Moire fringe-based mechanical arm motion precision detection system and method

Abstract

The invention provides a mechanical arm motion precision detection system and method based on moire fringes, and relates to the field of image data processing, and the system comprises a grating projection module which is used for determining grating parameters based on motion information of a mechanical arm and projecting a grating pattern to a detection surface of the mechanical arm based on the grating parameters, the grating parameters at least comprise grating density, grating angle and grating type; the image acquisition module is used for acquiring continuous frames of moire fringe images in the movement process of the mechanical arm; and the data processing module is used for carrying out phase calculation on the continuous frames of moire fringe images and resolving the displacement error, the rotation angle error and the vibration error of the mechanical arm, and the motion detection precision of the mechanical arm is improved.

Description

Technical Field

[0001] The present invention relates to the field of image data processing, and particularly to a robotic arm motion accuracy detection system and method based on Moiré fringes. Background Art

[0002] Robotic arms are widely used in industrial manufacturing, medical equipment, and scientific research experiments. Their motion accuracy directly affects work efficiency and product quality. With the rapid development of intelligent manufacturing, the accuracy requirements for robotic arms are also increasing day by day.

[0003] Existing robotic arm motion accuracy detection methods have gradually revealed limitations in certain application scenarios. For example, contact measurement methods use encoders or displacement sensors to directly measure the motion accuracy of robotic arms. Although contact methods have a certain degree of reliability, their measurement process is affected by friction, mechanical installation errors, and motion inertia, which may lead to a decrease in measurement accuracy. In addition, contact measurement may cause damage to the surface of the robotic arm, limiting its application in high-precision scenarios. Another example is that laser interferometry is a typical non-contact measurement technology that can achieve high-precision detection at the nanometer level. However, the laser interferometer system is complex, costly, and sensitive to environmental vibrations, air disturbances, and temperature changes, making it difficult to meet the stability requirements of complex industrial environments. Especially in multi-degree-of-freedom measurements, laser interferometry requires multiple sets of equipment to work together, increasing the complexity of the system. Another example is to analyze the displacement and deformation of the robotic arm using the texture features of optical projection. The operation is simple, but its detection accuracy is limited by the resolution of the projection system and is easily affected by environmental light interference, limiting its application in dynamic detection scenarios.

[0004] Therefore, it is necessary to provide a robotic arm motion accuracy detection system and method based on Moiré fringes to improve the accuracy of robotic arm motion detection. Summary of the Invention

[0005] The present invention provides a robotic arm motion accuracy detection system based on Moiré fringes, including: a grating projection module for determining grating parameters based on the motion information of the robotic arm and projecting a grating pattern onto the detection surface of the robotic arm based on the grating parameters, where the grating parameters at least include grating density, grating angle, and grating type; an image acquisition module for obtaining continuous-frame Moiré fringe images through a plurality of image acquisition devices during the motion of the robotic arm; and a data processing module for performing phase calculation on the continuous-frame Moiré fringe images to solve the displacement error, rotation angle error, and vibration error of the robotic arm.

[0006] Further, the grating projection module determines the grating parameters based on the motion information of the robotic arm, including: determining the grating density based on the detection accuracy requirement and the motion speed of the robotic arm.

[0007] Further, the grating projection module determines grating parameters based on the motion information of the robotic arm, including: if the motion type of the robotic arm is translational motion, determining the grating type as a linear grating; if the motion type of the robotic arm is rotational motion, determining the grating type as a radial grating; if the motion type of the robotic arm is composite motion, dynamically switching the grating type according to the magnitude of the motion components.

[0008] Further, the grating projection module determines grating parameters based on the motion information of the robotic arm, including: if the motion type of the robotic arm is translational motion, determining the grating angle as perpendicular to the motion direction; if the motion type of the robotic arm is rotational motion, determining the grating center to be aligned with the rotation axis.

[0009] Further, the data processing module performs phase calculation on consecutive frames of moiré fringe images to solve the displacement error, rotation angle error, and vibration error of the robotic arm, including: when the motion type is translational motion, converting each moiré fringe image from the spatial domain to the frequency domain, extracting the phase information of the moiré fringe, generating a phase map, converting the phase map to a continuous phase through a phase unwrapping algorithm, and calculating the displacement error based on the continuous phase.

