A traffic tunnel inspection system based on a robot dog platform based on AI technology

Abstract

The present invention discloses a traffic tunnel inspection system using a robot dog platform based on AI technology, which relates to the field of data analysis technology. It solves the technical problems of being difficult to adapt to long cross-sea channels, difficult to ensure timely analysis and processing of data, and difficult to predict potential failures or degradations. The edge data management node can receive and process data from the robot dog in real time, reduce data transmission delays, and improve response speed. By processing and analyzing data at the edge, the data processing pressure on the cloud is reduced, making the entire system more efficient. The edge data management node can perform security analysis based on the data collected by the robot dog, and timely discover and warn of potential security risks. The security level analysis model can intuitively understand the security status of each grid area, which helps managers quickly identify potential risk areas and thus rationally allocate resources and energy.

Description

Technical Field

[0001] The present invention belongs to the field of data analysis, and specifically is a traffic tunnel inspection system using a robot dog platform based on AI technology. Background Art

[0002] With the acceleration of urbanization and the increasing complexity of transportation networks, the safety and maintenance of transportation corridors, such as the inspection of tunnels, bridges, and roads, has become increasingly important. Traditional inspection methods rely heavily on manual labor, which is subject to inefficiency, error-proneness, and data processing lags. Furthermore, the complex environments and confined spaces of many transportation corridors make manual inspections challenging, such as the inability to promptly identify potential safety hazards and maintenance needs. Efficient inspections are particularly difficult for cross-sea corridors with complex regional structures, such as the Shenzhen-Zhongshan Link.

[0003] The general robot dog platform channel inspection system analyzes and calculates data on a unified server, which makes it difficult to adapt to long cross-sea channels and ensure timely analysis and processing of data. At the same time, there is a lack of targeted analysis of various areas of the channel, making it difficult to predict potential failures or degradations. Summary of the Invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a traffic tunnel inspection system using a robot dog platform based on AI technology, which is used to solve the technical problems of being difficult to adapt to long cross-sea channels, difficult to ensure timely analysis and processing of data, and difficult to predict potential failures or degradations.

[0005] To solve the above problems, a first aspect of the present invention provides a traffic tunnel inspection system based on a robot dog platform of AI technology, comprising: an edge terminal and a cloud management center;

[0006] The edge end includes:

[0007] Area allocation module: This module obtains channel data for the cross-sea channel area and divides the target area into wind tower channel area, pipe gallery channel area, and external channel area. It then deploys robot dogs that comply with different channel management rules in different areas and divides different areas into grid areas for different detection types.

[0008] Data collection module: Set different data collection strategies for different grids. The robot dogs in each grid area collect data according to the data collection strategies of different grid areas and send the data to the edge data management node;

[0009] Edge data management nodes: Edge data management nodes are set up at every safe communication distance in the cross-sea channel area. These edge data management nodes perform scheduling control by communicating with robot dogs within the safe communication distance. Based on the data collected by the robot dogs in the grid area, they conduct security analysis on different types of grids and issue security alerts for grids whose security analysis falls below the threshold.

[0010] The cloud management center includes:

[0011] Model building module: Based on the security analysis results of grids in different areas, a security analysis model for different grid areas is established, and the security level analysis of each grid area is performed;

[0012] Robot dog scheduling module: Based on the fault degree analysis results of the grid area and the detection progress of the grid in the grid area, support scheduling of robot dogs in different grid areas is carried out through the edge data management node.

[0013] Optionally, in an example of the above aspect, the area allocation module includes:

[0014] Area division unit: This unit obtains channel data of the cross-sea channel area, including channel detection area data, channel accident records, channel type, and channel map data. Based on the channel detection area data and channel map data, it divides the channel connected to the wind tower into the wind tower channel area, the channel accommodating various pipelines into the pipeline corridor channel area, and the channel accommodating the pedestrian and vehicle traffic area into the external channel area.

[0015] Robot dog allocation unit: According to the detection area area in the channel detection area data of the wind tower channel area, the pipe gallery channel area and the external channel area, robot dogs that comply with the channel management rules of different areas are set in different areas;

[0016] Grid division unit: According to the channel data of each area, the wind tower channel area is divided into equipment grid area, channel grid area and wind tower monitoring grid area; the corridor channel area is divided into channel grid area, equipment grid area and pipe network detection grid area; the external channel area is divided into channel grid area and equipment grid area.

[0017] Optionally, in an example of the above aspect, the data collection module sets different data collection strategies for different grids, including:

[0018] For all grid areas in the wind tower access area, pipe gallery access area, and external access area, basic operating data is tested, including wind speed, temperature, humidity, high-definition image data and infrared image data of the access area;

[0019] For the equipment grid areas in the wind tower access area, pipe gallery access area, and external access area, the data collection strategy is as follows:

[0020] Additionally collect high-definition images and infrared images of the equipment in the equipment grid area, as well as infrared images and electrical parameters of the equipment wiring points;

[0021] For the channel grid areas of the wind tower channel area and the pipe gallery channel area, the data collection strategy is as follows:

[0022] Additional collection of smoke conditions in the channel grid area;

[0023] For the pipe network detection grid area in the pipe gallery area, the data collection strategy is as follows:

[0024] Additional collection of high-definition images and polarization images of cables and pipelines in the pipeline network inspection grid area;

[0025] For the channel grid area in the external channel region, the data collection strategy is:

[0026] The salt spray concentration of the channels in the channel grid area is additionally collected.

