Rail transit operation safety event monitoring system and method based on Internet of Things

Abstract

The invention discloses a rail transit operation safety event monitoring system and method based on the Internet of Things, and relates to the technical field of rail transit operation safety event monitoring, and the system comprises a plurality of sensor units, a plurality of edge preliminary analysis modules, a plurality of ad hoc network modules, and a rail transit operation central safety analysis module. According to the invention, the side cloud is combined with a main framework to realize acquisition of sensors and visual data, a plurality of sensor units can realize real-time monitoring of different monitoring items, the visual monitor has a generalization monitoring capability, and visual identification of different abnormal features is realized based on a visual feature comparison algorithm; the edge preliminary analysis module is arranged, through preliminary screening of the edge preliminary analysis module, the data processing amount of the rail transit operation central safety analysis module can be greatly reduced, the computing power requirement and the hardware arrangement cost of a server are reduced, and finally the rail transit operation central safety analysis module only carries out safety evaluation on data with large risks. And the safety assessment efficiency is high.

Description

Technical Field

[0001] The present invention relates to the technical field of rail transit operation safety event monitoring, and in particular to a rail transit operation safety event monitoring system and method based on the Internet of Things. Background Art

[0002] The system architecture of the existing rail transit operation safety incident monitoring system includes sensor module, network communication module, rail transit operation safety assessment and analysis module and monitoring display module.

[0003] The sensor module usually includes various sensors (such as temperature, humidity, vibration, displacement, video, etc.) and communication modules (such as Wi-Fi, Bluetooth, ZigBee, etc.). The network communication module is responsible for transmitting the data collected by the sensor module to the rail transit operation safety assessment and analysis module deployed in the cloud. The rail transit operation safety assessment and analysis module analyzes and processes the collected data, and finally displays it through the monitoring display module to realize safety event monitoring.

[0004] The above technical solutions have the following defects: sensor data island phenomenon and lack of effective data fusion. The monitoring system has a slow response speed. A large amount of sensor data, especially visual data, is collected in real time to the operation safety assessment and analysis module, resulting in huge computing power requirements for the operation safety assessment and analysis module, and it is unable to process a large amount of data in real time. The hardware equipment has high energy consumption and is centralized, with high maintenance costs and high failure rates.

[0005] Therefore, how to design a rail transit operation safety incident monitoring system based on the Internet of Things that can efficiently monitor rail transit operation safety incidents, require relatively low hardware equipment computing power, equalize energy consumption, and reduce failure rate is a technical problem that needs to be solved urgently. Summary of the invention

[0006] In order to solve the above technical problems, the present invention provides a rail transit operation safety event monitoring system and method based on the Internet of Things. The following technical solutions are adopted:

[0007] The rail transit operation safety event monitoring system based on the Internet of Things includes multiple sensor units, multiple edge preliminary analysis modules, multiple self-organizing network modules and a rail transit operation central safety analysis module. The multiple sensor units divide the rail transit operation safety area into multiple regional grids based on rail transit safety operation related parameters in the sensor monitoring area. The multiple edge preliminary analysis modules and the multiple self-organizing network modules are respectively arranged at the physical center of the regional grid. The sensor unit is communicated with the edge preliminary analysis module of the regional grid. The edge preliminary analysis module performs preliminary analysis and screening on the rail transit safety operation related parameters of the sensor unit based on an abnormal threshold, caches the rail transit safety operation related parameters below the abnormal threshold for a set time, packages any rail transit safety operation related parameters that are higher than or equal to the abnormal threshold and the threshold exceeding result, and forwards them to the rail transit operation central safety analysis module through the self-organizing network module at the physical center of the same regional grid. The rail transit operation central safety analysis module sorts the abnormal events in the rail transit operation safety area based on the threshold exceeding result, performs safety assessment and analysis on the abnormal events based on the safety event assessment algorithm of data fusion according to the sorting result, and outputs the assessment result.

[0008] Optionally, the sensor unit includes a temperature sensor, a vibration sensor, a displacement sensor and a visual monitor, and within the grid area, the temperature sensor, the vibration sensor, the displacement sensor and the visual monitor are respectively connected to the edge preliminary analysis module at the physical center of the grid area by direct connection via communication lines.

