A fault recovery method and system for railway video monitoring system

Abstract

The present invention relates to a fault recovery method and system for a railway video monitoring system, the method comprising: collecting operation data of a video access node; performing neural learning on the operation data to obtain operation characteristics, and performing fault prediction on the video access node based on the operation characteristics to obtain an early warning fault node; establishing a fault takeover node, taking over the early warning fault node through the fault takeover node; troubleshooting the early warning fault node, and continuing to record the fault takeover node by restoring the normal early warning fault node to complete the fault recovery of the railway video monitoring system. The present invention can significantly improve the fault prediction capability, automation level and compatibility of the railway monitoring system, reduce the operation and maintenance cost, and provide a strong guarantee for the continuity and stability of the railway video monitoring system.

Description

Technical Field

[0001] The present invention relates to the technical field of video node management, and in particular to a fault recovery method and system for a railway video monitoring system. Background Art

[0002] Railway integrated video surveillance system is designed to meet the needs of railway traffic command, production organization, passenger service, operation monitoring, emergency rescue and public security prevention. It uses audio and video monitoring, communication, computer network, cloud computing, cloud storage and other technologies, combined with intelligent analysis technology, and has the functions of audio and video line-of-sight acquisition, transmission, control, display, storage and processing. With the increasing development of railway systems, video surveillance systems are gradually expanding their use in the fields of high-speed railway transportation command, production operations, operation management, public security maintenance, etc., and have become a key scientific and technological guarantee for railway safety production, improving efficiency and strengthening management.

[0003] In the current three-level architecture of the railway integrated video surveillance system, the video regional node cascades many video access nodes. In actual operation, there is a certain probability that the video access node and the regional node will be disconnected, resulting in the video regional node being unable to view all the real-time videos and recordings of the video access node, which in turn will also cause the video core node to be unable to view all the real-time videos and recordings of the video access node. Summary of the invention

[0004] The present invention provides a fault recovery method and system for a railway video monitoring system, which are used to solve the defects of the prior art.

[0005] The present invention provides a fault recovery method for a railway video monitoring system, comprising:

[0006] S1: Collect the operation data of the video access node;

[0007] S2: performing neural learning on the operation data to obtain operation characteristics, and performing fault prediction on the video access node based on the operation characteristics to obtain a warning fault node;

[0008] S3: Establish a fault takeover node, and take over the warning fault node through the fault takeover node;

[0009] S4: troubleshoot the early warning fault node, and resume recording the fault takeover node by restoring the normal early warning fault node to complete the fault recovery of the railway video surveillance system.

[0010] According to a fault recovery method for a railway video surveillance system provided by the present invention, the operating data in step S1 includes: device status, load information, and alarm logs of video access nodes.

[0011] According to a fault recovery method for a railway video surveillance system provided by the present invention, step S2 further comprises:

[0012] S21: Marking the operation data for a fault event to obtain marked data;

[0013] S22: extracting abnormal time of the fault abnormal value of the operation data based on the marking data to obtain time data;

[0014] S23: The operation data, the marking data and the time data are combined into a data set, and a neural network is trained through the data set to obtain a prediction network;

[0015] S24: Perform fault prediction on the video access node through the prediction network to obtain a warning fault node.

[0016] According to a fault recovery method for a railway video surveillance system provided by the present invention, the time data in step S22 includes: abnormal time points, abnormal time point ratios, and abnormal durations.

[0017] According to a fault recovery method for a railway video surveillance system provided by the present invention, in step S3, the fault takeover node maintains a heartbeat mechanism with the video area node to which the video access node belongs.

[0018] According to a fault recovery method for a railway video monitoring system provided by the present invention, the video access node and the video area node are connected via a consistent connection protocol.

[0019] According to a fault recovery method for a railway video surveillance system provided by the present invention, the connection protocol includes: user management information, authority management information, equipment resource management information, network management information, real-time status information of video access nodes, load information of video access nodes, control configuration commands of video access nodes, streaming media transmission of video access nodes, and alarm information of video access nodes.

[0020] According to a fault recovery method for a railway video surveillance system provided by the present invention, in step S3, when the predicted fault node does not exist in the video access node, the fault takeover node runs at no load.

