Method and device for detecting defects in an elevator rope

Abstract

This application provides a method and apparatus for detecting elevator cable defects. The method includes: acquiring a multimodal defect perception dataset through an integrated cable defect detection probe module, preprocessing it to obtain a defect perception association dataset, performing feature extraction and dynamic fusion processing based on the defect perception association dataset to obtain a defect perception fusion feature vector, and inputting this vector into an edge detection terminal and a cloud-based preset model for processing to obtain first cable defect detection data and second cable defect detection data. A preliminary cable defect warning is issued based on the first cable defect detection data. The first and second cable defect detection data are compared to further evaluate the cable defects. The remaining lifespan is predicted using a preset digital twin model. This application achieves intelligent detection of elevator cable defects by acquiring and fusing multimodal data, performing edge and cloud-based analysis and detection, and combining digital twin model simulation analysis.

Description

Technical Field

[0001] This application relates to the field of elevator safety monitoring technology, and more specifically, to a method and apparatus for detecting defects in elevator cables. Background Technology

[0002] Elevator cables are typically made of steel wire rope. Defect detection is crucial for ensuring the safe operation of elevators. Traditional cable safety inspections rely primarily on manual visual inspection and manual instruments, which are inefficient, subjective, have a high rate of missed detections, cannot detect internal defects, have long inspection cycles, and cannot achieve real-time monitoring. Traditional automated inspections mainly rely on single-modal automated inspections, such as traditional magnetic leakage or simple vision. Single magnetic leakage inspection is not good at detecting cracks or shallow wear parallel to the cable axis, and the magnetic signal is easily interfered with by vibration and oil. Single vision inspection cannot detect internal damage and is easily affected by light and surface cleanliness. Currently, traditional inspection technologies lack in-depth analysis and predictive capabilities. Defect detection is mainly limited to the level of determining the presence or absence of defects, and cannot accurately classify defect types, quantify their severity, or predict cable performance degradation and remaining service life based on historical data trends, making it difficult to achieve true predictive maintenance.

[0003] Effective technical solutions are urgently needed to address the above problems. Summary of the Invention

[0004] The purpose of this application is to provide a method and device for detecting elevator cable defects. It can achieve intelligent detection of elevator cable defects by acquiring and processing multimodal data, analyzing and detecting from the edge and cloud, and combining digital twin model simulation analysis.

[0005] This application also provides a method for detecting defects in elevator cables, including the following steps: The multimodal defect perception dataset is obtained by the integrated cable defect detection probe module, and then preprocessed to obtain the defect perception association dataset. Based on the defect-aware associated dataset, feature extraction and dynamic fusion processing are performed to obtain a defect-aware fusion feature vector. The defect perception fusion feature vector is input into a preset multi-task parallel defect detection model for processing to obtain the first cable defect detection data. The cable defect detection data is analyzed and processed to obtain cable defect evaluation parameters, which are then compared with preset cable defect early warning thresholds. Based on the threshold comparison results, a cable defect early warning is output.

[0006] Optionally, in the elevator cable defect detection method described in this application, the step of acquiring a multimodal defect perception dataset through an integrated cable defect detection probe module and preprocessing it to obtain a defect perception association dataset includes: The multimodal defect perception dataset of a preset cable segment is obtained through the integrated cable defect detection probe module, including TMR magnetic sensing data, multispectral linear array image slices and cable perception point cloud data. The TMR magnetic sensing data is filtered, differentially amplified, and synthesized to obtain a comprehensive leakage magnetic vector. The multispectral linear array image slices are subjected to image stitching, geometric correction, image enhancement and noise reduction processing to obtain a multispectral linear array unfolded image; The cable sensing point cloud data is subjected to cylindrical model fitting and diameter deviation sequence, ellipticity and surface quality analysis to obtain cable sensing geometric feature sequence data. The integrated magnetic leakage vector, multispectral linear array unfolded image, and cable sensing geometric feature sequence data are spatiotemporally aligned, and data blocks are divided and data association processing is performed according to preset cable segments to obtain a defect sensing association dataset.