[0010] Further, the data processing module performs phase calculation on consecutive frames of moiré fringe images to solve the displacement error, rotation angle error, and vibration error of the robotic arm, including: when the motion type is rotational motion, converting each moiré fringe image from the spatial domain to the frequency domain, extracting the phase information of the moiré fringe, generating a phase map, converting the phase map to a continuous phase through a phase unwrapping algorithm, and calculating the rotation angle error based on the continuous phase.

[0011] Further, the data processing module is also used for: performing error compensation for the multi-degree-of-freedom motion of the robotic arm through multi-channel data fusion technology.

[0012] Further, the data processing module performs error compensation for the multi-degree-of-freedom motion of the robotic arm through multi-channel data fusion technology, including: preprocessing consecutive frames of moiré fringe images to generate preprocessed consecutive frames of moiré fringe images, where the preprocessing at least includes data alignment, data normalization, and filtering; performing data association, state estimation, and error compensation on the preprocessed consecutive frames of moiré fringe images.

[0013] Further, the system further includes: a display module for visualizing the displacement error, rotation angle error, and vibration error of the robotic arm.

[0014] The present invention provides a method for detecting the motion accuracy of a robotic arm based on Moiré fringes, which is applied to the above-mentioned system for detecting the motion accuracy of a robotic arm based on Moiré fringes. Based on the motion information of the robotic arm, grating parameters are determined, and a grating pattern is projected onto the detection surface of the robotic arm based on the grating parameters. Among them, the grating parameters at least include grating density, grating angle, and grating type; during the motion of the robotic arm, continuous-frame Moiré fringe images are acquired by a plurality of image acquisition devices; phase calculations are performed on the continuous-frame Moiré fringe images to solve the displacement error, rotation angle error, and vibration error of the robotic arm.

[0015] Compared with the prior art, the system and method for detecting the motion accuracy of a robotic arm based on Moiré fringes provided by the present invention at least have the following beneficial effects: 1. By accurately determining the grating parameters (such as grating density, angle, and type) based on the motion information of the robotic arm, the detection requirements under different motion states can be adapted, so that the projected grating pattern can better fit the actual motion situation, which helps to obtain more accurate and clear Moiré fringe images and lay a foundation for subsequent high-precision error calculation.

[0016] 2. The data processing module can perform phase calculations on the continuous-frame Moiré fringe images, which can not only solve the displacement error and rotation angle error, but also solve the vibration error, comprehensively evaluate the motion accuracy of the robotic arm, and can understand the motion performance of the robotic arm more deeply compared with the traditional single-dimensional detection method.

[0017] 3. For the translational motion, rotational motion, and composite motion of the robotic arm, appropriate grating types (linear grating, radial grating, or dynamic switching) and grating angles are determined respectively, so that the system can flexibly adapt to various complex robotic arm motion scenarios and has wide applicability.

[0018] 4. For composite motion, the grating type can be dynamically switched according to the magnitude of the motion components. This dynamic adjustment mechanism can better capture various motion characteristics of the robotic arm during the composite motion process and improve the accuracy and reliability of detection.

[0019] 5. Through multi-channel data fusion technology, error compensation for the multi-degree-of-freedom motion of the robotic arm is performed, which means that the motion accuracy information of the robotic arm can be obtained in real time during the motion of the robotic arm, potential problems can be discovered in time, and a basis for the real-time control and adjustment of the robotic arm is provided, improving production efficiency and safety.

[0020] 6. The system adopts advanced data processing methods such as phase calculation, phase unwrapping algorithm, and multi-channel data fusion technology, which can quickly and accurately process and analyze a large number of Moiré fringe images, greatly improving the detection efficiency and reducing manual intervention and errors.