[0027] Optionally, in an example of the above aspect, the edge data management node performs scheduling control by communicating with a robot dog within a safe communication distance, including the following steps:

[0028] The edge data management node establishes a communication connection with the robot dog within a safe communication distance;

[0029] After receiving the robot dog scheduling control instructions between different grid areas in the same area sent by the robot dog scheduling module, the data collection strategy of the corresponding block is sent to the corresponding robot dog according to the scheduling control instructions, and an instruction to go to the corresponding block for data collection is sent.

[0030] Optionally, in an example of the above aspect, the edge data management node performs security analysis on different types of grids based on data collected by a robot dog in the grid area, and issues a security alert for grids whose security analysis is below a threshold, including the following steps:

[0031] Obtain high-definition image data, infrared image data, and polarization images of historical inspection data of cross-sea channels where surface anomalies are present, as well as high-definition image data, infrared image data, and polarization images where no surface anomalies are present, and mark the locations of surface anomalies;

[0032] The deep learning model is trained using image data from historical inspection data to identify and annotate abnormal locations in high-definition image data, infrared image data, and polarization images, and calculate the area of ​​the abnormal area.

[0033] Obtain historical data from high-definition image data, infrared image data, and polarization images of each area during normal operation. Use the trained deep learning model to detect the area of ​​the abnormal area and take the average value as the maximum abnormal fault area threshold for the high-definition image data, infrared image data, and polarization image data.

[0034] For all grid areas in the wind tower access area, pipe gallery access area, and external access area, basic operating data is tested during the preset test time interval, and basic grid safety analysis is performed using the following formula:

[0035]

[0036] Among them, S is the basic safety analysis value of the grid, Sv is the grid wind speed safety analysis value, Sth is the grid temperature and humidity safety analysis value, Sp is the grid image anomaly safety analysis value, w1, w2 and w3 are the corresponding weights of the grid wind speed safety analysis value, grid temperature and humidity safety analysis value and grid image anomaly safety analysis value respectively; v is the average wind speed of the corresponding grid within the preset detection time interval, v0 is the abnormal wind speed threshold of the corresponding grid area, T0 is the average temperature of the corresponding grid within the preset detection time interval, H0 is the average humidity of the corresponding grid within the preset detection time interval, σt is the standard deviation of the temperature of the corresponding grid within the preset detection time interval, σh is the standard deviation of the humidity of the corresponding grid within the preset detection time interval, fai, fbi and fci are the abnormal areas detected by the high-definition image data, infrared image data and polarization image data in the grid within the preset detection time interval. If no abnormality is detected or there is no corresponding detection data for the corresponding grid, the corresponding abnormal area is set to zero. famax, fbmax and fcmax are the maximum abnormal fault area thresholds detected by the high-definition image data, infrared image data and polarization image data;

[0037] Based on the results of the basic grid security analysis, additional security analysis is conducted for different areas;

[0038] The grid's basic security analysis results and additional security analysis results are weighted averaged to obtain the grid's final security analysis results, and security alerts are issued for grids whose security analysis results are below the threshold.

[0039] Optionally, in an example of the above aspect, performing additional security analysis on different regions includes the following steps:

[0040] For the equipment grid area in each region, obtain additional high-definition images and infrared images of the equipment in the equipment grid area, as well as infrared images and electrical parameters of the equipment wiring points, and perform additional safety analysis of the equipment using the following formula:

[0041]

[0042] Wherein, Sep is the additional safety analysis value of the equipment, fej is the area of ​​the abnormal region in the j-th image data additionally collected in the equipment grid area of ​​each region, femaxj is the maximum abnormal fault area threshold of the image type corresponding to the j-th image data additionally collected, σek is the standard deviation of the k-th electrical parameter additionally collected in the equipment grid area of ​​each region within the preset detection time interval, and Tek is the mean value of the k-th electrical parameter additionally collected in the equipment grid area of ​​each region within the preset detection time interval;

[0043] For the channel grid areas of the wind tower channel area and the pipe gallery channel area, obtain the smoke status of the channels in the additional collection channel grid area. If smoke is detected, immediately send a safety alarm signal to the corresponding grid; otherwise, no safety alarm signal is sent;

[0044] For the pipe network inspection grid area in the pipe corridor passageway, obtain high-definition images and polarization images of the cables and pipelines in the additional collection pipe network inspection grid area, and calculate the additional safety analysis value of the cables and pipelines using the same method as the calculation method of the image anomaly safety analysis value of the basic operation data grid;

[0045] For the channel grid area in the external channel area, obtain the salt spray concentration of the channels in the additional acquisition channel grid area, and calculate the ratio of the average value of the detected salt spray concentration within the preset detection time interval to the abnormal salt spray concentration threshold as the additional safety analysis value of the salt spray;

[0046] Add up all the additional safety analysis values ​​calculated in each grid. If the sum is zero, the additional safety analysis result of the corresponding grid is zero, and the additional safety analysis value of the grid does not participate in the subsequent calculation of the final safety analysis result;

[0047] If the added result is greater than zero, the reciprocal of the added result plus 1 is taken as the additional safety analysis result of the corresponding grid and is used in the subsequent calculation of the final safety analysis result.

