A method for measuring the unevenness of a steel wire rope based on a robot

Abstract

The present application relates to the technical field of nondestructive testing, and particularly relates to a steel wire unevenness measurement method based on a robot, which comprises the following steps: robot positioning and initial calibration, steel wire surface pretreatment, unevenness preliminary scanning, high-precision three-dimensional profile measurement, real-time preprocessing of collected original data, automatic generation of a detailed measurement report, autonomous resetting after the measurement is completed, and prediction of the remaining life of the steel wire according to the measurement result of this time. According to the method, multi-modal sensors and an adaptive robot platform are deeply fused, and through a unique algorithm process of global positioning, local precision measurement, data fusion and dynamic compensation, unmanned, full-coverage and high-precision measurement of the three-dimensional appearance of the steel wire surface is realized.

Description

[0001] This application is a divisional application of application filed on November 3, 2025, with application number 2025115935325 and invention title "A Robot-Based Method for Measuring the Unevenness of Steel Wire Rope". Technical Field

[0002] This invention relates to the field of nondestructive testing technology, and in particular to a robot-based method for measuring the unevenness of steel wire ropes. Background Technology

[0003] As a key load-bearing component, the surface unevenness, wear, and broken wires of steel wire ropes directly affect the safety of the overall structure. Traditional inspection methods mainly rely on manual visual inspection or handheld inspection instruments. Manual inspection is slow, and the inspection results depend on the experience of the personnel. Handheld equipment is difficult to achieve continuous and full-coverage measurement, and it is easy to miss local defects. It is also impossible to accurately assess the depth / height and volume of unevenness such as dents and bulges.

[0004] Currently, some existing technologies disclose wire rope detection devices based on machine vision, but most of them are for stationary or short indoor ropes and lack deep integration with robot motion control. There is an urgent need to overcome the shortcomings of existing technologies and provide a fully automatic, high-precision, high-reliability wire rope unevenness measurement method and system that can adapt to complex environments. Summary of the Invention

[0005] The purpose of this invention is to propose a robot-based method for measuring the unevenness of wire ropes in order to solve the problems in the prior art.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a robot-based method for measuring the unevenness of steel wire ropes, comprising the following steps: Step S1, Robot positioning and initial calibration: The robot automatically identifies the starting point of the wire rope and performs precise positioning through the integrated vision system, which captures images of the wire rope. Step S2: Wire rope surface pretreatment. The robot is equipped with a non-contact cleaning device to clean the surface of the wire rope and remove impurities. Step S3, preliminary scan of unevenness: The robot uses a multi-sensor array to move quickly along the steel cable to perform a preliminary scan and identify obvious uneven areas. Step S4: High-precision three-dimensional contour measurement. The robot equipped with the scanning module moves slowly along the steel wire rope to obtain detailed three-dimensional surface contour data. Step S5: Real-time data preprocessing. During the robot's movement, the collected raw data is preprocessed in real time. Step S6: Automatically generate a detailed measurement report, including an unevenness distribution map, defect location coordinates, safety assessment, and recommended measures; Step S7: After the robot completes the measurement, it automatically resets and predicts the remaining life of the wire rope based on the measurement results.

[0007] The beneficial effects of the technical solution provided by this invention include at least the following: This invention achieves unmanned, full-coverage, and high-precision measurement of the three-dimensional morphology of steel wire rope surfaces by deeply integrating multimodal sensors with an autonomous navigation and adaptive robot platform and employing a unique algorithm process of global positioning, local precision measurement, data fusion, and dynamic compensation.

[0008] The robot system of this invention can automatically and accurately locate the starting end of the steel wire rope to be measured and establish a robot measurement coordinate system based on the axis of the steel wire rope. Combining the wide-angle global vision and the micro-range crosshair laser, it achieves a seamless connection from macro-navigation to micro-precision measurement and has strong resistance to environmental interference.

[0009] This invention combines three physical principles—ultrasound, vortex cyclone, and electrostatics—in a multimodal synergistic manner, targeting the three stages of adhesion, purging, and suspended residue, respectively, forming a complete cleaning chain with efficiency far exceeding that of a single method. It introduces a high-precision spectral confocal sensor for in-situ quality inspection and forms a closed-loop control to ensure reliable pretreatment results, providing a high-quality surface guarantee for subsequent preliminary scanning of unevenness and avoiding misjudgments caused by incomplete cleaning. The entire process is completely non-contact, avoiding methods such as brushing that may scratch the wire rope or cause secondary pollution.

[0010] This invention combines visible light surface texture imaging and near-infrared transmission imaging, enabling simultaneous detection of surface geometric irregularities and subsurface material defects. This significantly improves the comprehensiveness and reliability of the initial scan. By employing a statistical process control method, the threshold can be dynamically adjusted based on the background texture of the wire rope itself, effectively overcoming interference caused by uneven global illumination or natural bending of the rope and significantly reducing the false alarm rate.

