Railway power distribution station intelligent operation and maintenance monitoring and analyzing method and system

Abstract

The application belongs to the technical field of railway operation and maintenance monitoring, and provides a railway power distribution station intelligent operation and maintenance monitoring analysis method and system, which solves the problem that the fault root is difficult to accurately identify due to insufficient multi-source data collaborative processing capability. The method comprises the following steps: acquiring communication message data and fiber core attenuation value, calculating response delay sequence and message retransmission times; calculating attenuation gradient based on the fiber core attenuation value to obtain an attenuation gradient spectrum and an attenuation mutation sequence, and obtaining a protocol layer event set and a physical layer event set based on the response delay sequence, the message retransmission times and the attenuation gradient spectrum; obtaining a first deviation degree sequence through correlation analysis, and obtaining a second deviation degree sequence through correlation after time alignment of the attenuation mutation sequence; dividing the second deviation degree sequence into continuous windows, and dividing intervals in each window; determining the window influence degree based on the response delay and the retransmission times in the interval, determining the fault type and the channel score, and realizing communication link fault positioning and health assessment.

Description

Technical Field

[0001] This application belongs to the technical field of railway operation and maintenance monitoring, and in particular relates to an intelligent operation and maintenance monitoring and analysis method and system for railway substations. Background Technology

[0002] As a key node in the railway power supply system, the operational reliability of railway substations directly affects railway transportation safety. With the advancement of railway informatization, intelligent operation and maintenance monitoring and analysis technology for substations has received increasing attention. By integrating multiple monitoring methods, real-time perception and health assessment of equipment operating status can be achieved, providing technical support for the transformation of railway power supply systems from traditional planned maintenance to condition-based maintenance.

[0003] Currently, railway power distribution substation operation and maintenance monitoring typically uses a layered data acquisition method to obtain operational data. One type of data is collected on the response time and retransmission records during the communication interaction process of the equipment, while the other type is collected on the characteristic parameters of the transmission medium. However, at the data processing level, existing technologies usually send these two types of data into independent processing flows, generate analysis results from each, and then simply overlay them. This makes it difficult to determine whether the abnormality is due to the degradation of the transmission medium or the failure of the terminal equipment by means of the inherent correlation between the data when communication indicators are abnormal. Consequently, the correlation information hidden in the data processing results is lost, and the location of the root cause of the fault needs to rely on human experience for secondary analysis. Summary of the Invention

[0004] The purpose of this application is to provide a method and system for intelligent operation and maintenance monitoring and analysis of railway substations, so as to solve the problem that the insufficient multi-source data collaborative processing capability in the existing technology makes it difficult to accurately identify the root cause of the fault.

[0005] To address the aforementioned technical problems, firstly, this application provides an intelligent operation and maintenance monitoring and analysis method for railway substations, comprising: Obtain communication message data from the power distribution station and the fiber core attenuation value of the communication optical fiber. The communication message data includes the response timestamp and message retransmission flag. Calculate the response delay sequence between adjacent instructions based on the response timestamp, count the number of message retransmissions in each preset statistical window based on the message retransmission flag, and calculate the attenuation gradient based on the filtered fiber core attenuation value to obtain the attenuation gradient map and attenuation mutation sequence. The response delay sequence, message retransmission count, and attenuation gradient map are compared with the response window and retransmission threshold determined according to the preset communication protocol and the preset attenuation gradient baseline to obtain the protocol layer event set and the physical layer event set. Correlation analysis is performed based on the protocol layer event set and the physical layer event set to obtain the first deviation sequence. After aligning the first deviation sequence with the decay mutation sequence in time, an association analysis was performed to obtain a second deviation sequence including mutation point markers; The second deviation sequence is divided into multiple continuous windows along the time axis. Within each window, multiple intervals are divided according to the attenuation gradient. The window impact is determined based on the response delay and message retransmission count within each interval. The fault type is determined based on the correlation between the frequency of abrupt change point markers and the attenuation gradient within each window. The channel score is calculated based on the attenuation gradient and window impact of each window.

[0006] Optionally, the response delay sequence between adjacent commands is calculated based on the response timestamp, the number of message retransmissions within each preset statistical window is counted based on the message retransmission flag, and the attenuation gradient is calculated based on the filtered fiber core attenuation value to obtain the attenuation gradient map and attenuation abrupt change sequence, including: Arrange the response timestamps in chronological order, calculate the time difference between two adjacent response timestamps, and arrange the time differences in chronological order to obtain the response delay sequence; The message retransmission flag is divided according to the statistical window, and the message retransmission flag is accumulated and counted in each statistical window to obtain the message retransmission count. A preset sliding window is slid along the fiber length direction, and the average value of the fiber core attenuation value within the sliding window is repeatedly calculated to obtain a filtered attenuation sequence distributed along the fiber length direction. Calculate the difference between adjacent attenuation values ​​in the filtered attenuation sequence, divide the difference by the distance between the corresponding adjacent attenuation values ​​to obtain the attenuation change rate, and arrange the attenuation change rates according to the fiber length direction to obtain the attenuation gradient map. The positions in the attenuation gradient map where the attenuation rate exceeds a preset threshold are arranged along the fiber length to form an attenuation mutation sequence.

[0007] Optionally, the response delay sequence, message retransmission count, and attenuation gradient map are compared with the response window and retransmission threshold determined according to a preset communication protocol, as well as a preset attenuation gradient baseline, to obtain a protocol layer event set and a physical layer event set, including: When the response delay in the response delay sequence is greater than the upper limit of the response window, the time point corresponding to the response delay is determined as the timeout moment, and all timeout moments are arranged in chronological order to obtain the timeout event sequence; When the retransmission count value of each statistical window in the message retransmission count is greater than the retransmission threshold, the end time of the statistical window is determined as the retransmission abnormal time, and all retransmission abnormal times are arranged in chronological order to obtain the retransmission event sequence. The timeout event sequence and the retransmission event sequence are merged in chronological order to obtain the protocol layer event set. Each event in the protocol layer event set includes the occurrence time and the event type, which is either timeout or retransmission exception. When the attenuation change rate of the attenuation gradient map is greater than the attenuation gradient baseline, the fiber position corresponding to the attenuation change rate is determined as the attenuation anomaly position, and all attenuation anomaly positions are arranged in order of fiber length to obtain the physical layer event set.

[0008] Optionally, a correlation analysis is performed based on the protocol layer event set and the physical layer event set to obtain the first deviation sequence, including: Arrange the occurrence times of each timeout event and each retransmission anomaly event in the protocol layer event set in chronological order to obtain the protocol layer time point sequence. Arrange the sampling times corresponding to each attenuation anomaly position in the physical layer event set in chronological order to obtain the physical layer time point sequence. The first count value is obtained by counting the number of moments in the protocol layer time point sequence within a consecutive preset time segment on the time axis. The second count value is obtained by counting the number of moments in the physical layer time point sequence within a consecutive preset time segment. The first and second count values ​​corresponding to each time segment are added together to obtain the deviation value of each time segment. All deviation values ​​are then arranged in the order of the time segments to obtain the first deviation sequence.

