Multi-mode sensing fusion intelligent comprehensive safety detection system for conveying belt

Abstract

The invention relates to the technical field of industrial production, in particular to a multi-mode sensing fusion intelligent comprehensive safety detection system for a conveying belt, which comprises a sensing module, a data processing module, an intelligent analysis module and a linkage module, and is characterized in that the data processing module is connected with the sensing module, and the intelligent analysis module is connected with the data processing module; the linkage module is connected with the intelligent analysis module; the sensing module is used for collecting belt surface and internal state data in real time; the data processing module is used for preprocessing the original data; the intelligent analysis module performs foreign matter recognition, tear detection and ore volume calculation based on a deep learning model in combination with preprocessed data (such as YOLO and U-Net), and performs fault prediction at the same time; the linkage module is used for receiving a processing result of the intelligent analysis module and triggering alarm, shutdown or broadcast prompt to realize closed-loop control; in this way, the technical problems that in the prior art, a belt detection system is single in means, high in false alarm rate and high in maintenance cost are solved.

Description

Technical Field

[0001] The present invention relates to the field of industrial production technology, and in particular to a multi-modal sensing fusion intelligent comprehensive safety detection system for transport belts. Background Art

[0002] With the rapid development of industrial production, conveyor belts, as important equipment for material transportation, have been widely used in mining, metallurgy, chemical industry and other fields. In order to ensure the safe and stable operation of conveyor belts and improve their operating efficiency, real-time monitoring and analysis of the surface and internal state of the belts is particularly important. In recent years, with the continuous advancement of sensing technology, data processing technology and intelligent analysis technology, conveyor belt detection systems have gradually developed in the direction of intelligence and modularization.

[0003] However, the belt detection system in the prior art has the problems of single means, high false alarm rate and high maintenance cost. Summary of the invention

[0004] The purpose of the present invention is to provide a multi-modal sensor fusion transport belt intelligent comprehensive safety detection system, aiming to solve the technical problems of the belt detection system in the prior art, such as single means, high false alarm rate and high maintenance cost.

[0005] To achieve the above-mentioned purpose, the present invention adopts a multi-modal sensing fusion transport belt intelligent comprehensive safety detection system, including a perception module, a data processing module, an intelligent analysis module and a linkage module, wherein the data processing module is connected to the perception module, the intelligent analysis module is connected to the data processing module, and the linkage module is connected to the intelligent analysis module;

[0006] The sensing module is used to collect belt surface and internal status data in real time;

[0007] The data processing module is used to pre-process the original data;

[0008] The intelligent analysis module performs foreign body recognition, tear detection, ore volume calculation, and fault prediction based on the deep learning model combined with pre-processed data (such as YOLO and U-Net);

[0009] The linkage module receives the processing result of the intelligent analysis module, triggers an alarm, shuts down or broadcasts a prompt, and realizes closed-loop control.

[0010] The sensing module includes a laser stripe detector, an infrared camera, an X-ray flaw detection device, a 3D laser radar and an industrial camera;

[0011] The laser stripe detector, the infrared camera, the X-ray flaw detection device, the 3D laser radar and the industrial camera are all connected to the data processing module.

[0012] Among them, the data processing module includes an embedded industrial computer and an edge computing unit, and the embedded industrial computer and the edge computing unit are both connected to the perception module.

[0013] Wherein, the multi-modal sensing fusion transport belt intelligent comprehensive safety detection system further includes a time synchronization module, and the time synchronization module is connected to the perception module;

[0014] The time synchronization module uses PTP (Precision Time Protocol) to unify the clocks of each sensor with an error of <1ms.

[0015] Among them, the multi-modal sensor fusion transport belt intelligent comprehensive safety detection system also includes a space calibration module, and the space calibration module is also connected to the perception module.

[0016] Wherein, the laser stripe detector is deployed just below the belt unloading port.

[0017] Wherein, the infrared camera is used to monitor the belt surface temperature field.

[0018] Among them, the space calibration module establishes the spatial coordinate system conversion relationship of the laser, camera, and infrared equipment based on the checkerboard calibration method to achieve data space alignment.