[0011] This invention integrates multiple error sources such as robot vibration, thermal deformation, and kinematic errors to establish a comprehensive compensation model, which improves the absolute accuracy of data at the system level. This surpasses simple sensor denoising and improves the nonlocal mean filtering algorithm, enabling it to distinguish between noise and real surface features, and avoiding the smoothing effect of traditional filters on minor defects. Attached Figure Description

[0012] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 The diagram illustrates the method steps provided in this embodiment of the invention. Detailed Implementation

[0014] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a robot-based method for measuring the unevenness of wire ropes according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0015] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0016] The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0017] The following description, in conjunction with the accompanying drawings, details a specific scheme for a robot-based method for measuring the unevenness of wire ropes provided by this invention. Example

[0018] Please see Figure 1 The diagram illustrates a flowchart of a robot-based method for measuring the unevenness of a wire rope, according to an embodiment of the present invention. The method includes the following steps: Step S1, Robot positioning and initial calibration: The robot automatically identifies the starting point of the wire rope and performs precise positioning through the integrated vision system, which captures images of the wire rope. Step S1 further includes the following sub-steps: S1-1: Select a six-degree-of-freedom high-precision industrial robot. Fix a high-resolution global CCD camera above the measurement scene. The field of view covers the entire area of ​​the steel wire rope to be measured. Use the fixed global camera to acquire two-dimensional images of the scene. Use image processing technology to identify the approximate position and direction of the steel wire rope, and provide navigation information for the robot to move to the approximate starting point. A global camera captures scene photos containing steel wire ropes, and a deep learning-based semantic segmentation model is applied. This semantic segmentation model is specially trained to quickly identify the type of steel wire rope and effectively distinguish the steel wire rope from the complex background. The segmented wire rope region is skeletonized and extracted to obtain the pixel-level centerline of the wire rope. The centerline is then fitted to a straight line using the least squares method to obtain its equation in the image coordinate system. Global camera hand-eye calibration parameters are then used to transform the image centerline into a coarse three-dimensional straight line L in the robot's base coordinate system. g This straight line defines the approximate direction and position of the wire rope, and the control system guides the robot to move to L. g A safe preparatory position above the starting end.

[0019] S1-2 After the robot is in place, the crosshair laser integrated at the end is projected onto the surface of the steel wire rope. The resulting laser line is deformed. By analyzing the shape of the deformed laser line, the three-dimensional contour of the steel wire rope surface can be calculated, and then the central axis can be accurately calculated. The crosshair laser at the robot's end effector is activated, projecting two laser lines, one horizontal and one vertical, onto the steel cable. A macro CMOS camera simultaneously captures images of the projection area. Since the steel cable is cylindrical, the horizontal laser line becomes an arched arc, and the vertical laser line becomes a curve that undulates with the surface of the steel cable. The horizontal laser arched arc in the image is extracted and regarded as the outer contour of the steel cable on that cross-section. An ellipse fitting algorithm is used to fit this contour, and the center of the fitted ellipse is the approximate center point of the cross-section.

[0020] The vertical laser line is extracted. The robot's end effector moves a short distance along the approximate direction of the steel wire rope, continuously acquiring multiple frames of images during this process. The centerline of the vertical laser line in each frame is extracted. The center points of multiple cross-sections obtained from the horizontal laser line at different locations are fused with the point cloud extracted from the vertical laser line. A random sampling consensus algorithm is used to fit a spatial straight line to the fused 3D point set, ultimately obtaining a high-precision steel wire rope axis L. p Spatial line fitting is usually defined by a point P0(x0,y0,z0) and a direction vector V(a,b,c). The goal of the RANSAC algorithm is to find P0 and V that minimize the sum of the distances from all interior points to the line.

[0021] The formula for the distance from a point to a line is: d=|(P i -P0)×V| / |V| Where × represents the cross product of vectors, and the optimal L is obtained by minimizing the sum of d for all interior points. p .

[0022] S1-3, Dynamically establish and verify the robot's measurement coordinate system, and use the precisely calibrated wire rope axis L p As a benchmark, the robot's measurement coordinate system is dynamically defined, and the accuracy of this coordinate system is verified: The Z-axis of the coordinate system, i.e., the measurement direction, is related to the L-axis.p The direction vectors coincide with V, and the positive direction is the direction of measurement forward movement; Let the origin O be L p The preset point that is closest to the origin of the robot's base coordinate system is usually near the beginning of the wire rope; The X-axis is defined as the direction perpendicular to the Z-axis and pointing directly above the surface of the wire rope; The Y-axis is determined by the right-hand rule, thus completing the construction of the coordinate system {M}; The homogeneous transformation matrix from the robot's base coordinate system {B} to the newly defined measurement coordinate system {M} is precisely calculated. All sensor data will be uniformly transformed to the {M} coordinate system for processing. The coordinate transformation relationship is established. The robot moves a small distance along the newly defined Z-axis, and the process of the previous step is repeated. The axis L of a small section of the wire rope is calculated again. v Comparing L v With L p The direction and position deviations are checked. If the deviation is within the threshold, the calibration is successful; otherwise, fine-tuning iterations are performed.