[0009] By adopting the above technical solution, the edge cloud is combined with the main architecture to realize the collection of sensor and visual data. Multiple sensor units can realize real-time monitoring of different monitoring items, such as the temperature value in the distribution box of important rail transit facilities, the vibration value and displacement value of rail transit tracks, etc. The visual monitor has a generalized monitoring capability and can realize visual recognition of different abnormal features based on the visual feature comparison algorithm. Although it does not have the monitoring accuracy of sensors, it can detect some moving pedestrians or animals entering places where pedestrians are prohibited from entering, water accumulation, whether the indicator lights are normal, etc.

[0010] These monitoring items are forwarded to the central safety analysis module of rail transit operation through the self-organizing network module at the physical center of the same regional grid. The purpose of dividing the monitored area into regional grids is to maintain a stable wireless communication effect within the physical range. The monitored area of ​​rail transit operation is generally more complex. If the entire communication network is directly connected by communication lines, the complexity and stability of its wiring are poor. Therefore, a short-line direct connection within the regional grid is adopted, and then a regional grid and a self-organizing network module are used for unified forwarding. This networking form can more stably and reliably transmit the edge sensor data to the central safety analysis module of rail transit operation efficiently;

[0011] Before the sensor data is transmitted, an edge preliminary analysis module is set up in units of regional grids, which can perform preliminary screening based on abnormal thresholds. In actual operation, the abnormal proportion of monitoring data is not large in most cases. Therefore, the preliminary screening of the edge preliminary analysis module will greatly reduce the data processing volume of the rail transit operation central safety analysis module, reduce the computing power requirements of the server and the hardware layout cost. Finally, the rail transit operation central safety analysis module only performs security assessments on data with higher risks. The security assessment is efficient, and the monitoring energy consumption is averaged to the grid area, reducing the impact of small area failures on the stability of the entire monitoring system.

[0012] Optionally, the edge preliminary analysis module includes a first cache, a data analysis chip, a visual analysis chip, a second cache and a memory, the first cache is respectively communicated with the temperature sensor, vibration sensor, displacement sensor and visual monitor of the sensor unit within the grid area, the data analysis chip and the visual analysis chip are respectively communicated with the first cache and the memory, and the data stored in the first cache is compared with the abnormal threshold based on the time series, and any rail transit safety operation related parameters and threshold excess results that are higher than or equal to the abnormal threshold are packaged and stored in the second cache, and the second cache is communicated with the data input end of the self-organizing network module at the physical center of the grid in the same area.

[0013] By adopting the above technical solution, the data received by the edge preliminary analysis module is first stored in the first buffer, and the data analysis chip and the visual analysis chip can be synchronously called for analysis. The data analysis chip mainly performs analysis based on the set abnormal threshold, so the computing power requirement is relatively low. The analysis of the visual analysis chip is mainly based on the abnormal feature database stored in the memory for similarity comparison. The precision and accuracy of the comparison here are not very high, and it can also be implemented with a low-cost visual analysis chip, thereby achieving the goal of significantly reducing the maximum computing power of the entire rail transit operation safety event monitoring system, reducing the high hardware cost brought by high-computing power servers, and evenly allocating the concentrated high power consumption of high-computing power servers.

[0014] Optionally, multiple self-organizing network modules realize self-organizing network based on LoRaWAN technology, and multiple self-organizing network modules are wirelessly networked with the rail transit operation central safety analysis module respectively. When the self-organizing network module cannot directly communicate wirelessly with the rail transit operation central safety analysis module, it is forwarded through the adjacent self-organizing network module.

[0015] By adopting the above technical solutions, wireless ad hoc networks within the long-distance grid area are realized based on LoRaWAN. Nodes with LoRaWAN compatible modules are selected, with sufficient processing power, storage and power. The node software is configured to support the LoRaWAN protocol and ad hoc networking functions.

[0016] By utilizing the gateway relay function of LoRaWAN, when data cannot be sent wirelessly directly to the central server, multi-hop communication between nodes is allowed through the gateway.

[0017] Configure the nodes to support self-organizing network capabilities, enabling them to automatically discover neighbor nodes.

[0018] Optionally, the rail transit operation central safety analysis module includes a central self-organizing network module, a computer server and a rail transit monitoring screen. The computer server is wirelessly connected to multiple self-organizing network modules through the central self-organizing network module. The computer server performs safety assessment and analysis on abnormal events based on a security event assessment algorithm based on data fusion according to the sorting results, and outputs the assessment results through the rail transit monitoring screen.

[0019] By adopting the above technical solution, a backup module can be set up in the central self-organizing network module in specific applications to avoid failures that cause data to be unable to be received. The computer server receives abnormal data packets from the central self-organizing network module and performs security assessment and analysis on abnormal events based on the security event assessment algorithm of data fusion, and finally outputs the assessment results through the rail transit monitoring large screen to realize intelligent security event assessment.

