the microscopic axial deformation of the spindle under variable load conditions, which is transmitted to the inside of the photoelectric encoder. The constant is a sensitivity constant, reflecting the linear propagation slope of the attenuation of interference fringe contrast caused by the spatial gap variation. By multiplying these two, the algorithm establishes a linear approximation model under the first-order Taylor expansion, extracting the purely optical attenuation caused by mechanical deflection. In the specific application scenario, when the calculated gap modulation amount... for And the pre-calibrated sensitivity constant for When, the product of this term is This value physically represents the proportion of the light intensity contrast reaching the photosensitive surface that decreases due to the widening of the gap.

[0045] The second term of the expression The problem addressed is the quantification of the destructive effect of thermal redshift on optical measurement standards, among which... The center wavelength of the light-emitting diode. This is the temperature drift coefficient. For real-time junction temperature rise data, molecular part Following the physical laws governing the change of semiconductor energy bands with temperature, the actual center emission wavelength of the light-emitting element under the current thermal accumulation state was calculated; denominator The grating constant represents the spatial geometric period of adjacent light-transmitting slits. According to the grating diffraction equation in fundamental physics, the fringe pitch distribution is directly related to the incident light wavelength. By dividing the actual emitted wavelength by the constant grating constant, the calculation process converts the wavelength shift at the nanoscale into a relative phase drift ratio with respect to a fixed spatial period.

[0046] In specific operating conditions, when the center wavelength for Temperature drift coefficient for Temperature rise data for At that time, the actual transmission wavelength shift was Divide by the preset Grating constant Then, the value of this item was obtained as follows: The expression uses addition to fuse mechanical and thermal terms. Its underlying logic is that, under small perturbations, the contrast spatial expansion caused by axial clearance and the diffraction angle spatial expansion caused by wavelength drift are considered independent error bases within the phase space of the light intensity distribution, conforming to the principle of linear superposition of errors in physics. The final sum of the two terms yields the value... steady-state equivalent This computational process not only strips away the specific dimensions (micrometers and nanometers) with different physical units and transforms them into dimensionless characteristic parameters, but also solves the technical problem that macroscopic mechanical deformation and microscopic optical wavelength changes cannot be uniformly measured in the same dimension. This allows for the real-time output of objective physical indicators characterizing the degree of optical-mechanical coupling distortion at a lower computational cost, without relying on a large and computationally expensive three-dimensional ray tracing simulation model. This establishes a steady-state numerical benchmark for further decoupling at the electrical level.

[0047] like Figure 4 As shown, in one specific embodiment, S3 includes: S31, acquiring two signals with a preset phase difference angle based on the rotation data, using them as a first signal and a second signal respectively, and treating the first signal and the second signal as orthogonal signals. Here, the preset angle is... Electrical angle.

[0048] S32. Using the first signal as the abscissa and the second signal as the ordinate, fit a characteristic ellipse in the phase plane; obtain the major axis length and minor axis length based on the characteristic ellipse, calculate the ratio of the square of the minor axis length to the square of the major axis length, subtract the ratio from the result, and then take the square root to obtain the trajectory features. The characteristic ellipse refers to the closed trajectory formed by the two signals in a two-dimensional rectangular coordinate system; the trajectory features characterize the degree of orthogonality degradation of the signal phases, and their specific calculation process is as follows:

[0049] in, Trajectory features, dimensionless; The length of the minor axis, in units of ; The length of the major axis, in units of .

[0050] S33. Obtain the arrival time of adjacent data frames based on the rotation data, and use the arrival time as the frame timestamp.

[0051] S34. Calculate the time difference between adjacent frame timestamps and use this time difference as the actual period; calculate the mean of the squares of the difference between the actual period and the nominal period, and use this mean of the squared differences as the jitter variance. Here, the nominal period is the set standard data transmission interval; the jitter variance is used to quantify the time dispersion caused by mechanical vibration interference of the internal clock.

