Underground cable on-line monitoring and fault early warning method and system

Abstract

The invention discloses a method and a system for online monitoring and fault early warning of an underground cable. The method comprises the following steps: acquiring monitoring data of each key part of the underground cable by using a sensor and preprocessing the monitoring data; transmitting the preprocessed monitoring data collected by each key part of the underground cable to a cloud computing platform, inputting the monitoring data belonging to different key part types into a fault recognition model matched with the corresponding key part type by using the cloud computing platform, and obtaining a fault recognition result of each key part in real time; according to the fault identification result, obtaining a fault risk assessment result of each key part without a fault, and correspondingly increasing the acquisition frequency of each key part with a fault risk and computing resources of a fault identification model matched with the type of the corresponding key part; and the physical position of each key part of the fault is inquired in combination with the GIS technology, and fault early warning prompt information is correspondingly generated. According to the invention, comprehensive monitoring of the underground cable can be realized, and faults can be timely and accurately found and early warning can be carried out.

Description

Technical Field

[0002] The present application relates to the technical field of online monitoring and fault warning of underground cables, and specifically to a method and system for online monitoring and fault warning of underground cables. Background Art

[0003] Underground cables play a vital role in modern urban power grids. With the continuous advancement of urbanization, the amount of underground cables is increasing day by day, becoming an important part of the power grid infrastructure. However, since underground cables are buried under the ground, their operating status is often difficult to control in real time, which puts higher requirements on the stable operation of the power grid. Therefore, the development of an efficient and reliable underground cable monitoring technology has become the focus of current research.

[0004] Existing cable monitoring solutions mostly use regular inspections and offline detection. The former relies on manual regular inspections, while the latter performs status checks by shutting down the equipment. In addition, there are also some online monitoring solutions in the existing technology that attempt to improve monitoring efficiency through real-time data collection, but these solutions can usually only monitor a single state parameter of the cable, such as temperature or current, and cannot comprehensively evaluate the overall operating status of the cable. Although this method can provide a certain degree of protection, its efficiency and accuracy are difficult to meet the requirements of modern society for power grid stability.

[0005] It can be seen that the current underground cable monitoring technology cannot achieve all-round and real-time monitoring of the cable operation status, especially in the monitoring of key parts such as cable joints and terminals. This leads to the late detection of potential faults, which in turn affects the overall stable operation of the power grid. Summary of the invention

[0006] In order to achieve comprehensive monitoring of underground cables, timely and accurately detect faults and issue early warnings, the present application provides a method and system for online monitoring and fault early warning of underground cables.

[0007] In a first aspect, the present application provides a method for online monitoring and fault warning of underground cables, comprising: The sensors are used to collect monitoring data of key parts of underground cables; the key parts include: cable joints, cable terminals and several preset positions of the underground cable body; the monitoring data include: electrical parameter data, vibration data and environmental data; Preprocessing the monitoring data collected from each key part of the underground cable; the preprocessing includes: integrating the monitoring data of the same key part type according to the key part type, sorting the monitoring data of the same key part type according to the preset priority, and fusing the monitoring data of the same key part; The pre-processed monitoring data collected from each key part of the underground cable is transmitted to the cloud computing platform, and the monitoring data belonging to different key parts types are input into the fault identification model matching the corresponding key part type by using the cloud computing platform to obtain the fault identification results of each key part in real time; For each key part whose fault identification result is not a fault, the monitoring data collected from each key part that is not a fault is input into the fault risk assessment model that matches the corresponding key part type to obtain the fault risk assessment results of each key part that is not a fault; according to the fault risk assessment results, the collection frequency of each key part whose fault risk assessment result is a fault risk and the computing resources of the fault identification model that matches the key part type of each key part whose fault risk assessment result is a fault risk are increased; for each key part whose fault identification result is a fault, the physical location of each key part of the fault is queried in combination with GIS technology and corresponding fault warning prompt information is generated.

[0008] By adopting the above solution, multimodal data collection is carried out on key parts of underground cables to achieve all-round online monitoring; the cloud computing platform is used to realize real-time analysis of online monitoring data, and the fault risk assessment model is used to dynamically adjust the collection frequency and computing resource allocation for non-faulty parts to enhance the ability to identify and handle faults; for identified faulty parts, GIS technology is used to accurately locate them, and early warning information containing the physical location of the key fault parts is generated, which improves the efficiency and accuracy of fault handling.

[0009] Preferably, the method further comprises: using a traveling wave ranging device installed at each key position of the underground cable to collect traveling wave signal data and transmit the data to a cloud computing platform; For each key part of the fault whose fault identification result is a fault, the corresponding traveling wave signal data of each key part of the fault is queried, and after the traveling wave signal of the fault is analyzed and identified using the deep learning algorithm, the traveling wave ranging technology is used to determine the distance between the fault point and each key part, and the specific physical location of the fault point is determined, and a fault warning prompt information containing the specific physical location of the fault is generated.