[0020] Optionally, the computer server includes a first computer, a second computer and a third computer. The first computer is communicatively connected to the data output end of the central self-organizing network module, and parses abnormal data packets sent by multiple self-organizing network modules in a time sequence, analyzes whether the abnormal data packets meet the emergency safety event standards based on set emergency safety event standards, sends abnormal data packets that meet the emergency safety event standards to the third computer, and sends abnormal data packets that do not meet the emergency safety event standards to the second computer in a time sequence. The second computer and the third computer respectively communicate with the rail transit monitoring large screen to interact with the evaluation results.

[0021] By adopting the above technical solution, three ordinary industrial control computers can be used to achieve relatively reliable and stable abnormal data packet processing. The first computer parses the abnormal data packets sent by multiple self-organizing network modules in a time sequence, and analyzes whether the abnormal data packets meet the emergency security event standards based on the set emergency security event standards. Those that meet the emergency security event standards can be sent to the third computer for immediate processing to avoid affecting the response time of the emergency event due to the processing time. The second computer is responsible for performing security assessments in a time sequence. The second computer and the third computer communicate and interact with the rail transit monitoring large screen to evaluate the results, and can also set sound and light alarms or network alarms to alarm security events.

[0022] The rail transit operation safety incident monitoring method adopts the rail transit operation safety incident monitoring system based on the Internet of Things to monitor and analyze the rail transit operation safety incidents within the regional scope, including the following steps:

[0023] Step 1: Mark the multiple regional grids of the rail transit operation safety area as A1, A2, ..., A n , where n is the number of regional grids, regional grid A n The monitoring temperature sensors, vibration sensors, displacement sensors and visual monitors in the area send signals to the grid A every two seconds. n The first buffer of the edge preliminary analysis module arranged at the physical center transmits temperature values, vibration values, displacement values ​​and visual images;

[0024] Step 2: The data analysis chip calls the temperature value, vibration value, and displacement value every two seconds, and determines whether they are abnormal based on the temperature abnormality threshold, vibration abnormality threshold, and displacement abnormality threshold;

[0025] The visual analysis chip matches the visual image with abnormal features based on the abnormal feature database stored in the memory. If there is an abnormal feature that matches successfully, it is determined that an abnormality exists;

[0026] If there is no abnormality in the data uploaded at the same time, the data uploaded at the same time will be packaged and cached for one week; otherwise, the data uploaded at the same time will be packaged, and the upload time and sensor number will be marked in the data packet header to form an abnormal data packet, which will be forwarded to the rail transit operation central safety analysis module through the self-organizing network module at the physical center of the grid in the same area;

[0027] Step 3, the first computer parses the abnormal data packets sent by the multiple ad hoc network modules in a time sequence, analyzes whether the abnormal data packets meet the emergency security event standards based on the set emergency security event standards, and directly sends the abnormal data packets that meet the emergency security event standards to the third computer for security event assessment;

[0028] Otherwise, the abnormal data packet is forwarded to the second computer according to the time sequence, and the second computer performs a security event assessment on the abnormal data packet based on the time sequence;

[0029] Step 4: The second computer and the third computer respectively send the security event assessment results to the rail transit monitoring screen, and the rail transit monitoring screen displays the security event assessment results.

[0030] Optionally, in step 3, the emergency safety event standard is: whether there is at least one monitoring value exceeding the abnormal threshold and whether there is at least one abnormal feature match successfully.

[0031] By adopting the above technical solution, generally speaking, for a single monitoring value exceeding the abnormal threshold, the urgency it brings is usually not high, and it can be evaluated through the normal safety assessment process. If the abnormal threshold exceeds the standard and the abnormal feature is successfully matched, the probability of a safety incident will be greatly increased. For example, if the temperature of the distribution cabinet exceeds the standard and smoke is detected by visual monitoring, then there is a high probability that the distribution cabinet has a circuit breaker and a fire, so emergency processing is required. Of course, in specific applications, the emergency safety event standard can be flexibly set according to actual conditions, such as higher abnormal thresholds.

