Method and system for online monitoring and fault early warning and positioning of overhead line

Abstract

The invention discloses an overhead line on-line monitoring and fault early warning positioning method and system. The method comprises the following steps: acquiring traveling wave signals by using sensors mounted at each measurement point of the overhead line; using a deep learning algorithm to identify a fault traveling wave signal and a fault type; using a traveling wave distance measurement technology and a GIS to obtain the distance of each fault point relative to the measurement point, matching the geographic information and the topological structure of the overhead line, and obtaining the initial position of each fault point; virtual simulation software is utilized to independently inject the fault into the initial position of each fault point obtained correspondingly for the traveling wave signal of each fault, and the traveling wave signal is simulated and generated; and for the traveling wave signal of each fault, comparing the similarity between the traveling wave signal of each fault and the traveling wave signal generated by corresponding simulation, determining the final fault position of the fault point according to a similarity result, and generating a fault early warning positioning prompt. According to the invention, the accuracy of online monitoring data of the overhead line can be improved, the timeliness of fault early warning is realized, and the fault positioning precision is improved.

Description

Technical Field

[0002] The present application relates to the technical field of power line monitoring, and in particular to a method and system for online monitoring and fault early warning location of overhead lines. Background Art

[0003] Overhead lines in power systems are an important part of power transmission and are crucial to ensuring the safe and stable operation of power systems. With the continuous expansion of power system scale and the increase in power demand, the monitoring and maintenance requirements for overhead lines are becoming increasingly higher.

[0004] At present, there are two main ways to monitor overhead lines in power systems: manual inspection and online monitoring. Among them, manual inspection is one of the traditional monitoring methods, which mainly relies on professionals to regularly inspect the lines, but this method is not only inefficient, but also difficult to detect potential faults in time. The other type is sensor-based online monitoring technology, which installs sensor equipment on overhead lines to monitor electrical parameters and environmental parameters in real time, and transmits data to the background for analysis and processing. This technology can improve monitoring efficiency and coverage to a certain extent, but there are still problems such as inaccurate monitoring data, untimely fault warnings, and low fault location accuracy.

[0005] In the actual application process, the monitoring means in the existing technology still need to be improved, especially in fault warning and accurate fault location. Conventional means are often difficult to achieve timely detection and location of potential faults. These problems seriously affect the safe and stable operation of overhead lines, increase the line trip rate, and reduce the overall power supply reliability of the power system. Summary of the invention

[0006] In order to improve the accuracy of online monitoring data of overhead lines, realize the timeliness of fault warning, and enhance the accuracy of fault location, the present application provides a method and system for online monitoring and fault warning location of overhead lines.

[0007] In a first aspect, the present application discloses a method for online monitoring and fault early warning location of overhead lines, comprising: The traveling wave signals are collected by using sensors installed at preset measuring points on each line of the overhead line; The deep learning algorithm is used to extract features from the collected traveling wave signals, identify the traveling wave signals of the faults and the fault types; the traveling wave ranging technology is used to determine the time when the traveling wave signal of each fault reaches the measurement point, and obtain the distance of each fault point relative to the measurement point; Using GIS technology, the distances of each fault point relative to the measurement point are matched with the geographic information and topological structure of the overhead line to obtain the preliminary location of each fault point; Using virtual simulation software, a virtual model of the overhead line is built according to the parameters of the actual overhead line; using the built-in function of the virtual simulation software, for each fault traveling wave signal, a fault of the corresponding identified fault type is injected separately at the preliminary position of each fault point obtained correspondingly, and using the virtual simulation software to set the traveling wave signal collection point according to the measurement point of the overhead line, the simulated generated traveling wave signal is correspondingly collected; For each fault traveling wave signal, compare its similarity with the corresponding simulated traveling wave signal, select the initial position of the fault point corresponding to the simulated traveling wave signal with the greatest similarity as the final fault position of the fault point, and generate fault warning location prompt information.

