Pantograph carbon slide plate detection system based on MR glasses

Abstract

The invention relates to a pantograph carbon slide plate detection system based on MR glasses. The pantograph carbon slide plate detection system comprises an MR glasses terminal, a data processing module and a user interaction module, the MR glasses terminal comprises an MR glasses body, a data acquisition assembly, a processor assembly and a display assembly. The data acquisition assembly, the processor assembly and the display assembly are all arranged on the MR glasses body; the processor assembly is respectively connected with the data acquisition assembly and the display assembly; the data processing module and the user interaction module are both carried on the processor assembly; and the data processing module comprises a data fusion algorithm, a thickness calculation model and a wear analysis model, and is used for processing and analyzing the carbon slide plate detection data acquired by the data acquisition assembly in real time to obtain a detection result. The detection system provided by the invention can realize non-contact real-time detection, wear distribution modeling and intelligent judgment of the thickness of the carbon contact strip.

Description

Technical Field

[0001] The present invention belongs to the technical field of rail transit detection, and in particular relates to a pantograph carbon slide detection system based on MR glasses. Background Art

[0002] The pantograph carbon plate is a core component of the EMU catenary system, and its thickness and wear directly affect the quality and safety of current collection. Traditional inspection relies on manual visual inspection or offline measurement using a single sensor (such as a laser rangefinder), which suffers from low efficiency, insufficient accuracy, and the inability to conduct real-time dynamic monitoring. In existing technologies, infrared thermal imaging is often used to detect temperature anomalies, but it does not integrate three-dimensional spatial data. While laser ranging can obtain local thickness, it lacks global wear distribution analysis. Furthermore, the inspection process relies on fixed equipment and is difficult to adapt to dynamic scenarios. Examples include: inspection systems based on two-dimensional machine vision; and fixed laser scanning devices.

[0003] shortcoming:

[0004] 1. The inspection process relies entirely on manual inspection, and the "1+2" mode has low efficiency (the average inspection time for level 1 repair is 2 hours, and the inspection time for level 2 repair is 7 hours);

[0005] 2. The defect location marking relies on a two-dimensional schematic diagram, with a spatial positioning error of ±15mm;

[0006] 3. Data collection and report generation are separated, resulting in information islands;

[0007] reason:

[0008] Traditional inspection relies heavily on manpower (accounting for over 60%), and intelligent sensing coverage is low (3D vision accounts for less than 15%);

[0009] Positioning is stuck in two dimensions, lacking millimeter-level digital twins (80% of the point cloud is missing);

[0010] Data silos are severe (interface standardization rate is 32%). The root cause is the lagging transformation to Industry 4.0, and the difficulty of matching the technical route with the requirements of precision and digital operation and maintenance of electric trains. Summary of the Invention

[0011] (1) Technical issues to be solved

[0012] In response to the existing technical problems, the present invention provides a pantograph carbon slide detection system based on MR glasses.

[0013] (2) Technical solution

[0014] In order to achieve the above objectives, the main technical solutions adopted by the present invention include:

[0015] A pantograph carbon slide plate detection system based on MR glasses, comprising: an MR glasses terminal, a data processing module and a user interaction module;

[0016] The MR glasses terminal includes: an MR glasses body, a data acquisition component, a processor component and a display component;

[0017] The data acquisition component, the processor component and the display component are all arranged on the MR glasses body;

[0018] The processor component is connected to the data acquisition component and the display component respectively;

[0019] The data processing module and the user interaction module are both mounted on the processor component;

[0020] The data processing module includes a data fusion algorithm, a thickness calculation model and a wear analysis model, which are used to process and analyze the carbon slide plate detection data collected by the data acquisition component in real time to obtain the detection results;

[0021] The user interaction module can display the detection results, alarm information and maintenance guidance information through the display interface of the MR glasses terminal.

[0022] Preferably, the data acquisition component includes: an infrared camera, a laser ranging module and a gyroscope;

[0023] The infrared camera, the laser ranging module and the gyroscope are all arranged on the MR glasses body;

[0024] The infrared camera, the laser ranging module and the gyroscope are all connected to the processor component and can send the collected carbon skateboard detection data to the processor component for real-time processing and analysis of the detection data through the data processing module to obtain detection results.