[0010] By adopting the above scheme, the traveling wave ranging device and the traveling wave ranging technology of each key part of the fault are used to obtain the physical position distance between the actual fault position and each key part, so as to achieve more accurate fault point positioning.

[0011] Preferably, it also includes: Determine the application scenarios of each key part according to the physical location of each key part; the application scenarios include: high-load operation application scenarios, frequent geological activities application scenarios, and frequent environmental changes application scenarios; In the process of using the cloud computing platform to input the monitoring data belonging to different key part types into the fault identification model that matches the corresponding key part type, the feature weight of the matching fault identification model is dynamically adjusted according to the application scenario in the monitoring data of each key part, so that the feature weight corresponding to a type of monitoring data that meets the current application scenario is increased, and the fault identification model after the feature weight adjustment is used to perform fault identification on each key part input; among them, a type of monitoring data that meets the high-load operation application scenario is electrical parameter data, a type of monitoring data that meets the application scenario of frequent geological activities is vibration data, and a type of monitoring data that meets the application scenario of frequent environmental changes is environmental data.

[0012] By adopting the above solution, the application scenarios of key parts are determined according to their physical locations, and when using the cloud computing platform for fault identification, the feature weights of the fault identification model are dynamically adjusted according to different application scenarios, so that the fault identification model can adapt to the current application scenario more accurately, thereby improving the accuracy and efficiency of fault identification.

[0013] Preferably, it also includes: Determine whether the computing resources of the current cloud computing platform have reached a tense state; the tense state includes any variable among the CPU usage rate, memory usage rate and remaining storage capacity of the current cloud computing platform reaching a preset threshold value of the corresponding variable; If a tense situation is reached, it is judged based on the fault risk assessment results whether the key part types corresponding to the key parts with fault risks in the fault risk assessment results include all key part types. If all key part types are included, the elastic computing cluster technology is used to increase the computing resources of the cloud computing platform. If not, the computing resources of the fault identification model that matches the key part type corresponding to the key parts with fault risk in the fault risk assessment results are increased, while the computing resources of other fault identification models are reduced to balance the computing resources of the cloud computing platform.

[0014] By adopting the above solution, combined with the fault risk assessment results and the computing resource status of the current cloud computing platform, the fault identification model computing resource allocation corresponding to each key part type is automatically adjusted to achieve resource balance; or computing resources are increased through elastic computing cluster technology, further improving the system's processing power and reliability.

[0015] Preferably, the preprocessing of the monitoring data collected from key parts of underground cables includes: The local first edge computing device is used to integrate and process the monitoring data of the same key part type according to the key part type; the monitoring data belonging to the same key part type are respectively input into the matching local second edge computing device. The monitoring data of each key part type is equipped with a matching local second edge computing device. The local second edge computing device is used to sort and process the monitoring data of the same key part type according to the preset priority, and then the fusion processing of the monitoring data of the same key part is completed.

[0016] By adopting the above solution, multiple local edge devices are used to complete the monitoring data integration of the same key part type and the monitoring data fusion of the same key part, thereby improving the processing efficiency and accuracy of the monitoring data, ensuring the accuracy of subsequent fault identification and reducing the computing pressure of the cloud computing platform.

[0017] Preferably, it also includes: For each key part of the fault whose fault identification result is a fault, query the associated key parts of each key part of the fault; the associated key part refers to a key part whose physical position with the key part of the fault is less than a preset physical position, and whose own fault identification time and the average interval time between the fault identification time corresponding to the key part of the fault are less than a preset time; According to the query result, the acquisition frequency of the key parts associated with each key part of the fault is increased, and the computing resources of the fault identification model that matches the key part type corresponding to the key parts associated with each key part of the fault are increased.

[0018] By adopting the above scheme, the associated key parts of each key part of the fault will be affected by it. According to the query results, the collection frequency of the associated key parts and the computing resources of the corresponding fault identification model are increased, thereby achieving enhanced monitoring and analysis of potential faults and improving the accuracy and timeliness of fault warning.

[0019] Preferably, the several preset positions of the underground cable body include: an underground cable branching position, a position where the historical underground cable failure frequency is greater than a preset frequency, a position where the underground cable crosses or runs parallel to other pipelines, and a position where the underground cable bending rate is greater than a preset bending rate.

[0020] By adopting the above scheme, key monitoring is performed on cable branch locations, locations with a high historical cable failure frequency, locations where cables and pipelines cross or run parallel, and locations with a large cable bending rate, thereby effectively improving the accuracy and timeliness of fault warnings.