[0021] According to a fault recovery method for a railway video surveillance system provided by the present invention, in step S4, the video access node and the fault takeover node are connected via a consistent transmission protocol.

[0022] The present invention also provides a fault recovery system for a railway video surveillance system, which is used to execute a fault recovery method for a railway video surveillance system as described in any one of the above items, comprising:

[0023] Collection module: used to collect the operation data of the video access node;

[0024] Prediction module: used for performing neural learning on the operation data to obtain operation characteristics, and performing fault prediction on the video access node based on the operation characteristics to obtain a warning fault node;

[0025] Takeover module: used to establish a fault takeover node and control the fault takeover node to take over the warning fault node;

[0026] Recovery module: used to troubleshoot the early warning fault node and control the early warning fault node that has recovered to normal to resume recording the fault takeover node.

[0027] The present invention provides a fault recovery method and system for a railway video monitoring system, which collects the operation data of the video access node, including key information such as equipment status, load information, and alarm logs, and performs neural learning on it. In the neural learning process, more attention is paid to the fault time, and the operation characteristics are accurately extracted, so as to achieve fault prediction of the video access node. The data-driven prediction method of the present invention has higher accuracy and reliability, and can discover potential fault points in advance, reducing the possibility of faults. Once the fault node is predicted, the present invention can quickly take over the early warning fault node through the fault takeover node to ensure the continuity and stability of the railway video monitoring system. At the same time, after the early warning fault node is troubleshooted, the fault takeover node is continued by restoring the normal early warning fault node, so as to achieve rapid recovery of the fault and reduce the system downtime and business interruption risk.

[0028] The present invention can monitor the status of the video access node in real time by introducing a fault takeover node and a heartbeat mechanism, ensuring that a quick response and takeover can be performed when a fault occurs. In addition, the fault takeover node is connected to the video access node through a consistent connection protocol, ensuring the reliability and stability of data transmission, and further improving the reliability and stability of the entire railway video monitoring system. And when there is no predicted fault node in the video access node, the fault takeover node can run without load, and will not occupy additional system resources, which not only ensures the redundancy of the system, but also avoids the waste of resources and realizes the efficient use of resources.

[0029] In general, the present invention can timely discover and handle potential safety hazards by introducing neural learning and fault prediction technology, thereby enhancing the security of the system. In addition, the fault takeover node also ensures the data integrity of the system during the fault takeover process. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0031] Figure 1 A schematic flow chart of a fault recovery method for a railway video monitoring system provided by an embodiment of the present invention;

[0032] Figure 2 A schematic diagram of the structure of a fault recovery system for a railway video monitoring system provided by an embodiment of the present invention.

[0033] Reference numerals: 100, acquisition module; 200, prediction module; 300, takeover module; 400, recovery module. DETAILED DESCRIPTION

[0034] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings in the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments, and they should not be understood as limitations on the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. In the description of the present invention, it should be understood that the terms used are only for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0035] Combine the following Figure 1 to Figure 2 Embodiments of the present invention are described.

[0036] The present invention provides a fault recovery method for a railway video monitoring system, comprising:

[0037] S1: Collect the operation data of the video access node.

[0038] The operation data in step S1 includes: device status, load information, and alarm log of the video access node.

[0039] The purpose of collecting the above-mentioned operating data in step S1 is to comprehensively monitor the operating status of the video access node and promptly discover potential fault risks. By subsequently inputting these data into the subsequent neural learning model, a model that can accurately predict faults can be trained, so that it can quickly take over and recover when a fault occurs, ensuring the continuity and stability of the railway video surveillance system.

[0040] S2: Performing neural learning on the operation data to obtain operation characteristics, and performing fault prediction on the video access node based on the operation characteristics to obtain a warning fault node.

[0041] Wherein, step S2 further comprises:

[0042] S21: Mark the operation data for a fault event to obtain marked data.

[0043] In step S21, the operation data is first checked to see whether it contains information such as node shutdown and restart of the video access node, and the event logs of node unresponsiveness and node recovery response are obtained. In addition, a series of system startup operations such as service initialization of the video area node and startup of various basic processes are also obtained, which can be represented as a node failure, and the time point when the failure occurred is recorded for subsequent analysis and application.