[0007] Optionally, in the elevator cable defect detection method described in this application, the step of performing feature extraction and dynamic fusion processing based on the defect-aware association dataset to obtain a defect-aware fusion feature vector includes: Feature extraction is performed based on the defect-aware association dataset to obtain magnetic feature vectors, visual feature vectors, and geometric feature vectors. The magnetic feature vector, visual feature vector, and geometric feature vector are analyzed and processed by a self-attention module to obtain the weight values ​​of the magnetic feature vector, visual feature vector, and geometric feature vector. The magnetic feature vector, visual feature vector, and geometric feature vector, along with their corresponding weight values, are dynamically weighted and fused to obtain a defect-aware fusion feature vector.

[0008] Optionally, in the elevator cable defect detection method described in this application, the step of inputting the defect perception fusion feature vector into a preset multi-task parallel defect detection model for processing to obtain first cable defect detection data includes: The defect perception fusion feature vector is input into the preset edge end multi-task parallel defect detection model of the edge detection end for processing to obtain the first cable defect detection data corresponding to the preset cable segment. The first cable defect detection data includes defect category feature data, corresponding defect quantification indicators, and defect location data; The defect quantification indicators include the number of broken wires in the cable and the cable wear rate.

[0009] Optionally, in the elevator cable defect detection method described in this application, the step of analyzing and processing the first cable defect detection data to obtain cable defect evaluation parameters, comparing them with a preset cable defect early warning threshold, and outputting a cable defect early warning based on the threshold comparison result includes: Based on the defect category feature data, query the preset weight value list and the preset defect evaluation correction coefficient mapping table to obtain the cable breakage weight value, cable wear weight value and defect evaluation correction coefficient; Based on the cable breakage weight value and the cable wear weight value, the number of cable breakage wires and the cable wear rate are weighted and summed to obtain the initial evaluation parameters of cable defects. The initial evaluation parameters of the cable defect are corrected according to the defect evaluation correction coefficient to obtain the cable defect evaluation parameters. The cable defect evaluation parameters are compared with the preset cable defect early warning threshold. If the cable defect evaluation parameters are less than the preset cable defect warning threshold, then the output cable defect warning is no warning. If the cable defect evaluation parameter is greater than or equal to the preset cable defect warning threshold, then the cable defect warning is output as a defect warning.

[0010] Optionally, the elevator cable defect detection method described in this application further includes: Compare the defect category feature data of adjacent preset cable segments; If they are of the same category, the corresponding number of broken wires and cable wear rate are accumulated and weighted summation is performed to obtain the initial evaluation correction parameters for cable defects. If they are different categories, count the number of defect categories; The number of defect categories is compared with a preset warning threshold for the number of defect categories; If the number of defect categories is less than the preset defect category warning threshold, continuous monitoring will be performed. If the number of defect categories is greater than or equal to the preset defect category warning threshold, a defect warning will be output.

[0011] Optionally, the elevator cable defect detection method described in this application further includes: Multiple preset cable segment defect perception association datasets are input into a preset cloud multi-task parallel defect detection model in the cloud for processing to obtain the second cable defect detection data. Compare the first cable defect detection data with the second cable defect detection data; If there is a conflict in the defect category, obtain the confidence level corresponding to the defect category feature data, and output the defect category feature data with the higher confidence level. If there is no conflict in the defect category, obtain the cable breakage deviation rate and cable wear deviation rate corresponding to the preset cable segment cable breakage number and cable wear rate; If both the cable breakage deviation rate and the cable wear deviation rate are less than the preset deviation rate warning threshold, no warning will be output. If the cable breakage deviation rate or cable wear deviation rate is greater than or equal to the preset deviation rate warning threshold, an early warning will be issued.