[0021] 7. During the multi-channel data fusion process, preprocessing (data alignment, data normalization, and filtering) and fusion processing (data association, state estimation, and error compensation) are performed on the continuous-frame Moiré fringe images obtained by multiple image acquisition devices, which can effectively eliminate the noise and interference in the images, improve the accuracy and reliability of the measurement results, and ensure the stable operation of the detection system in complex environments. Description of the Drawings

[0022] This specification will be further described by way of exemplary embodiments, which will be described in detail through the drawings. These embodiments are not restrictive. In these embodiments, the same numbers represent the same structures, where: Figure 1 is a schematic block diagram of a robotic arm motion accuracy detection system based on Moiré fringes according to some embodiments of this specification; Figure 2 is a schematic flowchart of a method for detecting the motion accuracy of a robotic arm based on Moiré fringes according to some embodiments of this specification. Detailed Embodiments

[0023] To more clearly illustrate the technical solutions of the embodiments of this specification, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some examples or embodiments of this specification. For those of ordinary skill in the art, without creative efforts, this specification can also be applied to other similar scenarios based on these drawings. Unless obvious from the language context or otherwise stated, the same reference numerals in the figures represent the same structures or operations.

[0024] Figure 1 is a schematic block diagram of a robotic arm motion accuracy detection system based on Moiré fringes according to some embodiments of this specification. As Figure 1 shown, a robotic arm motion accuracy detection system based on Moiré fringes may include a grating projection module, an image acquisition module, and a data processing module. A robotic arm motion accuracy detection system based on Moiré fringes can be applied to various scenarios. For example, it can be applied to the production line detection of industrial robotic arms to monitor the motion accuracy of the robotic arms in real time and ensure the product processing quality. Another example is that during the robot assembly and calibration process, by detecting the motion accuracy of the robotic arms, the device performance can be optimized. Another example is for dynamic measurement in high-precision mechanical motion experiments to provide high-precision data support for research. The following is a specific explanation of each module.

[0025] The grating projection module is used to adopt a projection Moiré structure based on the motion information of the robotic arm. This structure includes a set of high-precision grating patterns and the grating patterns projected by DLP. Among them, the high-precision grating patterns are placed on the moving parts of the robotic arm (such as joints or linear guides). The density and angle of the grating patterns are designed according to the motion mode. Specifically, linear gratings are used to measure linear motion, and radial gratings are used for high-precision detection of rotational motion. To further improve the accuracy, the DLP projection module generates projection fringes with the same shape but different periods according to different forms of gratings. When the projection fringes are projected onto the grating surface, Moiré fringes will appear.

[0026] Specifically, the grating projection module uses an adjustable grating generator to project high-precision grating patterns through Digital Light Processing (DLP) technology, including linear gratings and radial gratings. The linear grating consists of a series of equally spaced and parallel stripes. The stripes can be alternately bright and dark stripes, or grids with alternating light-transmitting and light-blocking regions. The stripes of the radial grating radiate outward along the radial direction with the center point as the origin, forming a pattern similar to "sun rays". The stripe spacing increases with the increase of the radius (equal angular spacing) or remains constant (equal linear spacing).

[0027] The grating projection module can dynamically adjust the grating pattern according to the motion information of the robotic arm and the detection accuracy requirements.

[0028] For example, the grating projection module can determine the grating density based on the detection accuracy requirements and the motion speed of the robotic arm.

[0029] Specifically, the input parameters are: the motion speed of the robotic arm and the detection accuracy requirements.

[0030] Adjustment logic: When the motion speed of the robotic arm is high, reduce the grating density (increase the grating period) to avoid blurring of the Moiré fringe image. When the motion speed of the robotic arm is low, increase the grating density (decrease the grating period) to increase the detection sensitivity. Dynamically adjust the grating density according to the detection accuracy requirements to ensure that the accuracy of phase resolution meets the requirements.

[0031] Only as an example, the grating density can be calculated according to the following formula:

[0032] Among them, d is the grating period, k is the proportionality coefficient, v is the motion speed of the robotic arm, and P is the detection accuracy requirement. The detection accuracy (P) refers to the minimum size change that the detection system can distinguish or measure, and the unit is micrometer (μm). It directly reflects the ability of the system to identify subtle differences in target features.

[0033] As an example only, the robot arm performs precision detection under high-speed translation (v=1m / s), and the detection accuracy requirement is P=10μm. According to the movement speed (v=1m / s) and the detection accuracy (P=10μm), the grating density is calculated:

[0034] Among them, translational motion refers to the movement of the end of the robot arm in a straight line in a single direction (such as X, Y, Z axis).

[0035] For example, the robot arm performs precision detection under low-speed rotation (angular velocity ω=0.1rad / s), and the detection accuracy requirement is P=0.001μm. Calculate the grating density:

[0036] Among them, rotational motion refers to the rotation of the robot arm joint around the axis.