[0008] By adopting the above scheme, traveling wave signals are collected at preset measurement points on each line to achieve comprehensive and accurate monitoring; deep learning algorithms are used to identify the traveling wave signals of faults and the type of faults to achieve timely fault warnings; traveling wave ranging technology is used to determine the location of the fault point, and GIS technology is combined to accurately match the geographic information of the fault point to enhance the reliability of the system; virtual simulation software is used to simulate fault signals, and the fault point is located by comparing the similarities, and the initially determined fault point is relocated to avoid positioning errors caused by multiple branch lines.

[0009] Preferably, it also includes: Use sensors to collect environmental data of the lines where each measuring point is located; The time when the traveling wave signal of the determined fault reaches the measuring point and the environmental data of the line where the measuring point is located are input into the distance calculation model to obtain the distance of the fault point relative to the measuring point; the distance calculation model adopts a neural network model, and the input is the time when the traveling wave signal of the fault reaches the measuring point and the environmental data of the line where each measuring point is located, and the output is the distance of the fault point relative to the measuring point, which is generated through training of historical environmental data, the time when the traveling wave signal of historical faults reaches the measuring point, and the distance of the corresponding historical actual fault point relative to the measuring point.

[0010] By adopting the above scheme, the influence of environmental data on the transmission of traveling wave signals is taken into consideration, and the distance of the fault point relative to the measurement point is more accurately obtained by using the neural network model, thereby improving the accuracy of fault location.

[0011] Preferably, it also includes: Count the locations of fault points in the overhead line within a certain period of time, calculate and obtain the locations of fault points with a fault frequency greater than a preset frequency, and record them as high-frequency fault points; The overhead line is divided into several sections according to its length and type. The two ends of each section are used as preset measurement points to determine whether there are high-frequency fault points in the current section. If so, the section is divided to obtain several sub-sections so that the high-frequency fault points are evenly distributed in each sub-section, and the two ends of each sub-section are used as budgeted measurement points.

[0012] By adopting the above scheme, the overhead lines are divided into areas based on high-frequency fault points and measurement points are set, which enables accurate monitoring and early warning of high-frequency fault points, and improves the timeliness and effectiveness of fault early warning.

[0013] Preferably, it also includes: When selecting the initial position of the fault point corresponding to the simulated traveling wave signal with the largest similarity, the maximum similarity is compared with the preset similarity; if the maximum similarity is less than the preset similarity, the initial position of the fault point corresponding to the simulated traveling wave signal with the largest similarity is not directly selected as the final fault position of the fault point, but the optimized deep learning algorithm is selected to re-identify the traveling wave signal of the fault, and the similarity is re-compared until the maximum similarity is greater than the preset similarity.

[0014] By adopting the above scheme, considering that the traveling wave signal with fault is identified incorrectly, resulting in a low similarity in the end, we choose to optimize the deep learning algorithm and re-identify the traveling wave signal of the fault to ensure the reliability of fault warning positioning.

[0015] Preferably, it also includes: Acquire the geographic information and topological structure of the overhead line in real time; acquire the geographic information and topological structure of the overhead line at the time of collecting the traveling wave signal of the fault in advance, and match the acquired distance of the fault point relative to the measuring point with the geographic information and topological structure of the overhead line at the time of collecting the traveling wave signal of the fault.

[0016] By adopting the above scheme, taking into account that the overhead line may change, the geographic information and topological structure of the overhead line when the traveling wave signal is collected are obtained in real time to ensure the accuracy of matching and improve the accuracy of fault location.

[0017] Preferably, the traveling wave signal for each fault, separately injecting a fault whose fault type is the corresponding identified fault type at the initial position of each fault point obtained correspondingly, comprises: Fault injection is performed according to the sequence of collected traveling wave signals of each fault, and fault injection is completed at the preliminary position of each fault point obtained corresponding to the traveling wave signal of the fault by parallel processing. For traveling wave signals of multiple faults with the same sequence, multiple servers with virtual models of the same overhead line built based on virtual software are used to complete the fault injection of the preliminary position of the fault point obtained corresponding to the traveling wave signal of each fault.

