Railway line intelligent monitoring method and platform based on cloud edge collaboration

Abstract

The application discloses a kind of railway line intelligent monitoring method and platform based on cloud edge cooperation, belong to railway safety detection technical field, platform includes whole line monitoring center, station monitoring unit, network collection unit, line sensing acquisition unit and user monitoring terminal;Station monitoring unit obtains the foreign matter intrusion sensing signal that line sensing acquisition unit transmits through network collection unit in real time, and to sensing signal is mediated and preprocessed and the single-point abnormal state that single sensor is sensed is monitored and analyzed;To whole line monitoring center transmission is mediated and preprocessed after sensing signal and single-point abnormal state, to make whole line monitoring center response and based on deep learning algorithm to sensing signal and single-point abnormal state are monitored and analyzed railway line foreign matter intrusion risk early warning, and real-time transmission single-point abnormal state and foreign matter intrusion risk early warning.The application can improve the efficiency and accuracy of railway line monitoring, guarantee the safety of railway transportation.

Description

Technical Field

[0001] This invention belongs to the field of railway safety inspection technology, specifically relating to a cloud-edge collaborative intelligent monitoring method and platform for railway lines. Background Technology

[0002] Numerous safety hazards exist along railway lines, such as vandalism and natural disasters. These factors can lead to foreign objects encroaching on railway lines, thereby affecting the safety and normal operation of railway transportation. Traditional railway monitoring methods mainly rely on manual inspections and simple sensor monitoring, which have problems such as limited monitoring range, poor real-time performance, and insufficient data processing capabilities.

[0003] In recent years, cloud-edge collaborative technology has developed rapidly and has been widely applied in many fields. By combining the powerful data processing capabilities of cloud computing with the low latency and high real-time performance of edge computing, cloud-edge collaborative technology can effectively solve the data processing and real-time response problems in railway line monitoring. In railway line monitoring, cloud-edge collaborative technology can achieve real-time and comprehensive monitoring of the railway line, and promptly detect and warn of safety hazards such as foreign object intrusion. Currently, although some railway monitoring systems employ technologies such as video analysis and radar scanning, most of these systems suffer from insufficient data processing capabilities and limited monitoring range. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and provide a cloud-edge collaborative intelligent monitoring method and platform for railway lines, thereby improving the efficiency and accuracy of railway line monitoring and ensuring the safety of railway transportation.

[0005] To achieve the above objectives, the present invention is implemented using the following technical solution:

[0006] In a first aspect, the present invention provides a method for intelligent monitoring of railway lines based on cloud-edge collaboration, which is applied to a cloud-edge collaborative intelligent monitoring platform for railway lines. The platform includes a full-line monitoring center, station monitoring units, network aggregation units, line-side sensing and acquisition units, and user monitoring terminals.

[0007] The full-line monitoring center is interconnected with several station monitoring units and user monitoring terminals set up along the railway line through several network aggregation units. The station monitoring units are configured at stations along the railway line and are connected to several line sensing and acquisition units distributed along the railway line with at least one unit between adjacent stations through several network aggregation units. The line sensing and acquisition units are equipped with at least two different types of sensors for sensing foreign object intrusion.

[0008] And including the following steps:

[0009] The station monitoring unit acquires in real time the foreign object intrusion detection signals collected by various types of sensors from the line-side sensing and acquisition unit via the network aggregation unit.

[0010] And to modulate and preprocess the sensed signals and monitor and analyze the single-point abnormal state sensed by a single sensor.

[0011] The station monitoring unit transmits the pre-processed sensing signals and single-point anomaly statuses to the full-line monitoring center via the network aggregation unit.

[0012] This enables the entire monitoring center to respond and, based on deep learning algorithms, monitor and analyze the sensed signals and single-point anomalies, providing early warnings of foreign object intrusion risks along the railway line.

[0013] And according to preset emergency rules, send real-time warnings of single-point abnormal status and foreign object intrusion risk to the user monitoring terminal.

[0014] Furthermore, the station monitoring unit is connected to at least one of the line-of-line sensing and acquisition units distributed between the current station and the previous and / or the next station, for monitoring the abnormal status of the railway maintenance section corresponding to the current station.

[0015] The sensing and data acquisition unit along the railway line marks location symbols in a certain order, and the location symbols include codes and latitude and longitude.