[0037] Understandably, when moving at high speed, if the grating period is too small, the displacement of the robot arm will cause the moiré fringes to change rapidly, exceeding the sampling frequency of the camera and causing image blur. By increasing the grating period, the rate of change of the moiré fringes is reduced, ensuring that the camera can clearly capture the fringe information. When moving at low speed, the displacement of the robot arm is small, and the grating density needs to be increased to enhance the accuracy of phase resolution. Denser grating fringes can provide richer phase information, thereby improving the detection system's ability to resolve small deformations. By dynamically adjusting the grating density, the system can achieve optimal performance under different motion speeds and detection accuracy requirements. Increase the period at high speed to avoid blurring, and reduce the period at low speed to enhance sensitivity, thereby ensuring that the accuracy of phase resolution always meets the requirements. This strategy not only improves detection accuracy and system stability, but also enhances the adaptability and cost-effectiveness of the system, providing strong support for dynamic detection tasks.

[0038] In some embodiments, the grating projection module determines the grating parameters based on the motion information of the robot arm, including: If the motion type of the robot arm is translational motion, determine that the grating type is linear grating; If the motion type of the robot arm is rotational motion, determine that the grating type is a radial grating; If the motion type of the robot arm is compound motion, the grating type is dynamically switched according to the size of the motion component.

[0039] In some embodiments, the grating projection module determines the grating parameters based on the motion information of the robot arm, including: If the motion type of the robot arm is translational motion, the grating angle is determined to be perpendicular to the motion direction to maximize the contrast of the moiré fringes; If the motion type of the robotic arm is rotational motion, align the center of the grating with the axis of rotation to improve the sensitivity of rotational detection.

[0040] Only as an example, the robotic arm performs precision detection under high-speed translational motion (v = 1 m / s). The detection accuracy requirement is P = 10 μm. The motion type of the robotic arm is translational motion, and the motion direction θ = 30°. Then the grating angle α = θ + 90° = 30° + 90° = 120°. A linear grating is used, with a grating period of 10 mm and a grating angle of 120°. The system can clearly capture the Moiré fringes in high-speed translational motion and ensure that the detection accuracy reaches 10 μm.

[0041] Also as an example, the robotic arm performs precision detection under low-speed rotational motion (angular velocity ω = 0.1 rad / s). A radial grating is used, and the center of the grating is aligned with the axis of rotation. The system can accurately detect the angular changes in low-speed rotational motion, and the detection accuracy reaches 0.001°.

[0042] Also as an example, the grating projection module can dynamically switch the grating type according to the magnitude of the motion components based on the following process: S11. Determine the magnitude of the motion components: Translational component: The linear velocity of the end of the robotic arm along the X, Y, and Z axes , where is the linear velocity along the X axis, is the linear velocity along the Y axis, is the linear velocity along the Z axis; Rotational component: The angular velocity ω (unit: rad / s) of the robotic arm joint around the axis; S12. Weight calculation: Define the normalized translational component weight and the rotational component weight , where and are the maximum allowable translational speed and angular velocity.

[0043] S13. Dynamically switch the grating type: When the translational component weight is greater than or equal to , determine that the grating type is a linear grating; When the translational component weight is less than , determine that the grating type is a radial grating.

[0044] It is understandable that in translational motion (such as robotic arm grasping, planar scanning), linear gratings can accurately capture linear displacements, avoiding fringe blurring or phase errors caused by mismatches between the grating period and the motion speed. In rotational motion (such as robotic arm joint rotation, curved surface scanning), radial gratings can efficiently measure angular changes, reducing detection deviations caused by mismatches between the grating direction and the rotation axis. By switching the grating type in real time, the system can adapt to different motion scenarios, ensuring that the detection accuracy is always in an optimal state. In complex motion (such as translation + rotation), traditional fixed gratings are prone to error accumulation due to the coupling of motion components. The dynamic switching mechanism can separate the translational and rotational components, reducing the impact of interference. For example, when the robotic arm is translating at high speed with a small rotation, the system can preferentially select a linear grating and compensate for the rotational component through a filtering algorithm to avoid distortion of the detection results. The same set of grating projection modules can be applied to multiple tasks (such as assembly, inspection, scanning) without the need to redesign the hardware for different tasks. For example, in industrial assembly, a linear grating is used when the robotic arm performs a grasping task; a radial grating is switched to when performing a screw-tightening task.