[0025] Preferably, the infrared camera is capable of capturing thermal distribution images of the carbon slide surface, so as to identify abnormal temperature characteristics of the wear area through the data processing module;

[0026] The laser ranging module can obtain real-time distance data of multiple points on the surface of the carbon slide in an array scanning mode;

[0027] The gyroscope can continuously track the spatial position and attitude angle of the MR glasses;

[0028] The data processing module can dynamically calibrate the detection position through the spatial position and attitude angle of the MR glasses combined with the preset carbon skateboard three-dimensional model.

[0029] Preferably, the laser ranging module includes a laser ranging sensor;

[0030] The laser ranging sensor is connected to the processor component;

[0031] The gyroscope and the laser ranging sensor can detect the user's spatial position data, attitude angle data and motion state data in real time;

[0032] The data processing module can compare the real-time detected user's spatial position data, posture angle data and motion state data with the preset target position data and posture data range to determine whether the user is in the correct working position and posture.

[0033] Preferably, the data acquisition component and the processor component transmit data via the WebSocket protocol to support simultaneous acquisition and analysis by multiple sensors.

[0034] Preferably, the data processing module can be equipped with a Python script and acquire sensor data in real time through the SPI interface of the MR glasses;

[0035] The NumPy of the data processing module is used to process and analyze sensor data in real time to detect temperature changes of the carbon slide plate;

[0036] The data processing module can use the machine learning model SVM to determine the wear status.

[0037] Preferably, the data processing module can also use an improved Kalman filter to perform data cleaning on the collected carbon slide detection data to remove noise and abnormal values in the sensor data.

[0038] Preferably, the processor component further comprises: a database;

[0039] The database can store detection result data and preset target position data and posture data.

[0040] Preferably, the data processing module can use the laser ranging module to obtain 25 measurement points on the surface of the carbon skateboard in a 5×5 array scanning mode, and combine the infrared thermal imaging of the infrared camera to capture local hot spots, and then use Kalman filtering to fuse the spatial coordinates and temperature data to accurately locate the wear area.

[0041] Preferably, the user interaction module can provide visual guidance, tactile prompts and intelligent maintenance guidance to the user through the MR glasses terminal;

[0042] Visual guidance includes: superimposing a virtual arrow on the display interface to guide the target detection position;

[0043] Tactile prompts: Vibration alerts are sent through bone conduction of MR glasses, forming multimodal collaborative feedback with visual alerts;

[0044] Intelligent maintenance guidance: When an abnormality is detected, the knowledge base is automatically retrieved to generate AR maintenance guidance animation.

[0045] (3) Beneficial effects

[0046] The beneficial effects of the present invention are:

[0047] The present application provides a multi-sensor fusion system based on MR glasses, which realizes non-contact real-time detection of carbon skateboard thickness, wear distribution modeling and intelligent judgment through infrared imaging, laser ranging, gyroscope spatial positioning and intelligent algorithms.

[0048] Traditional manual inspection takes 30 minutes per inspection. This application uses AR guidance to compress the inspection process to within 5 minutes and can generate a digital report simultaneously.

[0049] Hardware cost optimization: By utilizing the native sensors of MR glasses (no external equipment required), this application reduces the hardware cost by more than 70% compared to the traditional laser profiler + thermal imager solution (costing approximately $20k). BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A schematic structural diagram of a pantograph carbon slide detection system based on MR glasses provided by the present invention;

[0051] Figure 2 A schematic diagram of the use process of a pantograph carbon slide detection system based on MR glasses provided by the present invention. DETAILED DESCRIPTION

[0052] In order to better explain the present invention and facilitate understanding, the present invention is described in detail below through specific implementation methods in conjunction with the accompanying drawings.

[0053] like Figure 1-Figure 2 As shown: This embodiment discloses a pantograph carbon slide detection system based on MR glasses, including: an MR glasses terminal, a data processing module and a user interaction module. The MR glasses terminal includes: an MR glasses body, a data acquisition component, a processor component and a display component; the data acquisition component, the processor component and the display component are all arranged on the MR glasses body;

[0054] The processor component is connected to the data acquisition component and the display component respectively; the data processing module and the user interaction module are both mounted on the processor component; the data processing module includes a data fusion algorithm, a thickness calculation model and a wear analysis model, which are used to perform real-time processing and analysis of the carbon skateboard detection data collected by the data acquisition component to obtain detection results; the user interaction module can display the detection results, alarm information and maintenance guidance information through the display interface of the MR glasses terminal.