[0009] Secondly, this application provides an intelligent operation and maintenance monitoring and analysis system for railway substations, comprising: The acquisition module is used to acquire the communication message data of the power distribution station and the fiber core attenuation value of the communication optical fiber. The communication message data includes the response timestamp and the message retransmission flag. The calculation module is used to calculate the response delay sequence between adjacent instructions based on the response timestamp, count the number of message retransmissions in each preset statistical window based on the message retransmission flag, and calculate the attenuation gradient based on the filtered fiber core attenuation value to obtain the attenuation gradient map and attenuation mutation sequence. The comparison module is used to compare the response delay sequence, message retransmission count, and attenuation gradient map with the response window and retransmission threshold determined according to the preset communication protocol and the preset attenuation gradient baseline, respectively, to obtain the protocol layer event set and the physical layer event set. Based on the protocol layer event set and the physical layer event set, correlation analysis is performed to obtain the first deviation sequence. The analysis module is used to perform correlation analysis after aligning the first deviation sequence with the decay mutation sequence in time, so as to obtain a second deviation sequence including mutation point markers. The determination module is used to divide the second deviation sequence into multiple continuous windows along the time axis, divide multiple intervals within each window according to the attenuation gradient, determine the window impact degree based on the response delay and message retransmission count within each interval, determine the fault type based on the correlation between the frequency of abrupt change point markers and the attenuation gradient within each window, and calculate the channel score based on the attenuation gradient and window impact degree of each window.

[0010] Thirdly, this application provides an electronic device, comprising: Memory, used to store computer programs; A processor is used to execute computer programs to implement the steps of an intelligent operation and maintenance monitoring and analysis method for railway substations as described in the first aspect above.

[0011] Fourthly, this application provides a computer-readable storage medium storing a computer program, which, when executed by a processor, can implement the steps of the intelligent operation and maintenance monitoring and analysis method for railway substations as described in the first aspect above.

[0012] The intelligent operation and maintenance monitoring and analysis method for railway substations proposed in this application has the following beneficial effects: This application first acquires data to lay the foundation for analysis; then it extracts response delay, retransmission count, attenuation gradient, and mutation location to form quantitative features; next, it uses threshold comparison to locate abnormal events and correlates the degree of deviation between protocol layer and physical layer events; then, it performs time alignment and marks physical mutation correlations, divides windows and partitions them according to attenuation gradient, and finally determines the impact by combining response delay and retransmission count, distinguishes fault types based on the correlation between mutation marking frequency and attenuation gradient, and calculates channel scores by fusing attenuation gradient and impact, ultimately achieving multi-source data collaborative processing to identify the root cause of the fault.

[0013] Furthermore, this application obtains a response delay sequence by subtracting the response timestamps sequentially; then, it accumulates the message retransmission flags according to a statistical window to obtain the message retransmission count; next, it slides a window along the fiber length and takes the average to obtain a filter attenuation sequence; then, it calculates the ratio of the difference between adjacent filter attenuation values ​​to the distance to obtain the attenuation change rate and arranges them into an attenuation gradient map; finally, it extracts the positions where the attenuation change rate exceeds a preset threshold to form an attenuation mutation sequence. This application can transform raw data into multi-dimensional features reflecting communication response, retransmission behavior, and spatial changes in fiber attenuation. Attached Figure Description

[0014] To more clearly illustrate the technical solutions of the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 A flowchart illustrating an intelligent operation and maintenance monitoring and analysis method for railway substations provided in this application embodiment; Figure 2 A flowchart illustrating a method for generating channel scores provided in an embodiment of this application; Figure 3 A schematic diagram of an attenuation gradient spectrum provided in an embodiment of this application; Figure 4 A schematic diagram illustrating the alignment of protocol layer and physical layer events and the generation of a first deviation sequence, provided in an embodiment of this application; Figure 5 A schematic diagram illustrating a continuous window and gradient interval division provided in an embodiment of this application; Figure 6 A schematic diagram of the structure of an intelligent operation and maintenance monitoring and analysis system for railway substations provided in this application embodiment; Figure 7 This is a schematic diagram of the hardware structure of an electronic device provided in one embodiment of this application. Detailed Implementation

[0016] To address the problem that existing methods process communication behavior analysis and physical layer state awareness independently, making it difficult to determine the root cause of anomalies, this application proposes an intelligent operation and maintenance monitoring and analysis method for railway substations. This method integrates communication interaction data and transmission medium state data into a unified analysis framework. By extracting response features, retransmission features, and attenuation change features of the transmission medium from both types of data, it establishes correlations between multi-dimensional features and distinguishes the root cause of anomalies based on the degree of correlation.

[0017] Specifically, this application extracts response delay and retransmission count based on communication interaction data, and then extracts attenuation gradient and abrupt change location based on transmission medium status data. The processing results of the two types of data are time-aligned and correlated. Then, by statistically analyzing response delay and retransmission behavior in different attenuation change intervals, it identifies whether the anomaly is dominated by transmission medium degradation or terminal equipment failure. Thus, it can accurately distinguish the root cause of the fault without human experience intervention and achieve a quantitative assessment of the health status of the communication link.

[0018] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are merely some embodiments of the present application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] To address the problems of existing technologies, embodiments of this application provide a method and system for intelligent operation and maintenance monitoring and analysis of railway substations, a computer storage medium, and a computer program product. The method for intelligent operation and maintenance monitoring and analysis of railway substations provided in this application embodiment will be described below first.

[0020] Figure 1This illustration shows a flowchart of an intelligent operation and maintenance monitoring and analysis method for railway substations according to an embodiment of this application. Figure 1 As shown.

[0021] S101. Obtain the communication message data of the power distribution station and the fiber core attenuation value of the communication optical fiber. The communication message data includes the response timestamp and message retransmission flag. In this sub-step, communication message data refers to the instruction data packets encapsulated and transmitted between the dispatching terminal and the terminal equipment within the substation according to a specific communication protocol. The response timestamp is the recorded time elapsed from when the dispatching terminal issues an instruction to when it receives a response from the terminal equipment within the substation; this time record is used to quantify the response speed of the communication link. The message retransmission flag is a repeat transmission marker automatically recorded by the system when instruction transmission fails; this marker reflects the stability of the communication link. The fiber core attenuation value is a measure of energy loss when an optical signal is transmitted through an optical fiber medium; this value is distributed along the length of the optical fiber and is used to characterize the physical transmission state of the optical fiber medium.

[0022] In this embodiment, a data capture point is set up on the network link between the dispatching terminal and the terminal equipment in the substation. All instruction interaction data passing through this capture point is mirrored in real time. The mirrored data is then parsed according to the frame structure of the communication protocol to extract the response timestamp and message retransmission flag from each instruction interaction cycle. For example, in a certain capture, the response timestamp of 120 milliseconds and the message retransmission flag of 0 are extracted from the response frame of the first instruction; the response timestamp of 350 milliseconds and the message retransmission flag of 1 are extracted from the response frame of the second instruction; and the response timestamp of 85 milliseconds and the message retransmission flag of 0 are extracted from the response frame of the third instruction.

[0023] Simultaneously, a monitoring device connected to one end of the optical fiber transmits light pulses into the fiber and receives the light signals reflected from various points along the fiber. The time difference between the time of light pulse transmission and the time of receiving the reflected light signal is recorded. The distance between the reflection point and the monitoring device is determined by multiplying the time difference by the speed of light in the optical fiber. The intensity values ​​of the transmitted light pulse and the received reflected light signal are also recorded. For each sampling point, the intensity values ​​of the transmitted light pulse and the reflected light signal are substituted into the formula: Calculate the fiber core attenuation value, where, This represents the fiber core attenuation value at the sampling point, and l represents the distance between the sampling point and the monitoring device. This indicates the intensity value of the emitted light pulse. This indicates the intensity value of the received reflected light signal.