[0045] An image acquisition module for capturing moiré fringe images facing the grating during the motion of the robotic arm.

[0046] Specifically, the image acquisition module is equipped with a high-resolution industrial camera, which captures moiré fringe images on the surface of the robotic arm in combination with a lens group. The camera has a high acquisition frequency and can meet the dynamic detection requirements of the fast motion of the robotic arm.

[0047] A data processing module for calculating the phase of consecutive frames of moiré fringe images and resolving the displacement error, rotation angle error, and vibration error of the robotic arm.

[0048] For example, when the motion type is translational motion, the data processing module can transform each moiré fringe image from the spatial domain to the frequency domain, extract the phase information of the moiré fringe, generate a phase map, convert the phase map to a continuous phase through a phase unwrapping algorithm, and calculate the displacement error based on the continuous phase.

[0049] Specifically, before extracting the phase information of the moiré fringe, it is necessary to preprocess the consecutive frames of moiré fringe images, which specifically includes: 1. Data alignment Time alignment: Synchronize the time of consecutive frames of moiré fringe images obtained by multiple image acquisition devices to ensure that the timestamps of the data are consistent.

[0050] Spatial alignment: Register the spatial positions of consecutive frames of moiré fringe images obtained by multiple image acquisition devices to ensure that the consecutive frames of moiré fringe images are in the same coordinate system.

[0051] 2. Data filtering Filter the Moiré fringe image with high noise (such as Kalman filtering or low-pass filtering) to improve the signal-to-noise ratio of the Moiré fringe image.

[0052] Perform a two-dimensional fast Fourier transform calculation on the preprocessed Moiré fringe image to obtain the frequency spectrum diagram of the image. The peaks in the frequency spectrum diagram correspond to the main frequency components of the Moiré fringe. In the frequency spectrum diagram, extract the main frequency component of the Moiré fringe through a band-pass filter to remove the low-frequency background and high-frequency noise.

[0053] Perform an inverse FFT (Inverse Fast Fourier Transform, IFFT) on the filtered frequency spectrum to obtain a complex image containing phase information.

[0054] Calculate the phase diagram through the real and imaginary parts of the complex image:

[0055] where is the complex image, is the imaginary part of the complex image, is the real part of the complex image, is the phase diagram.

[0056] Since the phase diagram is a wrapped phase (the value range is -π to π), it is necessary to convert it to a continuous phase through a phase unwrapping algorithm to calculate the displacement and rotation error of the robotic arm. The goal of phase unwrapping is to eliminate the 2π jumps in the phase diagram and restore the continuous phase distribution. The phase unwrapping algorithm can include: Quality-Guided Phase Unwrapping: Determine the unwrapping path according to the phase quality map (such as the variance of the phase derivative). Least-Squares Phase Unwrapping: Solve the continuous phase by minimizing the phase gradient error.

[0057] Steps of quality-guided phase unwrapping: Calculate the phase quality map: Calculate the phase quality of each pixel according to the variance of the phase derivative or the phase consistency.

[0058] Determine the unwrapping path: Start from the pixel with the highest phase quality and gradually unwrap the pixels with lower phase quality.

[0059] Phase unwrapping: Accumulate the 2π jumps along the unwrapping path to restore the continuous phase.

[0060] For a linear grating, the relationship between the displacement d and the phase Φ is:

[0061] where p is the grating period.

[0062] For a radial grating, the relationship between the rotation angle θ and the phase Φ is:

[0063] where r is the radius of the radial grating.

[0064] Calculate the displacement errors of the robotic arm in the X, Y, and Z directions according to the phase change of the linear grating.

[0065] Specifically, the displacement error calculation: Reference phase and measured phase: Reference phase: When the robotic arm is in the initial position, collect the Moiré fringe image and calculate the phase diagram to obtain the reference phase .

[0066] Measured phase: When the robotic arm moves to the target position, collect the Moiré fringe image and calculate the phase diagram to obtain the measured phase .