[0048] Optionally, in an example of the above aspect, the model building module establishes a security level analysis model for different grid areas based on the security analysis results of grids in different areas, and performs a security level analysis on each grid area, including the following steps:

[0049] Obtain historical detection data of different grid areas in different regions before a fault occurs, as well as historical detection data of different grid areas that have been operating normally for a long time, and label the historical detection data with pre-fault data and normal data;

[0050] The deep learning model trained by the edge data management node detects the area of ​​abnormal regions in the image data, adds it to the corresponding historical data to replace the original image data, and calculates the security analysis results of the grid of historical data;

[0051] According to the historical data of different grid areas, the LSTM-fully connected hybrid model is trained separately to predict whether the safety analysis results of the grid are pre-fault data or normal data.

[0052] Optionally, in an example of the above aspect, based on historical data of different grid areas, LSTM-fully connected hybrid models are trained separately to predict whether the safety analysis result of the grid is pre-fault data or normal data, including the following steps:

[0053] Using historical data from different grid areas, data from any time period is used as input, and the corresponding data from the next time period is used as output. LSTM multimodal time series models are trained separately to predict future detection data and grid security analysis results based on historical data.

[0054] The output of the LSTM multimodal time series model is equipped with a feature processing layer and a connection layer. The feature processing layer calls the edge data management node to calculate the future grid security analysis results based on the prediction results of future detection data as feature values, and retains the labels of pre-fault data and normal data in the estimated value.

[0055] The characteristic values ​​of the future grid safety analysis results and the predicted values ​​of the safety analysis results output by the LSTM multimodal time series model are input into the fully connected classifier through the connection layer, and the safety analysis results of the grid are predicted as pre-fault data or normal data in different grid areas.

[0056] Optionally, in an example of the above aspect, the robot dog scheduling module performs support scheduling for robot dogs in different grid areas through the edge data management node based on the fault degree analysis results of the grid area and the detection progress of the grid in the grid area, including the following steps:

[0057] Obtain the results of the fault degree analysis of the grid area and the inspection progress of the grids in the grid area;

[0058] If the fault level analysis result of a grid area is normal, the robot dogs in different grid areas will not be dispatched;

[0059] If the result of the fault degree analysis in the grid area is the pre-fault data, then the detection progress of the grid in the grid area is checked to see if it is greater than 50%. If so, no scheduling is performed;

[0060] If not, the detection progress of the grid areas in the same area will be checked, and the grid areas with normal analysis results and detection progress greater than 50% will be selected as candidate grid areas. If there are no candidate grid areas, no scheduling will be performed;

[0061] If there is an alternative grid area, the robot dog in the grid area closest to the grid area before the failure is selected, and the support scheduling command and the robot dog number are sent to the edge data management node.

[0062] Compared with the prior art, the present invention has the following beneficial effects:

[0063] The present invention can receive and process data from the robot dog in real time through the edge data management node, reduce the delay in data transmission, and improve the response speed. By processing and analyzing data at the edge, the data processing pressure on the cloud is reduced, making the entire system more efficient; the edge data management node can perform security analysis based on the data collected by the robot dog, and promptly discover and warn of potential security risks; since the data is processed on the edge side, the risk of data leakage during transmission is reduced, and the security of the data is improved.

[0064] This invention uses a safety analysis model to intuitively understand the safety status of each grid zone, helping managers quickly identify potential risk areas and effectively allocate resources and attention. Edge data management nodes schedule robot dogs for support based on fault severity and detection progress, ensuring timely and effective monitoring of critical areas. Through model analysis and data node scheduling, problems can be quickly located and nearby robot dogs mobilized for support. Based on the grid zone's safety and fault severity analysis results, the robot dog deployment and scheduling strategy can be dynamically adjusted to avoid idle or over-concentrated resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0066] Figure 1 Schematic diagram of the system framework of the present invention. DETAILED DESCRIPTION

[0067] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0068] See also Figure 1 , the first aspect of the present invention provides a traffic tunnel inspection system of a robot dog platform based on AI technology, including: an edge terminal and a cloud management center;

[0069] The edge end includes:

[0070] Area allocation module: This module obtains channel data for the cross-sea channel area and divides the target area into wind tower channel area, pipe gallery channel area, and external channel area. It then deploys robot dogs that comply with different channel management rules in different areas and divides different areas into grid areas for different detection types.

[0071] Data collection module: Set different data collection strategies for different grids. The robot dogs in each grid area collect data according to the data collection strategies of different grid areas and send the data to the edge data management node;

[0072] Edge data management nodes: Edge data management nodes are set up at every safe communication distance in the cross-sea channel area. These edge data management nodes perform scheduling control by communicating with robot dogs within the safe communication distance. Based on the data collected by the robot dogs in the grid area, they conduct security analysis on different types of grids and issue security alerts for grids whose security analysis falls below the threshold.

[0073] The cloud management center includes:

[0074] Model building module: Based on the security analysis results of grids in different areas, a security analysis model for different grid areas is established, and the security level analysis of each grid area is performed;

[0075] Robot dog scheduling module: Based on the fault degree analysis results of the grid area and the detection progress of the grid in the grid area, support scheduling of robot dogs in different grid areas is carried out through the edge data management node.

[0076] Specifically, in this embodiment, the wind tower channel area includes the wind turbine generator set, the transmission tower base, and the connecting corridor, where a high-wind-speed-resistant robot dog is deployed;

[0077] Pipeline corridor area: Explosion-proof and waterproof robot dogs are deployed inside pipeline corridors and maintenance passages on the seabed used to accommodate pipelines or cables;

[0078] External passage area: and emergency passage, deploy all-terrain patrol robot dogs;

[0079] Edge data management node, hardware configuration

[0080] Computing unit: NVID IA Jetson AGX Xavier, 32TOPS AI computing power;

[0081] Storage system: RAID 1 dual solid-state drives, 1TB x 2, read and write speed 550MB / s,

[0082] Communication module: supports 5G / satellite dual-link redundancy, uplink and downlink rates ≥ 1Gbps.