[0024] This application, by acquiring communication message data and fiber core attenuation values, can lay a data foundation for subsequent analysis, thereby ensuring the collection of key information required for operation and maintenance monitoring from both the application layer and the physical layer.

[0025] S102. Calculate the response delay sequence between adjacent instructions based on the response timestamp, count the number of message retransmissions in each preset statistical window based on the message retransmission flag, and calculate the attenuation gradient based on the filtered fiber core attenuation value to obtain the attenuation gradient map and attenuation mutation sequence. In one specific implementation, step S102 includes: Step 1021: Arrange the response timestamps in chronological order, calculate the time difference between two adjacent response timestamps, and arrange the time differences in chronological order to obtain the response delay sequence; In this sub-step, the response delay sequence refers to a set of time difference data obtained by subtracting the response timestamp of each instruction from the timestamp of the corresponding instruction issuance time. Each value in the sequence represents the complete round-trip delay of a single instruction interaction from issuance to feedback, which is used to accurately reflect the application layer response performance of the communication link.

[0026] In this embodiment, all response timestamps extracted in step S101 are arranged in chronological order to form a response timestamp sequence. Then, starting from the first response timestamp in the sequence, the difference between each subsequent response timestamp and the preceding response timestamp is calculated sequentially. Each calculated difference is then arranged in chronological order to obtain the final response delay sequence. For example, two differences are calculated from three response timestamps; the first difference is... milliseconds, the second difference is milliseconds, then and Arranged in order, they form the response delay sequence.

[0027] Step 1022: Divide the message retransmission flags according to the statistical window, and accumulate the message retransmission flag count in each statistical window to obtain the message retransmission count; In this sub-step, the message retransmission count refers to the cumulative count obtained by adding the message retransmission flags within a preset statistical window. This count is used to quantify the frequency of retransmissions occurring on the communication link within a specific time period. The preset statistical window refers to a fixed-length time segment divided on the time axis, with adjacent windows being consecutive and non-overlapping. For example, a statistical window of 5 seconds can be used, and the entire monitoring period can be divided into multiple consecutive time segments. The first statistical window covers seconds 0 to 5, the second statistical window covers seconds 5 to 10, and so on.

[0028] In this embodiment, all message retransmission flags are arranged in chronological order, and consecutive preset statistical windows are set on the timeline, with each preset statistical window having a fixed time length and adjacent windows being contiguous. Then, each message retransmission flag is assigned to its corresponding preset statistical window based on its occurrence time, and the number of message retransmissions within each preset statistical window is accumulated to obtain the message retransmission count for that window. For example, the number of retransmissions accumulated within the first preset statistical window is... The second window accumulates the results. The sum is obtained in the third window. These counts are then arranged in window order to form a sequence of message retransmission counts.

[0029] Step 1023: Slide a preset sliding window along the fiber length direction and repeatedly calculate the average value of the fiber core attenuation value within the sliding window to obtain the filtered attenuation sequence distributed along the fiber length direction. In this sub-step, the filtered attenuation sequence refers to a new numerical sequence obtained by performing a moving average on the original fiber core attenuation value along the fiber length. Each value in this sequence represents the attenuation level in a region near a certain location on the fiber, used to filter out local random fluctuations and retain the overall trend of attenuation change. The preset sliding window refers to a fixed-length interval set along the fiber length, which is used to limit the range of fiber core attenuation values ​​participating in the averaging calculation each time.

[0030] In this embodiment, the fiber core attenuation values ​​are arranged according to the fiber length direction to form a fiber core attenuation value sequence. Then, a preset sliding window with a fixed length is set on the fiber core attenuation value sequence. The window is placed at the beginning position of the fiber core attenuation value sequence. Then, the arithmetic mean of all fiber core attenuation values ​​in the window is calculated. When the window is placed at the beginning or end of the sequence and the window coverage exceeds the actual length of the sequence, the missing values ​​in the window are filled by a symmetrical filling method. That is, the adjacent sampled values ​​are mirrored and copied with the boundary value axis, and the calculation result is used as the filtered value corresponding to the center position of the window.

[0031] Next, the preset sliding window is moved step by step along the fiber length direction, and each time it moves by a fixed step distance. For example, the arithmetic mean of all fiber core attenuation values ​​in the window is calculated repeatedly after each movement until the window covers the end of the entire fiber core attenuation value sequence. Then, all the calculated average values ​​are arranged in order according to the fiber length direction to form a filtered attenuation sequence. The fixed step distance can be set according to the sampling interval of the fiber monitoring equipment and the required data smoothness.

[0032] For example, if an optical fiber monitoring device collects a fiber core attenuation value every 1 meter, then the distance between adjacent sampling points is 1 meter. The length of a preset sliding window is then set to 5 meters, meaning the window contains 5 consecutive fiber core attenuation value sampling points. Next, the starting position of the preset sliding window is placed at the first sampling point along the fiber's length, and the average of the 5 fiber core attenuation values ​​within the window is calculated to obtain the first filtered value. Then, the window is moved a fixed step distance of 1 meter along the length of the fiber, so that the window covers the second to sixth sampling points, and the average value of these 5 fiber core attenuation values ​​is calculated to obtain the second filtered value. This process continues, moving 1 meter at a time and repeatedly calculating the average fiber core attenuation value within the window until the window covers the last sampling point.

[0033] Step 1024: Calculate the difference between adjacent attenuation values ​​in the filter attenuation sequence, divide the difference by the distance between the corresponding adjacent attenuation values ​​to obtain the attenuation change rate, and arrange the attenuation change rates according to the fiber length direction to obtain the attenuation gradient map. In this sub-step, the attenuation gradient map refers to a data set formed by sequentially arranging the attenuation change rates at various locations along the fiber length. Each attenuation change rate represents the degree of change between two adjacent values ​​in the filtered attenuation sequence, characterizing how quickly the fiber attenuation changes along its length. The attenuation change rate is the ratio obtained by dividing the difference between two adjacent attenuation values ​​in the filtered attenuation sequence by the distance between the corresponding locations of these two attenuation values.

[0034] In this embodiment, two adjacent values ​​are sequentially extracted from the filter attenuation sequence, and the first value is set to correspond to the fiber position. The second value corresponds to the fiber optic location. Then calculate the difference between the second and first values, and divide the difference by 1 / 2. and The attenuation change rate of the interval is obtained by calculating the distance between the two points. Following this method, the attenuation change rate between all adjacent positions is calculated sequentially along the fiber length. These calculated attenuation change rates are then arranged in order along the fiber length to form an attenuation gradient map. For example, the first attenuation change rate calculated from the filtered attenuation sequence is... The second one is The third one is Then, these values ​​are arranged in order to form the attenuation gradient map.

[0035] Step 1025: Arrange the positions in the attenuation gradient map where the attenuation change rate exceeds the preset change rate threshold according to the fiber length direction to form an attenuation mutation sequence.