[0067] Phase difference calculation: Calculate the phase difference of each pixel:

[0068] Displacement calculation: Calculate the displacement of each pixel according to the phase difference:

[0069] Average displacement error: Average the displacements of all pixels to obtain the average displacement error of the robotic arm:

[0070] where M is the number of rows of pixels in the Moiré fringe image, and N is the number of columns of pixels in the Moiré fringe image.

[0071] For multiple image acquisition devices to obtain consecutive frames of Moiré fringe images, fuse multi-channel data by the least squares method to improve the displacement detection accuracy.

[0072] For another example, when the motion type is rotational motion, the data processing module can transform each Moiré fringe image from the spatial domain to the frequency domain, extract the phase information of the Moiré fringe, generate a phase diagram, convert the phase diagram to a continuous phase through a phase unwrapping algorithm, and calculate the rotation angle error based on the continuous phase.

[0073] For a radial Moiré pattern, use the polar coordinate fast Fourier transform to transform the Moiré fringe image from the spatial domain to the frequency domain, extract the phase information of the Moiré fringe, and generate a phase diagram to improve the detection accuracy of rotational motion.

[0074] Calculate the rotation angle error of the robotic arm based on the phase change of the radial grating.

[0075] For a robotic arm with multiple degrees of freedom, convert the rotation error into Euler angles or quaternion representation through coordinate transformation.

[0076] Perform time-frequency analysis (such as short-time Fourier transform or wavelet transform) on the displacement and rotation errors to extract the vibration frequency and amplitude of the robotic arm. Identify the vibration source through spectrum analysis to provide data support for the vibration control of the robotic arm.

[0077] As an example only, displacement detection of a linear grating: Input: Moiré fringe image of the linear grating.

[0078] Processing steps: Image preprocessing: Denoising, enhancement, cropping.

[0079] FFT calculation: Extract the main frequency component of the Moiré fringe.

[0080] Phase extraction: Calculate the wrapped phase map.

[0081] Phase unwrapping: Restore the continuous phase.

[0082] Displacement calculation: Calculate the displacement of the robotic arm according to the phase change.

[0083] Output: Displacements of the robotic arm in the X, Y, and Z directions.

[0084] Also as an example, rotation detection of a radial grating: Input: Moiré fringe image of the radial grating.

[0085] Processing steps: Image preprocessing: Denoising, enhancement, cropping.

[0086] Polar coordinate FFT calculation: Extract the main frequency component of the rotational motion.

[0087] Phase extraction: Calculate the wrapped phase map.

[0088] Phase unwrapping: Restore the continuous phase.

[0089] Rotation calculation: Calculate the rotation angle of the robotic arm according to the phase change.

[0090] Output: Rotation angle of the robotic arm.

[0091] The data processing module is also used for: Perform error compensation for the multi-degree-of-freedom motion of the robotic arm through multi-channel data fusion technology.

[0092] Specifically include: Preprocess the continuous-frame moiré fringe images to generate the preprocessed continuous-frame moiré fringe images, where the preprocessing includes at least data alignment, data normalization, and filtering. Perform data association, state estimation, and error compensation on the preprocessed continuous-frame moiré fringe images.

[0093] 1. Multi-channel data sources The input data for multi-channel data fusion includes: Raster data: Linear raster data: used to detect the translational motion (in the X, Y, and Z directions) of the robotic arm or radial raster data: used to detect the rotational motion of the robotic arm (rotation of the robotic arm joints around the axis).

[0094] Time series data: Continuous-frame moiré fringe images during the motion of the robotic arm, used to analyze dynamic errors and vibration characteristics.

[0095] 2. Data preprocessing Before data fusion, it is necessary to preprocess the multi-channel data to ensure the consistency and reliability of the data.

[0096] (1) Data alignment Time alignment: Synchronize the time of the continuous-frame moiré fringe images obtained by multiple image acquisition devices to ensure that the timestamps of the data are consistent.

[0097] Spatial alignment: Perform spatial registration on the multi-continuous-frame moiré fringe images to ensure that the data is in the same coordinate system.

[0098] (2) Data normalization Normalize the data of different channels to the same dimension and range for subsequent fusion calculations.

[0099] For phase data, convert the wrapped phase (-π to π) to continuous phase.

[0100] (3) Data filtering Filter the data with large noise (such as Kalman filtering or low-pass filtering) to improve the signal-to-noise ratio of the data.