[0021] In a second aspect, the present application provides a system for online monitoring and fault warning of underground cables, comprising: The monitoring data acquisition module is used to collect monitoring data of key parts of underground cables using sensors; the key parts include: cable joints, cable terminals and several preset positions of the underground cable body; the monitoring data include: electrical parameter data, vibration data and environmental data; The monitoring data preprocessing module is used to preprocess the monitoring data collected from each key part of the underground cable; the preprocessing includes: integrating the monitoring data of the same key part type according to the key part type, sorting the monitoring data of the same key part type according to the preset priority, and fusing the monitoring data of the same key part; The fault identification module is used to transmit the pre-processed monitoring data collected from each key part of the underground cable to the cloud computing platform, and use the cloud computing platform to input the monitoring data belonging to different key parts into the fault identification model matching the corresponding key part type, so as to obtain the fault identification results of each key part in real time; The fault warning module is used to input the monitoring data collected from each key part that is not faulty into the fault risk assessment model that matches the corresponding key part type for each key part whose fault identification result is not faulty, so as to obtain the fault risk assessment results of each key part that is not faulty; according to the fault risk assessment results, increase the collection frequency of each key part whose fault risk assessment results are that there is a fault risk and the computing resources of the fault identification model that matches the key part type for each key part whose fault risk assessment results are that there is a fault risk; for each key part whose fault identification result is faulty, use GIS technology to query the physical location of each key part of the fault and generate corresponding fault warning prompt information.

[0022] By adopting the above solution, comprehensive monitoring of underground cables can be achieved, faults can be discovered in a timely and accurate manner, and early warnings can be issued.

[0023] In a third aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the method as described above.

[0024] In a fourth aspect, the present application provides a computer device, comprising a memory, a processor, and a program stored and executable on the memory, wherein the program implements the steps of the above method when executed by the processor.

[0025] In summary, this application has the following beneficial effects: 1. Comprehensive monitoring of electrical parameter data, vibration data and environmental data of key parts of underground cables, including cable joints, cable terminals and preset positions of cable bodies, has been achieved, which has improved the timeliness and accuracy of fault discovery. By using the cloud computing platform for fault identification, the fault identification results of key parts can be obtained in real time, and the monitoring frequency and computing resources of the fault identification model can be dynamically adjusted according to the fault risk assessment results of different key parts, which has improved the timeliness and accuracy of fault warning. By combining GIS technology and traveling wave ranging technology, the fault point can be quickly located and warning information containing the fault location can be generated. 2. Multiple local edge computing devices are used to integrate monitoring data of the same key part type and fuse monitoring data of the same key part, which improves the efficiency of monitoring data processing, facilitates the corresponding parameter input and identification of subsequent fault identification models, improves the accuracy of fault identification and reduces the computing pressure of the cloud computing platform. 3. Based on the different application scenarios of underground cables, the weights of the fault identification models corresponding to the monitoring data in each application scenario are adjusted accordingly to achieve more accurate identification. At the same time, the resource usage status of the cloud computing platform is considered, and according to the fault prediction and evaluation results, the computing resources corresponding to different fault identification models are dynamically balanced or the computing resources are increased by using elastic computing cluster technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a flow chart of the method for online monitoring and fault early warning of underground cables described in a specific embodiment; Figure 2 It is a diagram of monitoring data collection and transmission in the method for online monitoring and fault early warning of underground cables described in a specific embodiment; Figure 3 It is a structural schematic diagram of the system for online monitoring and fault warning of underground cables in a specific embodiment. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0028] like Figure 1 As shown, the embodiment of the present application discloses a method for online monitoring and fault warning of underground cables, and the specific steps are as follows: S1. Use sensors to collect monitoring data of key parts of underground cables.

[0029] Specifically, in order to achieve comprehensive monitoring of underground cables, various sensors or other collection devices installed at key locations of the underground cables are used to collect monitoring data.

[0030] Considering the factors affecting the stability of underground cables, underground cable joints, as the weaker link in underground cable lines, are easily affected by external factors and fail; underground cable terminals, as the part where cables are connected to equipment, have performance that affects the safe operation of the entire underground cable line; there are a large number of locations in the cable body that are easily affected by external factors and fail; therefore, the key parts of the underground cable mainly include: cable joints of underground cables, cable terminals of underground cables, and several preset positions of the underground cable body. Among them, the several preset positions of the underground cable body include: underground cable branching locations, locations where the historical underground cable failure frequency is greater than the preset frequency, locations where underground cables cross or run parallel with other pipelines, and locations where the underground cable bending rate is greater than the preset bending rate.