[0019] The multi-modal sensing fusion intelligent comprehensive safety detection system for conveyor belts of the present invention, when in use, continuously scans the surface of the conveyor belt through the laser stripe detector in the sensing module, generates an infrared image of the conveyor belt through the infrared camera, detects damages such as perforations, broken ropes, corrosion, deformation of steel ropes and poor joints on the conveyor belt through the X-ray flaw detection device in the sensing module, and accurately finds the location where the damage occurs, scans the ore image on the conveyor belt in real time through the 3D laser radar in the sensing module, and monitors the area where the ore passes in real time through the industrial camera in the sensing module. At the same time, the data The processing module pre-processes the raw data collected by the perception module (such as point cloud reconstruction, image noise reduction, temperature calibration, etc.), and then the intelligent analysis module performs foreign body recognition, tear detection, ore volume calculation, etc. on the pre-processed data based on a deep learning model (such as YOLO, U-Net), and performs fault prediction in combination with an expert system. Finally, the linkage module receives the processing results of the intelligent analysis module, connects with the production control system through PLC, triggers alarms, shutdowns or broadcast prompts, and realizes closed-loop control. In this way, the technical problems of the belt detection system in the prior art, such as single means, high false alarm rate, and high maintenance cost, are solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0021] Figure 1 It is a principle block diagram of the multi-modal sensing fusion transport belt intelligent comprehensive safety detection system of the present invention.

[0022] Figure 2 It is a schematic diagram of laser stripe tear detection and infrared tear detection of the present invention.

[0023] Figure 3 It is a schematic diagram of deviation detection of the present invention.

[0024] Figure 4 It is a schematic diagram of foreign body identification and ore analysis of the present invention.

[0025] 1-Perception module, 2-Data processing module, 3-Intelligent analysis module, 4-Linkage module, 5-Time synchronization module, 6-Spatial calibration module, 7-Laser stripe detector, 8-Infrared camera, 9-X-ray flaw detection device, 10-3D laser radar, 11-Industrial camera, 12-Embedded industrial computer, 13-Edge computing unit. DETAILED DESCRIPTION

[0026] Embodiments of the present invention are described in detail below. Examples of the embodiments are shown in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, but should not be construed as limiting the present invention.

[0027] See also Figure 1 to Figure 4 ,in Figure 1 It is a principle block diagram of the multi-modal sensing fusion transport belt intelligent comprehensive safety detection system of the present invention. Figure 2 It is a schematic diagram of laser stripe tear detection and infrared tear detection of the present invention. Figure 3 It is a schematic diagram of deviation detection of the present invention. Figure 4 It is a schematic diagram of foreign body identification and ore analysis of the present invention.

[0028] The present invention provides a multi-modal sensing fusion intelligent comprehensive safety detection system for conveyor belts, comprising a perception module 1, a data processing module 2, an intelligent analysis module 3, a linkage module 4, a time synchronization module 5 and a space calibration module 6. The perception module 1 comprises a laser stripe detector 7, an infrared camera 8, an X-ray flaw detection device 9, a 3D laser radar 10 and an industrial camera 11; the data processing module 2 comprises an embedded industrial computer 12 and an edge computing unit 13. The above scheme solves the technical problems of the belt detection system in the prior art, such as a single means, a high false alarm rate and a high maintenance cost.

[0029] For this specific implementation, the sensing module 1 is used to collect belt surface and internal state data in real time;

[0030] The data processing module 2 is used to pre-process the original data;

[0031] The intelligent analysis module 3 performs foreign body recognition, tear detection, ore volume calculation, and fault prediction based on a deep learning model combined with pre-processed data (such as YOLO and U-Net);

[0032] The linkage module 4 receives the processing result of the intelligent analysis module 3, triggers an alarm, shuts down or broadcasts a prompt, and implements closed-loop control.

[0033] The data processing module 2 is connected to the perception module 1, the intelligent analysis module 3 is connected to the data processing module 2, and the linkage module 4 is connected to the intelligent analysis module 3;

[0034] The laser stripe detector 7, the infrared camera 8, the X-ray flaw detection device 9, the 3D laser radar 10 and the industrial camera 11 are all connected to the data processing module 2;

[0035] The laser stripe detector 7 is deployed just below the belt unloading port;

[0036] The infrared camera 8 is used to monitor the belt surface temperature field;

[0037] In specific use, the laser stripe detector 7 in the perception module 1 continuously scans the surface of the conveyor belt, the infrared camera 8 generates an infrared image of the conveyor belt, the X-ray flaw detection device 9 in the perception module 1 detects damages such as perforations, broken ropes, corrosion, deformation of steel ropes and poor joints on the conveyor belt, and accurately finds the location where the damage occurs, the 3D laser radar 10 in the perception module 1 scans the ore image on the conveyor belt in real time, and the industrial camera 11 in the perception module 1 monitors the area where the ore passes in real time. At the same time, the data processing module 2 processes the perception module The raw data collected by block 1 is preprocessed (such as point cloud reconstruction, image noise reduction, temperature calibration, etc.), and then the intelligent analysis module 3 performs foreign body recognition, tear detection, ore volume calculation, etc. on the preprocessed data based on a deep learning model (such as YOLO, U-Net), and combines the expert system to predict faults. Finally, the linkage module 4 receives the processing results of the intelligent analysis module 3, connects with the production control system through PLC, triggers alarms, shutdowns or broadcast prompts, and realizes closed-loop control. In this way, the technical problems of the belt detection system in the prior art, such as single means, high false alarm rate and high maintenance cost, are solved.