[0023] In unstructured environments, the robotic system can automatically and accurately locate the starting end of the steel wire rope under test and establish a robot measurement coordinate system based on the wire rope axis. Combining the wide-angle global vision and the micro-range crosshair laser, it achieves seamless integration from macroscopic navigation to microscopic precision measurement. It has strong resistance to environmental interference. One laser line is used for cross-sectional center positioning, and another for axial guidance. Through multi-frame data fusion and RANSAC fitting, the spatial axis can be more accurately restored. The coordinate system is dynamically generated according to the actual spatial posture of the steel wire rope before each measurement, perfectly adapting to the uncertainties such as bending and sag of the steel wire rope. It provides a unique and accurate benchmark for subsequent steps and adds a verification closed loop to ensure the reliability of the initial calibration. It guarantees the measurement accuracy of the entire system from the first step and avoids error propagation and accumulation.

[0024] Step S2: Wire rope surface pretreatment. The robot is equipped with a non-contact cleaning device to clean the surface of the wire rope and remove impurities. Step S2 further includes the following sub-steps: The S2-1, a six-DOF high-precision industrial robot, integrates a pre-processing actuator at its end effector. Its core is a ring-shaped processing head that can circle the steel wire rope 360 ​​degrees without contacting it. An ultrasonic transducer array is installed on the inner wall of this ring head to generate low-frequency ultrasonic waves. Multiple precision-designed tangential vortex cyclone nozzles are installed on the ring head and connected to a high-pressure air source to generate a high-speed rotating air knife. A ring-shaped electrostatic adsorption ring with a high-voltage electrostatic generator adsorbs the detached particles. A miniature spectral confocal sensor integrated within the ring head monitors surface cleanliness in real time. The high-pressure air source, ultrasonic generator, and high-voltage electrostatic generator all move with the robot via pipelines. Oil stains can cause laser lines to produce special diffuse reflection and color changes, water films can produce specular reflection spots, and rust and dust can change the texture features of the surface. In the image saved in the previous step, the clarity, color space features and texture features of the laser line outline are extracted. Sharpness of the laser line outline: Calculate the average gradient value G using the image gradient operator. avg Oil and water films can reduce gradient values.

[0025] Color space characteristics: In the HSV color space, analyze the saturation (S) and lightness (V) of the laser line area. Oil stains usually cause specific changes in S and V.

[0026] Texture features: Local Binary Pattern (LBP) analysis was used to analyze the texture uniformity of the regions on both sides of the laser line.

[0027] Construct a pollution index C i The index is a weighted sum of the above features: C i =w1*(1-G avg norm)+w2*ΔS+w3*ΔV+w4*LBP Where w1~w4 represent weighting coefficients, and G is determined experimentally. avg `norm` represents the normalized gradient value; `ΔS` represents the saturation difference between the current sample and the clean sample after normalization to 0-1 using the Min-Max formula; `ΔV` represents the brightness channel difference between the current sample and the clean sample after normalization to 0-1 using the Min-Max formula; `LBP` is the texture entropy, based on C... i The values ​​and dominant characteristics are normalized to 0-1 using the Min-Max formula. Based on the pollution index, the pollution level is divided into light, moderate, and heavy, and the pollution type is preliminarily determined. S2-2, intelligent matching and execution of multimodal pretreatment parameters, dynamically adjusts the intensity, action time and execution order of different cleaning modes for different pollution combinations, presets an expert knowledge base, inputs pollution type and pollution degree, and finally outputs the optimal combination of pretreatment parameters, including ultrasonic power, cyclone pressure, electrostatic voltage and treatment time; For example, for heavy oil stains, high ultrasonic power + medium cyclone pressure + low electrostatic voltage is used. First, the ultrasonic waves break up the oil film, then the cyclone blows away the large droplets, and finally the electrostatic adsorption is used to adsorb the residual oil mist. For light dust, low ultrasonic power or off + high cyclone pressure + medium electrostatic voltage is used to directly blow away the dust with the cyclone as the main force, and electrostatics prevent dust from being stirred up.

[0028] The robot, carrying a ring-shaped processing head, moves at a constant speed along the Z-axis of the measurement coordinate system defined in step one. Simultaneously, it activates the corresponding pre-processing module based on the aforementioned parameter combination. Low-frequency ultrasonic waves generate cavitation effects in the medium on the surface of the steel wire rope, forming and bursting microbubbles, generating a powerful impact force that peels contaminants from the substrate. The intensity of this effect is related to power and frequency. High-pressure gas is ejected from a tangential nozzle, forming a high-speed rotating vortex field within the annular cavity. This air knife can generate extremely high local shear force with relatively low overall gas consumption, efficiently sweeping away the peeled contaminants. Its purging efficiency is related to the cyclone pressure and the airflow rotation angular velocity. High pressure is applied to the electrostatic adsorption ring, generating a strong electrostatic field around it. Micron-sized and submicron-sized charged or uncharged particles that are blown up by the cyclone but not completely carried away will be attracted to the inner wall of the electrostatic ring due to induction charging or polarization, completely moving them away from the measurement area.