[0016] Furthermore, the line-of-sight sensing and acquisition unit is equipped with a camera, radar, and fiber optic sensor;

[0017] The cameras and / or radars are arranged at intervals along the railway line and facing the railway tracks and protective fences to collect radar data and video data.

[0018] The fiber optic sensors are installed on the bottom of the protective fence along both sides of the railway line to collect fiber optic vibration and stress data.

[0019] Furthermore, monitoring and analyzing single-point abnormal states sensed by individual sensors includes the following steps:

[0020] The station monitoring unit acquires radar data, video data, fiber optic vibration and stress data collected and transmitted by several sensing and acquisition units along the line connected to the current station where the station monitoring unit is located;

[0021] After preprocessing the fiber optic vibration and stress data, the station monitoring unit judges the vibration intensity and stress intensity according to preset threshold rules:

[0022] If the vibration intensity exceeds the first warning threshold and the duration exceeds the first time threshold and / or if the stress intensity exceeds the second warning threshold and the duration exceeds the second time threshold, then it is determined to be in a single-point abnormal state.

[0023] When a single point of abnormality occurs, the radar data and video data are retrieved and displayed in real time at the station monitoring unit. The radar data, video data, fiber optic vibration and stress data, single point of abnormality, and location information are associated with the location marker and then transmitted to the full-line monitoring center.

[0024] Furthermore, monitoring and analysis of the risk of foreign object intrusion along railway lines will be conducted, including the following steps:

[0025] The station monitoring unit sends transmission data to the full-line monitoring center so that the full-line monitoring center can respond and perform target object detection and analysis based on deep learning algorithm to fuse the radar data and video data collected by the individual sensing and acquisition units along the line to determine whether there are foreign objects; if there are, it is determined as a foreign object intrusion risk warning.

[0026] If not, the system monitors and analyzes the data transmitted by other sensing and acquisition units along the line connected to the current station within the same time period. If two or more sensing and acquisition units along the line are determined to be in a single-point abnormal state, it is determined to be a foreign object intrusion risk warning.

[0027] Furthermore, the preset emergency rules are as follows:

[0028] If, within the same time period, only one of the several sensing and acquisition units connected to the current station is determined to be in a single-point abnormal state, the entire line monitoring center sends the single-point abnormal state with a matching location symbol and a preset first emergency maintenance plan to the user monitoring terminal in real time, so that maintenance workers can conduct fixed-point inspections of the railway line protective fence according to the location symbol.

[0029] When several of the aforementioned sensing and acquisition units connected to the current station within the same time period determine that there is a risk warning of foreign object intrusion, the entire line monitoring center sends a foreign object intrusion risk warning with a matching location symbol and a preset second emergency maintenance plan to the user monitoring terminal in real time. This plan is used by maintenance workers to remove foreign objects from the railway line protective fence according to the location symbol.

[0030] Secondly, the present invention provides a cloud-edge collaborative intelligent monitoring platform for railway lines, including a full-line monitoring center, station monitoring units, network aggregation units, line-side sensing and acquisition units, and user monitoring terminals.

[0031] The line sensing and acquisition unit is equipped with at least two different types of sensors for sensing foreign object intrusion. The station monitoring unit is configured at railway stations along the line and is connected to several line sensing and acquisition units distributed along the railway line and configured at least one between adjacent stations through several network aggregation units. It is used to adjust and preprocess the sensing signals collected by each individual sensor in the line sensing and acquisition unit and monitor the single-point abnormal state corresponding to each individual sensor.

[0032] The entire line monitoring center is interconnected with several station monitoring units and user monitoring terminals set up along the railway through several network aggregation units. It is used to monitor and analyze the sensing signals and individual foreign object intrusion status transmitted by each of the station monitoring units based on deep learning algorithms, and to issue early warnings of foreign object intrusion risks along the railway. It also sends single-point abnormal status and foreign object intrusion risk warnings to the user monitoring terminals in real time.

[0033] Furthermore, the station monitoring unit is connected to at least one of the line-of-line sensing and acquisition units distributed between the current station and the previous and / or the next station, for monitoring the abnormal status of the railway maintenance section corresponding to the current station.

[0034] The sensing and data acquisition unit along the railway line marks location symbols in a certain order, and the location symbols include codes and latitude and longitude.

[0035] Furthermore, the line-of-sight sensing and acquisition unit is equipped with a camera, radar, and fiber optic sensor;

[0036] The cameras and / or radars are arranged at intervals along the railway line and facing the railway tracks and protective fences to collect radar data and video data.