[0038] Secondly, the embedded industrial computer 12 and the edge computing unit 13 are both connected to the perception module 1, and the embedded industrial computer 12 and the edge computing unit 13 are used to pre-process the original data (such as point cloud reconstruction, image denoising, temperature calibration, etc.).

[0039] At the same time, the time synchronization module 5 is connected to the perception module 1;

[0040] The time synchronization module 5 uses PTP (Precision Time Protocol) to unify the clocks of each sensor with an error of <1ms.

[0041] In addition, the space calibration module 6 is also connected to the perception module 1.

[0042] The spatial calibration module 6 establishes the spatial coordinate system conversion relationship among the laser stripe detector 7, the infrared camera 8, the X-ray flaw detection device 9, the 3D laser radar 10 and the industrial camera 11 through the checkerboard calibration method to achieve data space alignment.

[0043] At the same time, the dynamic weight fusion algorithm based on Kalman filtering is used to optimize the confidence of anomaly detection.

[0044] The present invention can realize 3D laser point cloud and ore flow volume calculation, laser stripe and infrared dual-mode tear detection, X-ray flaw detection and internal damage identification, and AI-driven foreign body identification and ore analysis;

[0045] Specifically, the 3D laser point cloud and mine flow volume calculation technology are as follows:

[0046] Technical implementation: A line laser scanner (wavelength 905nm) is used to vertically illuminate the belt surface, and the laser stripe deformation is captured by a high-speed CMOS camera, combined with triangulation to generate a high-precision 3D point cloud.

[0047] Algorithm flow: point cloud filtering (voxel grid downsampling) → surface reconstruction (Poisson reconstruction algorithm) → mine flow volume integration (based on Voxel grid segmentation).

[0048] Real-time guarantee: FPGA hardware accelerates point cloud processing, single frame processing time <50ms, meeting the real-time requirements of belt speed ≤5m / s.

[0049] Laser stripe and infrared dual-mode tear detection technology is as follows:

[0050] The laser stripe detector 7 is deployed just below the belt unloading port, and the laser stripes are projected onto the belt surface, and the longitudinal tear is analyzed by stripe continuity. If the stripe is broken or deviated beyond the threshold (±3mm), an alarm is triggered.

[0051] The infrared camera 8 uses an uncooled infrared focal plane array (resolution 640×480) to monitor the belt surface temperature field. At the moment of tearing, the local temperature rise (ΔT≥15°C) due to friction is detected, and the damage point is located by the temperature gradient algorithm in the time and space domain.

[0052] Dual-mode collaboration: Laser detection provides geometric deformation data, infrared detection supplements thermodynamic characteristics, and through DS evidence theory fusion judgment, the false alarm rate is reduced to less than 0.1%.

[0053] X-ray flaw detection and internal damage identification technologies are as follows:

[0054] Hardware configuration: A transmission imaging system consisting of a micro X-ray source (energy 60kV) and a flat-panel detector can penetrate the multi-layer structure of the belt (rubber + steel rope).

[0055] Damage identification: Based on Faster R-CNN model training, the data set contains 10 types of defects such as perforation, broken rope, corrosion, etc., with a detection accuracy of 99.2%.

[0056] The steel rope deformation detection uses a skeleton extraction algorithm (Zhang-Suen thinning) to calculate the curvature anomaly points.

[0057] AI-driven foreign body identification and ore analysis technologies are as follows:

[0058] Foreign body detection: The high frame rate industrial camera 11 (2000fps) captures ore flow images and classifies them in real time (metal foreign bodies, large pieces of ore, etc.) through the lightweight MobileNetV3 model, supporting the ONNX inference engine.

[0059] Ore particle size statistics: Combine 3D point cloud and RGB image, use Euclidean clustering (DBSCAN) to segment ore individuals, calculate equivalent diameter distribution, and output particle size curve (D50, D90).

[0060] In addition, a camera is deployed directly above the running belt to capture video of the belt's running direction, and the "belt edge-baseline" comparison method is used to detect deviation.

[0061] The multi-modal sensing fusion intelligent comprehensive safety detection system for conveyor belts of the present invention is used. When in use, the laser stripe detector 7 in the sensing module 1 continuously scans the surface of the conveyor belt, the infrared camera 8 generates an infrared image of the conveyor belt, the X-ray flaw detection device 9 in the sensing module 1 detects damages such as perforations, broken ropes, corrosion, deformation of steel ropes, and poor joints on the conveyor belt, and accurately finds the location where the damage occurs, the 3D laser radar 10 in the sensing module 1 scans the ore image on the conveyor belt in real time, and the industrial camera 11 in the sensing module 1 monitors the area where the ore passes in real time. At the same time, The data processing module 2 pre-processes the raw data collected by the perception module 1 (such as point cloud reconstruction, image noise reduction, temperature calibration, etc.), and then the intelligent analysis module 3 performs foreign body recognition, tear detection, ore volume calculation, etc. on the pre-processed data based on a deep learning model (such as YOLO, U-Net), and performs fault prediction in combination with an expert system. Finally, the linkage module 4 receives the processing result of the intelligent analysis module 3, connects with the production control system through PLC, triggers an alarm, shutdown or broadcast prompt, and realizes closed-loop control. In this way, the technical problems of the belt detection system in the prior art, such as a single means, a high false alarm rate, and a high maintenance cost, are solved.