[0044] S22: Extracting abnormal time of the fault abnormal value of the operation data based on the marking data to obtain time data.

[0045] Since transient anomalies are a common situation in computer systems, they may quickly complete self-repair and return to normal working status. However, if the monitoring value anomaly persists, the possibility of affecting node operation will be significantly increased. Therefore, the present invention combines time data when selecting data for the prediction model.

[0046] The time data in step S22 includes: abnormal time point, abnormal time point ratio and abnormal duration.

[0047] The present invention adds the time point of the fault abnormal value, the ratio of the abnormal time point of the indicator in the past 120 time points of the abnormal time point (i.e., the number of abnormal times / 120), and the current abnormal duration to the time data, which are used to monitor the overall abnormal situation of the abnormal value. The reason why 120 time points are used as the statistical window capacity is that according to statistics, the time when all indicators are most recently classified as abnormal values ​​before all node failures is about 90% within 1 hour of the time when the failure occurs. Therefore, it is approximately rounded to 120 time points to achieve the goal of easy understanding and operation.

[0048] S23: The operation data, the marking data and the time data are combined into a data set, and a neural network is trained through the data set to obtain a prediction network.

[0049] In order to use machine learning methods to establish a fault anomaly prediction model, it is necessary to convert the operating feature data into a vector representation, where each vector represents the node status at a monitoring record time point. Therefore, in step S23, all relevant operating data are converted into operating feature vectors, and correspond to the abnormal marking results by time to form a vector data set for modeling and testing. After the vector data set is established, it is used as the input for training the model, and the abnormal marking will be used as the output of the model. Then, based on this, a fault anomaly prediction model for classifying normal and abnormal nodes is established.

[0050] S24: Perform fault prediction on the video access node through the prediction network to obtain a warning fault node.

[0051] S3: Establish a fault takeover node, and take over the warning fault node through the fault takeover node.

[0052] Wherein, in step S3, when the predicted fault node does not exist in the video access node, the fault takeover node runs without load.

[0053] The video access node and the video area node are connected via a consistent connection protocol.

[0054] The connection protocol includes: user management information, authority management information, device resource management information, network management information, real-time status information of video access nodes, load information of video access nodes, control configuration commands of video access nodes, streaming media transmission of video access nodes, and alarm information of video access nodes.

[0055] When all video access nodes in the system are normal, they are in an unloaded state and only maintain a heartbeat mechanism with the video area nodes. The video area nodes and the video access nodes are connected through a unified public protocol to achieve decoupling between the video area nodes and the access nodes, so as to ensure that when the video area nodes and the video access nodes are on different platforms, they can be connected synchronously as long as the public communication protocol is followed.

[0056] Wherein, in step S3, the failover node maintains a heartbeat mechanism with the video area node to which the video access node belongs.

[0057] Based on the heartbeat mechanism, when a predicted video node is about to fail, the video area node automatically senses the failure and controls the fault takeover node in the unloaded state to take over the video device belonging to the failed video access node.

[0058] S4: troubleshoot the early warning fault node, and resume recording the fault takeover node by restoring the normal early warning fault node to complete the fault recovery of the railway video surveillance system.

[0059] Wherein, in step S4, the video access node is connected to the failover node via a consistent transmission protocol.

[0060] When the failed video access node returns to normal, it first notifies the video area node, which then controls the failed takeover node to stop taking over the service. The video area node controls the restored video access node to re-record the video recording during the failure period from the failed takeover node. When the recording synchronization (re-recording) operation between the failed takeover node and the restored video access node is performed, since video access nodes on different platforms are involved, a common transmission protocol also needs to be defined.

[0061] like Figure 2 As shown, the present invention also provides a fault recovery system for a railway video monitoring system, comprising:

[0062] Collection module 100: used to collect operation data of video access nodes;

[0063] Prediction module 200: used to perform neural learning on the operation data to obtain operation characteristics, and perform fault prediction on the video access node based on the operation characteristics to obtain a warning fault node;

[0064] The takeover module 300 is used to establish a fault takeover node and control the fault takeover node to take over the warning fault node;

[0065] Restoration module 400: used to troubleshoot the early warning fault node and control the early warning fault node that has recovered to normal to resume recording the fault takeover node.