[0012] Optionally, the elevator cable defect detection method described in this application further includes: The defect perception fusion feature vector, defect location data, number of broken cable wires and cable wear rate are input into a preset cable digital twin model for finite element analysis and stress analysis to obtain stress data of a preset cable segment. Acquire operating load data, operating speed, and number of start-stop cycles within a preset time period; The stress data, operating load data, operating speed, and number of start-stop cycles are used to perform fatigue damage analysis through a preset cable digital twin model to obtain the predicted remaining life. If the predicted remaining lifespan is less than the preset remaining warning lifespan, a warning is output.

[0013] Secondly, this application provides a device for detecting defects in elevator cables, the device comprising: The integrated cable defect detection probe module is used to acquire TMR magnetic sensing data, multispectral linear array image slices, and cable sensing point cloud data. The data analysis and processing module includes an edge detection unit and a cloud unit, which are used to perform cable defect analysis and detection at the edge and in the cloud according to a preset model; The data transmission module is used for data transmission between modules. The user-side display module is used to display the cable analysis and testing results.

[0014] Optionally, in the elevator cable defect detection system described in this application, the integrated cable defect detection probe module includes: It includes a TMR magnetic sensing array unit, a multispectral linear array camera unit, and a laser contour scanning unit, used to acquire TMR magnetic sensing data, multispectral linear array image slices, and cable sensing point cloud data.

[0015] As can be seen from the above, the elevator cable defect detection method and device provided in this application achieve intelligent detection of elevator cable defects by acquiring and processing multimodal data, analyzing and detecting from the edge end and the cloud, and combining digital twin model simulation analysis.

[0016] Other features and advantages of this application will be set forth in the following description and will be apparent in part from the description or may be learned by practicing embodiments of this application. The objectives and other advantages of this application may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A flowchart illustrating a method for detecting elevator cable defects provided in an embodiment of this application; Figure 2 A flowchart illustrating the process of obtaining a defect-aware association dataset for a method of detecting elevator cable defects provided in this application embodiment; Figure 3 A high-rise flowchart illustrating a method for detecting elevator cable defects provided in this application embodiment; Figure 4 This is a diagram of an integrated detection probe module for elevator cable defects, provided in an embodiment of this application. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0020] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0021] Please refer to Figure 1 , Figure 1This is a flowchart illustrating a method for detecting elevator cable defects according to some embodiments of this application. This method for detecting elevator cable defects is used in terminal devices, such as computers and mobile terminals. The method for detecting elevator cable defects includes the following steps: S11. Obtain a multimodal defect perception dataset through the cable defect integrated detection probe module, and perform preprocessing to obtain a defect perception association dataset; S12. Perform feature extraction and dynamic fusion processing based on the defect-aware associated dataset to obtain a defect-aware fusion feature vector; S13. Input the defect perception fusion feature vector into a preset multi-task parallel defect detection model for processing to obtain the first cable defect detection data. S14. Analyze and process the first cable defect detection data to obtain cable defect evaluation parameters, compare them with the preset cable defect early warning threshold, and output cable defect early warning based on the threshold comparison result.

[0022] It should be noted that, in order to achieve accurate multimodal fusion judgment of cable defects, firstly, the magnetic sensor and camera are organically integrated through spatiotemporal synchronization and dynamic feature fusion. Then, the accuracy is improved through edge lightweight rapid diagnosis and cloud-based in-depth analysis. Finally, the traditional state description is optimized into lifespan prediction through digital twin model, thereby accurately detecting cable defects and providing timely early warning.