[0055] The data acquisition components in this embodiment include: an infrared camera, a laser ranging module and a gyroscope.

[0056] The infrared camera, the laser ranging module, and the gyroscope are all disposed on the MR glasses body. The infrared camera, the laser ranging module, and the gyroscope are all connected to the processor component and are capable of transmitting collected carbon slide plate detection data to the processor component, which processes and analyzes the detection data in real time through the data processing module to obtain detection results.

[0057] Specifically, the infrared camera captures thermal images of the carbon slide's surface, enabling the data processing module to identify temperature anomalies in worn areas. The laser ranging module acquires real-time distance data from multiple points on the carbon slide's surface using an array scanning pattern. The gyroscope continuously tracks the spatial position and attitude angle of the MR glasses. The data processing module dynamically calibrates the detection position based on the spatial position and attitude angle of the MR glasses and a pre-set three-dimensional model of the carbon slide.

[0058] In this embodiment, the laser ranging module includes a laser ranging sensor connected to the processor assembly. The gyroscope and the laser ranging sensor are capable of detecting the user's spatial position data, attitude angle data, and motion state data in real time. The data processing module is capable of comparing the real-time detected user spatial position data, attitude angle data, and motion state data with preset target position data and attitude data ranges to determine whether the user is in the correct working position and posture.

[0059] Specifically, the data acquisition component and the processor component perform data transmission via the WebSocket protocol to support simultaneous data acquisition and analysis by multiple sensors.

[0060] Using the WebSocket protocol, it supports simultaneous data collection and analysis from multiple sensors. Sensor data is transmitted to the data processing module, enabling low-latency, full-duplex, real-time communication. Data is transmitted directly after a WebSocket connection is established. Efficiency: Lightweight data frames are suitable for high-frequency sensor data transmission. Supports a variety of terminals, including browsers, mobile devices, and embedded devices.

[0061] The data processing module described in this embodiment can be equipped with Python scripts and obtain sensor data in real time through the SPI interface of the MR glasses.

[0062] In actual applications, by writing Python scripts, sensor data is obtained in real time through the SPI interface of MR glasses.

[0063] The NumPy of this data processing module is used to process and analyze sensor data in real time to detect temperature changes of the carbon slide. Abnormal temperature increases may indicate increased wear; the data processing module can use the machine learning model SVM to determine the wear status.

[0064] The data processing module in this embodiment can also use an improved Kalman filter to perform data cleaning on the collected carbon slide detection data to remove noise and abnormal values in the sensor data.

[0065] The data processing module also uses a gyroscope, real-time data processing algorithms, and a multimodal feedback mechanism to monitor the user's position, posture, and motion status in real time. This includes the use of edge computing and asynchronous programming. Machine learning models are trained multiple times to determine status.

[0066] The processor component in this embodiment further includes: a database; the database can store detection result data and preset target position data and posture data.

[0067] The analysis results can be fed back to the MR glasses or detection system for real-time display or storage in a database. The analysis results can be visualized through the MR glasses, providing real-time feedback and warnings. Sound or vibration notifications indicate abnormal conditions, and WebSocket integration enables real-time data transmission. A virtual arrow is superimposed on the MR glasses to indicate the target location. Posture deviations are color-coded (green for normal, red for abnormal) and displayed through the MR glasses' AR interface and tactile prompts.

[0068] The data processing module can use the laser ranging module to obtain 25 measurement points on the surface of the carbon skateboard in a 5×5 array scanning mode, and combine the infrared thermal imaging of the infrared camera to capture local hotspots, and then use Kalman filtering to fuse the spatial coordinates and temperature data to accurately locate the wear area.

[0069] The user interaction module described in this embodiment can provide visual guidance, tactile prompts and intelligent maintenance guidance to the user through the MR glasses terminal; visual guidance includes: superimposing a virtual arrow on the display interface to indicate the target detection position; tactile prompts: emitting a vibration alarm through bone conduction of the MR glasses, and forming multimodal collaborative feedback with the visual alarm; intelligent maintenance guidance: when an abnormality is detected, the knowledge base is automatically retrieved to generate an AR maintenance guidance animation.