[0083] An edge data management node is set up at each safe communication distance in the cross-sea channel area. The edge data management node performs scheduling control by communicating with the robot dogs within the safe communication distance. Based on the data collected by the robot dogs in the grid area, it conducts security analysis on different types of grids and issues security alerts for grids whose security analysis is below the threshold.

[0084] Edge data management nodes can receive and process data from the robot dogs in real time, reducing data transmission delays and improving response speed. Data processing and analysis at the edge reduces data processing pressure on the cloud, making the entire system more efficient. Edge data management nodes can perform security analysis based on the data collected by the robot dogs, promptly identifying and issuing warnings about potential security risks. Because data is processed at the edge, the risk of data leakage during transmission is reduced, improving data security.

[0085] The edge data management node configuration allows for flexible adjustments based on the specific conditions of the cross-sea channel, adapting to different safe communication distances and grid areas. As technology evolves and demand grows, new edge data management nodes and robot dogs can be added to expand the system's scale and functionality.

[0086] The model building module establishes safety analysis models for different grid areas based on the safety analysis results of grids in different areas, and conducts safety analysis on each grid area; the robot dog scheduling module supports and schedules robot dogs in different grid areas through edge data management nodes based on the fault degree analysis results and the detection progress of grids in the grid area.

[0087] The safety analysis model provides an intuitive understanding of the safety status of each grid zone, helping managers quickly identify potential risk areas and effectively allocate resources and attention. Edge data management nodes dispatch robot dogs based on fault severity and detection progress, ensuring timely and effective monitoring of critical areas and improving overall safety management efficiency.

[0088] In the event of a failure or security incident, model analysis and data node scheduling can quickly locate the problem and mobilize nearby robot dogs for support, shortening emergency response time. Based on the grid area's safety and fault severity analysis, the robot dog deployment and scheduling strategy is dynamically adjusted to avoid idle resources or excessive concentration. Through rational resource allocation, each grid area is properly monitored and supported, improving resource utilization efficiency.

[0089] Through continuous monitoring and scheduling, potential safety hazards can be discovered and dealt with in a timely manner, reducing the probability of accidents and improving overall safety.

[0090] In one embodiment of the present invention, the area allocation module includes:

[0091] Area division unit: This unit obtains channel data of the cross-sea channel area, including channel detection area data, channel accident records, channel type, and channel map data. Based on the channel detection area data and channel map data, it divides the channel connected to the wind tower into the wind tower channel area, the channel accommodating various pipelines into the pipeline corridor channel area, and the channel accommodating the pedestrian and vehicle traffic area into the external channel area.

[0092] Robot dog allocation unit: According to the detection area area in the channel detection area data of the wind tower channel area, the pipe gallery channel area and the external channel area, robot dogs that comply with the channel management rules of different areas are set in different areas;

[0093] Grid division unit: According to the channel data of each area, the wind tower channel area is divided into equipment grid area, channel grid area and wind tower monitoring grid area; the corridor channel area is divided into channel grid area, equipment grid area and pipe network detection grid area; the external channel area is divided into channel grid area and equipment grid area.

[0094] Specifically, in this embodiment, according to the detection area area in the channel detection area data of the wind tower channel area, the pipe gallery channel area, and the external channel area, robot dogs that comply with the channel management rules of different areas are set in different areas;

[0095] Wind tower access areas are typically narrow and may experience high wind speeds and changes in wind direction.

[0096] Wind tower access area channel management rules:

[0097] Wind tower equipment needs to be inspected, maintained, and troubleshooted.

[0098] The robot dog should have good stability and wind resistance.

[0099] The robot dog is set to:

[0100] A robot dog with a compact structure and light weight is selected to reduce the impact on wind resistance.

[0101] The robot dog is equipped with a high-definition camera, infrared thermal imager sensor, Doppler radar (wind speed) and temperature and humidity sensor, as well as a smoke sensor and equipment grid area current and voltage detector to carry out detailed inspections of wind tower equipment.

[0102] Pipeline corridors are typically located underground, with narrow, long, and enclosed spaces. This can lead to electromagnetic interference and difficulties in wireless signal transmission.

[0103] Pipeline corridor area channel management rules:

[0104] It is necessary to inspect and troubleshoot pipelines, cables and other facilities.

[0105] The robot dog should have good wireless communication and anti-interference capabilities.

[0106] The robot dog is set to:

[0107] Choose a robot dog with strong wireless communication capabilities and good anti-interference performance.

[0108] Equipped with high-performance wireless base stations and roaming airborne terminals to ensure wireless signal coverage and stable transmission within the corridor.

[0109] The robot dog is equipped with a high-definition camera, an infrared thermal imaging sensor, a temperature and humidity sensor, a smoke sensor, a polarized light camera, and an equipment grid area current and voltage detector, so as to conduct real-time monitoring and data analysis of pipelines, cables and other facilities.

[0110] The external passage area is usually relatively open, but may face different climatic conditions and terrain conditions such as sandy, rocky, and snowy terrain.