[0029] S2-3 utilizes a miniature spectral confocal sensor integrated into the pretreatment head to perform point measurements at the same location immediately after cleaning. This sensor determines the precise distance to the surface by analyzing the spectrum of reflected light and analyzes the intensity of reflected light, which can indirectly assess the surface cleanliness. If the standard is not met, the robot control system will record the coordinates of the location and may automatically return to the point after completing a trip. Based on the real-time measured pollution characteristics, it will dynamically adjust the pretreatment parameters for secondary treatment until the standard is met. For example, if the reflectivity is low, it may indicate oil residue, and the ultrasonic treatment time will be increased.

[0030] Multimodal synergy combines three physical principles—ultrasound, vortex cyclone, and electrostatics—to target the three stages of adhesion, purging, and suspended residue, forming a complete cleaning chain with efficiency far exceeding that of a single method. High-precision spectral confocal sensors are introduced for in-situ quality inspection, forming a closed-loop control system to ensure reliable pretreatment results. This provides a high-quality surface guarantee for subsequent preliminary scanning of unevenness, avoiding misjudgments caused by incomplete cleaning. The entire process is completely non-contact, avoiding methods such as brushing that may scratch the wire rope or cause secondary pollution.

[0031] Step S3: Preliminary scan of unevenness. The robot uses a multi-sensor array to move quickly along the steel cable to perform a preliminary scan and identify obvious uneven areas. Step S3 further includes the following sub-steps: The S3-1 is a six-DOF high-precision industrial robot. Its end effector is equipped with a scanning module. The core sensors of the scanning module include: a high-frame-rate linear CMOS camera for high-speed surface texture acquisition; a near-infrared (NIR) linear camera for detecting potential damage below the surface; a lightweight, low-friction encoder-integrated wheel assembly that maintains contact with the steel cable to accurately measure the robot's displacement along the cable; dual-sided high-brightness LED linear light sources; a near-infrared backlight source for transmission imaging; and a high-performance FPGA board for real-time synchronization and preprocessing of sensor data. S3-2 utilizes the differences in the interaction between light and steel wire rope in different wavelength bands to obtain complementary surface and subsurface information. Visible light imaging reflects macroscopic geometry and surface texture, while near-infrared light is more sensitive to density changes caused by internal corrosion of materials. The robot's end effector moves at a relatively high, constant speed along the Z-axis of the measurement coordinate system defined in step one. The speed is maintained by real-time feedback from the encoder wheel assembly. Each minute movement of the encoder wheel assembly generates a trigger pulse, which synchronously triggers both the visible light array camera and the NIR line array camera to acquire one line of images. The processing unit precisely controls the timing to ensure strict spatial alignment of the two image streams. After continuous acquisition, two two-dimensional image data streams (I) are obtained. v (x,y) and I n (x,y), where the x-direction represents the axial direction of the wire rope and the y-direction represents the circumferential direction of the rope; S3-3, Real-time anomaly detection based on image sequences. A normal steel wire rope surface should exhibit a continuous and regular texture pattern under uniform illumination. Any unevenness will disrupt this pattern, manifesting as abrupt changes in image grayscale, texture, or edge features. The average grayscale gradient AGG, local binary mode variance LBPV, and transmittance anomaly index TAI based on NIR images are calculated in parallel on the FPGA for each image row or small image block.

[0032] Since the wire rope itself may have slow diameter changes or slight bending, a fixed threshold is not applicable. Therefore, a statistical process control method based on a sliding window is adopted to calculate the mean μ and standard deviation σ of each feature value in the current window and set the control limit. For example, anomaly criteria, if the feature value f of the current row... c Satisfy |f c If -μ|>k*σ, then the point is marked as a "potential outlier", where k is a constant that can be adjusted according to requirements.

[0033] Only when multiple features exceed the threshold simultaneously, or when a certain feature is severely out of bounds, is the area ultimately determined to be a suspicious segment, greatly reducing the possibility of false alarms.

[0034] S3-4 aggregates consecutive or closely spaced anomalies into suspicious regions. For each region, it calculates the start and end positions, the average value of the characteristic deviations of all anomalies within the region, and preliminarily determines whether the anomaly is a surface geometric defect or an internal material defect based on the characteristics that trigger it. Finally, it outputs a structured list.

[0035] By combining visible light surface texture imaging and near-infrared transmission imaging, it is possible to simultaneously detect surface geometric irregularities and subsurface material defects, which greatly improves the comprehensiveness and reliability of the initial scan. By adopting a statistical process control method, the threshold can be dynamically adjusted according to the background texture of the wire rope itself, effectively overcoming the interference caused by uneven global illumination or natural bending of the rope, and significantly reducing the false alarm rate.