[0052] In this embodiment, it should be noted that in S31, acquiring the underlying electrical analog signal characterizing the dynamic performance of high-speed operation solves the problem that conventional detection only focuses on digital output and ignores the attenuation of the front-end analog link. In a high-frequency oscillation environment of 15,000 revolutions per minute, mechanical shock directly affects the photoelectric receiving array and its signal conditioning circuit. By using a dual-channel high-speed analog-to-digital converter, bypassing the internal digital comparator, two analog quadrature signals with a phase difference of 90 electrical degrees from the output of the photoelectric receiving array are simultaneously acquired. By acquiring the most original first and second signals, high-frequency physical characteristics such as waveform distortion, amplitude attenuation, and phase jitter caused by vibration are preserved. This data contains true information about the encoder's dynamic response capability under harsh operating conditions. This operational logic delves the detection probe into the underlying electrical system before analog-to-digital conversion, providing unalgorithm-modified raw samples for subsequent quantification of dynamic phase shift and analysis of orthogonality degradation mechanisms.

[0053] In S32, the problem of accurately measuring the disruption of electrical orthogonality caused by mechanical vibration was solved. Specifically, the calculation expression... A geometric evaluation model independent of the time-domain variable of rotational speed is proposed for electrical signal distortion under high-frequency dynamic operating conditions. The design of its calculation process is based on the two-dimensional phase plane analysis theory in signal processing.

[0054] Parameters in the expression The length of the major axis of the characteristic ellipse. The length of the minor axis of the characteristic ellipse. Under ideal, interference-free operation, the two analog orthogonal signals output by the photoelectric receiving array should have equal amplitudes and a 90-degree electrical phase difference. Mapping these two signals onto the horizontal and vertical axes of a Cartesian coordinate system, their vector synthesis trajectory should be a standard circle with equal lengths for the major and minor axes. However, under high-speed rotation accompanied by severe mechanical vibration, the receiving array plane inside the encoder will experience high-frequency micro-tilts, leading to differences in the luminous flux received by different photosensitive units. This results in amplitude mismatch and non-orthogonal phase shift between the two signals, causing the synthesized trajectory in the phase plane to degenerate from a circle to an ellipse. The calculation process first executes... The ratio calculation involves squaring the amplitude data of the voltage signal. Physically, this converts the voltage signal into a dimensionless proportion related to energy or power, effectively eliminating common-mode interference caused by fluctuations in the overall luminous intensity of the light source. Subsequently, the following steps are performed: The subtraction operation uses the ideal circular state (where the ratio is one) as the zero point reference to calculate the energy difference between the current state and the ideal state. Finally, the square root operation is performed on this difference, which strictly conforms to the definition of eccentricity of an ellipse in analytic geometry, and remaps the deviation of the energy dimension back to the first-order linear scale.

[0055] In practical milling machine spindle applications, when the acquired electrical signal is fitted to the major axis length of a characteristic ellipse... for minor axis length for When, the ratio of their squares is Subtracting that ratio from one gives the value. Trajectory features obtained after square root processing for This computational process addresses a core technical problem: existing time-domain waveform analysis methods heavily rely on the specific rotational speed of the axis. Every minute fluctuation in speed causes stretching or compression of the time-domain waveform period, resulting in signal distortion from high-frequency vibrations intertwining with period changes caused by speed variations, making it difficult to separate them at the underlying logic level. This expression, however, constructs a two-dimensional phase plane and calculates the geometric eccentricity, shifting the analysis dimension from a dynamic one-dimensional waveform containing time variables to static two-dimensional geometric features purely determined by signal amplitude and phase, thus decoupling the interference of speed variables on error assessment. The calculated dimensionless trajectory features... It can objectively and stably quantify the degree of orthogonality degradation of the analog-to-digital conversion front end caused by pure mechanical shock, providing electrical characteristic parameters independent of operating speed for subsequent determination of the decoding error of the interpolation subdivision circuit.