[0111] External channel area channel management rules:

[0112] External facilities need to be inspected, monitored and handled in an emergency.

[0113] The robot dog should have good driving ability and the ability to work in all weather conditions.

[0114] The robot dog is set to:

[0115] Choose a robot dog with tires or tracks to adapt to various road conditions; it has waterproof, dustproof and temperature-resistant properties to ensure that it can still work normally under various conditions.

[0116] The robot dog is equipped with a high-definition camera, an infrared thermal imager sensor, a temperature and humidity sensor, a Doppler radar for detecting wind speed, a salt spray concentration sensor, and a current and voltage detector for the equipment grid area to improve the robot dog's environmental perception ability and positioning accuracy.

[0117] In one embodiment of the present invention, the data collection module sets different data collection strategies for different grids, including:

[0118] For all grid areas in the wind tower access area, pipe gallery access area, and external access area, basic operating data is tested, including wind speed, temperature, humidity, high-definition image data and infrared image data of the access area;

[0119] For the equipment grid areas in the wind tower access area, pipe gallery access area, and external access area, the data collection strategy is as follows:

[0120] Additionally collect high-definition images and infrared images of the equipment in the equipment grid area, as well as infrared images and electrical parameters of the equipment wiring points;

[0121] For the channel grid areas of the wind tower channel area and the pipe gallery channel area, the data collection strategy is as follows:

[0122] Additional collection of smoke conditions in the channel grid area;

[0123] For the pipe network detection grid area in the pipe gallery area, the data collection strategy is as follows:

[0124] Additional collection of high-definition images and polarization images of cables and pipelines in the pipeline network inspection grid area;

[0125] For the channel grid area in the external channel region, the data collection strategy is:

[0126] The salt spray concentration of the channels in the channel grid area is additionally collected.

[0127] In one embodiment of the present invention, the edge data management node performs scheduling control by communicating with a robot dog within a safe communication distance, including the following steps:

[0128] The edge data management node establishes a communication connection with the robot dog within a safe communication distance;

[0129] After receiving the robot dog scheduling control instructions between different grid areas in the same area sent by the robot dog scheduling module, the data collection strategy of the corresponding block is sent to the corresponding robot dog according to the scheduling control instructions, and an instruction to go to the corresponding block for data collection is sent.

[0130] In one embodiment of the present invention, the edge data management node performs security analysis on different types of grids based on data collected by a robot dog in the grid area, and issues a security alert for grids whose security analysis score is below a threshold, including the following steps:

[0131] Obtain high-definition image data, infrared image data, and polarization images of historical inspection data of cross-sea channels where surface anomalies are present, as well as high-definition image data, infrared image data, and polarization images where no surface anomalies are present, and mark the locations of surface anomalies;

[0132] The deep learning model is trained using image data from historical inspection data to identify and annotate abnormal locations in high-definition image data, infrared image data, and polarization images, and calculate the area of ​​the abnormal area.

[0133] Obtain historical data from high-definition image data, infrared image data, and polarization images of each area during normal operation. Use the trained deep learning model to detect the area of ​​the abnormal area and take the average value as the maximum abnormal fault area threshold for the high-definition image data, infrared image data, and polarization image data.

[0134] For all grid areas in the wind tower access area, pipe gallery access area, and external access area, basic operating data is tested during the preset test time interval, and basic grid safety analysis is performed using the following formula:

[0135]

[0136] Wherein, S is the basic safety analysis value of the grid, Sv is the grid wind speed safety analysis value, Sth is the grid temperature and humidity safety analysis value, Sp is the grid image anomaly safety analysis value, w1, w2 and w3 are the corresponding weights of the grid wind speed safety analysis value, grid temperature and humidity safety analysis value and grid image anomaly safety analysis value respectively; v is the mean wind speed of the corresponding grid within the preset detection time interval, v0 is the abnormal wind speed threshold of the corresponding grid area, T0 is the mean temperature of the corresponding grid within the preset detection time interval, H0 is the mean humidity of the corresponding grid within the preset detection time interval, σt is the standard deviation of the temperature of the corresponding grid within the preset detection time interval, σh is the standard deviation of the humidity of the corresponding grid within the preset detection time interval, fai, fbi and fci are the abnormal areas detected by the high-definition image data, infrared image data and polarization image data in the grid within the preset detection time interval. If no abnormality is detected or there is no corresponding detection data for the corresponding grid, the corresponding abnormal area is set to zero. famax, fbmax and fcmax are the maximum abnormal fault area thresholds detected by the high-definition image data, infrared image data and polarization image data;

[0137] Based on the results of the basic grid security analysis, additional security analysis is conducted for different areas;

[0138] The grid's basic security analysis results and additional security analysis results are weighted averaged to obtain the grid's final security analysis results, and security alerts are issued for grids whose security analysis results are below the threshold.

[0139] Specifically, in this embodiment, for grids that only detect basic operating data, the basic safety analysis results of the grid are used as the final safety analysis results of the grid. For grids that perform additional safety analysis, the basic safety analysis results of the grid and the additional safety analysis are weighted averaged. This embodiment performs basic safety analysis and additional safety analysis by statistically analyzing data from a large number of well-operating grid areas and faulty grid areas. The weight of the basic safety analysis results of the grid is set to 0.75, and the weight of the additional safety analysis results is set to 0.25. At the same time, after experiments, the final safety analysis results of the grid are obtained. The safety analysis results of the well-operating grid areas are all greater than 0.75, so the threshold for the safety analysis is set to 0.75.