[0023] Please refer to Figure 2 , Figure 2 This is a flowchart illustrating the process of obtaining a defect-aware associated dataset in a method for detecting elevator cable defects according to some embodiments of this application. According to embodiments of the present invention, the step of acquiring a multimodal defect-aware dataset through an integrated cable defect detection probe module and performing preprocessing to obtain a defect-aware associated dataset includes: S21. Obtain a multimodal defect perception dataset of a preset cable segment through the integrated cable defect detection probe module, including TMR magnetic sensing data, multispectral linear array image slices and cable perception point cloud data. S221. The TMR magnetic sensing data is filtered, differentially amplified, and data synthesized to obtain a comprehensive leakage magnetic vector. S222. Perform image stitching, geometric correction, image enhancement and noise reduction on the slices of the multispectral linear array image to obtain a multispectral linear array unfolded image. S223. Perform cylindrical model fitting and diameter deviation sequence, ellipticity and surface quality analysis on the cable sensing point cloud data to obtain cable sensing geometric feature sequence data. S23. The integrated magnetic leakage vector, multispectral linear array unfolded image and cable sensing geometric feature sequence data are spatiotemporally aligned, and data blocks are divided and data association processing is performed according to the preset cable segments to obtain a defect sensing association dataset.

[0024] It should be noted that TMR magnetic sensing data representing the original voltage signal is acquired through a TMR magnetic sensing array unit. This data is then filtered using a high-pass filter, a band-stop filter, and a low-pass filter. The signals from the symmetrically arranged sensors on the circumference are differentially analyzed and amplified to obtain the leakage magnetic field signal caused by local defects. The leakage magnetic field components detected along the cable axis are vector-synthesized to obtain the axial leakage magnetic field signal, and the leakage magnetic field components along the cable radial direction are vector-synthesized to obtain the radial leakage magnetic field signal, thus obtaining the comprehensive leakage magnetic field vector. A predetermined number of one-dimensional image slices are acquired using a multispectral linear array camera and stitched together end-to-end in space to form a two-dimensional elongated surface unfolded image. Feature point matching or optical flow methods are used to align adjacent slices at the sub-pixel level to correct slight image misalignment caused by cable torsion or vibration. The Retinex algorithm or homomorphic filtering is used to eliminate the interference caused by linear light sources. The system analyzes the illumination gradient between two sides, where the brightness is high and the brightness is low, and fuses images from different wavelengths to output a high-contrast, seamlessly stitched, and uniformly illuminated cable surface unfolding map. Three-dimensional cable sensing point cloud data is acquired using a laser contour scanning unit. An optimal cylindrical model is fitted using the least squares method, and the difference between the actual measured radius and the preset nominal cable radius is calculated. A negative value indicates wear. The point cloud is projected onto a plane perpendicular to the cylinder axis, and an ellipse is fitted. The difference between the major and minor axes is used to quantify the ellipticity of the cable, representing indentations or local plastic deformation. The radial distance from each measurement point to the fitted cylindrical surface is calculated, and the root mean square is used to represent the surface micro-irregularities or local pits, thus obtaining the cable sensing geometric feature sequence data. Finally, a unified spatiotemporal label is used for spatiotemporal alignment, and the data is divided and associated according to the segmented cable sections to determine the cable location represented by the data.

[0025] According to an embodiment of the present invention, the step of performing feature extraction and dynamic fusion processing based on the defect-aware association dataset to obtain a defect-aware fusion feature vector includes: Feature extraction is performed based on the defect-aware association dataset to obtain magnetic feature vectors, visual feature vectors, and geometric feature vectors. The magnetic feature vector, visual feature vector, and geometric feature vector are analyzed and processed by a self-attention module to obtain the weight values ​​of the magnetic feature vector, visual feature vector, and geometric feature vector. The magnetic feature vector, visual feature vector, and geometric feature vector, along with their corresponding weight values, are dynamically weighted and fused to obtain a defect-aware fusion feature vector.

[0026] It should be noted that, based on the axial and radial magnetic flux leakage signals, several layers of convolution and global pooling are used to obtain a fixed-dimensional magnetic feature vector. Based on the multispectral linear array unfolded image, a lightweight convolutional neural network is used for feature extraction and adaptive pooling to obtain a fixed-dimensional visual feature vector. Based on the cable perception geometric feature sequence data, it is directly flattened into a one-dimensional vector to obtain a fixed-dimensional geometric feature vector. Then, through the attention fusion module, the relationship between features is extracted by depthwise separable convolution to generate dynamic attention weights (obtained by dynamic analysis based on input features, with no fixed weights). After normalization, dynamic weighted fusion processing is performed to obtain the defect perception fusion feature vector.