[0031] In order to obtain more accurate multi-modal monitoring data, various types of high-precision sensors are used, such as various types of high-precision electrical sensors, various types of high-precision environmental sensors, and high-precision optical fiber sensors (or cable anti-breakage monitoring devices), etc., to collect and obtain multi-modal data. The monitoring data includes: electrical parameter data (current, voltage, resistance, etc.), vibration data (vibration signal), and environmental data (temperature, humidity, pH, gas, etc.); In addition, in order to further realize all-round monitoring and obtain underground cable installation information, the physical location information of key parts of underground cables can be determined in combination with the GIS system. The underground cable laying pipeline information can also be collected, and sensors can be used to collect environmental data, vibration data and other information corresponding to the underground cable laying pipelines at the locations of key parts of underground cables.

[0032] S2. Pre-process the monitoring data collected from key parts of underground cables.

[0033] Specifically, the preprocessing includes: data cleaning, data compensation, data integration and fusion processing; wherein, the data integration and fusion processing includes: classifying and aggregating the monitoring data of the same key parts type, such as: aggregating the monitoring data of cable joints belonging to the same underground cables, aggregating the monitoring data of cable terminals belonging to the same underground cables, etc.; the monitoring data of the same key parts type are sorted according to the preset priority, and the preset priority can be set manually. The priority can be set from high to low according to the historical fault frequency of each key part, and the same fault frequency can be sorted according to the physical location of each key part. Whether it corresponds to an important area, the order of the important areas is followed, and the higher the level, the higher the priority; such as: the cable joints in the underground cable include X001, X002, X001 to X00n, n is the number of cable joints, and they are sorted according to the preset priority: X00n, X002, ... X001; the monitoring data of the same key parts are time synchronized, and the data are fused to generate real-time monitoring data sets for each key part, such as: the monitoring data set of the cable joint marked as X001 at time t1 is {V1, I1, R1, X1, T1, S1, PH1}, where X1 is the vibration signal.

[0034] In order to further quickly complete the preprocessing of monitoring data, the local edge computing device is selected to preprocess the monitoring signal to reduce the subsequent computing resources occupied by the cloud computing platform. Specifically, the local first edge computing device is used to integrate and process the monitoring data of the same key part type according to the key part type; The monitoring data belonging to the same key part type are respectively input into the matching local second edge computing device. The monitoring data of each key part type is provided with a matching local second edge computing device, such as a second edge computing device matching the underground cable joint, a second edge computing device matching the underground cable terminal, and a second edge computing device matching the preset position of the underground cable body, etc.; after the monitoring data of the same key part type are sorted and processed according to the preset priority by the local second edge computing device, the monitoring data fusion processing of the same key part is completed, that is, the monitoring devices of different key part types can be pre-processed separately and synchronously, thereby improving the efficiency of the pre-processing.

[0035] S3. The monitoring data collected from each key part of the underground cable after pre-processing is transmitted to the cloud computing platform. The cloud computing platform is used to input the monitoring data belonging to different key part types into the fault identification model that matches the corresponding key part type, and the fault identification results of each key part are obtained in real time.

[0036] Specifically, in order to obtain more abundant computing resources to complete the analysis of the monitoring data of each key part in real time, the monitoring data collected from each key part of the underground cable after preprocessing is transmitted to the cloud computing platform; Figure 2 As shown, each key part can use DTU or fusion terminal to transmit the pre-processed monitoring data.

[0037] In order to ensure the accuracy of fault identification, taking into account the differences in monitoring data collected from each different type of key parts, a fault identification model suitable for fault identification of a specific type of key parts is matched accordingly; specifically, a first fault identification model is set for the cable joint of the underground cable, a second fault identification model is set for the cable terminal of the underground cable, and a third fault identification model is set for several preset positions of the underground cable body; the above models all use neural networks, the first fault identification model inputs the monitoring data of a certain cable joint type of the underground cable at a certain moment, the output of the model corresponds to the fault identification result of the cable joint, that is, fault or no fault, which is generated by training with monitoring data of cable joint types of several historical underground cables marked with faults or no faults; the second fault identification model inputs the monitoring data of a certain cable terminal type of the underground cable at a certain moment, the output of the model corresponds to the fault identification result of the cable terminal, that is, fault or no fault, which is generated by training with monitoring data of cable terminal types of several historical underground cables marked with faults or no faults; the third fault identification model inputs the monitoring data of the cable body type at the preset position of the underground cable at a certain moment, the output of the model corresponds to the fault identification result of the cable body at the preset position, that is, fault or no fault, which is generated by training with monitoring data of cable body types at several historical underground cable preset positions marked with faults or no faults.

[0038] The cloud computing platform is used to input the monitoring data belonging to different key part types into the fault identification model that matches the corresponding key part type. That is, the monitoring data sets of each key part type transmitted in real time are input one by one into the fault identification model corresponding to the key part type, and the fault identification results of each key part type are obtained in real time.