[0036] Step S4: High-precision three-dimensional contour measurement. The robot equipped with the scanning module moves slowly along the steel wire rope to obtain detailed three-dimensional surface contour data. Step S4 further includes the following sub-steps: The S4-1 high-precision industrial robot end effector, equipped with a dedicated scanning module, also includes several ring-shaped high-precision spectral confocal probes, a miniaturized Mirau-type white light interferometer objective lens, and a high-speed galvanometer system. Based on the suspicious area segments generated in the previous step, the robot's motion path and measurement mode are dynamically planned to achieve optimal allocation of measurement resources. The robot path planning algorithm sorts all suspicious areas according to their positional order and calculates the shortest movement path to avoid unnecessary round trips. Areas with high anomaly intensity are marked as fine measurement areas, where the robot will move at low speed and simultaneously activate the spectral confocal array and the white light interferometer module. Areas with low anomaly intensity are marked as rapid survey areas, where the robot moves at higher speed and only activates the spectral confocal array for low-density data acquisition, used for contour verification and transition area measurement. The field of view and sampling density of the white light interferometer scan are preset according to the anomaly intensity. High-resolution areas use a small field of view and high-density scanning to capture details. S4-2, the robot moves to the starting point of the first suspicious area, and multiple spectral confocal probes work simultaneously. As the robot moves along the Z-axis, it rapidly acquires the three-dimensional contour point cloud P of the measurement strip on the surface of the wire rope. ccd (x, y, z), a confocal probe is selected and designated as the master probe, whose measured value d m It is sent to the control system in real time, and the control system according to d m and the preset ideal working distance d i The focal deviation Δd is calculated as follows: Δd=d m -d i ; Based on the PID control algorithm, the PZT active focusing mechanism is driven in real time to adjust the Z-axis position of the white light interferometer objective, ensuring that its focus always tracks the surface undulations of the wire rope. u(t) = K p *e(t)+K i *∫e(t)dt+K d *de(t) / dt Where u(t) represents the control voltage output to PZT, e(t) represents the current focus deviation Δd, and K p K i K d These represent the proportional, integral, and differential coefficients, respectively. The mechanism ensures that white light interferometry is always in optimal working condition, avoiding signal attenuation or measurement failure caused by defocusing.

[0037] White light interferometry measures surface morphology by analyzing the characteristics of interference fringes generated by a broadband light source as a function of optical path difference. A robot moves to a predetermined fine measurement point, triggering a white light interferometric scan. A PZT-driven reference mirror moves N steps at equal steps Δz. At each step, a hyperspectral camera acquires an interferogram I. n (x,y), where n=1,2,...,N; For each pixel (x, y), its light intensity value changes with the position of the reference mirror to form an interference signal envelope. The phase is calculated using a five-step phase shift algorithm. Due to the periodicity of the phase, the obtained phase is a wrapped phase. Phase unwrapping is required to obtain a continuous phase distribution, and finally, a small area of ​​three-dimensional topography data with nanometer-level resolution is obtained. Accurate registration and fusion of multi-source point cloud data: All data are converted to the measurement coordinate system established in step one. The common features of the two point clouds in the overlapping area are used to perform fine registration using the iterative nearest point algorithm, eliminating the systematic errors caused by robot positioning errors and sensor installation deviations. The 3D topographic data of the fine measurement area replaces the spectral confocal cloud data of the corresponding location; the spectral confocal cloud data of the rapid survey area is retained, and finally a complete, seamless 3D contour model with different precision in different areas is generated.

[0038] Dynamic optimization of measurement paths and parameters concentrates high-cost, high-precision measurement resources only where they are most needed, greatly improving overall efficiency. The introduction of closed-loop focus control based on PID feedback solves the defocusing problem when performing high-precision optical measurements on undulating surfaces, ensuring the stability of white light interferometric measurement conditions, which is key to obtaining reliable data. Through precise registration algorithms, point clouds from different sources and with different accuracies are fused into a unified, high-quality dataset.

[0039] Step S5: Real-time data preprocessing. During the robot's movement, the collected raw data is preprocessed in real time. Step S5 further includes the following sub-steps: S5-1, streaming data reception and cache management, the preprocessing system receives and processes the data generated in the previous step in a streaming manner to avoid data accumulation. The system divides the continuous point cloud stream into data blocks according to a fixed length or fixed time interval. When one piece of data is being processed, the next piece of data is stored in another cache area to achieve seamless connection. Each data block is bound to its corresponding robot pose, timestamp, sensor ID metadata. S5-2, the robot's end effector integrates an inertial measurement unit (IMU). The acceleration and angular velocity measured by the IMU are double-integrated to estimate the end effector displacement caused by vibration. Temperature sensors (T_sensors) are placed at key locations on the robot and sensors. Based on the material's coefficient of thermal expansion α, the thermal deformation δ is calculated. δ=α*L*(T s -T c ) Where L represents the feature length, T s T represents temperature sensor data. c The temperature during system calibration is represented by the robot's encoder feedback and kinematic model. The small deviation between the theoretical pose and the actual commanded pose is calculated, and the preceding error terms are combined to form a comprehensive error vector for each point P in the current data block. o Compensation is performed on (x,y,z).