[0032] Optionally, in step 3, the security incident assessment includes the following steps:

[0033] The monitoring data provided by each sensor is X m , the abnormal feature items whose matching similarity of abnormal features is greater than the similarity threshold are treated as monitoring items and are uniformly processed. The labels are arranged after the last sensor monitoring value. The security event evaluation value is Y, and the total number of abnormal features whose sensor monitoring values ​​and matching similarity are greater than the similarity threshold is recorded as m;

[0034] Y=X1·W1·α1+X2·W2·α2+…+X m ·W m α m ;

[0035] Where W m is the weight of the mth monitoring value, α m is the monitoring item setting weight of the mth monitoring value;

[0036]

[0037] By adopting the above technical solution, IF(W m <0,0,W m ) is to reduce W less than 0 m The weight is set to 0. If the monitoring value does not exceed the threshold, it is considered to have no impact on the assessment of the security incident.

[0038] In summary, the present invention includes at least one of the following beneficial technical effects:

[0039] The present invention can provide a rail transit operation safety event monitoring system and method based on the Internet of Things, which adopts the edge-cloud combined main architecture to realize the collection of sensor and visual data. Multiple sensor units can realize real-time monitoring of different monitoring items. The visual monitor has a generalized monitoring capability and can realize visual recognition of different abnormal features based on the visual feature comparison algorithm.

[0040] These monitoring items are forwarded to the rail transit operation central safety analysis module through the self-organizing network module at the physical center of the same regional grid. The monitored area is divided into regional grids to maintain a stable wireless communication effect within the physical range, which can more stably and reliably transmit the edge sensor data to the rail transit operation central safety analysis module efficiently;

[0041] Before the sensor data is transmitted, an edge preliminary analysis module is set up based on the regional grid, which can perform preliminary screening based on the abnormal threshold. The preliminary screening of the edge preliminary analysis module can greatly reduce the data processing volume of the rail transit operation central safety analysis module, reduce the computing power requirements of the server and the hardware layout cost. Finally, the rail transit operation central safety analysis module only performs security assessment on data with higher risks. The security assessment is efficient, and the monitoring energy consumption is averaged to the grid area, reducing the impact of small area failures on the stability of the entire monitoring system. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a schematic diagram of the connection principle of electrical components of the rail transit operation safety event monitoring system based on the Internet of Things of the present invention;

[0043] Figure 2 It is a schematic diagram of the hardware connection principle within a regional grid of the rail transit operation safety event monitoring system based on the Internet of Things of the present invention.

[0044] Explanation of the accompanying drawings: 11. Temperature sensor; 12. Vibration sensor; 13. Displacement sensor; 14. Visual monitor; 2. Edge preliminary analysis module; 21. First cache; 22. Data analysis chip; 23. Visual analysis chip; 24. Second cache; 25. Memory; 3. Self-organizing network module; 4. Central safety analysis module for rail transit operation; 41. Central self-organizing network module; 42. Computer server; 421. First computer; 422. Second computer; 423. Third computer; 43. Rail transit monitoring screen. DETAILED DESCRIPTION

[0045] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0046] The embodiments of the present invention disclose a rail transit operation safety event monitoring system and method based on the Internet of Things.

[0047] Reference Figure 1 and Figure 2 Embodiment 1, a rail transit operation safety event monitoring system based on the Internet of Things, including multiple sensor units, multiple edge preliminary analysis modules 2, multiple self-organizing network modules 3 and a rail transit operation central safety analysis module 4, multiple sensor units respectively divide the rail transit operation safety area into multiple regional grids based on the rail transit safety operation associated parameters in the sensor monitoring area, multiple edge preliminary analysis modules 2 and multiple self-organizing network modules 3 are respectively arranged at the physical center of the regional grid, the sensor unit is communicated with the edge preliminary analysis module 2 of the regional grid, the edge preliminary analysis module 2 performs preliminary analysis and screening on the rail transit safety operation associated parameters of the sensor unit based on the abnormal threshold, caches the rail transit safety operation associated parameters below the abnormal threshold for a set time, packages any rail transit safety operation associated parameters that are higher than or equal to the abnormal threshold and the threshold exceeding result, and forwards them to the rail transit operation central safety analysis module 4 through the self-organizing network module 3 at the physical center of the same regional grid, the rail transit operation central safety analysis module 4 sorts the abnormal events in the rail transit operation safety area based on the threshold exceeding result, and performs safety assessment and analysis on the abnormal events based on the safety event assessment algorithm of data fusion according to the sorting result, and outputs the assessment result.

[0048] In embodiment 2, the sensor unit includes a temperature sensor 11, a vibration sensor 12, a displacement sensor 13 and a visual monitor 14, and within the grid area, the temperature sensor 11, the vibration sensor 12, the displacement sensor 13 and the visual monitor 14 are respectively connected to the edge preliminary analysis module 2 at the physical center of the grid area by direct connection via communication lines.