[0018] By adopting the above scheme, the fault injection of multiple preliminary fault point positions corresponding to the traveling wave signal of each fault is processed in parallel, and the fault injection of the preliminary fault point positions corresponding to the traveling wave signals of multiple faults at the same time is processed in a distributed manner, thereby improving the efficiency and accuracy of the fault injection.

[0019] Preferably, it also includes: Receive fault feedback information for the final fault location. If the ratio of accurate feedback for the final fault location to all feedback in the fault feedback information is greater than the preset ratio value, no operation is performed. Otherwise, the deep learning algorithm is continuously optimized to identify the traveling wave signals of the fault on the subsequently collected traveling wave signals until the ratio of accurate feedback for the final fault location to all feedback in the fault feedback information is greater than the preset ratio value.

[0020] By adopting the above solution, the feedback fault location information is used to determine whether the current fault location is accurate. If it is inaccurate, the recognition algorithm is further optimized to improve the accuracy of the location.

[0021] In a second aspect, the present application provides a system for online monitoring and fault early warning location of overhead lines, comprising: The traveling wave signal acquisition module is used to collect traveling wave signals using sensors installed at preset measuring points on each overhead line; the fault traveling wave signal identification module is used to extract features from the collected traveling wave signals using a deep learning algorithm, and identify the traveling wave signals of the fault and the type of fault; The module for determining the preliminary position of the fault point is used to determine the time when the traveling wave signal of each fault reaches the measuring point by using the traveling wave ranging technology, and obtain the distance of each fault point relative to the measuring point; using the GIS technology, the distance of each fault point relative to the measuring point is matched with the geographic information and topological structure of the overhead line to obtain the preliminary position of each fault point; The simulated traveling wave signal acquisition module is used to build a virtual model of the overhead line according to the parameters of the actual overhead line using the virtual simulation software; using the built-in function of the virtual simulation software, for each fault traveling wave signal, a fault of the corresponding identified fault type is injected separately at the preliminary position of each fault point obtained correspondingly, and the virtual simulation software is used to set the traveling wave signal acquisition points according to the measurement points of the overhead line to correspondingly acquire the simulated generated traveling wave signal; The module for determining the final position of the fault point is used to compare the similarity of the traveling wave signal of each fault with the corresponding simulated traveling wave signal, select the initial position of the fault point corresponding to the simulated traveling wave signal with the greatest similarity as the final fault position of the fault point, and generate fault warning location prompt information.

[0022] By adopting the above solution, deep learning algorithms, traveling wave ranging technology, GIS technology, and multi-branch line fault simulation technology are used to jointly achieve the timeliness of fault warning and improve the accuracy of fault location.

[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. Collect traveling wave signals through multiple sensors, and use deep learning algorithms to extract features from the collected traveling wave signals. Adjust feature weights based on the signal quality of the traveling wave signals, and use fusion features to better identify the traveling wave signals of faults and the types of faults, so as to achieve timely fault warning. Combine traveling wave ranging technology and GIS to complete the preliminary location of the fault point, and use virtual simulation software to simulate fault injection based on the preliminary location of the fault point. Compare the simulated traveling wave signals with the collected traveling wave signals, and re-locate the preliminary determined fault point location to avoid positioning errors caused by multiple branch lines and improve the accuracy of fault location. 2. Determine the high-frequency fault points in the overhead line based on the monitored fault location information, and determine the measurement points of the overhead line considering the high-frequency fault points, so as to realize the accurate monitoring and early warning of the high-frequency fault points in the overhead line, and improve the timeliness and effectiveness of the fault early warning; 3. Real-time acquisition of geographic information and topological structure of overhead lines, ensuring that the distance of the acquired fault point relative to the measurement point is accurately matched with the geographic information and topological structure of the overhead lines, thus improving the accuracy of fault location. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a flow chart of the method for online monitoring and fault early warning location of overhead lines described in a specific embodiment; Figure 2 A diagram of a fault point determined by using traveling wave ranging technology in the method for online monitoring and fault early warning location of overhead lines described in a specific embodiment; Figure 3 It is a structural schematic diagram of the overhead line online monitoring and fault early warning and positioning system described 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 early warning location of overhead lines, and the specific steps include: S1. Collect the traveling wave signal of the overhead line.