[0140] At the same time, in this embodiment, for the grid of the wind tower channel area, w1, w2 and w3 are set to 0.4, 0.3 and 0.3 respectively; for the grid of the pipe gallery channel area, w1, w2 and w3 are set to 0.2, 0.5 and 0.3 respectively; for the grid of the external channel area, w1, w2 and w3 are set to 0.2, 0.3 and 0.5 respectively.

[0141] In one embodiment of the present invention, additional security analysis of different regions is performed, including the following steps:

[0142] For the equipment grid area in each region, obtain additional high-definition images and infrared images of the equipment in the equipment grid area, as well as infrared images and electrical parameters of the equipment wiring points, and perform additional safety analysis of the equipment using the following formula:

[0143]

[0144] Wherein, Sep is the additional safety analysis value of the equipment, fej is the area of ​​the abnormal region in the j-th image data additionally collected in the equipment grid area of ​​each region, femaxj is the maximum abnormal fault area threshold of the image type corresponding to the j-th image data additionally collected, σek is the standard deviation of the k-th electrical parameter additionally collected in the equipment grid area of ​​each region within the preset detection time interval, and Tek is the mean value of the k-th electrical parameter additionally collected in the equipment grid area of ​​each region within the preset detection time interval;

[0145] For the channel grid areas of the wind tower channel area and the pipe gallery channel area, obtain the smoke status of the channels in the additional collection channel grid area. If smoke is detected, immediately send a safety alarm signal to the corresponding grid; otherwise, no safety alarm signal is sent;

[0146] For the pipe network inspection grid area in the pipe corridor passageway, obtain high-definition images and polarization images of the cables and pipelines in the additional collection pipe network inspection grid area, and calculate the additional safety analysis value of the cables and pipelines using the same method as the calculation method of the image anomaly safety analysis value of the basic operation data grid;

[0147] For the channel grid area in the external channel area, obtain the salt spray concentration of the channels in the additional acquisition channel grid area, and calculate the ratio of the average value of the detected salt spray concentration within the preset detection time interval to the abnormal salt spray concentration threshold as the additional safety analysis value of the salt spray;

[0148] Add up all the additional safety analysis values ​​calculated in each grid. If the sum is zero, the additional safety analysis result of the corresponding grid is zero, and the additional safety analysis value of the grid does not participate in the subsequent calculation of the final safety analysis result;

[0149] If the added result is greater than zero, the reciprocal of the added result plus 1 is taken as the additional safety analysis result of the corresponding grid and is used in the subsequent calculation of the final safety analysis result.

[0150] In one embodiment of the present invention, the model building module establishes a security analysis model for different grid areas based on the security analysis results of grids in different areas, and performs security analysis on each grid area, including the following steps:

[0151] Obtain historical detection data of different grid areas in different regions before a fault occurs, as well as historical detection data of different grid areas that have been operating normally for a long time, and label the historical detection data with pre-fault data and normal data;

[0152] The deep learning model trained by the edge data management node detects the area of ​​abnormal regions in the image data, adds it to the corresponding historical data to replace the original image data, and calculates the security analysis results of the grid of historical data;

[0153] According to the historical data of different grid areas, the LSTM-fully connected hybrid model is trained separately to predict whether the safety analysis results of the grid are pre-fault data or normal data.

[0154] In one embodiment of the present invention, based on the historical data of different grid areas, LSTM-fully connected hybrid models are trained separately to predict whether the safety analysis results of the grid are pre-fault data or normal data, including the following steps:

[0155] Using historical data from different grid areas, data from any time period is used as input, and the corresponding data from the next time period is used as output. LSTM multimodal time series models are trained separately to predict future detection data and grid security analysis results based on historical data.

[0156] The output of the LSTM multimodal time series model is equipped with a feature processing layer and a connection layer. The feature processing layer calls the edge data management node to calculate the future grid security analysis results based on the prediction results of future detection data as feature values, and retains the labels of pre-fault data and normal data in the estimated value.

[0157] The characteristic values ​​of the future grid safety analysis results and the predicted values ​​of the safety analysis results output by the LSTM multimodal time series model are input into the fully connected classifier through the connection layer, and the safety analysis results of the grid are predicted as pre-fault data or normal data in different grid areas.

[0158] In one embodiment of the present invention, the robot dog scheduling module performs support scheduling for robot dogs in different grid areas through the edge data management node based on the fault degree analysis results of the grid area and the detection progress of the grid area, including the following steps:

[0159] Obtain the results of the fault degree analysis of the grid area and the inspection progress of the grids in the grid area;

[0160] If the fault level analysis result of a grid area is normal, the robot dogs in different grid areas will not be dispatched;

[0161] If the result of the fault degree analysis in the grid area is the pre-fault data, then the detection progress of the grid in the grid area is checked to see if it is greater than 50%. If so, no scheduling is performed;

[0162] If not, the detection progress of the grid areas in the same area will be checked, and the grid areas with normal analysis results and detection progress greater than 50% will be selected as candidate grid areas. If there are no candidate grid areas, no scheduling will be performed;

[0163] If there is an alternative grid area, the robot dog in the grid area closest to the grid area before the failure is selected, and the support scheduling command and the robot dog number are sent to the edge data management node.