[0062] Beneficial effects of the present invention:

[0063] 1. Full-dimensional detection coverage: from surface deformation (laser), internal defects (X-ray), thermodynamic characteristics (infrared) to dynamic mineral flow (3D point cloud), to achieve holographic perception of the health status of the belt.

[0064] 2. Modular design: Each detection unit is independently packaged, supports hot plug and distributed deployment, and is suitable for different belt lengths and working conditions.

[0065] 3. Edge-cloud collaboration: Local edge computing ensures real-time performance, and key data is uploaded to the cloud for long-term trend analysis and predictive maintenance modeling.

[0066] 4. Adaptability to harsh environments: The sensor has an IP67 protection level, and the X-ray module is equipped with a self-cleaning air curtain to ensure stable operation in dusty and humid environments.

[0067] Application examples:

[0068] Application data of a certain iron mine: After the system of the present invention is deployed, the belt tearing missed detection rate is reduced from 2.1% to 0.3%, and the ore volume measurement error is <1.5%.

[0069] The X-ray flaw detection module gave an early warning of wire rope joint fatigue 14 days in advance, avoiding unplanned downtime losses of approximately RMB 1.2 million.

[0070] Real-time test: From data collection to control command issuance, the end-to-end delay is ≤200ms, meeting the ISO 13849 safety standard.

[0071] What is disclosed above is only a preferred embodiment of the present invention, and it certainly cannot be used to limit the scope of rights of the present invention. Ordinary technicians in this field can understand that all or part of the processes of the above embodiment and equivalent changes made according to the claims of the present invention still fall within the scope of the invention.

Claims

1. A multi-modal sensor fusion transport belt intelligent comprehensive safety detection system, characterized in that: It includes a perception module, a data processing module, an intelligent analysis module and a linkage module, wherein the data processing module is connected to the perception module, the intelligent analysis module is connected to the data processing module, and the linkage module is connected to the intelligent analysis module; The sensing module is used to collect belt surface and internal status data in real time; The data processing module is used to pre-process the original data; The intelligent analysis module performs foreign body identification, tear detection, ore volume calculation, and fault prediction based on the deep learning model combined with preprocessed data; The linkage module receives the processing result of the intelligent analysis module, triggers an alarm, shuts down or broadcasts a prompt, and realizes closed-loop control.

2. The multi-modal sensing fusion intelligent comprehensive safety detection system for conveyor belts according to claim 1 is characterized in that: The sensing module includes a laser stripe detector, an infrared camera, an X-ray flaw detection device, a 3D laser radar and an industrial camera; The laser stripe detector, the infrared camera, the X-ray flaw detection device, the 3D laser radar and the industrial camera are all connected to the data processing module.

3. The multi-modal sensing fusion intelligent comprehensive safety detection system for conveyor belts according to claim 2 is characterized in that: The data processing module includes an embedded industrial computer and an edge computing unit, and both the embedded industrial computer and the edge computing unit are connected to the perception module.

4. The multi-modal sensing fusion intelligent comprehensive safety detection system for conveyor belts according to claim 3 is characterized in that: The multi-modal sensing fusion transport belt intelligent comprehensive safety detection system also includes a time synchronization module, which is connected to the perception module.

5. The multi-modal sensing fusion intelligent comprehensive safety detection system for conveyor belts according to claim 4 is characterized in that: The multi-modal sensing fusion transport belt intelligent comprehensive safety detection system also includes a space calibration module, and the space calibration module is also connected to the perception module.

6. The multi-modal sensing fusion intelligent comprehensive safety detection system for conveyor belts according to claim 5 is characterized in that: The laser stripe detector is deployed just below the belt unloading port.

7. The multi-modal sensing fusion intelligent comprehensive safety detection system for conveyor belts according to claim 6 is characterized in that: The infrared camera is used to monitor the belt surface temperature field.

8. The multi-modal sensing fusion intelligent comprehensive safety detection system for conveyor belts according to claim 7 is characterized in that: The space calibration module establishes the space coordinate system conversion relationship of the laser, camera and infrared equipment based on the checkerboard calibration method to achieve data space alignment.

Images