[0037] The fiber optic sensors are installed on the bottom of the protective fence along both sides of the railway line to collect fiber optic vibration and stress data.

[0038] Thirdly, the present invention provides a monitoring platform device, comprising:

[0039] Interconnected memory and processor,

[0040] The processor is used to execute program instructions stored in the memory to implement the cloud-edge collaborative intelligent monitoring method for railway lines as described in any one of the first aspects.

[0041] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: The technical solution of the present invention is based on cloud-edge collaboration technology. The station monitoring unit can acquire foreign object intrusion sensing signals from the sensing and acquisition units along the line in real time, and perform preprocessing and single-point abnormal state monitoring. The preprocessed data is transmitted to the monitoring center of the entire line through the network aggregation unit, and further analyzed and warned by deep learning algorithms. The cloud-edge collaboration technology between the station monitoring unit and the monitoring center of the entire line reduces the dependence on the central cloud computing resources, reduces the cost of data transmission and processing, and the hierarchical processing method greatly improves the efficiency and real-time performance of data processing, and can detect and warn of foreign object intrusion risks in a timely manner.

[0042] The sensing and acquisition unit along the railway line is equipped with various types of sensors, such as cameras, radar, and fiber optic sensors, which can sense environmental changes along the railway line from multiple angles and dimensions, providing more comprehensive and accurate monitoring data; it also uses deep learning algorithms to fuse and analyze multi-source data, more accurately determining whether there is a risk of foreign object intrusion.

[0043] Based on deep learning algorithms, the system monitors and analyzes sensing signals and single-point abnormal states, automatically identifies the risk of foreign object intrusion, and sends early warning information and emergency maintenance plans to the user's monitoring terminal in real time according to preset emergency rules, which greatly reduces manual intervention and improves the reliability and efficiency of monitoring.

[0044] The connection between the station monitoring unit and the sensing and acquisition units along the line, as well as the connection between the full-line monitoring center and the station monitoring units and user monitoring terminals, forms a complete monitoring network platform. This platform not only enables real-time data transmission and sharing, but also allows for precise location of monitoring data through location markers, facilitating rapid response and processing by maintenance personnel. Attached Figure Description

[0045] Figure 1 A flowchart of a cloud-edge collaborative intelligent monitoring method for railway lines provided in an embodiment of the present invention.

[0046] Figures 2-3 This is a block diagram of a cloud-edge collaborative intelligent monitoring platform for railway lines, provided as an embodiment of the present invention.

[0047] Figure 4 This is an architecture diagram of a cloud-edge collaboration technology provided in an embodiment of the present invention.

[0048] Figure 5 This is a schematic diagram of a station monitoring unit and a line-of-line sensing and acquisition unit provided in an embodiment of the present invention.

[0049] Figure 6 This is a schematic diagram of an optical fiber sensor installation provided in an embodiment of the present invention.

[0050] Figure 7 This is a schematic diagram of a radar and video system provided in an embodiment of the present invention.

[0051] Figures 8-9 This is a schematic diagram of radar data and video data provided in an embodiment of the present invention.

[0052] Figures 10-11 This is a schematic diagram of video data and stress data of an optical fiber sensor provided in an embodiment of the present invention.

[0053] In the picture: 1. Protective fence; 2. Fiber optic sensor; 3. Encryption net; 4. Camera; 5. Pole; 6. Radar. Detailed Implementation

[0054] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.

[0055] Example

[0056] like Figure 1 As shown, this embodiment of the invention provides a method for intelligent monitoring of railway lines based on cloud-edge collaboration, applied to a cloud-edge collaborative intelligent monitoring platform for railway lines. Figure 2 , Figure 3 and Figure 4 As shown, the platform includes a full-line monitoring center, station monitoring units, network aggregation units, line-side sensing and data acquisition units, and user monitoring terminals.

[0057] The monitoring center of the entire line is interconnected with several station monitoring units and user monitoring terminals set up along the railway line through several network aggregation units. The station monitoring units are configured at stations along the railway line and are connected to several line sensing and acquisition units distributed along the railway line with at least one unit between adjacent stations through several network aggregation units. The line sensing and acquisition units are equipped with at least two different types of sensors for sensing foreign object intrusion.

[0058] like Figure 1 As shown, the intelligent monitoring method along railway lines based on cloud-edge collaboration includes the following steps:

[0059] The station monitoring unit acquires in real time foreign object intrusion detection signals collected by various types of sensors from the sensing and acquisition units along the line, transmitted through the network aggregation unit.