[0056] In S33, the extraction of time-discrete features at the digital communication protocol layer addresses the technical blind spot where the impact of underlying hardware oscillations on software computation cycles could not be assessed. Mechanical shocks and high-frequency vibrations, while causing optical and electrical distortions, also affect the zero-crossing point determination and state machine toggle rate of the encoder's internal processing core. The timestamps of adjacent data frames arriving at the receiving terminal in the rotation data are extracted, and the time difference between these timestamps is calculated as the actual communication cycle. Comparing this to the set nominal 50-microsecond cycle allows for the detection of microsecond-level data arrival delays or advances. This logic maps physical-level instability to the digital time domain, making clock delays or data congestion at the protocol layer visible. Through this cross-layer timestamp alignment and extraction, a monitoring mechanism for measuring the penetration of mechanical disturbances into the digital logic layer is established, enriching the physical dimensions of overall error detection.

[0057] In S34, the jitter of protocol time is quantified by statistical variance, solving the problem that a single communication cycle deviation is insufficient to reflect the overall dynamic delay pattern. After acquiring a large number of actual communication cycles, the mean of the squared differences of all actual cycles deviating from the nominal 50-microsecond cycle within the statistical window is calculated. In the above scenario, the mean of this squared difference is 120 square nanoseconds, which is used as the jitter variance. This calculation logic avoids overreacting to single, occasional delays, uses variance characteristics to smooth random noise, and centrally reflects the macroscopic discrete trend of communication bus clock modulation under high-speed variable load conditions. This parameter with square time dimensions measures the encoder position data refresh lag effect caused by electromechanical disturbances. By establishing the jitter variance, the error monitoring chain from mechanical deformation to digital communication is improved, providing a data foundation for subsequently converting time hysteresis into spatial compensation.

[0058] like Figure 5 As shown, in one specific embodiment, S4 includes: S41, obtaining the scaling constant, electrical coefficients, and protocol coefficients; multiplying the steady-state equivalent by the scaling constant to obtain a first product; multiplying the trajectory features by the electrical coefficients to obtain a second product; multiplying the jitter variance by the protocol coefficients to obtain a third product; and adding the first, second, and third products to obtain a compensation parameter. The compensation parameter is a feedforward correction value used to offset measurement errors in the underlying drive loop, and its specific calculation process is as follows:

[0059] in, The compensation parameter is in arcseconds; It is a steady-state equivalent and is dimensionless. is a scaling constant used to convert dimensionless steady-state equivalents into angular displacements, in arcseconds; is an electrical coefficient used to convert trajectory characteristics into angular displacement, measured in arcseconds; Trajectory features, dimensionless; These are protocol coefficients used to convert time jitter into angular displacement, measured in arcseconds / ; The variance is the jitter, in units of .

[0060] Wherein, scaling constant Electrical coefficient With protocol coefficient The method for determining the values ​​is primarily based on multiple linear regression analysis of historical operating data under various working conditions. During the calibration process, the photoelectric encoder under test is coaxially mounted with the reference-level polyhedron and autocollimator, allowing the spindle to operate under a preset working condition spectrum covering different speeds and loads. Simultaneously, the steady-state equivalent, trajectory characteristics, jitter variance sequence calculated by this method, and the true angular displacement deviation output by the autocollimator are recorded. For example, calibration data from 500 consecutive sampling periods are extracted, and the true angular displacement deviation is used as the target vector. The three dimensionless or time-dimensional features extracted above are used to construct an independent variable matrix. The gradient descent algorithm is used to solve this matrix equation, iterating until the sum of squared residuals converges, thereby separating the independent influence weights of each physical field feature on the final position deviation. These three weights are set as the scaling constant, electrical coefficient, and protocol coefficient, respectively. This value selection logic ensures that the feedforward compensation parameters have accurate physical dimensions.

[0061] S42. Associate the compensation parameters with the control node. Here, associating with the control node means writing the calculated compensation parameters into the position loop control register of the servo driver to participate in the position control calculation of the next cycle.