[0049] The edge cloud is combined with the main architecture to realize the collection of sensor and visual data. Multiple sensor units can realize real-time monitoring of different monitoring items, such as the temperature value in the distribution box of important rail transit facilities, the vibration value and displacement value of rail transit tracks, etc. The visual monitor 14 has a generalized monitoring capability and can realize visual recognition of different abnormal features based on the visual feature comparison algorithm. Although it does not have the monitoring accuracy of sensors, it can detect some moving pedestrians or animals entering places where pedestrians are prohibited from entering, water accumulation, whether the indicator lights are normal, etc.;

[0050] These monitoring items are forwarded to the rail transit operation central safety analysis module 4 through the self-organizing network module 3 at the physical center of the same regional grid. The purpose of dividing the monitored area into regional grids is to maintain a stable wireless communication effect within the physical range. The monitored area of ​​rail transit operation is generally more complex. If the entire communication network is directly connected by communication lines, the complexity and stability of its wiring are poor. Therefore, a short-line direct connection within the regional grid is adopted, and then a regional grid and a self-organizing network module 3 are used for unified forwarding. The networking form can more stably and reliably transmit the edge sensor data to the rail transit operation central safety analysis module 4 efficiently;

[0051] Before the sensor data is transmitted, an edge preliminary analysis module 2 is set up in units of regional grids, which can perform preliminary screening based on abnormal thresholds. In actual operation, the abnormal proportion of monitoring data is not large in most cases. Therefore, through the preliminary screening of the edge preliminary analysis module 2, the data processing volume of the rail transit operation central safety analysis module 4 will be greatly reduced, and the computing power requirements and hardware layout costs of the server will be reduced. Finally, the rail transit operation central safety analysis module 4 only performs security assessments on data with higher risks. The security assessment efficiency is high, and the monitoring energy consumption is averaged to the grid area, reducing the impact of small area failures on the stability of the entire monitoring system.

[0052] Embodiment 3, the edge preliminary analysis module 2 includes a first buffer 21, a data analysis chip 22, a visual analysis chip 23, a second buffer 24 and a memory 25, the first buffer 21 is respectively communicated with the temperature sensor 11, the vibration sensor 12, the displacement sensor 13 and the visual monitor 14 of the sensor unit within the grid area, the data analysis chip 22 and the visual analysis chip 23 are respectively communicated with the first buffer 21 and the memory 25, the data stored in the first buffer 21 is compared with the abnormal threshold based on the time series, and any rail transit safety operation-related parameters and threshold exceeding results that are higher than or equal to the abnormal threshold are packaged and stored in the second buffer 24, and the second buffer 24 is communicated with the data input end of the self-organizing network module 3 at the physical center of the grid in the same area.

[0053] The data received by the edge preliminary analysis module 2 is first stored in the first buffer 21, and the data analysis chip 22 and the visual analysis chip 23 can be synchronously called for analysis. The data analysis chip 22 mainly performs analysis based on the set abnormal threshold, so the computing power requirement is relatively low. The analysis of the visual analysis chip 23 is mainly based on the abnormal feature database stored in the memory 25 for similarity comparison. The precision and accuracy of the comparison here are not very high, and the low-cost visual analysis chip 23 can also be used to implement it, thereby achieving the goal of significantly reducing the maximum computing power of the entire rail transit operation safety event monitoring system, reducing the high hardware cost brought by high-computing power servers, and evenly allocating the concentrated high power consumption of high-computing power servers.

[0054] In embodiment 4, multiple self-organizing network modules 3 realize self-organizing network based on LoRaWAN technology. Multiple self-organizing network modules 3 are wirelessly networked with the central safety analysis module 4 of rail transit operation respectively. When the self-organizing network module 3 cannot directly communicate with the central safety analysis module 4 of rail transit operation wirelessly, it is forwarded through the adjacent self-organizing network module 3.

[0055] To implement wireless ad hoc networking in a long-range grid area based on LoRaWAN, select nodes with LoRaWAN-compatible modules with sufficient processing power, storage, and power. Configure the node software to support the LoRaWAN protocol and ad hoc networking functions.

[0056] By utilizing the gateway relay function of LoRaWAN, when data cannot be sent wirelessly directly to the central server, multi-hop communication between nodes is allowed through the gateway.

[0057] Configure the nodes to support self-organizing network capabilities, enabling them to automatically discover neighbor nodes.