[0060] And to modulate and preprocess the sensed signals and monitor and analyze the single-point abnormal state sensed by a single sensor.

[0061] The station monitoring unit transmits the pre-processed sensing signals and individual point anomalies to the entire line monitoring center via the network aggregation unit.

[0062] This enables the monitoring center along the entire railway line to respond and, based on deep learning algorithms, monitor and analyze the sensed signals and individual point anomalies to provide early warnings of foreign object intrusion risks along the railway line.

[0063] It also sends real-time warnings of single-point abnormal status and foreign object intrusion risk to the user's monitoring terminal according to preset emergency rules.

[0064] In this embodiment, the station monitoring unit is connected to at least one line-of-line sensing and acquisition unit distributed between the current station and the previous and / or the next station, for monitoring the abnormal status of the railway maintenance section corresponding to the current station.

[0065] Along the railway line, the sensing and data acquisition units are marked with location symbols in a certain order. The location symbols include codes and latitude and longitude, which are used to identify the unique data acquisition unit and locate the monitoring position.

[0066] Typically, only one monitoring center is deployed along the entire railway line to conduct comprehensive monitoring and centralized dispatch management. Station monitoring units are deployed at each station along the line, serving as the initial monitoring center for each station. They perform adjustment and preprocessing of individual monitoring and sensing data, conduct simple comparative analysis of the data, and assess the abnormal states corresponding to the monitoring points, enabling station management and maintenance personnel to make preliminary inspection strategies.

[0067] Edge computing technology is used in the station monitoring unit to process and filter monitoring data from the front-end sensing and acquisition units along the line, and uploads the valid data to the full-line monitoring center. The cloud-edge collaborative computing resource scheduling technology mainly relies on edge computing technology to achieve effective scheduling of computing resources, effectively reducing the throughput pressure on the device end and the cloud, alleviating the load pressure on the data center, improving the data analysis response efficiency, and fully ensuring the stability and efficiency of system operation through the cloud-edge collaborative big data platform.

[0068] In order to meet the requirements of national information technology innovation and the "Technical Requirements for Intelligent Video Analysis and Related Equipment for Railway Safety Environment (Provisional)" and other standards, and to meet the development requirements of various system functions, this embodiment of the invention develops and adopts intelligent analysis edge terminals, which include two models: intelligent edge computing box (PG-ECD-BH01) and intelligent edge server (PG-ECD-HH02).

[0069] The hardware layer includes computing units, storage units, and network modules. The computing units process local data, utilizing domestically produced CPUs compatible with domestic IT innovation standards, as well as AI chips or inference cards specifically designed for AI computing. Storage units store data; in enclosures, solid-state drives (SSDs) are used, while server-grade SSDs are supported. Network modules connect to other devices or networks, employing Ethernet or fiber optic interfaces. Operating systems include domestic systems such as Kylin and Euler, offering efficient hardware resource management and underlying support. The edge computing platform provides computing resource management, data processing, and containerization / virtualization support, supporting Docker container deployment. The application layer includes data acquisition and preprocessing, local AI / ML inference, and network security and management. Data acquisition and preprocessing obtains raw data from devices and sensors, performing initial cleaning and transformation. Local AI / ML inference runs machine learning models locally for inference and computational analysis. Network security and management provide firewall, data encryption, and authentication functions to ensure the security of edge devices and data.

[0070] In the network aggregation unit, low-latency, high-bandwidth communication technologies, such as wired broadband, 5G, and Internet of Things (IoT), are employed to ensure efficient data transmission from edge devices to the cloud. This enables cloud-edge data synchronization and backup, ensuring no data loss even during network interruptions, and providing automatic data recovery and compensation mechanisms. Simultaneously, tasks are dynamically scheduled to the cloud or edge nodes based on network conditions and computing resources to ensure the overall system's efficient operation.

[0071] In this embodiment, machine learning and deep learning algorithms are widely used in the station monitoring unit and the overall monitoring center. For example, machine learning algorithms such as Support Vector Machine (SVM) and Random Forest (RF) are applied to learn and analyze historical and real-time data, improving the self-learning and predictive capabilities of the safety monitoring system. Deep learning algorithms such as Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN) are used to extract features and detect targets from video and radar data, improving the accuracy and real-time performance of foreign object intrusion detection.