[0066] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0067] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0068] The present invention provides a fault recovery method and system for a railway video surveillance system. When it is predicted that a video access node is about to fail, the present invention can automatically establish a fault takeover node and monitor the fault situation in real time through a heartbeat mechanism. Once a fault occurs, the fault takeover node can quickly take over the business of the faulty node to ensure the continuity of the video surveillance system. After the fault is eliminated, the normal warning fault node can automatically resume recording the fault takeover node to complete the fault recovery process. This automated fault takeover and recovery mechanism greatly reduces manual intervention and improves the response speed and recovery efficiency of the system.

[0069] The present invention collects the operation data of the video access node, including the equipment status, load information and alarm log, and combines the neural learning model to predict the fault. The present invention can accurately identify the potential fault risk and realize the early warning of the fault. This greatly reduces the system downtime and business interruption risk caused by the fault, and improves the stability and reliability of the railway video monitoring system; on this basis, through the automated fault prediction, takeover and recovery mechanism, it can significantly reduce the operation and maintenance cost of the railway video monitoring system, reduce the manual intervention and downtime caused by the fault, reduce the workload and maintenance cost of the maintenance personnel, and at the same time, by improving the stability and reliability of the system, reduce the business interruption and loss caused by the fault.

[0070] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A fault recovery method for a railway video monitoring system, characterized in that: include: S1: Collect the operation data of the video access node; The operation data in step S1 includes: device status, load information, and alarm log of the video access node; S2: learning the operation data through a neural network to obtain operation characteristics, and performing fault prediction on the video access node based on the operation characteristics to obtain a warning fault node; Step S2 further comprises: S21: Marking the operation data for a fault event to obtain marked data; S22: extracting abnormal time of the fault abnormal value of the operation data based on the marking data to obtain time data, wherein the time data includes: abnormal time point, abnormal time point ratio and abnormal duration; S23: The operation data, the marking data and the time data are combined into a data set, and a neural network is trained through the data set to obtain a prediction network; S24: performing fault prediction on the video access node through the prediction network to obtain a warning fault node; S3: Establish a fault takeover node, and take over the warning fault node through the fault takeover node; In step S3, the failover node maintains a heartbeat mechanism with the video area node to which the video access node belongs; S4: troubleshoot the early warning fault node, and resume recording the fault takeover node by restoring the normal early warning fault node to complete the fault recovery of the railway video surveillance system.

2. A fault recovery method for a railway video monitoring system according to claim 1, characterized in that: The video access node is connected to the video area node via a consistent connection protocol.

3. A fault recovery method for a railway video monitoring system according to claim 2, characterized in that: The connection protocol includes: user management information, authority management information, device resource management information, network management information, real-time status information of video access nodes, load information of video access nodes, control configuration commands of video access nodes, streaming media transmission of video access nodes, and alarm information of video access nodes.

4. A fault recovery method for a railway video monitoring system according to claim 1, characterized in that: In step S3, when the early warning fault node does not exist in the video access node, the fault takeover node runs without load.

5. A fault recovery method for a railway video monitoring system according to claim 1, characterized in that: In step S4, the video access node is connected to the failover node via a consistent transmission protocol.

6. A fault recovery system for a railway video surveillance system, used to execute a fault recovery method for a railway video surveillance system as claimed in any one of claims 1 to 5, characterized in that: include: Collection module: used to collect the operation data of the video access node; Prediction module: used for performing neural learning on the operation data to obtain operation characteristics, and performing fault prediction on the video access node based on the operation characteristics to obtain a warning fault node; Specifically used for marking the fault events of the operation data to obtain the marking data; Extracting abnormal time of fault abnormal values ​​of the operation data based on the marking data to obtain time data; The operation data, the marking data and the time data are combined into a data set, and a neural network is trained through the data set to obtain a prediction network; Performing fault prediction on the video access node through the prediction network to obtain a warning fault node; Takeover module: used to establish a fault takeover node and control the fault takeover node to take over the warning fault node; Recovery module: used to troubleshoot the early warning fault node and control the early warning fault node that has recovered to normal to resume recording the fault takeover node.

Images