[0040] Point clouds contain both realistic surface details and random noise. Traditional filters blur the details while denoising. This paper uses a parallel search based on a KD-tree to quickly estimate the normal vector of each point and calculate the curvature c of each point. i Set a curvature threshold to classify points into "high feature points" and "flat points": c i ≈λ min / (λ0+λ1+λ2), Where λ represents the eigenvalue of the covariance matrix, a larger search radius is used for strong smoothing filtering in the "flat point" region to effectively suppress noise, while a very small search radius is used for light filtering in the "high feature point" region, or filtering is skipped directly to preserve the sharp edges and details of the defects to the greatest extent. After the above filtering, obvious outliers are removed. The space is divided into a uniform voxel grid, with only one point retained in each voxel. Larger voxels are used in the rapid survey area, and smaller voxels are used in the fine measurement area to simplify multi-resolution data.

[0041] For large-scale point cloud data, a pipeline-style processing flow was designed. Combining double buffering and parallel computing, real-time preprocessing was achieved. Multiple error sources such as robot vibration, thermal deformation, and kinematic errors were integrated to establish a comprehensive compensation model, which improved the absolute accuracy of the data at the system level. This surpasses simple sensor denoising and improves the nonlocal mean filtering algorithm, enabling it to distinguish between noise and real surface features, and avoiding the smoothing effect of traditional filters on small defects.

[0042] Step S6: Automatically generate a detailed measurement report, including an unevenness distribution map, defect location coordinates, safety assessment, and recommended measures; In step S6, the preprocessed data from steps 3-5 are integrated using Python Pandas, and a heatmap and 3D model are generated using Matplotlib. The risk level is assessed based on the ISO 4309 standard algorithm. Finally, a PDF report is output by rendering a LaTeX template using Jinja2 and transmitted to the cloud using the MQTT protocol. A built-in data conflict detection mechanism ensures the reliability of the report. When data is missing, the original scan path log is retained, and conflicting data triggers a warning flag.

[0043] Step S7: After the robot completes the measurement, it automatically resets and predicts the remaining life of the wire rope based on the measurement results. In step S5, after the robot completes the measurement, it dynamically calculates the remaining lifespan using a built-in lifespan prediction algorithm. This algorithm is trained based on historical wear data and automatically triggers a three-level early warning system: a green maintenance reminder is sent when the remaining lifespan exceeds 30 days, a yellow warning is triggered between 15 and 30 days, and a red emergency alarm is activated within 15 days. At the same time, the maintenance time, defect location, and suggested measures are pushed to the cloud management platform through the NB-IoT module, realizing closed-loop management from detection to decision-making.

[0044] In this way, a robot-based method for measuring the unevenness of wire ropes can be achieved.