[0072] like Figure 5As shown, the local network of the station monitoring unit and the sensing and acquisition unit along the line employs interconnected components such as fiber optic sensors, fiber optic monitoring hosts, optical cables, and transmission devices. The fiber optic monitoring host, equipped with algorithm software, has the function of monitoring perimeter intrusion or foreign object intrusion. It can operate independently or interface with the station monitoring unit and the entire line monitoring center to exchange monitoring alarm information / data. Real-time monitoring of perimeter intrusion and foreign object intrusion is achieved. Fiber optic composite sensors and armored steel strands are installed in various perimeter intrusion and foreign object intrusion scenarios along the line, including perimeter boundaries, slopes, active networks, and passive networks. The fiber optic composite sensors are connected to the fiber optic monitoring host via optical cables. The monitoring host emits detection light through the optical cable to the fiber optic composite sensor, and detects the wavelength of the light reflected back by the fiber optic grating, thereby achieving real-time monitoring and alarm location of perimeter intrusion and foreign object intrusion scenarios.

[0073] In this embodiment, the application scenario is the throat of the railway line, bridge and tunnel entrances, roadbed sections, and under overpasses. The main monitoring targets are personnel and large animals intruding onto the track, and foreign objects (falling rocks, landslides, debris flows, thrown objects, and other debris) encroaching on the track at the throat of the railway line, bridge and tunnel entrances, roadbed sections, and under overpasses. Specifically, for example, perimeter intrusion monitoring, targeting perimeter security along high-speed railways and subway lines and stations; foreign object intrusion monitoring, targeting foreign object interference monitoring such as active nets, passive nets, retaining walls, and anti-throw nets; train trajectory monitoring, etc. Figure 10 and Figure 11 The image shows people climbing and disturbing the area.

[0074] The station monitoring unit also requires the analysis and processing of raw data, including data cleaning, transformation, enhancement, noise reduction, and filtering.

[0075] This involves processing raw or primary data through annotation, transforming it into information that can be recognized by machines or algorithms. Data annotation is a crucial step in constructing multi-source datasets, directly impacting the effectiveness of subsequent model training. Data annotation needs to consider accuracy, consistency, and efficiency. Image annotation typically uses LabelMe, LabelImg, X-AnyLabeling, VoTT, and our self-developed annotation platform; point cloud annotation generally uses tools such as CloudCompare and Sustechpoints; and fiber optic annotation uses our self-developed annotation tools.

[0076] like Figure 3 As shown, the sensing and acquisition unit along the line is equipped with cameras, radar, and fiber optic sensors.

[0077] Cameras and / or radars are positioned at intervals along the railway line, facing the railway tracks and protective fence, to collect radar and video data, such as... Figure 7 As shown.

[0078] Fiber optic sensors are installed at the bottom of the reinforced mesh of the protective fence along both sides of the railway line to collect data on fiber optic vibration and stress. Figure 6 As shown. In addition, fiber optic sensors can also be laid on structures such as protective nets, anti-throw nets, and active / passive nets at tunnel entrances, slopes, and bridge crossings.

[0079] It should be noted that, in this embodiment, in addition to radar data, video data, and fiber optic vibration and stress data, other sensor data can also be fused, such as meteorological data (temperature, humidity, wind speed, etc.) and geological data (seismic waves, groundwater level, etc.). Through multi-source data fusion, the environmental and safety conditions along the railway line can be assessed more comprehensively.

[0080] In this embodiment, monitoring and analyzing a single-point abnormal state sensed by a single sensor includes the following steps:

[0081] The station monitoring unit acquires radar data, video data, fiber optic vibration and stress data collected and transmitted by several sensing and acquisition units along the line connected to the current station where the station monitoring unit is located;

[0082] After preprocessing the fiber optic vibration and stress data, the station monitoring unit judges the vibration intensity and stress intensity according to preset threshold rules:

[0083] If the vibration intensity exceeds the first warning threshold and the duration exceeds the first time threshold and / or if the stress intensity exceeds the second warning threshold and the duration exceeds the second time threshold, then it is determined to be in a single-point abnormal state.

[0084] When a single point of abnormality occurs, radar and video data are retrieved and displayed in real time at the station monitoring unit. The radar data, video data, fiber optic vibration and stress data, single point of abnormality status and location symbols are associated and then transmitted to the monitoring center of the entire line.

[0085] It should be noted that the station monitoring unit can dynamically adjust the early warning threshold based on real-time environment and historical data, thereby improving the accuracy and timeliness of early warnings.