[0062] In this embodiment, it should be noted that in S41, the final decoupling of the multiphysics field is achieved, solving the technical problem of the inability to uniformly compensate for errors of different dimensions due to conflicting error dimensions. Specifically, the calculation expression... It constructs a mathematical hub for multi-physics error dimensionality reduction and closed-loop compensation. Its operation process achieves dimensional normalization of cross-domain error parameters by introducing transformation coefficients with specific physical dimensions.

[0063] Furthermore, the design logic of the expression is based on the multivariable feedforward compensation principle in control engineering, performing parallel scale mapping for disturbance characteristics at different physical levels. The first part of the calculation... The problem addressed is the conversion of optical steady-state distortion into servo control parameters, transforming the values ​​into... dimensionless steady-state equivalent With pre-calibrated Arcsecond scaling constant Multiply, and we get The first product of arcseconds, this step, gives the abstract fringe degradation ratio a physical meaning in terms of spatial angle, making the purely optical attenuation phenomenon have a data basis for the mechanical position ring to use.

[0064] Part Two Calculations To address high-frequency electrical distortion, the value is... dimensionless trajectory characteristics With the setting Arcsecond electrical coefficient Multiply, we get The underlying logic of the second product of arcseconds lies in using the transfer function coefficients identified through experiments to linearly map the geometric eccentricity in the phase plane into the actual interpolation subdivision angle deviation.

[0065] Part Three Operations Aimed at solving the spatial equivalence problem of time-domain communication delay, the numerical value is... jitter variance With settings Protocol coefficient per arcsecond per square nanosecond Multiply to obtain The third product of arcseconds, based on the principle of kinematic equivalence, transforms the discrete variance of the arrival time of the data frame into the position feedback hysteresis at a specific rotational speed, thus realizing the data association between the digital communication protocol layer and the physical position layer.

[0066] The expression uses a summation operation to aggregate these three products. The mathematical basis for this is that although the steady-state fringe distortion caused by optomechanical thermal coupling, the electrical orthogonal degradation caused by high-frequency vibration, and the protocol jitter caused by processor processing delay have significantly different physical origins, their contributions to the final position decoding output error can be approximated as mutually independent orthogonal components within a local small signal range, satisfying the linear superposition condition. By performing the addition, the total is calculated to be... Arcsecond compensation parameters The most critical technical challenge addressed by this computational process lies in resolving the data interaction between the underlying physical sensor data stream and the upper-level control stream. It unifies heterogeneous error indicators from different dimensions into a single, simple angular displacement variable through rigorous dimensional alignment and scaling. This allows the servo control to directly inject the compensation parameter into the adder of the underlying control node without needing to analyze complex internal multiphysics mechanisms, thus preemptively canceling multi-source coupled disturbances and improving the driver's feedforward response capability under complex variable load conditions.

[0067] In S42, the execution end of the entire accuracy detection and error compensation method is located, solving the problem that the calculated deviation value cannot take effect in the motor's underlying control loop. The compensation parameter of 10.854 arcseconds calculated in the previous step is associated with the control node, specifically by writing this value into the feedforward control register of the servo drive via the industrial communication bus. When the next high-speed position loop control cycle of the drive arrives, the control algorithm subtracts the planned angle position command from the original position fed back by the encoder and adds this compensation parameter to the deviation. This execution logic forms a complete feedforward control closed loop, enabling the servo electromagnetic circuit to adaptively cancel the errors caused by the combined disturbances of mechanical, optical, electrical, and protocol layers before receiving a lagging or distorted position signal. This transforms theoretical deduction into a control intervention method, improving the dynamic measurement fidelity and robustness of the encoder under complex working conditions.

[0068] The present invention also provides a precision detection system for an optical encoder, the system comprising: The acquisition module is configured to acquire rotational data and underlying parameters under the same working condition. The underlying parameters include deflection data and spectral data. The mapping module is configured to obtain the gap modulation amount based on the deflection data, obtain the dispersion characteristics based on the spectral data, and obtain the steady-state equivalent based on the gap modulation amount and the dispersion characteristics. The extraction module is configured to obtain orthogonal signals and frame timestamps based on rotation data, obtain trajectory features based on orthogonal signals, and obtain jitter variance based on frame timestamps. The execution module is configured to obtain compensation parameters based on steady-state equivalents, trajectory characteristics, and jitter variance, and associate the compensation parameters with the control node.