[0058] Embodiment 5, the central safety analysis module 4 for rail transit operation includes a central self-organizing network module 41, a computer server 42 and a rail transit monitoring screen 43. The computer server 42 is wirelessly connected to a plurality of self-organizing network modules 3 through the central self-organizing network module 41. The computer server 42 performs safety assessment and analysis on abnormal events based on a safety event assessment algorithm fused with data according to the sorting results, and outputs the assessment results through the rail transit monitoring screen 43.

[0059] The central self-organizing network module 41 can also be equipped with a backup module in specific applications to avoid failures that cause data to fail to be received. The computer server 42 receives abnormal data packets from the central self-organizing network module 41, performs security assessment analysis on abnormal events based on the security event assessment algorithm of data fusion, and finally outputs the assessment results through the rail transit monitoring screen 43 to realize intelligent security event assessment.

[0060] Embodiment 6, the computer server 42 includes a first computer 421, a second computer 422 and a third computer 423, the first computer 421 is communicatively connected with the data output terminal of the central self-organizing network module 41, and analyzes the abnormal data packets sent by multiple self-organizing network modules 3 in a time sequence, analyzes whether the abnormal data packets meet the emergency safety event standards based on the set emergency safety event standards, sends the abnormal data packets that meet the emergency safety event standards to the third computer 423, and sends the abnormal data packets that do not meet the emergency safety event standards to the second computer 422 in a time sequence, and the second computer 422 and the third computer 423 communicate and interact with the rail transit monitoring screen 43 to evaluate the results.

[0061] The use of three ordinary industrial control computers can achieve relatively reliable and stable abnormal data packet processing. The first computer 421 parses the abnormal data packets sent by multiple self-organizing network modules 3 in a time sequence, and analyzes whether the abnormal data packets meet the emergency safety event standards based on the set emergency safety event standards. Those that meet the emergency safety event standards can be sent to the third computer 423 for immediate processing to avoid affecting the response time of the emergency event due to the processing time. The second computer 422 is responsible for performing security assessments in a time sequence. The second computer 422 and the third computer 423 communicate and interact with the rail transit monitoring screen 43 to obtain assessment results, and can also set up sound and light alarms or network alarms to alarm security events.

[0062] Embodiment 7, a rail transit operation safety event monitoring method, using a rail transit operation safety event monitoring system based on the Internet of Things to monitor and analyze rail transit operation safety events within a region, includes the following steps:

[0063] Step 1: Mark the multiple regional grids of the rail transit operation safety area as A1, A2, ..., A n , where n is the number of regional grids, regional grid A n The monitoring temperature sensor 11, vibration sensor 12, displacement sensor 13 and visual monitor 14 in the area grid A are sent to the area grid A every two seconds. n The first buffer 21 of the edge preliminary analysis module 2 arranged at the physical center transmits temperature values, vibration values, displacement values ​​and visual images;

[0064] Step 2, the data analysis chip 22 calls the temperature value, the vibration value, and the displacement value every two seconds, and determines whether they are abnormal based on the temperature abnormality threshold, the vibration abnormality threshold, and the displacement abnormality threshold;

[0065] The visual analysis chip 23 performs abnormal feature matching on the visual image based on the abnormal feature database stored in the memory 25. If there is an abnormal feature that matches successfully, it is determined that an abnormality exists;

[0066] If there is no abnormality in the data uploaded at the same time, the data uploaded at the same time will be packaged and cached for one week; otherwise, the data uploaded at the same time will be packaged, and the upload time and sensor number will be marked in the data packet header to form an abnormal data packet, which will be forwarded to the rail transit operation central safety analysis module 4 through the self-organizing network module 3 at the physical center of the grid in the same area;

[0067] Step 3, the first computer 421 analyzes the abnormal data packets sent by the multiple ad hoc network modules 3 in time sequence, analyzes whether the abnormal data packets meet the emergency security event standards based on the set emergency security event standards, and directly sends the abnormal data packets that meet the emergency security event standards to the third computer 423 for security event assessment;

[0068] Otherwise, the abnormal data packet is forwarded to the second computer 422 according to the time sequence, and the second computer 422 performs a security event assessment on the abnormal data packet based on the time sequence;

[0069] Step 4: The second computer 422 and the third computer 423 respectively send the security event assessment results to the rail transit monitoring screen 43, and the rail transit monitoring screen 43 displays the security event assessment results.

[0070] In Example 8, in step 3, the emergency safety event standard is: whether there is at least one monitoring value exceeding the abnormal threshold and whether there is at least one abnormal feature match successfully.