[0027] According to an embodiment of the present invention, the step of inputting the defect-aware fusion feature vector into a preset multi-task parallel defect detection model for processing to obtain first cable defect detection data includes: The defect perception fusion feature vector is input into the preset edge end multi-task parallel defect detection model of the edge detection end for processing to obtain the first cable defect detection data corresponding to the preset cable segment. The first cable defect detection data includes defect category feature data, corresponding defect quantification indicators, and defect location data; The defect quantification indicators include the number of broken wires in the cable and the cable wear rate.

[0028] It should be noted that a pre-set model is deployed at the edge to achieve rapid assessment of the cable's initial evaluation and detection. Based on the shared backbone network, high-level shared features are extracted from the fused features. Three task heads are designed for classification, quantization, and localization. The multimodal fused feature sequence of the entire cable, i.e., multiple defect-aware fused feature vectors, is analyzed and processed through a 1D fully convolutional network to obtain the defect category feature data of each cable segment, as well as the corresponding defect quantification index and defect location data. Among them, the defect categories include normal, single-point broken wire, uniform wear, dense broken wire, local wear, and indentation deformation. The defect category feature data are represented by different values.

[0029] According to an embodiment of the present invention, the step of analyzing and processing the first cable defect detection data to obtain cable defect evaluation parameters, comparing them with a preset cable defect early warning threshold, and outputting a cable defect early warning based on the threshold comparison result includes: Based on the defect category feature data, query the preset weight value list and the preset defect evaluation correction coefficient mapping table to obtain the cable breakage weight value, cable wear weight value and defect evaluation correction coefficient; Based on the cable breakage weight value and the cable wear weight value, the number of cable breakage wires and the cable wear rate are weighted and summed to obtain the initial evaluation parameters of cable defects. The initial evaluation parameters of the cable defect are corrected according to the defect evaluation correction coefficient to obtain the cable defect evaluation parameters. The cable defect evaluation parameters are compared with the preset cable defect early warning threshold. If the cable defect evaluation parameters are less than the preset cable defect warning threshold, then the output cable defect warning is no warning. If the cable defect evaluation parameter is greater than or equal to the preset cable defect warning threshold, then the cable defect warning is output as a defect warning.

[0030] It should be noted that, in order to improve the accuracy of the assessment, the quantitative indicators are further modified. First, those skilled in the art pre-construct a list of preset weight values ​​and a mapping table of preset defect assessment correction coefficients based on historical experience, and then dynamically fine-tune them according to specific applications. The weights for cable breakage and cable wear are different for different defect categories. For example, if the defect category is normal, the weight value for cable breakage is 0.1 and the weight for cable wear is 0.9. The initial assessment parameters for cable defects are determined by weighted summation. Then, the parameters are corrected according to the defect assessment correction coefficients corresponding to the defect categories. The more severe the defect, the larger the defect assessment correction coefficient. For example, the coefficient for normal is 1, and the coefficient for dense breakage is 1.5. The cable defect assessment parameters are obtained by multiplying the defect assessment correction coefficients by the initial assessment parameters for cable defects. Finally, the threshold comparison is used to determine whether an early warning needs to be output, so as to achieve rapid judgment at the edge.

[0031] According to an embodiment of the present invention, it further includes: Compare the defect category feature data of adjacent preset cable segments; If they are of the same category, the corresponding number of broken wires and cable wear rate are accumulated and weighted summation is performed to obtain the initial evaluation correction parameters for cable defects. If they are different categories, count the number of defect categories; The number of defect categories is compared with a preset warning threshold for the number of defect categories; If the number of defect categories is less than the preset defect category warning threshold, continuous monitoring will be performed. If the number of defect categories is greater than or equal to the preset defect category warning threshold, a defect warning will be output.