[0069] In one specific implementation, the acquisition module is further configured to: acquire continuous frames based on the underlying raw data and use the continuous frames as rotation data; acquire stator current, acquire electromagnetic torque based on torque constant and stator current, acquire deformation based on elastic modulus, geometric dimensions and electromagnetic torque, and use the deformation as deflection data; acquire voltage drop data, acquire temperature rise data based on voltage drop data, and use the temperature rise data as spectral data.

[0070] In one specific implementation, the mapping module is further configured to: obtain the gap displacement based on the deflection data and flange distance, and use the gap displacement as the gap modulation amount; obtain the offset based on the temperature rise data and temperature drift coefficient, and use the offset as the dispersion feature; and add the product of the gap modulation amount and the sensitivity constant, and the ratio of the sum of the center wavelength and the offset to the grating constant to obtain the steady-state equivalent.

[0071] To enable those skilled in the art to fully understand and implement the technical solutions described in this specification, the following section, in conjunction with a specific application scenario, provides a detailed deduction and data analysis of the entire process of the accuracy detection method and system for photoelectric encoders.

[0072] Taking a precision CNC milling machine spindle operating under high-speed variable load conditions as an example, the spindle's rated speed is 15000 r / min, and the nonlinear disturbances of the physical environment are significant in this scenario. At this time, the system executes S1, continuously receiving undecoded continuous frame sequences as rotation data through the high-speed synchronous serial interface of S11. In S12, the system synchronously acquires the stator current of the spindle servo motor as 18.5A, and combines this with the fixed torque constant and rotor elastic modulus to calculate the dynamic deformation of the rotating shaft system under the current cutting stress, forming deflection data. Simultaneously, in S13, the system monitors a 0.12V decrease in the forward voltage drop of the LED inside the photoelectric encoder, and based on this, calculates the real-time junction temperature rise as 20K, which is then input as spectral data into the underlying model.

[0073] The system then proceeds to step S2, where it begins establishing the boundary conditions for the steady-state space error. In step S21, the system uses the acquired deflection data and the preset flange distance of the photoelectric encoder to perform a geometric mapping, calculating the axial gap displacement between the code disk surface and the photoelectric receiving array, and establishing the gap modulation amount. The value is 2.5 μm. In S22, the system combines a junction temperature rise of 20 K with the factory-calibrated temperature drift coefficient to extract the numerical offset of the center emission wavelength of the LED, which is used as a characterization of the optical dispersion feature. These two sub-steps transform the invisible mechanical and thermal perturbations into measurable physical boundary parameters.

[0074] In S23, the system calls the core computational logic of the steady-state space moiré fringe distortion equivalent and executes the formula. The system loads pre-calibrated parameters: optical sensitivity constant. It is 0.02 center wavelength 850nm, temperature drift coefficient The grating constant is 0.2 nm / K. The wavelength is 20000 nm. Substituting the real-time acquired data into the calculation, the mechanical part is 0.02 × 2.5 = 0.05; the optical part is (850 + 0.2 × 20) / 20000, which is 854 / 20000 = 0.0427. Adding the two together yields the steady-state equivalent. The value is 0.0927. This dimensionless parameter accurately quantifies the degree of combined distortion of interference fringe contrast attenuation and pitch physical drift under the combined effects of the current 2.5μm mechanical deformation and 20K temperature rise.

[0075] Subsequently, the system executes S3 to extract dynamic high-frequency distortion features. In S31 and S32, under a strong vibration environment of 15000 r / min, the system simultaneously acquires two simulated orthogonal signals with a 90-degree phase difference and fits a characteristic ellipse in a two-dimensional phase plane. Physical measurements and fitting results show that the major axis length of this characteristic ellipse is... 150mV, minor axis length The value is 120mV. The vector composite trajectory, which should have been a standard circle, has undergone significant morphological degradation due to the disruption of electrical phase orthogonality caused by high-frequency mechanical shock.