[0071] Generally speaking, if a single monitoring value exceeds the abnormal threshold, the urgency it brings is usually not high and can be evaluated through the normal safety assessment process. If the abnormal threshold exceeds the standard and the abnormal feature is successfully matched, the probability of a safety incident will increase significantly. For example, if the temperature of the distribution cabinet exceeds the standard and smoke is detected by visual monitoring, there is a high probability that the distribution cabinet has a circuit breaker and a fire has occurred, so emergency handling is required. Of course, in specific applications, the emergency safety event standard can be flexibly set according to actual conditions, such as a higher abnormal threshold.

[0072] In Example 9, in step 3, the security incident assessment includes the following steps:

[0073] The monitoring data provided by each sensor is X m , the abnormal feature items whose matching similarity of abnormal features is greater than the similarity threshold are treated as monitoring items and are uniformly processed. The labels are arranged after the last sensor monitoring value. The security event evaluation value is Y, and the total number of abnormal features whose sensor monitoring values ​​and matching similarity are greater than the similarity threshold is recorded as m;

[0074] Y=X1·W1·α1+X2·W2·α2+…+X m ·W m αm ;

[0075] Where W m is the weight of the mth monitoring value, α m is the monitoring item setting weight of the mth monitoring value;

[0076]

[0077] IF(W m <0,0,W m ) is to reduce W less than 0 m The weight is set to 0. If the monitoring value does not exceed the threshold, it is considered to have no impact on the assessment of the security incident.

[0078] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. The rail transit operation safety event monitoring system based on the Internet of Things is characterized by: The invention comprises a plurality of sensor units, a plurality of edge preliminary analysis modules (2), a plurality of self-organizing network modules (3) and a rail transit operation central safety analysis module (4); the plurality of sensor units respectively divide the rail transit operation safety area into a plurality of regional grids based on rail transit safety operation related parameters in the sensor monitoring area; the plurality of edge preliminary analysis modules (2) and the plurality of self-organizing network modules (3) are respectively arranged at the physical center position of the regional grid; the sensor units are connected in communication with the edge preliminary analysis modules (2) of the regional grids; the edge preliminary analysis modules (2) analyze the rail transit safety operation related parameters of the sensor units based on abnormal thresholds; A preliminary analysis and screening is performed, rail transit safety operation related parameters that are below the abnormal threshold are cached for a set time, any rail transit safety operation related parameters that are higher than or equal to the abnormal threshold and the threshold exceeding result are packaged, and forwarded to the rail transit operation central safety analysis module (4) through the self-organizing network module (3) at the physical center of the grid in the same area. The rail transit operation central safety analysis module (4) sorts the abnormal events in the rail transit operation safety area based on the threshold exceeding result, and performs safety assessment analysis on the abnormal events based on the safety event assessment algorithm of data fusion according to the sorting result, and outputs the assessment result.

2. The rail transit operation safety event monitoring system based on the Internet of Things according to claim 1 is characterized in that: The sensor unit comprises a temperature sensor (11), a vibration sensor (12), a displacement sensor (13) and a visual monitor (14), and within the range of a grid area, the temperature sensor (11), the vibration sensor (12), the displacement sensor (13) and the visual monitor (14) are respectively connected to the edge preliminary analysis module (2) at the physical center of the grid area by direct connection via communication lines.

3. The rail transit operation safety event monitoring system based on the Internet of Things according to claim 2 is characterized in that: The edge preliminary analysis module (2) comprises a first buffer (21), a data analysis chip (22), a visual analysis chip (23), a second buffer (24) and a memory (25); the first buffer (21) is respectively connected to the temperature sensor (11), the vibration sensor (12), the displacement sensor (13) and the visual monitor (14) of the sensor unit within the grid area; the data analysis chip (22) and the visual analysis chip (23) are respectively connected to the first buffer (21) and the memory (25); an abnormal threshold comparison is performed on the data stored in the first buffer (21) based on the time series; and any rail transit safety operation related parameter and threshold exceeding result that is higher than or equal to the abnormal threshold is packaged and stored in the second buffer (24); the second buffer (24) is connected to the data input end of the self-organizing network module (3) at the physical center of the grid in the same area.

4. The rail transit operation safety event monitoring system based on the Internet of Things according to claim 3 is characterized in that: A plurality of self-organizing network modules (3) realize self-organizing network based on LoRaWAN technology, and the plurality of self-organizing network modules (3) are wirelessly networked with the rail transit operation central safety analysis module (4) respectively, and when the self-organizing network module (3) cannot directly communicate with the rail transit operation central safety analysis module (4) wirelessly, the communication is forwarded through an adjacent self-organizing network module (3).