[0101] 3. Multi-channel data fusion algorithm The core algorithms of multi-channel data fusion include data association, state estimation, and error compensation. The following are the specific steps: (1) Data association Target matching: Associate the data of different channels to the same target (such as the end effector of the robotic arm).

[0102] Feature extraction: Extract features (such as displacement, angle, phase change, etc.) from the data of each channel.

[0103] Similarity calculation: Calculate the similarity between data of different channels to determine the correlation relationship of the data.

[0104] (2)State estimation Definition of state vector: Define the state vector s of the robotic arm, including position, instantaneous linear velocity, instantaneous angular velocity, rotation angle, etc.:

[0105] Among them, is the state vector, is the absolute position of the end of the robotic arm in the X-axis direction, is the absolute position of the end of the robotic arm in the Y-axis direction, is the absolute position of the end of the robotic arm in the Z-axis direction, is the rotation angle of the end of the robotic arm around the X-axis, is the rotation angle of the end of the robotic arm around the Y-axis, is the rotation angle of the end of the robotic arm around the Z-axis, is the instantaneous linear velocity of the end of the robotic arm in the X-axis direction, is the instantaneous linear velocity of the end of the robotic arm in the Y-axis direction, is the instantaneous linear velocity of the end of the robotic arm in the Z-axis direction, is the instantaneous angular velocity of the end of the robotic arm around the X-axis, is the instantaneous angular velocity of the end of the robotic arm around the Y-axis, is the instantaneous angular velocity of the end of the robotic arm around the Z-axis.

[0106] State prediction: According to the motion model of the robotic arm (such as the rigid body motion model), predict the state at the next moment:

[0107] Among them, is the state transition matrix, is the process noise.

[0108] State update: Update the state estimation according to the multi-channel observation data:

[0109] Among them, is the observation vector, is the observation matrix, is the Kalman gain.

[0110] (3)Error compensation Residual calculation: Calculate the residual between the observation data and the predicted state:

[0111] Error compensation: Adjust the state estimation according to the residuals to compensate for systematic errors (such as sensor biases, installation errors, etc.).

[0112] As an example only, synchronous detection of translation and rotation: Input: Linear grating data: Displacements of the robotic arm in the X, Y, and Z directions. Radial grating data: Rotation angles of the robotic arm about the X, Y, and Z axes.

[0113] Processing steps: Data alignment: Register the continuous-frame Moiré fringe images acquired by multiple image acquisition devices to the same coordinate system, and extract displacement and angle features from each image.

[0114] State prediction: Predict the state of the next moment according to the motion model of the robotic arm.

[0115] State update: Update the state estimation according to the observation data.

[0116] Error compensation: Compensate for systematic errors and output accurate displacements and angles.

[0117] Output: 6-degree-of-freedom motion state of the robotic arm (x, y, z, θ x , θ y , θ z ).

[0118] In some embodiments, a robotic arm motion accuracy detection system based on Moiré fringes may further include a display module for visualizing the displacement error, rotation angle error, and vibration error of the robotic arm.

[0119] Specifically, the display module can display the displacement error, rotation angle error, and vibration error of the robotic arm in the form of real-time curves or 3D models. The data can be exported in a standard format for quality analysis and equipment calibration.

[0120] Figure 2 is a schematic flowchart of a method for detecting the motion accuracy of a robotic arm based on Moiré fringes according to some embodiments of the present specification, as Figure 2 shown. In some embodiments, a method for detecting the motion accuracy of a robotic arm based on Moiré fringes may include the following processes: Based on the motion information of the robotic arm, determine the grating parameters, and project the grating pattern onto the detection surface of the robotic arm based on the grating parameters, where the grating parameters at least include the grating density, grating angle, and grating type; During the motion of the robotic arm, acquire continuous-frame Moiré fringe images; Perform phase calculation on the continuous-frame Moiré fringe images to solve the displacement error, rotation angle error, and vibration error of the robotic arm.

[0121] A method for detecting the motion accuracy of a robotic arm based on Moiré fringes can be applied to a system for detecting the motion accuracy of a robotic arm based on Moiré fringes, which will not be elaborated here.

[0122] Finally, it should be understood that the embodiments described in this specification are only used to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, by way of example and not limitation, alternative configurations of the embodiments of this specification can be regarded as consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly introduced and described in this specification.