[0036] In this sub-step, the attenuation mutation sequence refers to a data set formed by arranging the location information of the attenuation change rate exceeding the preset change rate threshold extracted from the attenuation gradient map in the order of the fiber length direction. Each position in this sequence represents a node on the fiber where the attenuation changes drastically.

[0037] In this embodiment, a threshold value for the rate of change is preset. This threshold value is used to define the criteria for determining a drastic change in attenuation. For example, in a conventional single-mode fiber monitoring scenario, this threshold value is set to 0.15 dB / km. Then, each attenuation rate of change in the attenuation gradient map obtained in step 1024 is compared with this threshold value. When the value of a certain attenuation rate of change is greater than the threshold value, the fiber position corresponding to that attenuation rate of change is extracted and used as an element in the attenuation mutation sequence. Then, all attenuation rates of change are compared sequentially along the fiber length direction. The positions corresponding to all attenuation rates of change that meet the conditions, along with the start and end times of the full-length scan performed by the fiber monitoring device, are used as time stamps and arranged sequentially along the fiber length direction to form the attenuation mutation sequence. For example, in the attenuation gradient map, the position located along the fiber length direction... The rate of decay change at a location exceeds the threshold. If the rate of decay change at a certain location also exceeds the threshold, then these locations are arranged in order to obtain a decay mutation sequence.

[0038] For details, please refer to Figure 3 , Figure 3 The horizontal axis represents the position along the fiber length direction, and the vertical axis represents the attenuation change rate. Curve 301 is the attenuation gradient curve formed by the attenuation change rates arranged along the fiber length direction, which is the attenuation gradient spectrum obtained in step 1024; the dashed line 302 is the preset attenuation gradient baseline; and the dot 303 is the position of the attenuation change rate exceeding the preset change rate threshold.

[0039] And by Figure 3 As can be seen, the attenuation gradient curve 301 can intuitively reflect the rate of change of fiber attenuation along the length direction. When the attenuation gradient curve 301 shows a significant peak at a certain position and the attenuation change rate corresponding to the peak exceeds the preset change rate threshold, the position is extracted as the attenuation mutation position 303. All attenuation mutation positions 303 arranged along the fiber length direction constitute the attenuation mutation sequence in step 1025. The reason for using a graph to represent the attenuation change rate is that a single attenuation value is difficult to reflect the degradation trend of attenuation along the space, while the attenuation gradient graph can amplify local degradation into a peak that can be identified by the threshold, thereby providing a basis for the subsequent extraction of physical layer events.

[0040] This application enables multi-dimensional feature extraction of communication behavior and optical fiber status by transforming raw data into response delay sequences, message retransmission counts, attenuation gradient maps, and attenuation mutation sequences.

[0041] S103. The response delay sequence, message retransmission count and attenuation gradient map are compared with the response window and retransmission threshold determined according to the preset communication protocol and the preset attenuation gradient baseline to obtain the protocol layer event set and the physical layer event set. Correlation analysis is performed based on the protocol layer event set and the physical layer event set to obtain the first deviation sequence. In one specific implementation, step S103 includes: Step 1031: When the response delay in the response delay sequence is greater than the upper limit of the response window, the time point corresponding to the response delay is determined as the timeout moment, and all timeout moments are arranged in chronological order to obtain the timeout event sequence. In this embodiment, the preset communication protocol specifies the maximum allowed time range for a normal response. The upper limit of this range is the upper limit of the response window. For example, when using the IEC 60870-5-104 protocol, this protocol specifies that the maximum waiting time for the terminal within the facility to respond after the dispatcher issues an instruction is 200 milliseconds. Then, each response delay value is sequentially extracted from the response delay sequence and compared with the upper limit of the response window. When the response delay value is greater than the upper limit of the response window, the time point corresponding to the response delay is recorded and determined as the timeout moment. All marked timeout moments are arranged in chronological order to form a timeout event sequence. For example, extracting from the response delay sequence... , , Equal values, among which, and If the corresponding response delay is greater than the upper limit of the response window, then record it. and corresponding time point and The final timeout event sequence is obtained. , ].

[0042] Step 1032: When the retransmission count value of each statistical window in the message retransmission count is greater than the retransmission threshold, the end time of the statistical window is determined as the retransmission abnormal time, and all retransmission abnormal times are arranged in chronological order to obtain the retransmission event sequence. In this embodiment, a retransmission threshold is preset, such as 10 times. Then, the retransmission count value of each preset statistical window is sequentially extracted from the message retransmission count of each window, and the count value is compared with the retransmission threshold. When the retransmission count value is greater than the retransmission threshold, the end time of the preset statistical window is recorded, and the end time is determined as the retransmission abnormal time. Then, all the marked retransmission abnormal times are arranged in chronological order to form a retransmission event sequence. For example, in three preset statistical windows, the retransmission count value of the first window is... The second window is greater than the retransmission threshold. The third window is greater than the retransmission threshold. If the value is less than the retransmission threshold, then record the end times of the first and second windows. and The retransmission event sequence is obtained as [ , ].

[0043] Step 1033: Merge the timeout event sequence and the retransmission event sequence in chronological order to obtain the protocol layer event set. Each event in the protocol layer event set includes the occurrence time and event type, and the event type is timeout or retransmission exception. In this sub-step, the protocol layer event set refers to the event set formed by merging the timeout event sequence and the retransmission event sequence. Each event in this set includes the occurrence time and event type, which are used to uniformly characterize abnormal situations in application layer communication behavior.

[0044] In this embodiment, the timeout event sequence and the retransmission event sequence are merged according to the chronological order of their occurrence, while maintaining the original temporal order during merging. Events occurring at the same time can be arranged in any order. Each event, in addition to recording its occurrence time, also needs to be accompanied by an event type identifier, where the event type for timeout events is "timeout" and the event type for retransmission exception events is "retransmission exception". The resulting set of merged events is the protocol layer event set. For example, the timeout event sequence contains timestamps... and The retransmission event sequence includes time. After merging them in chronological order, the protocol layer event set is obtained as [( ,time out),( (retransmission) ,time out)].

[0045] Step 1034: When the attenuation change rate of the attenuation gradient map is greater than the attenuation gradient baseline, the fiber position corresponding to the attenuation change rate is determined as the attenuation anomaly position, and all attenuation anomaly positions are arranged in order of fiber length to obtain the physical layer event set.

[0046] In this embodiment, a preset attenuation gradient baseline, such as 0.05 dB / km, is used to determine whether the attenuation change rate exceeds the normal range. Then, each attenuation change rate is sequentially extracted from the attenuation gradient map and compared with the attenuation gradient baseline. When the attenuation change rate is greater than the attenuation gradient baseline, the fiber position corresponding to that attenuation change rate is recorded, and this position is identified as an attenuation anomaly location. All marked attenuation anomaly locations are then arranged sequentially along the fiber length to form a physical layer event set. For example, extracting from the attenuation gradient map... , , Equal values, among which, and If the attenuation gradient is greater than 0.05 dB / km from the baseline, then record. Corresponding position and Corresponding position The physical layer event set is obtained as , .