[0039] S4. Adjust the computing resources of the cloud computing platform and generate fault warning prompts based on the fault identification results of each key part.

[0040] In order to accurately and timely determine whether each key part has faults or not, in addition to fault identification based on real-time collected monitoring data, it is also necessary to pay more attention to monitoring the key parts with fault risks, and the monitoring data collection frequency of the key parts with fault risks can be adjusted accordingly, and the computing resources of the fault identification model can be adjusted accordingly. Specifically: for each key part whose fault identification result is not a fault, the monitoring data collected from each key part without a fault is input into the fault risk assessment model that matches the type of the corresponding key part, and the fault risk assessment results of each key part without a fault are obtained; wherein, a first fault risk assessment model is set for the cable joint of the underground cable, a second fault risk assessment model is set for the cable terminal of the underground cable, and a third fault risk assessment model is set for several preset positions of the underground cable body; the above fault risk assessment models all use neural networks, and the input of the model is the monitoring data of the key parts of the corresponding type, and the model output is the fault risk assessment result of each key part, that is, whether there is a fault risk or not.

[0041] According to the fault risk assessment results, the collection frequency of each key part with fault risk as a result of the fault risk assessment is increased. Specifically, the collection frequency can be set to increase to 120%-150% of the original collection frequency, which can be set manually. Due to the increase in the collection frequency, the subsequent monitored data will increase, and correspondingly, better computing resources need to be provided for the analysis of these data. Therefore, the computing resources of the fault identification model that matches the key part type of each key part with fault risk assessment results are increased. Specifically, the size of the increased computing resources can be set.

[0042] For the key parts of the fault identified as faults, it can be shown that faults already exist in the key parts. Therefore, the physical locations of the key parts of the faults are queried in combination with GIS technology, and corresponding fault warning prompt information is generated, including the positioning information of the key parts of the faults.

[0043] In addition, considering that the key parts currently determined to have faults may be caused by other fault parts (such as the cable body near a key part of the cable fault), in order to more accurately determine the location of the fault, the traveling wave ranging method is chosen, specifically including: using cable fault locating devices (such as traveling wave ranging devices) installed at key parts of underground cables to collect traveling wave signal data information (including collection time information) and transmit it to the cloud computing platform; among them, the pre-processing of the traveling wave signal data can be sent to the third local edge design for filtering processing.

[0044] For each key part of the fault whose fault identification result is a fault, the corresponding traveling wave signal data of each key part of the fault is queried, and after analyzing and identifying the traveling wave signal of the fault using a deep learning algorithm (such as building an identification model of the traveling wave signal of the fault), the traveling wave ranging technology is used to determine the distance between the fault point and each key part, determine the specific physical location of the fault point, and generate fault warning prompt information including the specific physical location of the fault.

[0045] By adopting the scheme described in the above embodiment, different types of sensors are deployed at key locations of the cable to comprehensively collect monitoring data, and the online monitoring method of cloud computing is combined to perform real-time analysis and processing of a large amount of monitoring data, thereby identifying faults in a timely manner, accurately locating the fault point, and improving the efficiency of fault handling.

[0046] In a specific embodiment, considering that different application scenarios pay different attention to different types of monitoring data, some scenarios pay more attention to the monitoring of power parameters, and some scenarios pay more attention to the monitoring of environmental parameters. Considering different scenarios, the corresponding adjustment of the weight of monitoring data analysis is considered, and the method further includes: The application scenarios of each key part are determined according to the physical location of each key part; the application scenarios include: high-load operation application scenarios, frequent geological activity application scenarios and frequent environmental change application scenarios; specifically, the characteristics of each location are queried according to the geographical location of each key part (such as: areas with high power demand and cable load; or areas with earthquakes, geological subsidence activities or trains and EMUs, and locations with large changes in soil moisture) to determine the application scenario.

[0047] In the process of using the cloud computing platform to input the monitoring data belonging to different key parts types into the fault identification model that matches the corresponding key parts types, the feature weight of the matching fault identification model is dynamically adjusted for the application scenario in each key part monitoring data, so that the feature weight corresponding to a type of monitoring data that meets the current application scenario is increased, and the fault identification model after the feature weight adjustment is used to perform fault identification on each key part input. Among them, a type of monitoring data that meets the high-load operation application scenario is electrical parameter data, a type of monitoring data that meets the application scenario of frequent geological activities is vibration data, and a type of monitoring data that meets the application scenario of frequent environmental changes is environmental data; Among them, the specific amount of feature weight increase can be set manually, and in order to ensure that the sum of the total feature weights of various types of monitoring data is 1, when increasing the feature weight corresponding to a certain type of monitoring data, the feature weights corresponding to the remaining types of monitoring data are reduced in addition to reducing the current type of monitoring data that needs to be added. Therefore, if it meets multiple application scenarios, the feature weight corresponding to the monitoring data in each application scenario will be increased accordingly, but if it meets all application scenarios, the original feature weight ratio will remain unchanged.