[0086] The station monitoring unit combines historical data to analyze the spatiotemporal correlation of data from multiple sensors at different locations. By analyzing the changing trends of sensor data at different locations over different times, it can determine whether there are potential safety hazards.

[0087] Furthermore, based on the severity of the foreign object intrusion risk, the early warning system is divided into multiple levels, such as low risk, medium risk, and high risk. Different levels of early warning trigger different emergency response measures. When a single point of abnormality is detected, the abnormality level is initially assessed using multi-level early warning thresholds. If it is confirmed as low risk, an early warning message is sent and regular inspections are arranged; if it is confirmed as high risk, the station monitoring unit immediately activates the emergency response and notifies maintenance personnel to handle the situation on-site.

[0088] In this embodiment, monitoring and analyzing the risk warning of foreign object encroachment along the railway line includes the following steps:

[0089] The station monitoring unit sends data to the full-line monitoring center to enable the full-line monitoring center to respond and perform target detection and analysis based on deep learning algorithms to fuse radar data and video data collected by individual sensing and acquisition units along the line to determine whether there are foreign objects; if there are, it is determined as a foreign object intrusion risk warning.

[0090] If not, the system monitors and analyzes the data transmitted by other sensing and acquisition units along the line connected to the current station within the same time period. If two or more sensing and acquisition units along the line are determined to be in a single-point abnormal state, it is determined to be a foreign object intrusion risk warning.

[0091] In this embodiment, the preset emergency rule is as follows: if only one of the several sensing and acquisition units connected to the current station is determined to be in a single-point abnormal state within the same time period, the monitoring center of the entire line will send the single-point abnormal state with a location symbol and the preset first emergency maintenance plan to the user monitoring terminal in real time, so that maintenance workers can carry out fixed-point inspections of the protective fence along the railway line according to the location determined by the location symbol.

[0092] Furthermore, if several sensing and data collection units connected to the current station within the same time period determine that there is a risk of foreign object intrusion, the entire line monitoring center will send a foreign object intrusion risk warning with a matching location symbol and a preset second emergency maintenance plan to the user monitoring terminal in real time. This plan is used by maintenance workers to remove foreign objects from the railway line protective fence according to the location symbol. When necessary, emergency response involves the collaborative work of multiple departments, such as railway operation departments, maintenance departments, and public security departments. By establishing a cross-departmental emergency response mechanism, it is ensured that all relevant departments can work together quickly and effectively.

[0093] In addition, the platform can provide scientific decision-making suggestions, automatically generate emergency maintenance plans, and make adjustments according to the actual situation.

[0094] like Figure 2 and Figure 3The present invention also provides a cloud-edge collaborative intelligent monitoring platform for railway lines, including a full-line monitoring center, station monitoring units, network aggregation units, line-side sensing and acquisition units, and user monitoring terminals.

[0095] The line sensing and acquisition unit is equipped with at least two different types of sensors for sensing foreign object intrusion. The station monitoring unit is configured at railway stations along the line and is connected to several line sensing and acquisition units distributed along the railway line and configured at least one between adjacent stations through several network aggregation units. It is used to adjust and preprocess the sensing signals collected by each individual sensor in the line sensing and acquisition unit and monitor the single-point abnormal state corresponding to each individual sensor.

[0096] The monitoring center along the entire line is interconnected with several station monitoring units and user monitoring terminals set up along the railway through several network aggregation units. It is used to monitor and analyze the sensing signals transmitted by each station monitoring unit and the status of individual foreign object intrusion based on deep learning algorithms, and to issue early warnings of foreign object intrusion risks along the railway line. It also sends single-point abnormal status and foreign object intrusion risk warnings to user monitoring terminals in real time.

[0097] In this embodiment, as Figure 3 As shown, the entire monitoring center consists of servers, switches, network security equipment, and UPS equipment. Station monitoring units consist of servers, hard disk recorders, switches, network security equipment, and UPS equipment. Network aggregation units consist of switches and UPS equipment. Along-line sensing and acquisition units consist of fiber optic monitoring hosts, fiber optic composite sensors, vibration fibers, intelligent image acquisition equipment, and intelligent electronic sentinels. User terminals consist of hosts, monitors, keyboards, mice, speakers, and firewalls.