[0047] Step 1035: Arrange the occurrence times of each timeout event and each retransmission anomaly event in the protocol layer event set in chronological order to obtain the protocol layer time point sequence; arrange the sampling times corresponding to each attenuation anomaly position in the physical layer event set in chronological order to obtain the physical layer time point sequence. In this embodiment of the application, the occurrence time of each event is extracted sequentially from the protocol layer event set, and these times are arranged in chronological order to form a protocol layer time point sequence; then, the sampling time corresponding to each attenuation anomaly location is extracted sequentially from the physical layer event set, wherein the sampling time refers to the specific time point when the physical layer scan covers the specific location, which is determined by the scan start time of the fiber optic monitoring device plus the propagation time of the scanning beam to the location, and these times are arranged in chronological order to form a physical layer time point sequence.

[0048] Step 1036: Count the number of moments in the protocol layer time point sequence within a continuous preset time segment on the time axis to obtain the first count value; Count the number of moments in the physical layer time point sequence within a continuous preset time segment to obtain the second count value. In this sub-step, a preset time segment refers to a fixed-length time interval divided on the time axis, where adjacent segments are contiguous and do not overlap. The first count value refers to the number of time points in the protocol layer time point sequence falling within that segment within a single preset time segment. The second count value refers to the number of time points in the physical layer time point sequence falling within that segment within a single preset time segment.

[0049] In this embodiment, continuous preset time segments are set on the timeline, and each preset time segment has a fixed duration, for example, each segment is set to 5 seconds. The first segment covers 0 to 5 seconds, the second segment covers 5 to 10 seconds, and so on. For each preset time segment, the number of time points in the protocol layer time point sequence that fall into the segment is counted to obtain a first count value; the number of time points in the physical layer time point sequence that fall into the segment is counted to obtain a second count value. For example, in the first time segment, if 2 points in the protocol layer time point sequence fall into the segment, the first count value is 2; if 1 point in the physical layer time point sequence falls into the segment, the second count value is 1.

[0050] Step 1037: Add the first count value and the second count value corresponding to each time segment to obtain the deviation value of each time segment, and arrange all the deviation values ​​in the order of the time segments to obtain the first deviation sequence.

[0051] In this sub-step, the deviation value refers to the sum of the first and second count values ​​corresponding to each preset time segment. This value is used to quantify the degree of combination between protocol layer anomalies and physical layer anomalies within that time segment. The first deviation sequence refers to the data sequence formed by arranging the deviation values ​​of all preset time segments in the order of the time segments.

[0052] In this embodiment, for each preset time segment, the first count value and the second count value are added together to obtain the deviation value of that time segment; then, the deviation values ​​of all time segments are arranged in chronological order to form a first deviation sequence. For example, if the first count value of the first time segment is 2 and the second count value is 1, the deviation value is 3; if the first count value of the second time segment is 0 and the second count value is 0, the deviation value is 0; if the first count value of the third time segment is 1 and the second count value is 2, the deviation value is 3; then, these values ​​are arranged in order to obtain the first deviation sequence 3, 0, 3.

[0053] For details, please refer to Figure 4 , Figure 4 Two types of events are aligned along the same timeline. Sequence 401, composed of solid dots, represents the protocol layer time point sequence of step 1035, while sequence 402, composed of hollow squares, represents the physical layer time point sequence of step 1035. The timeline is divided into multiple pre-defined time segments 403 connected end-to-end by vertical dashed lines. For each pre-defined time segment 403, the number of protocol layer time points falling within that segment is counted to obtain a first count value, and the number of physical layer time points falling within that segment is counted to obtain a second count value. The first count value and the second count value are added together to obtain the deviation value of that segment. The deviation values ​​of each segment are arranged in chronological order to form the [missing information - likely a timeline or data structure]. Figure 4The first deviation sequence 404 is represented by the rectangular bar below.

[0054] And by Figure 4 It can be seen that projecting protocol layer events and physical layer events onto a unified time axis and counting them by segments can aggregate abnormal information scattered in different processing flows into a single deviation value in the time dimension, thereby providing a unified time reference for subsequent correlation analysis with decay mutation sequences.

[0055] This application establishes a quantitative correlation between abnormal events in the protocol layer and the physical layer through threshold comparison and event statistics, and generates a first deviation sequence, which enables a numerical expression of the degree of abnormality.

[0056] S104. After aligning the first deviation sequence with the decay mutation sequence in time, perform correlation analysis to obtain the second deviation sequence including mutation point markers. In one specific implementation, step S104 includes: Step 1041: Determine the start and end times corresponding to each deviation value in the first deviation sequence as a time interval; In this embodiment, each deviation value is sequentially extracted from the first deviation sequence, and each deviation value corresponds to a preset time segment; then, the start and end times of the preset time segment are extracted, and a continuous time interval is determined using the start and end times as endpoints. For example, for the first deviation value in the first deviation sequence... The corresponding preset time segment start time is The end time is The determined time interval is from arrive The range.

[0057] Step 1042: Compare the sampling time of each decay mutation position in the decay mutation sequence with the time interval. When the sampling time exceeds the time interval, bind the deviation value with the mutation marker. When the sampling time does not exceed the time interval, bind the deviation value with the no mutation marker. In this sub-step, the mutation marker refers to additional information used to identify whether there is a decay mutation within the time interval corresponding to the deviation value. The mutation marker is bound when there is a decay mutation within the time interval, otherwise the no mutation marker is bound.

[0058] In this embodiment, each decay mutation position is sequentially extracted from the decay mutation sequence, and each decay mutation position corresponds to a sampling time. Then, each sampling time is compared with the time interval determined in step 1041, and it is determined whether the sampling time is located between the start and end times of the time interval. When the sampling time is greater than or equal to the start time and less than or equal to the end time, the sampling time is determined to fall within the time interval, and the corresponding deviation value is bound to the mutation marker. When the sampling time is less than the start time or greater than the end time, the sampling time is determined to not fall within the time interval, and the corresponding deviation value is bound to the no-mutation marker.

[0059] Step 1043: Arrange the bound deviation values ​​in chronological order of their starting times to obtain the second deviation sequence.

[0060] In this embodiment, the bound deviation values ​​obtained in step 1042 are arranged in chronological order of their corresponding time intervals, maintaining the original time order during the arrangement. Each element consists of both the deviation value and the bound marker. The data set formed by arranging all elements in order is the second deviation sequence. For example, the first deviation value... The second one was bound to a mutation marker. The third one was bound to a non-mutation marker. With the mutation marker attached, the second deviation sequence can be represented as [( ,mutation),( (No mutation) ,mutation)].

[0061] This application enables precise temporal correlation between application layer deviation and physical layer mutation information by temporally aligning and tagging the first deviation sequence with the decay mutation sequence.

[0062] S105. Divide the second deviation sequence into multiple continuous windows along the time axis. Within each window, divide it into multiple intervals according to the attenuation gradient. Determine the window impact based on the response delay and message retransmission count within each interval. Determine the fault type based on the correlation between the frequency of mutation point markers and the attenuation gradient within each window. Calculate the channel score based on the attenuation gradient and window impact of each window.

[0063] In one specific implementation, such as Figure 2 As shown, step S105 includes: Step 1051: Divide the second deviation sequence into multiple continuous windows according to the time axis, extract the decay gradient values ​​corresponding to each time window from the decay gradient map, and divide the extracted decay gradient values ​​into multiple gradient intervals according to their magnitude. In this sub-step, a continuous window refers to dividing the second deviation sequence into multiple fixed-length intervals that are consecutive and non-overlapping along the time axis. A gradient interval refers to dividing the extracted decay gradient values ​​within each time window into multiple continuous numerical ranges based on their magnitude, with each range corresponding to a gradient interval.