[0048] In a specific embodiment, in order to more accurately identify the fault conditions of each key part, the resources corresponding to the fault identification model are set and adjusted. In order to further optimize the adjustment strategy of the computing resources, the method further includes: Determine whether the computing resources of the current cloud computing platform have reached a tense state; the tense state includes any variable among the CPU utilization rate, memory utilization rate and remaining storage capacity of the current cloud computing platform reaching a preset threshold value of the corresponding variable; wherein the computing resources of the cloud computing platform will be integrated into a unified resource pool, recording the specific content and quantity of the specific CPU, memory and storage capacity.

[0049] If a tense situation is reached, it is judged based on the fault risk assessment results, whether the key part types corresponding to the key parts with fault risks in the fault risk assessment results include all key part types; if all key part types are included, the elastic computing cluster technology is used to increase the computing resources of the cloud computing platform, that is, automatically expand the computing resources.

[0050] If not included, the computing resources of the fault identification model that matches the key part type corresponding to each key part with fault risk assessment results at risk of failure are increased, while the computing resources of other fault identification models are reduced to balance the computing resources of the cloud computing platform.

[0051] In a specific embodiment, considering that the critical part of a fault often affects the critical parts associated with it, the monitoring of the critical parts associated with the critical part of the fault will be increased to timely discover the fault. The method also includes: for each critical part whose fault identification result is a fault, querying the critical parts associated with each critical part of the fault; the associated critical part refers to a critical part whose physical position with the critical part of the fault is less than the preset physical position, and whose own fault identification moment and the average interval between the fault identification moment corresponding to the critical part of the fault are less than the preset time, that is, the distance is close and once one fails, the other will fail in a very short time.

[0052] According to the query result, the acquisition frequency of the associated key parts of each key part of the fault is increased, and the computing resources of the fault identification model that matches the key part type corresponding to the associated key parts of each key part of the fault are increased; and the associated faults are increased.

[0053] like Figure 3 As shown, the embodiment of the present application discloses a system for online monitoring and fault warning of underground cables, specifically comprising: a monitoring data acquisition module 101, which is used to collect monitoring data of key parts of underground cables using sensors; the key parts include: cable joints, cable terminals and several preset positions of the underground cable body; the monitoring data includes: electrical parameter data, vibration data and environmental data; The monitoring data preprocessing module 102 is used to preprocess the monitoring data collected from each key part of the underground cable; the preprocessing includes: integrating the monitoring data of the same key part type according to the key part type, sorting the monitoring data of the same key part type according to the preset priority, and fusing the monitoring data of the same key part; The fault identification module 103 is used to transmit the pre-processed monitoring data collected from each key part of the underground cable to the cloud computing platform, and use the cloud computing platform to input the monitoring data belonging to different key parts into the fault identification model matching the corresponding key part type, so as to obtain the fault identification results of each key part in real time; The fault warning module 104 is used to input the monitoring data collected from each key part that is not faulty into a fault risk assessment model that matches the corresponding key part type for each key part whose fault identification result is not faulty, so as to obtain the fault risk assessment results of each key part that is not faulty; according to the fault risk assessment results, increase the collection frequency of each key part whose fault risk assessment results are that there is a fault risk and the computing resources of the fault identification model that matches the key part type for each key part whose fault risk assessment results are that there is a fault risk; for each key part whose fault identification result is a fault, use GIS technology to query the physical location of each key part of the fault and generate corresponding fault warning prompt information.

[0054] In a specific embodiment, the system further includes: a traveling wave signal data acquisition module 105 for collecting traveling wave signal data using traveling wave ranging devices installed at key locations of underground cables and transmitting the data to a cloud computing platform; The fault warning module 104 is also used to query the corresponding traveling wave signal data of each key part of the fault for each key part whose fault identification result is a fault, use a deep learning algorithm to analyze and identify the traveling wave signal of the fault, and then use traveling wave ranging technology to determine the distance between the fault point and each key part, determine the specific physical location of the fault point, and generate fault warning prompt information including the specific physical location of the fault.