[0098] like Figure 3 As shown, the ring network access switches and cascade switches are configured with FE / GE optical ports and FE electrical ports, possessing network switching capabilities and supporting cascaded networking protocols. The cascade switches located at the field acquisition units are connected to the ring network access switches at the nearest network aggregation units via outdoor optical cables. These ring network access switches are then connected to the aggregation layer ring network aggregation switches at the monitoring units. Each monitoring unit is equipped with one ring network aggregation switch. The monitoring unit aggregation switches provide network access for devices such as hard disk recorders and servers, and are networked with the ring network access switches at the network aggregation units via optical cables. Each monitoring unit is securely interconnected with the professional communication data network through routers, firewalls, and other means to achieve data transmission with the monitoring center.

[0099] The monitoring unit and the monitoring center, and the monitoring center and the user terminal, are interconnected via a dedicated communication data network with a bandwidth of no less than 100 Mbps. The VPN private network of the fiber optic integrated monitoring system and the VPN private network of the integrated video system can be connected at stations along the line through network security equipment. The VPN private network of the fiber optic integrated monitoring system and the communication data network VPN private network of the line safety environment management platform cloud system are connected to some specific railway maintenance sections along the line, with a bandwidth of no less than 1 Gbps.

[0100] In this embodiment, the station monitoring unit is connected to at least one line-of-line sensing and acquisition unit distributed between the current station and the previous and / or the next station, for monitoring the abnormal status of the railway maintenance section corresponding to the current station;

[0101] The sensing and data acquisition units along the railway line are marked with location symbols in a certain order. The location symbols include codes and latitude and longitude.

[0102] The sensing and data acquisition unit along the route is equipped with cameras, radar, and fiber optic sensors.

[0103] Cameras and / or radars are positioned at intervals along the railway line and facing the railway tracks and protective fences to collect radar and video data.

[0104] Fiber optic sensors are installed on the bottom of the protective fence along both sides of the railway line to collect data on fiber optic vibration and stress.

[0105] Furthermore, embodiments of the present invention provide a monitoring platform device, comprising:

[0106] Interconnected memory and processor,

[0107] The processor is used to execute program instructions stored in the memory to implement the above-mentioned intelligent monitoring method for railway lines based on cloud-edge collaboration.

[0108] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for intelligent monitoring along railway lines based on cloud-edge collaboration, characterized in that, It is applied to a cloud-edge collaborative intelligent monitoring platform for railway lines, which includes a full-line monitoring center, station monitoring units, network aggregation units, line-side sensing and data acquisition units, and user monitoring terminals; The full-line monitoring center is interconnected with several station monitoring units and user monitoring terminals set up along the railway line through several network aggregation units. The station monitoring units are configured at stations along the railway line and are connected to several line sensing and acquisition units distributed along the railway line with at least one unit between adjacent stations through several network aggregation units. The line sensing and acquisition units are equipped with at least two different types of sensors for sensing foreign object intrusion. And including the following steps: The station monitoring unit acquires in real time the foreign object intrusion detection signals collected by various types of sensors from the line-side sensing and acquisition unit via the network aggregation unit. And to modulate and preprocess the sensed signals and monitor and analyze the single-point abnormal state sensed by a single sensor; The station monitoring unit transmits the pre-processed sensing signals and single-point anomaly statuses to the full-line monitoring center via the network aggregation unit. This enables the entire monitoring center to respond and, based on deep learning algorithms, monitor and analyze the sensed signals and single-point anomalies, providing early warnings of foreign object intrusion risks along the railway line. And send real-time warnings of single-point abnormal status and foreign object intrusion risk to the user monitoring terminal according to preset emergency rules; The station monitoring unit is connected to at least one of the line-of-line sensing and acquisition units distributed between the current station and the previous and / or next station, and is used to monitor the abnormal status of the railway maintenance section corresponding to the current station. The sensing and data acquisition unit along the railway line marks location symbols in a certain order, and the location symbols include codes and latitude and longitude. The sensing and acquisition unit along the line is equipped with a camera, radar and fiber optic sensor; The cameras and / or radars are arranged at intervals along the railway line and facing the railway tracks and protective fences to collect radar data and video data. The fiber optic sensor is laid on the bottom of the encrypted mesh of the protective fence on both sides of the railway line to collect fiber optic vibration and stress data. Monitoring and analyzing a single-point abnormal state sensed by a single sensor includes the following steps: The station monitoring unit acquires radar data, video data, fiber optic vibration and stress data collected and transmitted by several sensing and acquisition units along the line connected to the current station where the station monitoring unit is located; After preprocessing the fiber optic vibration and stress data, the station monitoring unit judges the vibration intensity and stress intensity according to preset threshold rules: If the vibration intensity exceeds the first warning threshold and the duration exceeds the first time threshold and / or if the stress intensity exceeds the second warning threshold and the duration exceeds the second time threshold, then it is determined to be in a single-point abnormal state. When a single point of abnormality is in a state, the radar data and video data are retrieved and displayed in real time at the station monitoring unit, and the radar data, video data, fiber optic vibration and stress data, single point of abnormality and location symbol are associated with the location symbol and then transmitted to the full-line monitoring center. Monitoring and analyzing the risk of foreign object intrusion along railway lines includes the following steps: The station monitoring unit sends transmission data to the full-line monitoring center so that the full-line monitoring center can respond and perform target object detection and analysis based on deep learning algorithm to fuse the radar data and video data collected by the individual sensing and acquisition units along the line to determine whether there are foreign objects; if there are, it is determined as a foreign object intrusion risk warning. If not, the system monitors and analyzes the data transmitted by other sensing and acquisition units along the line connected to the current station within the same time period. If two or more sensing and acquisition units along the line are determined to be in a single-point abnormal state, it is determined to be a foreign object intrusion risk warning.