[0045] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application 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 of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A robot-based method for measuring the unevenness of wire ropes, characterized in that, The method includes: Step S1, Robot positioning and initial calibration: The robot automatically identifies the starting point of the wire rope and performs precise positioning through the integrated vision system, which captures images of the wire rope. Step S2: Wire rope surface pretreatment. The robot is equipped with a non-contact cleaning device to clean the surface of the wire rope and remove impurities. Step S3, preliminary scan of unevenness: The robot uses a multi-sensor array to move quickly along the steel cable to perform a preliminary scan and identify obvious uneven areas. Step S4: High-precision three-dimensional contour measurement. The robot equipped with the scanning module moves slowly along the steel wire rope to obtain detailed three-dimensional surface contour data. Step S5: Real-time data preprocessing. During the robot's movement, the collected raw data is preprocessed in real time. Step S6: Automatically generate a detailed measurement report, including an unevenness distribution map, defect location coordinates, safety assessment, and recommended measures; Step S7: After the robot completes the measurement, it automatically resets and predicts the remaining life of the wire rope based on the measurement results. Step S3 further includes the following sub-steps: The S3-1 is a six-DOF high-precision industrial robot. Its end effector is equipped with a scanning module. The core sensors of the scanning module include: a high-frame-rate linear CMOS camera for high-speed surface texture acquisition; a near-infrared (NIR) linear camera for detecting potential damage below the surface; a lightweight, low-friction encoder-integrated wheel assembly that maintains contact with the steel cable to accurately measure the robot's displacement along the cable; dual-sided high-brightness LED linear light sources; a near-infrared backlight source for transmission imaging; and a high-performance FPGA board for real-time synchronization and preprocessing of sensor data. S3-2 utilizes the differences in the interaction between light and steel wire rope in different wavelength bands to obtain complementary surface and subsurface information. Visible light imaging reflects macroscopic geometry and surface texture, while near-infrared light is more sensitive to density changes caused by internal corrosion of materials. The robot end effector moves along the Z-axis of the measurement coordinate system at a relatively high constant speed. The movement speed is kept stable by real-time feedback from the encoder wheel assembly. Each time the encoder wheel assembly moves a tiny distance, it generates a trigger pulse. This pulse synchronously triggers the visible light array camera and the NIR line array camera to each acquire one line of images. The timing is precisely controlled by the processing unit to ensure that the two images are strictly aligned in space. After continuous acquisition, two two-dimensional image data streams are obtained respectively. S3-3 is a real-time anomaly detection based on image sequences. Under uniform lighting, the surface of a normal steel wire rope should exhibit a continuous and regular texture pattern. Any unevenness will destroy this pattern, manifesting as abrupt changes in image grayscale, texture, or edge features. The average grayscale gradient, local binary mode variance, and transmittance anomaly index based on NIR images are calculated in parallel on the FPGA for each image row or small image block. When multiple features exceed the threshold simultaneously, or when a certain feature is severely out of standard, it is finally determined to be a suspicious area segment, greatly reducing the possibility of false alarms. S3-4: Aggregate consecutive or closely spaced anomalous points into suspicious regions. For each region, calculate its start and end positions, and the average value of the characteristic deviations of all anomalous points within the region. Based on the characteristics that trigger the anomaly, make a preliminary judgment on whether it is a surface geometric defect or an internal material defect. Step S4 further includes the following sub-steps: The S4-1 high-precision industrial robot end effector is equipped with a dedicated scanning module, which also includes several high-precision spectral confocal probes distributed in a ring, a miniaturized Mirau-type white light interferometer objective lens, and a high-speed galvanometer system. Based on the generated suspicious area segments, the robot path planning algorithm sorts all the suspicious areas in order of position and calculates the shortest movement path to avoid empty round trips. S4-2, the robot moves to the starting point of the first suspicious area, and multiple spectral confocal probes work simultaneously. As the robot moves along the Z-axis, it rapidly acquires the three-dimensional contour point cloud P of the measurement strip on the surface of the wire rope. ccd (x, y, z), a confocal probe is selected and designated as the master probe, whose measured value d m It is sent to the control system in real time, and the control system according to d m and the preset ideal working distance d i The focal deviation Δd is calculated as follows: Δd=d m -d i ; Based on the PID control algorithm, the PZT active focusing mechanism is driven in real time to adjust the Z-axis position of the white light interferometer objective, ensuring that its focus always tracks the surface undulations of the wire rope. u(t)=K p *e(t)+K i *∫e(t)dt+K d *de(t) / dt Where u(t) represents the control voltage output to PZT, e(t) represents the current focus deviation Δd, and K p K i K d These represent the proportional, integral, and differential coefficients, respectively. This algorithm ensures that white light interferometry is always in optimal working condition, avoiding signal attenuation or measurement failure caused by defocusing. Step S1 further includes the following sub-steps: S1-1: Select a six-degree-of-freedom high-precision industrial robot. Fix a high-resolution global CCD camera above the measurement scene. The field of view covers the entire area of ​​the steel wire rope to be measured. Use the fixed global camera to acquire two-dimensional images of the scene. Use image processing technology to identify the approximate position and direction of the steel wire rope, and provide navigation information for the robot to move to the approximate starting point. S1-2 After the robot is in place, the crosshair laser integrated at the end is projected onto the surface of the steel wire rope. The resulting laser line is deformed. By analyzing the shape of the deformed laser line, the three-dimensional contour of the steel wire rope surface can be calculated, and then the central axis can be accurately calculated. S1-3, Dynamically establish and verify the robot's measurement coordinate system, and use the precisely calibrated wire rope axis L p As a benchmark, the robot's measurement coordinate system is dynamically defined, and its accuracy is verified. The homogeneous transformation matrix from the robot's base coordinate system {B} to the newly defined measurement coordinate system {M} is precisely calculated. All sensor data will be uniformly transformed to the {M} coordinate system for processing. The coordinate transformation relationship is established. The robot moves a small distance along the newly defined Z-axis, repeating the previous step, and the axis L of a small section of the wire rope is calculated again. v Comparing L v With L p The direction and position deviations are checked; if the deviations are within the threshold, the calibration is successful. Step S2 further includes the following sub-steps: The S2-1, a six-DOF high-precision industrial robot, integrates a pre-processing actuator at its end effector. Its core is a ring-shaped processing head that can circle the steel wire rope 360 ​​degrees without contacting it. An ultrasonic transducer array is installed on the inner wall of this ring head to generate low-frequency ultrasonic waves. Multiple precision-designed tangential vortex cyclone nozzles are mounted on the ring head to generate high-speed rotating air knives. A ring-shaped electrostatic adsorption ring with a high-voltage electrostatic generator adsorbs the detached particles. A miniature spectral confocal sensor integrated within the ring head monitors surface cleanliness in real time. The high-pressure air source, ultrasonic generator, and high-voltage electrostatic generator all move with the robot via pipelines. Oil stains can cause laser lines to produce special diffuse reflection and color changes, water films can produce specular reflection spots, and rust and dust can change the texture features of the surface. In the image saved by the CCD camera, the clarity, color space features and texture features of the laser line outline are extracted. Construct a pollution index C i The index is a weighted sum of the above features: C i =w1*(1-G avg norm)+w2*ΔS+w3*ΔV+w4*LBP Where w1~w4 represent weighting coefficients, and G is determined experimentally. avg `norm` represents the normalized gradient value; `ΔS` represents the saturation difference between the current sample and the clean sample after normalization to 0-1 using the Min-Max formula; `ΔV` represents the brightness channel difference between the current sample and the clean sample after normalization to 0-1 using the Min-Max formula; `LBP` is the texture entropy, based on C... i The values ​​and dominant characteristics are normalized to 0-1 using the Min-Max formula. Based on the pollution index, the pollution level is divided into light, moderate, and heavy, and the pollution type is preliminarily determined. S2-2, intelligent matching and execution of multimodal pretreatment parameters, dynamically adjusts the intensity, action time and execution order of different cleaning modes for different pollution combinations, presets an expert knowledge base, inputs pollution type and pollution degree, and finally outputs the optimal combination of pretreatment parameters, including ultrasonic power, cyclone pressure, electrostatic voltage and treatment time; S2-3 utilizes a miniature spectral confocal sensor integrated into the pretreatment head to perform point measurements at the same location immediately after cleaning. This sensor determines the precise distance to the surface by analyzing the spectrum of reflected light and simultaneously analyzes the intensity of reflected light, which can indirectly assess the surface cleanliness.