[0076] The system continues to call the calculation formula in S32. To assess the aforementioned degradation, the measured major and minor axis data are substituted into the equation. The calculation process involves subtracting the square of (120 / 150) from 1, which is equivalent to subtracting the square of 0.8 (0.64) from 1, resulting in 0.36. After taking the square root, the trajectory characteristics are obtained. The value is 0.6. This dimensionless trajectory characteristic mathematically isolates and quantifies the orthogonal deviation of the analog signal caused by the electromechanical response delay.

[0077] In S33 and S34, the system shifts the analysis dimension down to the digital communication protocol layer. The system extracts the arrival timestamps of adjacent frames in the rotation data and calculates multiple actual communication cycles. Compared to the system's nominal 50μs communication cycle, the actual cycles exhibit microsecond-level fluctuations. By calculating the mean of the squared differences between all actual cycles and the nominal cycle within the statistical window, the system extracts the jitter variance. The value is 120 ns². This data intuitively reflects the protocol layer time-domain discrepancy effect caused by underlying mechanical vibration, which, after causing a delay in electrical zero-crossing determination, further affects the internal processor clock.

[0078] The system then enters the final S4 stage for multiphysics decoupling and compensation. In S41, the system loads the calibrated scaling constant. 20 arcseconds, electrical coefficient For 5 arcseconds and protocol coefficients The value is 0.05 arcseconds / ns², and the formula is executed. The calculation process unfolds as follows: First product 0.0927 × 20 = 1.854 arcseconds; Second product 5 × 0.6 = 3.0 arcseconds; Third product 0.05 × 120 = 6.0 arcseconds. Adding these three together, the output is the dynamic angular displacement feedforward compensation parameter. The value is 10.854 arcseconds. In S42, the system associates this 10.854 arcsecond compensation parameter with the control node and writes it into the feedforward control register of the servo driver in real time, thus effectively canceling the composite error across the mechanical, optical, electrical and protocol layers in the underlying drive loop.

[0079] The preferred embodiments of this disclosure have been described in detail above with reference to the accompanying drawings. However, this disclosure is not limited to the specific details of the above embodiments. Within the scope of the technical concept of this disclosure, various simple modifications can be made to the technical solutions of this disclosure, and these simple modifications all fall within the protection scope of this disclosure.

[0080] It should also be noted that the various specific technical features described in the above embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, this disclosure will not describe the various possible combinations separately.

[0081] Furthermore, various different embodiments of this disclosure can be combined in any way, as long as they do not violate the spirit of this disclosure, they should also be regarded as the content disclosed in this disclosure.