2. The intelligent monitoring method for railway lines based on cloud-edge collaboration according to claim 1, characterized in that, The preset emergency rules are as follows: If, within the same time period, only one of the several sensing and acquisition units connected to the current station is determined to be in a single-point abnormal state, the entire line monitoring center sends the single-point abnormal state with a matching location symbol and a preset first emergency maintenance plan to the user monitoring terminal in real time, so that maintenance workers can conduct fixed-point inspections of the railway line protective fence according to the location symbol. When several of the aforementioned sensing and acquisition units connected to the current station within the same time period determine that there is a risk warning of foreign object intrusion, the entire line monitoring center sends a foreign object intrusion risk warning with a matching location symbol and a preset second emergency maintenance plan to the user monitoring terminal in real time. This plan is used by maintenance workers to remove foreign objects from the railway line protective fence according to the location symbol.

3. A cloud-edge collaborative intelligent monitoring platform for railway lines, characterized in that: The method for intelligent monitoring of railway lines based on cloud-edge collaboration as described in claim 1 or 2 includes a full-line monitoring center, station monitoring units, network aggregation units, line-side sensing and data acquisition units, and user monitoring terminals. The line sensing and acquisition unit is equipped with at least two different types of sensors for sensing foreign object intrusion. The station monitoring unit is configured at railway stations along the line and is connected to several line sensing and acquisition units distributed along the railway line and configured at least one between adjacent stations through several network aggregation units. It is used to adjust and preprocess the sensing signals collected by each individual sensor in the line sensing and acquisition unit and monitor the single-point abnormal state corresponding to each individual sensor. The entire line monitoring center is interconnected with several station monitoring units and user monitoring terminals set up along the railway through several network aggregation units. It is used to monitor and analyze the sensing signals and individual foreign object intrusion status transmitted by each of the station monitoring units based on deep learning algorithms, and to issue early warnings of foreign object intrusion risks along the railway. It also sends single-point abnormal status and foreign object intrusion risk warnings to the user monitoring terminals in real time.

4. The intelligent monitoring platform for railway lines based on cloud-edge collaboration according to claim 3, characterized in that, The station monitoring unit is connected to at least one of the line-of-line sensing and acquisition units distributed between the current station and the previous and / or next station, and is used to monitor the abnormal status of the railway maintenance section corresponding to the current station. The sensing and data acquisition unit along the railway line marks location symbols in a certain order, and the location symbols include codes and latitude and longitude.

5. The intelligent monitoring platform for railway lines based on cloud-edge collaboration according to claim 4, characterized in that, The sensing and acquisition unit along the line is equipped with a camera, radar and fiber optic sensor; The cameras and / or radars are arranged at intervals along the railway line and facing the railway tracks and protective fences to collect radar data and video data. The fiber optic sensors are installed on the bottom of the protective fence along both sides of the railway line to collect fiber optic vibration and stress data.

6. A monitoring platform device, characterized in that, include: Interconnected memory and processor, The processor is used to execute program instructions stored in the memory to implement the intelligent monitoring method for railway lines based on cloud-edge collaboration as described in any one of claims 1 to 2.

Images