[0164] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A traffic tunnel inspection system based on a robot dog platform of AI technology, characterized by: include: Edge and cloud management centers; The edge end includes: Area allocation module: This module obtains channel data for the cross-sea channel area and divides the target area into wind tower channel area, pipe gallery channel area, and external channel area. It then deploys robot dogs that comply with different channel management rules in different areas and divides different areas into grid areas for different detection types. Data collection module: Set different data collection strategies for different grids. The robot dogs in each grid area collect data according to the data collection strategies of different grid areas and send the data to the edge data management node; Edge data management nodes: Edge data management nodes are set up at every safe communication distance in the cross-sea channel area. These edge data management nodes perform scheduling control by communicating with robot dogs within the safe communication distance. Based on the data collected by the robot dogs in the grid area, they conduct security analysis on different types of grids and issue security alerts for grids whose security analysis falls below the threshold. The cloud management center includes: Model building module: Based on the security analysis results of grids in different areas, a security analysis model for different grid areas is established, and the security level analysis of each grid area is performed; Robot dog scheduling module: Based on the fault degree analysis results of the grid area and the detection progress of the grid in the grid area, support scheduling of robot dogs in different grid areas is carried out through the edge data management node.

2. The AI-based robot dog platform traffic tunnel inspection system according to claim 1 is characterized in that: Area allocation module, including: Area division unit: This unit obtains channel data of the cross-sea channel area, including channel detection area data, channel accident records, channel type, and channel map data. Based on the channel detection area data and channel map data, it divides the channel connected to the wind tower into the wind tower channel area, the channel accommodating various pipelines into the pipeline corridor channel area, and the channel accommodating the pedestrian and vehicle traffic area into the external channel area. Robot dog allocation unit: According to the detection area area in the channel detection area data of the wind tower channel area, the pipe gallery channel area and the external channel area, robot dogs that comply with the channel management rules of different areas are set in different areas; Grid division unit: According to the channel data of each area, the wind tower channel area is divided into equipment grid area, channel grid area and wind tower monitoring grid area; the corridor channel area is divided into channel grid area, equipment grid area and pipe network detection grid area; the external channel area is divided into channel grid area and equipment grid area.

3. The AI-based robot dog platform traffic tunnel inspection system according to claim 1 is characterized in that: The data acquisition module sets different data acquisition strategies for different grids, including: For all grid areas in the wind tower access area, pipe gallery access area, and external access area, basic operating data is tested, including wind speed, temperature, humidity, high-definition image data and infrared image data of the access area; For the equipment grid areas in the wind tower access area, pipe gallery access area, and external access area, the data collection strategy is as follows: Additionally collect high-definition images and infrared images of the equipment in the equipment grid area, as well as infrared images and electrical parameters of the equipment wiring points; For the channel grid areas of the wind tower channel area and the pipe gallery channel area, the data collection strategy is as follows: Additional collection of smoke conditions in the channel grid area; For the pipe network detection grid area in the pipe gallery area, the data collection strategy is as follows: Additional collection of high-definition images and polarization images of cables and pipelines in the pipeline network inspection grid area; For the channel grid area in the external channel region, the data collection strategy is: The salt spray concentration of the channels in the channel grid area is additionally collected.

4. The AI-based robot dog platform traffic tunnel inspection system according to claim 1 is characterized in that: The edge data management node performs scheduling control by communicating with the robot dog within a safe communication distance, including the following steps: The edge data management node establishes a communication connection with the robot dog within a safe communication distance; After receiving the robot dog scheduling control instructions between different grid areas in the same area sent by the robot dog scheduling module, the data collection strategy of the corresponding block is sent to the corresponding robot dog according to the scheduling control instructions, and an instruction to go to the corresponding block for data collection is sent.

5. The AI-based robot dog platform traffic tunnel inspection system according to claim 1 is characterized in that: The edge data management node performs security analysis on different types of grids based on the data collected by the robot dog in the grid area, and issues a security alert for grids whose security analysis is below a threshold, including the following steps: Obtain high-definition image data, infrared image data, and polarization images of historical inspection data of cross-sea channels where surface anomalies are present, as well as high-definition image data, infrared image data, and polarization images where no surface anomalies are present, and mark the locations of surface anomalies; The deep learning model is trained using image data from historical inspection data to identify and annotate abnormal locations in high-definition image data, infrared image data, and polarization images, and calculate the area of ​​the abnormal area. Obtain historical data from high-definition image data, infrared image data, and polarization images of each area during normal operation. Use the trained deep learning model to detect the area of ​​the abnormal area and take the average value as the maximum abnormal fault area threshold for the high-definition image data, infrared image data, and polarization image data. For all grid areas in the wind tower access area, pipe gallery access area, and external access area, basic operating data is tested during the preset test time interval, and basic grid safety analysis is performed using the following formula: Among them, S is the basic safety analysis value of the grid, Sv is the grid wind speed safety analysis value, Sth is the grid temperature and humidity safety analysis value, Sp is the grid image anomaly safety analysis value, w1, w2 and w3 are the corresponding weights of the grid wind speed safety analysis value, grid temperature and humidity safety analysis value and grid image anomaly safety analysis value respectively; v is the average wind speed of the corresponding grid within the preset detection time interval, v0 is the abnormal wind speed threshold of the corresponding grid area, T0 is the average temperature of the corresponding grid within the preset detection time interval, H0 is the average humidity of the corresponding grid within the preset detection time interval, σt is the standard deviation of the temperature of the corresponding grid within the preset detection time interval, σh is the standard deviation of the humidity of the corresponding grid within the preset detection time interval, fai, fbi and fci are the abnormal areas detected by the high-definition image data, infrared image data and polarization image data in the grid within the preset detection time interval. If no abnormality is detected or there is no corresponding detection data for the corresponding grid, the corresponding abnormal area is set to zero. famax, fbmax and fcmax are the maximum abnormal fault area thresholds detected by the high-definition image data, infrared image data and polarization image data; Based on the results of the basic grid security analysis, additional security analysis is conducted for different areas; The grid's basic security analysis results and additional security analysis results are weighted averaged to obtain the grid's final security analysis results, and security alerts are issued for grids whose security analysis results are below the threshold.