[0064] In this embodiment of the application, a fixed-length continuous time period is extracted sequentially from the second deviation sequence according to the time sequence, and each time period is used as a continuous window. The adjacent windows are connected end to end and do not overlap, resulting in multiple continuous windows. For example, the monitoring period starts from the 0th second, and 10 seconds are used as a continuous window. The first window covers the 0th to the 10th second, the second window covers the 10th to the 20th second, the third window covers the 20th to the 30th second, and so on.

[0065] For each continuous window, all attenuation gradient values ​​corresponding to that window time period are extracted from the attenuation gradient map. The extracted attenuation gradient values ​​are then sorted in ascending order, and the numerical range is divided into several continuous gradient intervals using an equidistant division method. Each gradient interval corresponds to a numerical range. For example, according to the fiber optic attenuation level standard, 0-0.10dB / km is set as the first gradient interval, 0.10dB / km-0.20dB / km is set as the second gradient interval, and >0.20dB / km is set as the third gradient interval.

[0066] For details, please refer to Figure 5 , Figure 5 The second deviation sequence is divided along the time axis into multiple consecutive, non-overlapping windows 501, such as window 1, window 2, and window 3. Within each consecutive window 501, based on the magnitude of the decay gradient extracted from the decay gradient map, it is further divided into multiple gradient intervals 502, such as gradient interval 1, gradient interval 2, and gradient interval 3. Figure 5 The triangle marker in the diagram represents the mutation point marker 503 that falls within the corresponding window.

[0067] And by Figure 5 It can be seen that the dual division method of first dividing the continuous window 501 according to the time axis, and then dividing the gradient interval 502 within each continuous window 501 according to the magnitude of the attenuation gradient, can classify the response delay and message retransmission number within the same time window into different gradient intervals 502 according to the magnitude of the attenuation gradient they are associated with. In this way, the cumulative delay value and cumulative retransmission value of each gradient interval 502 can be counted in subsequent steps, thereby quantifying the degree to which the communication behavior is affected by the attenuation gradient distribution.

[0068] Step 1052: Extract the response delay values ​​associated with the decay gradient values ​​falling into the gradient interval in the response delay sequence, and sum up all the response delay values ​​in each gradient interval to obtain the cumulative delay value for each gradient interval. In this embodiment, for each gradient interval within a continuous window, the response delay value corresponding to the decay gradient value falling within the gradient interval and within the window time period is found from the response delay sequence obtained in step 1021; then, all these response delay values ​​are retrieved and summed to obtain the cumulative delay value for the gradient interval. For example, for gradient interval 1, the associated response delay value is found to be... , , The accumulated delay value is then... + + .

[0069] Step 1053: Extract the retransmission count of the message that falls within the gradient interval and associate it with the attenuation gradient value. Then, sum up all the retransmission counts of the message in each gradient interval to obtain the cumulative retransmission value for each gradient interval. In this embodiment, for each gradient interval within each continuous window, the message retransmission count corresponding to the attenuation gradient value falling within the gradient interval and within the window time period is found from the message retransmission counts obtained in step 1022 for each preset statistical window; then, all these message retransmission counts are extracted and summed to obtain the cumulative retransmission value for the gradient interval. For example, for gradient interval 2, the associated message retransmission count is... , The retransmission accumulation value is then... + .

[0070] Step 1054: Perform a weighted sum of all accumulated delay values ​​and accumulated retransmission values ​​within each time window to obtain the window impact of each time window.

[0071] In this sub-step, the window influence degree is used to comprehensively reflect the overall degree to which communication behavior within the window is affected by the attenuation gradient distribution.

[0072] In this embodiment, weighting coefficients are set for the accumulated delay value and the accumulated retransmission value for each gradient interval. These weighting coefficients can be preset according to the actual scenario. Since the units of the accumulated delay value and the accumulated retransmission value are different, this application sets the weighting coefficients... Set as the normalization coefficient for dimensionless processing of the accumulated delay value, i.e. Take it as the reciprocal of the reference delay. For example, if the reference delay is 100 milliseconds, then... =0.01 / millisecond; and the weighting coefficient Take it as a dimensionless number, such as =1, thus ensuring that the window influence W is consistent in terms of dimensions; In a specific example, it can be set =0.01 / millisecond =1, so that the contributions of delay and retransmission factors to the window's influence are on the same order of magnitude; for each continuous window, the cumulative delay and retransmission values ​​of all gradient intervals within the window are weighted and summed according to their respective weights, and the calculation formula is as follows: ,in, Let m represent the window effect, and m be the number of gradient intervals. Let be the accumulated delay value for the i-th gradient interval. For its weight, This represents the accumulated retransmission value for the i-th gradient interval. Assign weights to them.

[0073] Step 1055: Arrange the mutation point markers in each time window of the second deviation sequence in chronological order and count them one by one to obtain the frequency of mutation points in each time window; In this sub-step, the mutation point frequency is used to characterize the frequency of physical layer decay mutation events within the window.

[0074] In this embodiment, all elements belonging to the continuous window range are extracted from the second deviation sequence; these elements are then arranged in chronological order, and each element is traversed to check whether it carries a mutation marker; the elements with mutation markers are then cumulatively counted, and the resulting count value is the mutation frequency of the continuous window. For example, if there are five elements in a window, and three of them carry mutation markers, then the mutation frequency is 3.

[0075] Step 1056: Calculate the average value of the decay gradient within each time window in the decay gradient map to obtain the average decay gradient for each time window. In this embodiment of the application, for each continuous window, all attenuation gradient values ​​corresponding to the time period of the window are extracted from the attenuation gradient map obtained in step 1024, and then these values ​​are summed and divided by the number of values ​​to obtain the average attenuation gradient of the window. The calculation formula is as follows: ,in, Let be the average decay gradient, and n be the number of decay gradient values ​​within the window. Let be the value of the j-th decay gradient.

[0076] Step 1057: Determine the correlation degree of each time window based on the product of the frequency of mutation points and the average attenuation gradient. When the correlation degree is greater than the preset correlation threshold, the fault type of the corresponding time window is determined as a physical channel fault. When the correlation degree is less than or equal to the preset correlation threshold, the fault type of the corresponding time window is determined as a device fault. In this sub-step, the correlation degree refers to the product of the frequency of abrupt change points and the average attenuation gradient, used to quantify the strength of the correlation between physical layer attenuation abrupt changes and abnormal communication behavior. The preset correlation threshold is a pre-set value used to determine whether the correlation degree meets the criteria for physical channel failure.

[0077] In this embodiment of the application, for each consecutive window, the frequency F of the abrupt change point is compared with the average decay gradient. Multiply by the product to obtain the relevance V of the window, then multiply this relevance by a preset relevance threshold. Its preset association threshold It is derived from the distribution pattern of the frequency of physical fault occurrences and the gradient product in historical operation and maintenance data. For example, it is set to 0.1 in a conventional single-mode fiber environment. If V > If V ≤ If so, the fault type of the window is determined to be a fault of the device itself.