[0032] It should be noted that, in order to further accurately assess the defects of the entire cable, the identification of the divided cable segments is comprehensively compared and evaluated. For those with the same defects, the results are accumulated to obtain the initial defect assessment correction parameters, which are then used to correct and evaluate the defects. For those with different defects, the number of defect categories is counted. The higher the number, the worse the cable quality, and the more timely the warning needs to be.

[0033] According to an embodiment of the present invention, it further includes: Multiple preset cable segment defect perception association datasets are input into a preset cloud multi-task parallel defect detection model in the cloud for processing to obtain the second cable defect detection data. Compare the first cable defect detection data with the second cable defect detection data; If there is a conflict in the defect category, obtain the confidence level corresponding to the defect category feature data, and output the defect category feature data with the higher confidence level. If there is no conflict in the defect category, obtain the cable breakage deviation rate and cable wear deviation rate corresponding to the preset cable segment cable breakage number and cable wear rate; If both the cable breakage deviation rate and the cable wear deviation rate are less than the preset deviation rate warning threshold, no warning will be output. If the cable breakage deviation rate or cable wear deviation rate is greater than or equal to the preset deviation rate warning threshold, an early warning will be issued.

[0034] It should be noted that after a quick preliminary judgment at the edge, in-depth analysis is then performed through a cloud model to obtain the second cable defect detection data. This data is then compared with the first cable defect detection data. If there is a conflict between the edge and cloud results, the corresponding confidence level is checked, and the result with the higher confidence level is determined as the defect detection result. If there is no conflict, the deviation rate of the number of broken cable strands and the cable wear rate between the two detection results is further evaluated, i.e., the absolute value of the difference between the two results and the edge detection result.

[0035] According to an embodiment of the present invention, it further includes: The defect perception fusion feature vector, defect location data, number of broken cable wires and cable wear rate are input into a preset cable digital twin model for finite element analysis and stress analysis to obtain stress data of a preset cable segment. Acquire operating load data, operating speed, and number of start-stop cycles within a preset time period; The stress data, operating load data, operating speed, and number of start-stop cycles are used to perform fatigue damage analysis through a preset cable digital twin model to obtain the predicted remaining life. If the predicted remaining lifespan is less than the preset remaining warning lifespan, a warning is output.

[0036] It should be noted that a parametric three-dimensional digital twin is created for each cable in the cloud based on the wire rope material, nominal diameter, twisting method, usage date, and rated load. Using fusion features and quantification results, the degradation of local mechanical properties is inversely calculated. For example, based on the number and location of broken wires, the stress concentration factor change of that section of the cable is simulated through finite element analysis. Then, the operating load data, operating speed, and number of start-stop cycles uploaded by the elevator IoT are used as input to drive the twin to perform virtual operation, simulating the fatigue accumulation process, thereby achieving life prediction. When the predicted life is about to be less than the preset remaining warning life, a warning is output in a timely manner.

[0037] Please refer to Figure 3 , Figure 3 This is a high-level flowchart of a method for detecting elevator cable defects in some embodiments of this application.

[0038] This invention also discloses a device for detecting defects in elevator cables, comprising: The integrated cable defect detection probe module is used to acquire TMR magnetic sensing data, multispectral linear array image slices, and cable sensing point cloud data. The data analysis and processing module includes an edge detection unit and a cloud unit, which are used to perform cable defect analysis and detection at the edge and in the cloud according to a preset model; The data transmission module is used for data transmission between modules. The user-side display module is used to display the cable analysis and testing results.

[0039] It should be noted that multimodal data is collected through the integrated cable defect detection probe module, and analyzed by models of different precision at the edge and cloud, achieving a balance between rapid and in-depth analysis, and displaying the analysis results.