[0055] In a specific embodiment, the system further includes: a fault identification model adjustment module 106, which is used to determine the application scenario of each key part according to the physical location of each key part; the application scenarios include: high-load operation application scenarios, frequent geological activity application scenarios and frequent environmental change application scenarios; in the process of using the cloud computing platform to input the monitoring data belonging to different key part types into the fault identification model corresponding to the key part type, the feature weight of the matching fault identification model is dynamically adjusted according to the application scenario in the monitoring data of each key part, so that the corresponding feature weight of a type of monitoring data that meets the current application scenario is increased, and the fault identification model after the feature weight adjustment is used to perform fault identification on each key part input; wherein, a type of monitoring data that meets the high-load operation application scenario is electrical parameter data, a type of monitoring data that meets the frequent geological activity application scenario is vibration data, and a type of monitoring data that meets the frequent environmental change application scenario is environmental data.

[0056] In a specific embodiment, the fault identification model adjustment module 106 in the system is also used to determine whether the computing resources of the current cloud computing platform have reached a tense state; the tense state includes that any variable among the CPU usage rate, memory usage rate and remaining storage capacity of the current cloud computing platform reaches a preset threshold value of the corresponding variable; if the tense state is reached, it is determined based on the fault risk assessment result whether the key part types corresponding to the key parts with fault risk added as a result of the fault risk assessment include all key part types; if all key part types are included, the computing resources of the cloud computing platform are increased by using elastic computing cluster technology; if not, the computing resources of the fault identification model matching the key part type corresponding to the key parts with fault risk assessment results as a result of the fault risk are increased, while the computing resources of other fault identification models are reduced to balance the computing resources of the cloud computing platform.

[0057] In a specific embodiment, the fault identification model adjustment module 106 in the system is also used to query the associated key parts of each key part of the fault for each key part of the fault whose fault identification result is a fault; the associated key part refers to a key part whose physical position with the key part of the fault is less than a preset physical position, and whose own fault identification moment and the average interval between the fault identification moment corresponding to the key part of the fault is less than a preset time; according to the query result, the collection frequency of the associated key parts of each key part of the fault is increased, and the computing resources of the fault identification model that matches the key part type corresponding to the associated key parts of each key part of the fault are matched.

[0058] The embodiment of the present application also discloses a computer-readable storage medium.

[0059] Specifically, the computer-readable storage medium stores a computer program that can be loaded by a processor and executes a method such as the above-mentioned underground cable online monitoring and fault warning method. The computer-readable storage medium includes, for example: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0060] The embodiment of the present application also discloses a computer device.

[0061] Specifically, the computer device includes a memory and a processor, and the memory stores a computer program that can be loaded by the processor and execute the above-mentioned method for online monitoring and fault warning of underground cables.

[0062] The above are all preferred embodiments of the present application, and are not intended to limit the protection scope of the present application. Any feature disclosed in this specification (including the abstract and drawings), unless otherwise stated, can be replaced by other equivalent or alternative features with similar purposes. That is, unless otherwise stated, each feature is only an example of a series of equivalent or similar features.

Claims

1. A method for online monitoring and fault early warning of underground cables, characterized in that: include: Use sensors to collect monitoring data from key parts of underground cables; The key parts include: cable joints, cable terminals and several preset positions of the underground cable body; the monitoring data include: electrical parameter data, vibration data and environmental data; Preprocessing the monitoring data collected from each key part of the underground cable; the preprocessing includes: integrating the monitoring data of the same key part type according to the key part type, sorting the monitoring data of the same key part type according to the preset priority, and fusing the monitoring data of the same key part; The pre-processed monitoring data collected from each key part of the underground cable is transmitted to the cloud computing platform, and the monitoring data belonging to different key parts types are input into the fault identification model matching the corresponding key part type by using the cloud computing platform to obtain the fault identification results of each key part in real time; For each key part whose fault identification result is not a fault, the monitoring data collected from each key part that is not a fault is input into the fault risk assessment model that matches the corresponding key part type to obtain the fault risk assessment results of each key part that is not a fault; according to the fault risk assessment results, the collection frequency of each key part whose fault risk assessment result is a fault risk and the computing resources of the fault identification model that matches the key part type of each key part whose fault risk assessment result is a fault risk are increased; for each key part whose fault identification result is a fault, the physical location of each key part of the fault is queried in combination with GIS technology and corresponding fault warning prompt information is generated.

2. The method for online monitoring and fault early warning of underground cables according to claim 1 is characterized in that: It also includes: using traveling wave ranging devices installed at key locations of underground cables to collect traveling wave signal data and transmit it to a cloud computing platform; For each key part of the fault whose fault identification result is a fault, the corresponding traveling wave signal data of each key part of the fault is queried, and after the traveling wave signal of the fault is analyzed and identified using the deep learning algorithm, the traveling wave ranging technology is used to determine the distance between the fault point and each key part, and the specific physical location of the fault point is determined, and a fault warning prompt information containing the specific physical location of the fault is generated.