Claims

1. A robotic arm motion accuracy detection system based on Moiré fringes, characterized in that, Including: A grating projection module, configured to determine grating parameters based on the motion information of the robotic arm, and project a grating pattern onto the detection surface of the robotic arm based on the grating parameters, where the grating parameters at least include grating density, grating angle, and grating type; An image acquisition module, configured to obtain continuous-frame moiré fringe images through a plurality of image acquisition devices during the motion of the robotic arm; A data processing module, configured to perform phase calculation on the continuous-frame moiré fringe images to solve the displacement error, rotation angle error, and vibration error of the robotic arm.

2. The robotic arm motion accuracy detection system based on Moiré fringes according to claim 1, wherein The grating projection module determines grating parameters based on the motion information of the robotic arm, including: Determining the grating density based on the detection accuracy requirement and the motion speed of the robotic arm.

3. The mechanical arm motion accuracy detection system based on Moiré fringes according to claim 2, characterized in that, The grating projection module determines grating parameters based on the motion information of the robotic arm, including: If the motion type of the robotic arm is translational motion, determining the grating type as a linear grating; If the motion type of the robotic arm is rotational motion, determining the grating type as a radial grating; If the motion type of the robotic arm is composite motion, dynamically switching the grating type according to the magnitude of the motion components.

4. A robotic arm motion accuracy detection system based on Moiré fringes according to claim 3, characterized in that, The grating projection module determines grating parameters based on the motion information of the robotic arm, including: If the motion type of the robotic arm is translational motion, determining the grating angle as perpendicular to the motion direction; If the motion type of the robotic arm is rotational motion, determining the grating center to be aligned with the rotation axis.

5. The robotic arm motion accuracy detection system based on Moiré fringes according to claim 3 or 4, characterized in that, The data processing module performs phase calculation on the continuous-frame moiré fringe images to solve the displacement error, rotation angle error, and vibration error of the robotic arm, including: When the motion type is translational motion, converting each moiré fringe image from the spatial domain to the frequency domain, extracting the phase information of the moiré fringe, generating a phase map, converting the phase map to a continuous phase through a phase unwrapping algorithm, and calculating the displacement error based on the continuous phase.

6. The robotic arm motion accuracy detection system based on Moiré fringes according to claim 5, characterized in that, The data processing module performs phase calculation on the continuous-frame moiré fringe images to solve the displacement error, rotation angle error, and vibration error of the robotic arm, including: When the motion type is rotational motion, converting each moiré fringe image from the spatial domain to the frequency domain, extracting the phase information of the moiré fringe, generating a phase map, converting the phase map to a continuous phase through a phase unwrapping algorithm, and calculating the rotation angle error based on the continuous phase.

7. The robotic arm motion accuracy detection system based on Moiré fringes according to claim 6, characterized in that, The data processing module is further configured to: Perform error compensation for the multi-degree-of-freedom motion of the robotic arm through a multi-channel data fusion technique.

8. The robotic arm motion accuracy detection system based on Moiré fringes according to claim 7, characterized in that, The data processing module performs error compensation for the multi-degree-of-freedom motion of the robotic arm through a multi-channel data fusion technique, including: Preprocessing the continuous-frame moiré fringe images to generate preprocessed continuous-frame moiré fringe images, where the preprocessing at least includes data alignment, data normalization, and filtering processing; Performing data association, state estimation, and error compensation on the preprocessed continuous-frame moiré fringe images.

9. A mechanical arm motion accuracy detection system based on Moiré fringes according to any one of claims 1-4, characterized in that, Further including: A display module, configured to visualize the displacement error, rotation angle error, and vibration error of the robotic arm.

10. A method for detecting the motion accuracy of a robotic arm based on Moiré fringes, characterized in that, Applied to a robotic arm motion accuracy detection system according to any one of claims 1-9, including: Determine grating parameters based on the motion information of the robotic arm, and project a grating pattern onto the detection surface of the robotic arm based on the grating parameters, where the grating parameters at least include grating density, grating angle, and grating type; During the movement of the robotic arm, obtain consecutive-frame moiré fringe images through multiple image acquisition devices; Perform phase calculation on the consecutive-frame moiré fringe images to solve the displacement error, rotation angle error, and vibration error of the robotic arm.

Images