[0040] Please refer to Figure 4 , Figure 4 This is a diagram of an integrated cable defect detection probe module of an elevator cable defect detection device according to some embodiments of this application. According to an embodiment of the present invention, the integrated cable defect detection probe module includes: It includes a TMR magnetic sensing array unit 41, a multispectral linear array camera unit 42, and a laser contour scanning unit 43, which are used to acquire TMR magnetic sensing data, multispectral linear array image slices, and cable sensing point cloud data.

[0041] It should be noted that by using magnetic sensors and cameras in parallel, spatiotemporal synchronous registration and feature-level dynamic fusion are achieved.

[0042] This invention discloses a method and apparatus for detecting elevator cable defects. By acquiring and processing multimodal data, analyzing and detecting from the edge and cloud, and combining digital twin model simulation analysis, intelligent detection of elevator cable defects is achieved.

[0043] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0044] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0045] In addition, in the various embodiments of the present invention, each functional unit can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0046] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0047] Alternatively, if the integrated units of this invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a readable storage medium. Based on this understanding, the technical solutions of the embodiments of this invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

Claims

1. A method for detecting defects in elevator cables, characterized in that, Includes the following steps: The multimodal defect perception dataset is obtained by the integrated cable defect detection probe module, and then preprocessed to obtain the defect perception association dataset. Based on the defect-aware associated dataset, feature extraction and dynamic fusion processing are performed to obtain a defect-aware fusion feature vector. The defect perception fusion feature vector is input into a preset multi-task parallel defect detection model for processing to obtain the first cable defect detection data. The cable defect detection data is analyzed and processed to obtain cable defect evaluation parameters, which are then compared with preset cable defect early warning thresholds. Based on the threshold comparison results, a cable defect early warning is output.

2. The method for detecting elevator cable defects according to claim 1, characterized in that, The multimodal defect perception dataset is acquired through the integrated cable defect detection probe module and preprocessed to obtain a defect perception association dataset, including: The multimodal defect perception dataset of a preset cable segment is obtained through the integrated cable defect detection probe module, including TMR magnetic sensing data, multispectral linear array image slices and cable perception point cloud data. The TMR magnetic sensing data is filtered, differentially amplified, and synthesized to obtain a comprehensive leakage magnetic vector. The multispectral linear array image slices are subjected to image stitching, geometric correction, image enhancement and noise reduction processing to obtain a multispectral linear array unfolded image; The cable sensing point cloud data is subjected to cylindrical model fitting and diameter deviation sequence, ellipticity and surface quality analysis to obtain cable sensing geometric feature sequence data. The integrated magnetic leakage vector, multispectral linear array unfolded image, and cable sensing geometric feature sequence data are spatiotemporally aligned, and data blocks are divided and data association processing is performed according to preset cable segments to obtain a defect sensing association dataset.

3. The method for detecting elevator cable defects according to claim 2, characterized in that, The step of performing feature extraction and dynamic fusion processing based on the defect-aware associated dataset to obtain a defect-aware fusion feature vector includes: Feature extraction is performed based on the defect-aware association dataset to obtain magnetic feature vectors, visual feature vectors, and geometric feature vectors. The magnetic feature vector, visual feature vector, and geometric feature vector are analyzed and processed by a self-attention module to obtain the weight values ​​of the magnetic feature vector, visual feature vector, and geometric feature vector. The magnetic feature vector, visual feature vector, and geometric feature vector, along with their corresponding weight values, are dynamically weighted and fused to obtain a defect-aware fusion feature vector.

4. The method for detecting elevator cable defects according to claim 3, characterized in that, The step of inputting the defect-aware fusion feature vector into a preset multi-task parallel defect detection model for processing to obtain the first cable defect detection data includes: The defect perception fusion feature vector is input into the preset edge end multi-task parallel defect detection model of the edge detection end for processing to obtain the first cable defect detection data corresponding to the preset cable segment. The first cable defect detection data includes defect category feature data, corresponding defect quantification indicators, and defect location data; The defect quantification indicators include the number of broken wires in the cable and the cable wear rate.