[0078] Step 1058: When the fault type is a physical channel fault, extract all the attenuation mutation positions that fall within the corresponding time window from the attenuation mutation sequence and arrange them according to the fiber length direction to obtain the fault location information. In this embodiment of the application, when step 1057 determines that the fault type of a certain continuous window is a physical channel fault, all attenuation change positions located within the window time period are extracted from the attenuation change sequence, and these positions are arranged in ascending order according to the fiber length direction to form fault location information.

[0079] Step 1059: Determine the window score for each time window based on the product of the average decay gradient and the window influence, and calculate the average of all window scores to obtain the channel score.

[0080] In this sub-step, the window score is the value obtained by multiplying the average decay gradient of each consecutive window by the window's influence, used to quantify the overall health of that window. The channel score is the average of all window scores, used to characterize the overall health level of the communication link throughout the entire monitoring period.

[0081] In this embodiment of the application, for each consecutive window, the average decay gradient is calculated. Multiplying the window's influence W by the window score gives the window score G. Then, summing the window scores G of all consecutive windows and dividing by the total number of windows gives the channel score. ,in, Rate the q-th window.

[0082] This application distinguishes fault types based on the correlation between the frequency of mutations within a window and the decay gradient, and calculates the window impact and channel score, which enables fault root cause localization and quantitative assessment of health status.

[0083] Figure 6 This is a schematic diagram illustrating a specific implementation of an intelligent operation and maintenance monitoring and analysis system for railway substations provided in this application. (Refer to...) Figure 6 The system may include: The acquisition module 61 is used to acquire the communication message data of the substation and the fiber core attenuation value of the communication optical fiber. The communication message data includes the response timestamp and the message retransmission flag. The calculation module 62 is used to calculate the response delay sequence between adjacent instructions based on the response timestamp, count the number of message retransmissions in each preset statistical window based on the message retransmission flag, and calculate the attenuation gradient based on the filtered fiber core attenuation value to obtain the attenuation gradient map and attenuation mutation sequence. The comparison module 63 is used to compare the response delay sequence, message retransmission count and attenuation gradient spectrum with the response window and retransmission threshold determined according to the preset communication protocol and the preset attenuation gradient baseline, respectively, to obtain the protocol layer event set and the physical layer event set. Based on the protocol layer event set and the physical layer event set, correlation analysis is performed to obtain the first deviation sequence. Analysis module 64 is used to perform correlation analysis after aligning the first deviation sequence with the decay mutation sequence in time, so as to obtain a second deviation sequence including mutation point markers; The determination module 65 is used to divide the second deviation sequence into multiple continuous windows along the time axis, divide multiple intervals within each window according to the attenuation gradient, determine the window impact degree based on the response delay and message retransmission count within each interval, determine the fault type based on the correlation between the frequency of mutation point markers and the attenuation gradient within each window, and calculate the channel score based on the attenuation gradient and window impact degree of each window.

[0084] The intelligent operation and maintenance monitoring and analysis system for railway substations in this application is used to implement the aforementioned intelligent operation and maintenance monitoring and analysis method for railway substations. Therefore, the specific implementation of the intelligent operation and maintenance monitoring and analysis system for railway substations can be found in the embodiment section of the intelligent operation and maintenance monitoring and analysis method for railway substations mentioned above. The specific implementation can be referred to the description of the corresponding embodiments, and will not be repeated here.

[0085] Figure 7A schematic diagram of the hardware structure of an electronic device provided in one embodiment of this application is shown.

[0086] The electronic device may include a processor 710 and a memory 720 storing computer program instructions.

[0087] Specifically, the processor 710 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0088] Memory 720 may include mass storage for data or instructions. For example, and not limitingly, memory 720 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 720 may include removable or non-removable (or fixed) media. Where appropriate, memory 720 may be internal or external to the integrated gateway disaster recovery device. In a particular embodiment, memory 720 is non-volatile solid-state memory.

[0089] Memory may include read-only memory (ROM), random access memory (RAM), disk storage media devices, optical storage media devices, flash memory devices, and electrical, optical, or other physical / tangible memory storage devices. Therefore, typically, memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to the first aspect of this disclosure.

[0090] The processor 710 reads and executes computer program instructions stored in the memory 720 to implement any of the intelligent operation and maintenance monitoring and analysis methods for railway substations in the above embodiments.

[0091] In one example, the electronic device may also include a communication interface 730 and a bus 740. Wherein, as... Figure 7 As shown, the processor 710, memory 720, and communication interface 730 are connected via bus 740 and communicate with each other.

[0092] The communication interface 730 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0093] Bus 740 includes hardware, software, or both, that couples components of an online data traffic metering device together. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 740 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, any suitable bus or interconnect is contemplated herein.

[0094] The electronic device can execute the intelligent operation and maintenance monitoring and analysis method for railway substations in the embodiments of this application, thereby realizing the intelligent operation and maintenance monitoring and analysis method for railway substations described in conjunction with the accompanying drawings.

[0095] Furthermore, in conjunction with the intelligent operation and maintenance monitoring and analysis method for railway substations in the above embodiments, this application embodiment can provide a computer-readable storage medium for implementation. This computer-readable storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any one of the intelligent operation and maintenance monitoring and analysis methods for railway substations in the above embodiments.

[0096] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0097] The functional blocks shown in the above-described structural diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0098] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0099] The aspects of this application have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by dedicated hardware performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0100] The above provides a detailed description of the intelligent operation and maintenance monitoring and analysis method and system for railway substations provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are merely for the purpose of helping to understand the method and its core ideas. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from its principles, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A method for intelligent operation and maintenance monitoring and analysis of railway substations, characterized in that, include: The communication message data of the substation and the fiber core attenuation value of the communication optical fiber are obtained. The communication message data includes a response timestamp and a message retransmission flag. Calculate the response delay sequence between adjacent instructions based on the response timestamp, count the number of message retransmissions in each preset statistical window based on the message retransmission flag, and calculate the attenuation gradient based on the filtered fiber core attenuation value to obtain the attenuation gradient map and attenuation mutation sequence. The response delay sequence, the number of message retransmissions, and the attenuation gradient map are compared with the response window and retransmission threshold determined according to the preset communication protocol and the preset attenuation gradient baseline, respectively, to obtain the protocol layer event set and the physical layer event set. Correlation analysis is performed based on the protocol layer event set and the physical layer event set to obtain the first deviation sequence. After aligning the first deviation sequence with the decay mutation sequence in time, a correlation analysis is performed to obtain a second deviation sequence including mutation point markers; The second deviation sequence is divided into multiple continuous windows along the time axis. Within each window, multiple intervals are divided according to the attenuation gradient. The window impact is determined based on the response delay and message retransmission count within each interval. The fault type is determined based on the correlation between the frequency of mutation point markers and the attenuation gradient within each window. The channel score is calculated based on the attenuation gradient of each window and the window impact.