2. The method for measuring the unevenness of a wire rope based on a robot as described in claim 1, characterized in that: In step S4, white light interferometry measures the surface morphology by analyzing the characteristics of the interference fringes generated by the broadband light source as a function of optical path difference. The robot moves to the predetermined fine measurement point and triggers white light interferometry scanning. At each step position, the hyperspectral camera acquires an interferogram. For each pixel (x, y), its light intensity value changes with the position of the reference mirror to form an interference signal envelope. The phase is calculated using a phase-shifting algorithm. Due to the periodicity of the phase, the obtained phase is a wrapped phase. Phase unwrapping is required to obtain a continuous phase distribution. Finally, a small area of ​​nanometer-resolution three-dimensional morphology data is obtained. For the accurate registration and fusion of multi-source point cloud data, all data are converted to the measurement coordinate system established in step one. The iterative nearest point algorithm is used to perform fine registration by utilizing the common features of the two point clouds in the overlapping area, thereby eliminating the systematic errors caused by robot positioning errors and sensor installation deviations.

3. The method for measuring the unevenness of a wire rope based on a robot as described in claim 1, characterized in that: Step S5 further includes the following sub-steps: S5-1, streaming data reception and cache management, the preprocessing system receives and processes the data generated in the previous step in a streaming manner to avoid data accumulation. The system divides the continuous point cloud stream into data blocks according to a fixed length or fixed time interval. When one piece of data is being processed, the next piece of data is stored in another cache area to achieve seamless connection. Each data block is bound to its corresponding robot pose, timestamp, sensor ID metadata. S5-2, the robot's end effector integrates an inertial measurement unit (IMU). The acceleration and angular velocity measured by the IMU are double-integrated to estimate the end effector displacement caused by vibration. Temperature sensors (T_sensors) are placed at key locations on the robot and sensors. Based on the material's coefficient of thermal expansion α, the thermal deformation δ is calculated. δ=α*L*(T s -T c ) Where L represents the feature length, T s T represents temperature sensor data. c The temperature during system calibration is represented by the robot's encoder feedback and kinematic model. The small deviation between the theoretical pose and the actual commanded pose is calculated, and the preceding error terms are combined to form a comprehensive error vector for each point P in the current data block. o Compensation is performed on (x,y,z).

4. The method for measuring the unevenness of a wire rope based on a robot as described in claim 3, characterized in that: In step S5, the point cloud contains both real surface details and random noise. Traditional filters blur the details while denoising. A parallel search based on a KD-tree is used to quickly estimate the normal vector of each point and calculate the curvature c of each point. i A curvature threshold is set to divide points into "high feature points" and "flat points". For "flat point" regions, a larger search radius is used for strong smoothing filtering to effectively suppress noise. For "high feature point" regions, a very small search radius is used for light filtering, or filtering is skipped directly, to preserve the sharp edges and details of defects to the greatest extent. After the above filtering, obvious outliers are removed. The space is divided into a uniform voxel grid, with only one point retained in each voxel. Larger voxels are used in the rapid survey area, and smaller voxels are used in the fine measurement area to simplify multi-resolution data.

5. The method for measuring the unevenness of a wire rope based on a robot as described in claim 1, characterized in that: In step S6, the preprocessed data from steps 3-5 are integrated using Python Pandas, and a heatmap and 3D model are generated using Matplotlib. The risk level is assessed based on the ISO 4309 standard algorithm. Finally, a PDF report is output by rendering a LaTeX template using Jinja2 and transmitted to the cloud using the MQTT protocol. A built-in data conflict detection mechanism ensures the reliability of the report. When data is missing, the original scan path log is retained, and conflicting data triggers a warning flag.

6. The method for measuring the unevenness of a wire rope based on a robot as described in claim 1, characterized in that: In step S7, after the robot completes the measurement, it dynamically calculates the remaining lifespan using a built-in lifespan prediction algorithm. This algorithm is trained based on historical wear data and automatically triggers a three-level early warning system: a green maintenance reminder is sent when the remaining lifespan exceeds 30 days, a yellow warning is triggered between 15 and 30 days, and a red emergency alarm is activated within 15 days. At the same time, the maintenance time, defect location, and suggested measures are pushed to the cloud management platform through the NB-IoT module, realizing closed-loop management from detection to decision-making.

Images