5. The method for detecting elevator cable defects according to claim 4, characterized in that, The step of analyzing and processing the first cable defect detection data to obtain cable defect evaluation parameters, comparing them with a preset cable defect early warning threshold, and outputting a cable defect early warning based on the threshold comparison result includes: Based on the defect category feature data, query the preset weight value list and the preset defect evaluation correction coefficient mapping table to obtain the cable breakage weight value, cable wear weight value and defect evaluation correction coefficient; Based on the cable breakage weight value and the cable wear weight value, the number of cable breakage wires and the cable wear rate are weighted and summed to obtain the initial evaluation parameters of cable defects. The initial evaluation parameters of the cable defect are corrected according to the defect evaluation correction coefficient to obtain the cable defect evaluation parameters. The cable defect evaluation parameters are compared with the preset cable defect early warning threshold. If the cable defect evaluation parameters are less than the preset cable defect warning threshold, then the output cable defect warning is no warning. If the cable defect evaluation parameter is greater than or equal to the preset cable defect warning threshold, then the cable defect warning is output as a defect warning.

6. The method for detecting elevator cable defects according to claim 4, characterized in that, Also includes: Compare the defect category feature data of adjacent preset cable segments; If they are of the same category, the corresponding number of broken wires and cable wear rate are accumulated and weighted summation is performed to obtain the initial evaluation correction parameters for cable defects. If they are different categories, count the number of defect categories; The number of defect categories is compared with a preset warning threshold for the number of defect categories; If the number of defect categories is less than the preset defect category warning threshold, continuous monitoring will be performed. If the number of defect categories is greater than or equal to the preset defect category warning threshold, a defect warning will be output.

7. The method for detecting elevator cable defects according to claim 2, characterized in that, Also includes: Multiple preset cable segment defect perception association datasets are input into a preset cloud multi-task parallel defect detection model in the cloud for processing to obtain the second cable defect detection data. Compare the first cable defect detection data with the second cable defect detection data; If there is a conflict in the defect category, obtain the confidence level corresponding to the defect category feature data, and output the defect category feature data with the higher confidence level. If there is no conflict in the defect category, obtain the cable breakage deviation rate and cable wear deviation rate corresponding to the preset cable segment cable breakage number and cable wear rate; If both the cable breakage deviation rate and the cable wear deviation rate are less than the preset deviation rate warning threshold, no warning will be output. If the cable breakage deviation rate or cable wear deviation rate is greater than or equal to the preset deviation rate warning threshold, an early warning will be issued.

8. The method for detecting elevator cable defects according to claim 2, characterized in that, Also includes: The defect perception fusion feature vector, defect location data, number of broken cable wires and cable wear rate are input into a preset cable digital twin model for finite element analysis and stress analysis to obtain stress data of a preset cable segment. Acquire operating load data, operating speed, and number of start-stop cycles within a preset time period; The stress data, operating load data, operating speed, and number of start-stop cycles are used to perform fatigue damage analysis through a preset cable digital twin model to obtain the predicted remaining life. If the predicted remaining lifespan is less than the preset remaining warning lifespan, a warning is output.

9. A device for detecting defects in elevator cables, characterized in that, include: The integrated cable defect detection probe module is used to acquire TMR magnetic sensing data, multispectral linear array image slices, and cable sensing point cloud data. The data analysis and processing module includes an edge detection unit and a cloud unit, which are used to perform cable defect analysis and detection at the edge and in the cloud according to a preset model; The data transmission module is used for data transmission between modules. The user-side display module is used to display the cable analysis and testing results.

10. The elevator cable defect detection system according to claim 9, characterized in that, The integrated cable defect detection probe module includes: It includes a TMR magnetic sensing array unit, a multispectral linear array camera unit, and a laser contour scanning unit, used to acquire TMR magnetic sensing data, multispectral linear array image slices, and cable sensing point cloud data.

Images