3. The method for online monitoring and fault early warning of underground cables according to claim 1 is characterized in that: Also includes: Determine the application scenarios of each key part according to its physical location; The application scenarios include: high-load operation application scenarios, frequent geological activity application scenarios, and frequent environmental change application scenarios; In the process of using the cloud computing platform to input the monitoring data belonging to different key part types into the fault identification model that matches the corresponding key part type, the feature weight of the matching fault identification model is dynamically adjusted according to the application scenario in the monitoring data of each key part, so that the feature weight corresponding to a type of monitoring data that meets the current application scenario is increased, and the fault identification model after the feature weight adjustment is used to perform fault identification on each key part input; among them, a type of monitoring data that meets the high-load operation application scenario is electrical parameter data, a type of monitoring data that meets the application scenario of frequent geological activities is vibration data, and a type of monitoring data that meets the application scenario of frequent environmental changes is environmental data.

4. The method for online monitoring and fault early warning of underground cables according to claim 1 is characterized in that: Also includes: Determine whether the computing resources of the current cloud computing platform have reached a critical state; The tense state includes any variable among the CPU usage rate, memory usage rate and remaining storage capacity of the current cloud computing platform reaching a preset threshold value of the corresponding variable; If a tense situation is reached, it is judged based on the fault risk assessment results whether the key part types corresponding to the key parts with fault risks in the fault risk assessment results include all key part types. If all key part types are included, the elastic computing cluster technology is used to increase the computing resources of the cloud computing platform. If not, the computing resources of the fault identification model that matches the key part type corresponding to the key parts with fault risk in the fault risk assessment results are increased, while the computing resources of other fault identification models are reduced to balance the computing resources of the cloud computing platform.

5. The method for online monitoring and fault early warning of underground cables according to claim 1 is characterized in that: The preprocessing of the monitoring data collected from key parts of underground cables includes: After the local first edge computing device is used to integrate and process the monitoring data of the same key part type according to the key part type; The monitoring data belonging to the same key part type are input into the matching local second edge computing device respectively, and each key part type of monitoring data is equipped with a matching local second edge computing device; after the monitoring data of the same key part type are sorted and processed according to the preset priority by the local second edge computing device, the fusion processing of the monitoring data of the same key part is completed.

6. The method for online monitoring and fault early warning of underground cables according to claim 1 is characterized in that: Also includes: For each key part of the fault identified as a fault, query the key parts associated with each key part of the fault; The associated key part refers to a key part whose physical position with respect to the key part of the fault is less than the preset physical position, and whose average interval between its own fault identification moment and the corresponding fault identification moment of the key part of the fault is less than the preset time length; According to the query result, the acquisition frequency of the key parts associated with each key part of the fault is increased, and the computing resources of the fault identification model that matches the key part type corresponding to the key parts associated with each key part of the fault are increased.

7. The method for online monitoring and fault early warning of underground cables according to claim 1 is characterized in that: The preset positions of the underground cable body include: underground cable branching positions, positions where the historical underground cable failure frequency is greater than a preset frequency, positions where the underground cable crosses or runs parallel to other pipelines, and positions where the underground cable bending rate is greater than a preset bending rate.

8. A system for online monitoring and fault warning of underground cables, characterized in that: include: Monitoring data acquisition module, used to collect monitoring data of key parts of underground cables using sensors; The key parts include: cable joints, cable terminals and several preset positions of the underground cable body; the monitoring data include: electrical parameter data, vibration data and environmental data; The monitoring data preprocessing module is used to preprocess the monitoring data collected from each key part of the underground cable; the preprocessing includes: integrating the monitoring data of the same key part type according to the key part type, sorting the monitoring data of the same key part type according to the preset priority, and fusing the monitoring data of the same key part; The fault identification module is used to transmit the pre-processed monitoring data collected from each key part of the underground cable to the cloud computing platform, and use the cloud computing platform to input the monitoring data belonging to different key parts into the fault identification model matching the corresponding key part type, so as to obtain the fault identification results of each key part in real time; The fault warning module is used to input the monitoring data collected from each key part that is not faulty into the fault risk assessment model that matches the corresponding key part type for each key part whose fault identification result is not faulty, so as to obtain the fault risk assessment results of each key part that is not faulty; according to the fault risk assessment results, increase the collection frequency of each key part whose fault risk assessment results are that there is a fault risk and the computing resources of the fault identification model that matches the key part type for each key part whose fault risk assessment results are that there is a fault risk; for each key part whose fault identification result is faulty, use GIS technology to query the physical location of each key part of the fault and generate corresponding fault warning prompt information.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the method according to any one of claims 1 to 7.

10. A computer device, characterized in that: The computer device comprises a memory, a processor and a program stored and executable on the memory, and the program implements the steps of the method according to any one of claims 1 to 7 when executed by the processor.

Images