2. The intelligent operation and maintenance monitoring and analysis method for railway substations according to claim 1, characterized in that, The steps include: calculating the response delay sequence between adjacent instructions based on the response timestamp; counting the number of message retransmissions within each preset statistical window based on the message retransmission flag; and calculating the attenuation gradient based on the filtered fiber core attenuation value to obtain the attenuation gradient map and attenuation abrupt change sequence. The response timestamps are arranged in chronological order, and the time difference between two adjacent response timestamps is calculated. The time difference values ​​are then arranged in chronological order to obtain the response delay sequence. The message retransmission flag is divided according to the statistical window, and the message retransmission flag is accumulated and counted in each statistical window to obtain the message retransmission count. A preset sliding window is slid along the fiber length direction, and the average value of the fiber core attenuation value within the sliding window is repeatedly calculated to obtain a filtered attenuation sequence distributed along the fiber length direction. Calculate the difference between adjacent attenuation values ​​in the filtered attenuation sequence, divide the difference by the distance between the corresponding adjacent attenuation values ​​to obtain the attenuation change rate, and arrange the attenuation change rates according to the fiber length direction to obtain the attenuation gradient map. The positions in the attenuation gradient map where the attenuation rate of change exceeds a preset rate of change threshold are arranged along the fiber length to form an attenuation mutation sequence.

3. The intelligent operation and maintenance monitoring and analysis method for railway substations according to claim 1, characterized in that, The step of comparing the response delay sequence, the number of message retransmissions, and the attenuation gradient map with the response window and retransmission threshold determined according to the preset communication protocol and the preset attenuation gradient baseline, respectively, to obtain the protocol layer event set and the physical layer event set, including: When the response delay in the response delay sequence is greater than the upper limit of the response window, the time point corresponding to the response delay is determined as the timeout moment, and all the timeout moments are arranged in chronological order to obtain the timeout event sequence; When the retransmission count value of each statistical window in the number of message retransmissions is greater than the retransmission threshold, the end time of the statistical window is determined as the retransmission abnormal time, and all the retransmission abnormal times are arranged in chronological order to obtain a retransmission event sequence. The timeout event sequence and the retransmission event sequence are merged in chronological order to obtain a protocol layer event set. Each event in the protocol layer event set includes the occurrence time and event type, and the event type is timeout or retransmission exception. When the attenuation change rate of the attenuation gradient map is greater than the attenuation gradient baseline, the fiber position corresponding to the attenuation change rate is determined as an attenuation anomaly position, and all the attenuation anomaly positions are arranged in order of fiber length to obtain a physical layer event set.

4. The intelligent operation and maintenance monitoring and analysis method for railway substations according to claim 1, characterized in that, The correlation analysis based on the protocol layer event set and the physical layer event set yields a first deviation sequence, including: Arrange the occurrence times of each timeout event and each retransmission anomaly event in the protocol layer event set in chronological order to obtain the protocol layer time point sequence. Arrange the sampling times corresponding to each attenuation anomaly position in the physical layer event set in chronological order to obtain the physical layer time point sequence. The number of moments in the protocol layer time point sequence within a consecutive preset time segment on the time axis is counted to obtain a first count value; the number of moments in the physical layer time point sequence within a consecutive preset time segment is counted to obtain a second count value. The first count value and the second count value corresponding to each time segment are added together to obtain the deviation value of each time segment. All the deviation values ​​are then arranged in the order of the time segments to obtain the first deviation sequence.

5. The intelligent operation and maintenance monitoring and analysis method for railway substations according to claim 1, characterized in that, The step of performing correlation analysis after aligning the first deviation sequence with the decay mutation sequence in time to obtain a second deviation sequence including mutation point markers includes: The start and end times corresponding to each deviation value in the first deviation sequence are defined as a time interval; The sampling time of each decay mutation position in the decay mutation sequence is compared with the time interval. When the sampling time exceeds the time interval, the deviation value is bound to the mutation marker. When the sampling time does not exceed the time interval, the deviation value is bound to the no-mutation marker. The bound deviation values ​​are arranged in chronological order according to the starting time to obtain the second deviation sequence.

6. The intelligent operation and maintenance monitoring and analysis method for railway substations according to claim 1, characterized in that, The step of dividing the second deviation sequence into multiple continuous windows along the time axis, further dividing each window into multiple intervals according to the attenuation gradient, and determining the window impact based on the response delay and message retransmission count within each interval includes: The second deviation sequence is divided into multiple continuous windows along the time axis. The decay gradient values ​​corresponding to each time window are extracted from the decay gradient map, and the extracted decay gradient values ​​are divided into multiple gradient intervals according to their magnitude. Extract the response delay values ​​associated with the decay gradient values ​​falling within the gradient intervals in the response delay sequence, and sum all the response delay values ​​in each gradient interval to obtain the cumulative delay value for each gradient interval; Extract the retransmission counts of the packets that fall within the gradient interval from the retransmission counts, and sum up all the retransmission counts of the packets in each gradient interval to obtain the retransmission sum value of each gradient interval. The window impact of each time window is obtained by weighted summing of all the accumulated delay values ​​and accumulated retransmission values ​​within each time window.

7. The intelligent operation and maintenance monitoring and analysis method for railway substations according to claim 1, characterized in that, The process of determining the fault type based on the correlation between the frequency of mutation point markers within each window and the attenuation gradient, and calculating the channel score based on the attenuation gradient of each window and the window influence, includes: The mutation point markers in each time window of the second deviation sequence are arranged in chronological order and then counted one by one to obtain the frequency of mutation points in each time window; Calculate the average value of the decay gradient within each time window in the decay gradient map to obtain the average decay gradient for each time window; The correlation degree of each time window is determined by multiplying the frequency of the mutation point with the average attenuation gradient. When the correlation degree is greater than the preset correlation threshold, the fault type of the corresponding time window is determined to be a physical channel fault. When the correlation degree is less than or equal to the preset correlation threshold, the fault type of the corresponding time window is determined to be a device fault. When the fault type is a physical channel fault, the fault location information is obtained by extracting all the attenuation mutation positions that fall within the corresponding time window from the attenuation mutation sequence and arranging them according to the fiber length direction. The window score for each time window is determined by multiplying the average decay gradient by the window influence, and the channel score is obtained by averaging all the window scores.

8. A smart operation and maintenance monitoring and analysis system for railway substations, characterized in that, include: The acquisition module is used to acquire the communication message data of the substation and the fiber core attenuation value of the communication optical fiber. The communication message data includes a response timestamp and a message retransmission flag. The calculation module is used to calculate the response delay sequence between adjacent instructions based on the response timestamp, count the number of message retransmissions in each preset statistical window based on the message retransmission flag, and calculate the attenuation gradient based on the filtered fiber core attenuation value to obtain the attenuation gradient map and attenuation mutation sequence. The comparison module is used to compare the response delay sequence, the number of message retransmissions, and the attenuation gradient map with the response window and retransmission threshold determined according to the preset communication protocol and the preset attenuation gradient baseline, respectively, to obtain the protocol layer event set and the physical layer event set. Based on the protocol layer event set and the physical layer event set, correlation analysis is performed to obtain the first deviation sequence. The analysis module is used to perform correlation analysis after aligning the first deviation sequence with the decay mutation sequence in time, so as to obtain a second deviation sequence including mutation point markers; The determination module is used to divide the second deviation sequence into multiple continuous windows along the time axis, divide multiple intervals within each window according to the attenuation gradient, determine the window impact degree based on the response delay and message retransmission count within each interval, determine the fault type based on the correlation between the frequency of abrupt change point markers and the attenuation gradient within each window, and calculate the channel score based on the attenuation gradient of each window and the window impact degree.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor is configured to execute the computer program to implement the steps of the intelligent operation and maintenance monitoring and analysis method for railway substations as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, enables the implementation of a railway substation intelligent operation and maintenance monitoring and analysis method as described in any one of claims 1 to 7.

Images