[0070] 1. Real-time and high efficiency of the entire process

[0071] Real-time processing on the device: Python is deployed directly on the MR glasses terminal. Combined with data fusion algorithms, thickness calculation models, and SVM machine learning models, this enables local processing of the entire "collection-analysis-feedback" process without relying on cloud servers, eliminating network latency.

[0072] Efficient communication architecture: The WebSocket protocol is used to support concurrent transmission of multiple sensors (such as infrared images, laser point clouds, and gyroscope attitude angles). The full-duplex communication bandwidth utilization rate reaches over 95%, meeting the real-time transmission requirements of more than 1,000 data points per second and avoiding the protocol overhead of traditional HTTP polling.

[0073] 2. Multimodal data fusion and high-precision detection

[0074] Multi-sensor collaborative analysis: The laser ranging module acquires 25 measurement points on the carbon slide surface in a 5×5 array scanning mode (resolution 0.01mm), combined with infrared thermal imaging (temperature sensitivity ±1°C) to capture local hotspots, and uses Kalman filtering to fuse spatial coordinates and temperature data to accurately locate the wear area (error <0.5mm).

[0075] 3. Intelligent analysis and adaptive learning

[0076] Machine learning based on edge computing: A lightweight SVM model (compressed to less than 50KB) is deployed on the MR glasses, and a classifier is trained using historical wear data to achieve real-time classification of the carbon slide wear status (normal / mild wear / severe wear) with an accuracy rate of >98%.

[0077] Adaptive filtering algorithm: To address laser ranging noise (typical noise amplitude ±0.05mm), an improved Kalman filter (state equation fusion device motion model) is used to increase the data signal-to-noise ratio to over 30dB, ensuring a thickness calculation error of <5μm.

[0078] 4. Innovation in machine-computer interaction and operation guidance

[0079] Multi-channel augmented reality feedback: Visual guidance: A virtual arrow (positioning accuracy ±0.5°) is superimposed on the AR interface to guide the target detection location, and the wear area is rendered using a heat map (color mapping response time <20ms). Tactile prompts: Vibration alerts (adjustable frequency range 50-200Hz) are issued through the MR glasses' bone conduction module, forming multimodal collaborative feedback with visual alerts (flashing red areas). Intelligent maintenance guidance: When an anomaly is detected, the knowledge base is automatically retrieved to generate AR maintenance guidance animations (such as a 3D demonstration of the carbon slide replacement steps).

[0080] 5. Improved Inspection Efficiency: Traditional manual inspections take 30 minutes per session. This invention compresses the inspection process to less than 5 minutes through AR guidance, and can simultaneously generate a digital report (compliant with EN 50317:2022 standards). Hardware Cost Optimization: Utilizing the native sensor of MR glasses (no external equipment required), this invention reduces hardware costs by over 70% compared to the traditional laser profiler + thermal imager solution (costing approximately $20,000).

[0081] In practical applications, such as Figure 2 As shown: The use process of the pantograph carbon slide plate detection system based on MR glasses provided in this embodiment is as follows:

[0082] 1. System Preparation Phase

[0083] 1. Hardware deployment: Wear HoloLens 2, install magnetic positioning tags, and deploy edge computing servers;

[0084] 2. Software configuration: Load the chassis 3D model library, maintenance standard database and Python.

[0085] 2. Maintenance process starts

[0086] 1. Start HoloLens2 to generate a dynamic mesh and perform a ToF environment scan;

[0087] 2. Automatically identify devices and calibrate spatial anchor points, 3D models and physical objects.

[0088] 3. Real-time maintenance monitoring

[0089] 1. Data collection: ultrasonic flaw detector detection equipment, synchronous tracking of gesture operation process;

[0090] 2. Exception handling: NumPy identifies defect characteristics in real time and intelligently determines the defect level.

[0091] 4. Feedback and Optimization

[0092] 1. Generate automatic reports with equipment health scores / life predictions and archive operation logs;

[0093] 2. Optimize the system and iterate the system through error analysis and personnel capability assessment.

[0094] The technical principles of the present invention have been described in conjunction with specific embodiments. These descriptions are intended solely to illustrate the principles of the present invention and are not to be construed in any way as limiting the scope of protection of the present invention. Based on the explanations herein, those skilled in the art will be able to devise other specific embodiments of the present invention without inventive effort, and such embodiments will fall within the scope of protection of the present invention.