5. The rail transit operation safety event monitoring system based on the Internet of Things according to claim 4 is characterized in that: The rail transit operation central safety analysis module (4) comprises a central self-organizing network module (41), a computer server (42) and a rail transit monitoring screen (43); the computer server (42) is wirelessly connected to a plurality of self-organizing network modules (3) through the central self-organizing network module (41); the computer server (42) performs safety assessment analysis on abnormal events based on a safety event assessment algorithm of data fusion according to the sorting results, and outputs the assessment results through the rail transit monitoring screen (43).

6. The rail transit operation safety event monitoring system based on the Internet of Things according to claim 5 is characterized in that: The computer server (42) comprises a first computer (421), a second computer (422) and a third computer (423), wherein the first computer (421) is communicatively connected to a data output terminal of a central self-organizing network module (41), and analyzes abnormal data packets sent by a plurality of self-organizing network modules (3) in a time sequence, analyzes whether the abnormal data packets meet the emergency safety event standard based on a set emergency safety event standard, sends abnormal data packets that meet the emergency safety event standard to the third computer (423), and sends abnormal data packets that do not meet the emergency safety event standard to the second computer (422) in a time sequence, and the second computer (422) and the third computer (423) communicate and interact with a large rail transit monitoring screen (43) to evaluate the results.

7. A method for monitoring rail transit operation safety events, characterized in that: The rail transit operation safety incident monitoring system based on the Internet of Things as described in claim 6 is used to monitor and analyze rail transit operation safety incidents within a region, comprising the following steps: Step 1: Mark the multiple regional grids of the rail transit operation safety area as A1, A2, ..., A n , where n is the number of regional grids, regional grid A n The monitoring temperature sensor (11), vibration sensor (12), displacement sensor (13) and visual monitor (14) in the area grid A send a signal to the area grid A every two seconds. n The first buffer (21) of the edge preliminary analysis module (2) arranged at the physical center transmits temperature values, vibration values, displacement values ​​and visual images; Step 2, the data analysis chip (22) calls the temperature value, the vibration value, and the displacement value every two seconds, and determines whether there is an abnormality based on the temperature abnormality threshold, the vibration abnormality threshold, and the displacement abnormality threshold; The visual analysis chip (23) performs abnormal feature matching on the visual image based on the abnormal feature database stored in the memory (25), and if there is an abnormal feature that is successfully matched, it is determined that an abnormality exists; If there is no abnormality in the data uploaded at the same time, the data uploaded at the same time are packaged and cached for a period of one week; otherwise, the data uploaded at the same time are packaged, and the upload time and sensor number are marked in the data packet header to form an abnormal data packet, which is forwarded to the rail transit operation central safety analysis module (4) through the self-organizing network module (3) at the physical center of the grid in the same area; Step 3, the first computer (421) analyzes the abnormal data packets sent by the plurality of ad hoc network modules (3) in a time sequence, analyzes whether the abnormal data packets meet the emergency security event standards based on the set emergency security event standards, and directly sends the abnormal data packets meeting the emergency security event standards to the third computer (423) for security event assessment; Otherwise, the abnormal data packet is forwarded to the second computer (422) according to the time sequence, and the second computer (422) performs a security event assessment on the abnormal data packet based on the time sequence; Step 4: The second computer (422) and the third computer (423) respectively send the security incident assessment results to the rail transit monitoring screen (43), and the rail transit monitoring screen (43) displays the security incident assessment results.

8. The rail transit operation safety event monitoring method according to claim 7, characterized in that: In step 3, the emergency safety event standard is: whether there is at least one monitoring value exceeding the abnormal threshold and at least one abnormal feature matching is successful.

9. The rail transit operation safety event monitoring method according to claim 7, characterized in that: In step 3, the security incident assessment includes the following steps: The monitoring data provided by each sensor is X m , the abnormal feature items whose matching similarity of abnormal features is greater than the similarity threshold are treated as monitoring items and are uniformly processed. The labels are arranged after the last sensor monitoring value. The security event evaluation value is Y, and the total number of abnormal features whose sensor monitoring values ​​and matching similarity are greater than the similarity threshold is recorded as m; Y=X1·W1·a1+X2·W2·a2+…+X m ·W m ·a m ; Where W m is the weight of the mth monitoring value, α m is the monitoring item setting weight of the mth monitoring value; IF(W m <0,0,W m )。

Images