6. The AI-based robot dog platform traffic tunnel inspection system according to claim 5 is characterized in that: For different regions, additional safety analysis is conducted in different areas, including the following steps: For the equipment grid area in each region, obtain additional high-definition images and infrared images of the equipment in the equipment grid area, as well as infrared images and electrical parameters of the equipment wiring points, and perform additional safety analysis of the equipment using the following formula: Wherein, Sep is the additional safety analysis value of the equipment, fej is the area of ​​the abnormal region in the j-th image data additionally collected in the equipment grid area of ​​each region, femaxj is the maximum abnormal fault area threshold of the image type corresponding to the j-th image data additionally collected, σek is the standard deviation of the k-th electrical parameter additionally collected in the equipment grid area of ​​each region within the preset detection time interval, and Tek is the mean value of the k-th electrical parameter additionally collected in the equipment grid area of ​​each region within the preset detection time interval; For the channel grid areas of the wind tower channel area and the pipe gallery channel area, obtain the smoke status of the channels in the additional collection channel grid area. If smoke is detected, immediately send a safety alarm signal to the corresponding grid; otherwise, no safety alarm signal is sent; For the pipe network inspection grid area in the pipe corridor passageway, obtain high-definition images and polarization images of the cables and pipelines in the additional collection pipe network inspection grid area, and calculate the additional safety analysis value of the cables and pipelines using the same method as the calculation method of the image anomaly safety analysis value of the basic operation data grid; For the channel grid area in the external channel area, obtain the salt spray concentration of the channels in the additional acquisition channel grid area, and calculate the ratio of the average value of the detected salt spray concentration within the preset detection time interval to the abnormal salt spray concentration threshold as the additional safety analysis value of the salt spray; Add up all the additional safety analysis values ​​calculated in each grid. If the sum is zero, the additional safety analysis result of the corresponding grid is zero, and the additional safety analysis value of the grid does not participate in the subsequent calculation of the final safety analysis result; If the added result is greater than zero, the reciprocal of the added result plus 1 is taken as the additional safety analysis result of the corresponding grid and is used in the subsequent calculation of the final safety analysis result.

7. The AI-based robot dog platform traffic tunnel inspection system according to claim 5 is characterized in that: The model building module establishes a safety level analysis model for different grid areas based on the safety analysis results of the grids in different areas, and performs safety level analysis on each grid area, including the following steps: Obtain historical detection data of different grid areas in different regions before a fault occurs, as well as historical detection data of different grid areas that have been operating normally for a long time, and label the historical detection data with pre-fault data and normal data; The deep learning model trained by the edge data management node detects the area of ​​abnormal regions in the image data, adds it to the corresponding historical data to replace the original image data, and calculates the security analysis results of the grid of historical data; According to the historical data of different grid areas, the LSTM-fully connected hybrid model is trained separately to predict whether the safety analysis results of the grid are pre-fault data or normal data.

8. The AI-based robot dog platform traffic tunnel inspection system according to claim 7 is characterized in that: Based on the historical data of different grid areas, the LSTM-fully connected hybrid model is trained separately to predict whether the grid security analysis results are pre-fault data or normal data. The following steps are included: Using historical data from different grid areas, data from any time period is used as input, and the corresponding data from the next time period is used as output. LSTM multimodal time series models are trained separately to predict future detection data and grid security analysis results based on historical data. The output of the LSTM multimodal time series model is equipped with a feature processing layer and a connection layer. The feature processing layer calls the edge data management node to calculate the future grid security analysis results based on the prediction results of future detection data as feature values, and retains the labels of pre-fault data and normal data in the estimated value. The characteristic values ​​of the future grid safety analysis results and the predicted values ​​of the safety analysis results output by the LSTM multimodal time series model are input into the fully connected classifier through the connection layer, and the safety analysis results of the grid are predicted as pre-fault data or normal data in different grid areas.

9. The AI-based robot dog platform traffic tunnel inspection system according to claim 1 is characterized in that: The robot dog scheduling module performs support scheduling for robot dogs in different grid areas through the edge data management node based on the fault degree analysis results of the grid area and the detection progress of the grid area, including the following steps: Obtain the results of the fault degree analysis of the grid area and the inspection progress of the grids in the grid area; If the fault level analysis result of a grid area is normal, the robot dogs in different grid areas will not be dispatched; If the result of the fault degree analysis in the grid area is the pre-fault data, then the detection progress of the grid in the grid area is checked to see if it is greater than 50%. If so, no scheduling is performed; If not, the detection progress of the grid areas in the same area will be checked, and the grid areas with normal analysis results and detection progress greater than 50% will be selected as candidate grid areas. If there are no candidate grid areas, no scheduling will be performed; If there is an alternative grid area, the robot dog in the grid area closest to the grid area before the failure is selected, and the support scheduling command and the robot dog number are sent to the edge data management node.

Images