Claims

1. A pantograph carbon slide detection system based on MR glasses, characterized in that: include: MR glasses terminal, data processing module and user interaction module; The MR glasses terminal includes: an MR glasses body, a data acquisition component, a processor component and a display component; The data acquisition component, the processor component and the display component are all arranged on the MR glasses body; The processor component is connected to the data acquisition component and the display component respectively; The data processing module and the user interaction module are both mounted on the processor component; The data processing module includes a data fusion algorithm, a thickness calculation model and a wear analysis model, which are used to process and analyze the carbon slide plate detection data collected by the data acquisition component in real time to obtain the detection results; The user interaction module can display the detection results, alarm information and maintenance guidance information through the display interface of the MR glasses terminal.

2. The pantograph carbon slide plate detection system based on MR glasses according to claim 1 is characterized in that: The data acquisition component includes: an infrared camera, a laser ranging module and a gyroscope; The infrared camera, the laser ranging module and the gyroscope are all arranged on the MR glasses body; The infrared camera, the laser ranging module and the gyroscope are all connected to the processor component and can send the collected carbon skateboard detection data to the processor component for real-time processing and analysis of the detection data through the data processing module to obtain detection results.

3. The pantograph carbon slide detection system based on MR glasses according to claim 2 is characterized in that ； The infrared camera can capture the thermal distribution image of the carbon slide surface, so as to identify the abnormal temperature characteristics of the wear area through the data processing module; The laser ranging module can obtain real-time distance data of multiple points on the surface of the carbon slide in an array scanning mode; The gyroscope can continuously track the spatial position and attitude angle of the MR glasses; The data processing module can dynamically calibrate the detection position through the spatial position and attitude angle of the MR glasses combined with the preset carbon skateboard three-dimensional model.

4. The pantograph carbon slide plate detection system based on MR glasses according to claim 3 is characterized in that: The laser ranging module includes a laser ranging sensor; The laser ranging sensor is connected to the processor component; The gyroscope and the laser ranging sensor can detect the user's spatial position data, attitude angle data and motion state data in real time; The data processing module can compare the real-time detected user's spatial position data, posture angle data and motion state data with the preset target position data and posture data range to determine whether the user is in the correct working position and posture.

5. The pantograph carbon slide plate detection system based on MR glasses according to claim 4 is characterized in that: The data acquisition component and the processor component transmit data via the WebSocket protocol to support simultaneous data collection and analysis by multiple sensors.

6. The pantograph carbon slide plate detection system based on MR glasses according to claim 5, characterized in that: The data processing module can be equipped with Python scripts and acquire sensor data in real time through the SPI interface of the MR glasses; The NumPy of the data processing module is used to process and analyze sensor data in real time to detect temperature changes of the carbon slide plate; The data processing module can use the machine learning model SVM to determine the wear status.

7. The pantograph carbon slide plate detection system based on MR glasses according to claim 6, characterized in that: The data processing module can also use an improved Kalman filter to perform data cleaning on the collected carbon slide detection data to remove noise and abnormal values in the sensor data.

8. The pantograph carbon slide plate detection system based on MR glasses according to claim 7, characterized in that: The processor component further includes: a database; The database can store detection result data and preset target position data and posture data.

9. The pantograph carbon slide plate detection system based on MR glasses according to claim 8, characterized in that: The data processing module can use the laser ranging module to obtain 25 measurement points on the surface of the carbon skateboard in a 5×5 array scanning mode, and combine the infrared thermal imaging of the infrared camera to capture local hotspots, and then use Kalman filtering to fuse the spatial coordinates and temperature data to accurately locate the wear area.

10. The pantograph carbon slide plate detection system based on MR glasses according to claim 9, characterized in that: The user interaction module can provide visual guidance, tactile prompts and intelligent maintenance guidance to the user through the MR glasses terminal; Visual guidance includes: superimposing a virtual arrow on the display interface to guide the target detection position; Tactile prompts: Vibration alerts are sent through bone conduction of MR glasses, forming multimodal collaborative feedback with visual alerts; Intelligent maintenance guidance: When an abnormality is detected, the knowledge base is automatically retrieved to generate AR maintenance guidance animation.

Images