[0082] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A method for accuracy detection of an optical encoder, characterized in that, The methods include: The system acquires rotational data and underlying parameters under the same operating condition. The underlying parameters include deflection data and spectral data. This process includes: acquiring continuous frames based on the original underlying data and using these continuous frames as rotational data; acquiring stator current, acquiring electromagnetic torque based on torque constant and stator current, acquiring deformation based on elastic modulus, geometric dimensions, and electromagnetic torque, and using the deformation as deflection data; and acquiring voltage drop data, acquiring temperature rise data based on the voltage drop data, and using the temperature rise data as spectral data. The gap modulation amount is obtained based on deflection data, and the dispersion characteristics are obtained based on spectral data. The steady-state equivalent is then obtained based on the gap modulation amount and dispersion characteristics, including: obtaining the gap displacement based on deflection data and flange distance, and using the gap displacement as the gap modulation amount; obtaining the offset based on temperature rise data and temperature drift coefficient, and using the offset as the dispersion characteristic; and adding the product of the gap modulation amount and the sensitivity constant, as well as the ratio of the sum of the center wavelength and the offset to the grating constant, to obtain the steady-state equivalent. The obtained steady-state equivalent is expressed as: ;in, It is a steady-state equivalent and is dimensionless. The sensitivity constant refers to the influence coefficient of the pre-calibrated optical gap on the contrast of the interference fringes, and its unit is 1000 ppm. ; This is the gap modulation amount, in units of ; The center wavelength refers to the nominal emission wavelength of the light-emitting diode, and the unit is 100 nm. ; This is the temperature drift coefficient, in units of... ; Temperature rise data, unit: ; The grating constant refers to the spacing between adjacent light-transmitting slits on the code disk of a photoelectric encoder, measured in units of... ; Two signals with a preset phase difference are obtained based on the rotation data and used as the first signal and the second signal, respectively. The first signal and the second signal are used as orthogonal signals. The arrival time of adjacent data frames is obtained based on the rotation data and used as the frame timestamp. Using the first signal as the abscissa and the second signal as the ordinate, a feature ellipse is fitted in the phase plane. The major and minor axis lengths are obtained from the feature ellipse. The ratio of the square of the minor axis length to the square of the major axis length is calculated. The square root of the result after subtracting the ratio is taken to obtain the trajectory features. The obtained trajectory features are represented as follows: ;in, Trajectory features, dimensionless; The length of the minor axis, in units of ; The length of the major axis, in units of ; Calculate the time difference between the timestamps of adjacent frames and use the time difference as the actual period; calculate the mean of the squares of the difference between the actual period and the nominal period and use it as the jitter variance. Obtain the scaling constant, electrical coefficient, and protocol coefficient; multiply the steady-state equivalent by the scaling constant to obtain the first product; multiply the trajectory feature by the electrical coefficient to obtain the second product; multiply the jitter variance by the protocol coefficient to obtain the third product; add the first, second, and third products to obtain the compensation parameter; the compensation parameter is expressed as: ;in, The compensation parameter is in arcseconds; It is a steady-state equivalent and is dimensionless. is a scaling constant used to convert dimensionless steady-state equivalents into angular displacements, in arcseconds; is an electrical coefficient used to convert trajectory characteristics into angular displacement, measured in arcseconds; Trajectory features, dimensionless; These are protocol coefficients used to convert time jitter into angular displacement, measured in arcseconds / ; The variance is the jitter, in units of ; Associate the compensation parameters with the control node.

2. A precision detection system for a photoelectric encoder, characterized in that, The system for implementing the accuracy detection method for an optical encoder as described in claim 1 includes: The acquisition module is configured to acquire rotational data and underlying parameters under the same working condition. The underlying parameters include deflection data and spectral data. The mapping module is configured to obtain the gap modulation amount based on the deflection data, obtain the dispersion characteristics based on the spectral data, and obtain the steady-state equivalent based on the gap modulation amount and the dispersion characteristics. The extraction module is configured to obtain orthogonal signals and frame timestamps based on the rotation data, obtain trajectory features based on the orthogonal signals, and obtain jitter variance based on the frame timestamps. The execution module is configured to obtain compensation parameters based on the steady-state equivalent, the trajectory features, and the jitter variance, and associate the compensation parameters with the control node.

3. The accuracy detection system for a photoelectric encoder according to claim 2, characterized in that, The acquisition module is also configured to: Continuous frames are obtained based on the underlying raw data, and the continuous frames are used as the rotation data. Obtain the stator current, obtain the electromagnetic torque based on the torque constant and the stator current, obtain the deformation based on the elastic modulus, geometric dimensions and the electromagnetic torque, and use the deformation as the deflection data; Obtain pressure drop data, obtain temperature rise data based on the pressure drop data, and use the temperature rise data as the spectral data.

4. The accuracy detection system for a photoelectric encoder according to claim 2, characterized in that, The mapping module is also configured to: The gap displacement is obtained based on the deflection data and flange distance, and the gap displacement is used as the gap modulation amount; The offset is obtained based on the temperature rise data and temperature drift coefficient, and the offset is used as the dispersion feature. The steady-state equivalent is obtained by adding the product of the gap modulation amount and the sensitivity constant, and the ratio of the sum of the center wavelength and the offset to the grating constant.

Images