[0029] Specifically, a number of measurement points are preset according to the topological structure, length and type of the overhead line, and a traveling wave signal acquisition sensor is installed at each measuring point. The installed sensors are used to collect traveling wave signals propagated in each line in the overhead line, that is, the traveling wave current generated at the moment of failure of the line fault point.

[0030] S2. Use deep learning algorithms to analyze the collected traveling wave signals to identify the faulty traveling wave signals and fault types.

[0031] Specifically, the collected traveling wave signal is preprocessed in advance; the preprocessing includes: signal filtering processing, that is, using signal processing technology to perform filter matching and filtering processing on the signal to suppress noise and interference.

[0032] Construct a deep learning neural network model, which can be a convolutional neural network or a recurrent neural network, and is recorded as the first neural network model. The input of the model is the collected traveling wave signal, and the output of the model is the traveling wave signal of the fault and the fault type, which is generated by training the fault traveling wave signal with historically labeled fault types; specifically, the collected traveling wave signal is feature extracted using the deep learning neural network model, and the traveling wave signal of the fault and the fault type are identified using the extracted features; the extracted features include: time domain features (mean, variance, peak, pulse factor, etc.) and frequency domain features (power spectrum, energy spectrum), etc.; the fault model includes: ground fault, short circuit fault, etc.

[0033] S3. Use the traveling wave ranging technology to determine the time when the traveling wave signal of each fault reaches the measurement point, and obtain the distance of each fault point relative to the measurement point.

[0034] Specifically, Figure 2 As shown, using traveling wave ranging technology, such as double-end traveling wave positioning technology, the time when the identified fault traveling wave signal reaches the corresponding measuring points M and N is determined, which are recorded as TM and TN respectively. The propagation speed of the traveling wave in the conductor is V (close to the speed of light). The distance between the fault point K generating the fault traveling wave signal and the measuring point M is calculated as: LMK=(LMN)+(TM-TN)×V / 2, and the distance between the fault point K and the measuring point N is calculated as: LNK=(LMN)-(TM-TN)×V / 2.

[0035] Since more than one fault traveling wave signal may be identified, the time for each fault traveling wave signal to reach the corresponding measurement point is determined, and then the distance of the fault point generating each fault traveling wave signal relative to the measurement point is calculated.

[0036] S4. Using GIS technology, the distance of each fault point relative to the measuring point is matched with the geographic information and topological structure of the overhead line to obtain the preliminary location of each fault point.

[0037] In order to further determine the specific location of the fault point, the geographic information and topological structure of the overhead line are obtained using GIS technology, and the distance of each fault point relative to the measurement point is matched with the geographic information and topological structure of the overhead line. The matching process includes: obtaining the matching geographical location of each measurement point of the overhead line, and determining the initial location of each fault point according to the distance of each fault point relative to the measurement point; Among them, since the overhead line may change due to line maintenance and modification, in order to ensure the accuracy of the matching, the geographic information and topological structure of the overhead line are obtained in real time, the time of obtaining the geographic information and topological structure of the overhead line is aligned with the time of collecting the traveling wave signal, the geographic information and topological structure of the overhead line at the time of collecting the traveling wave signal of the fault is obtained, and the distance of the acquired fault point relative to the measurement point is matched with the geographic information and topological structure of the overhead line at the time of collecting the traveling wave signal of the fault. In addition, considering that there are a large number of complex line structures in the overhead line, such as: multiple branches, ring networks, T-connections, etc., multiple branches will also result in the possibility of multiple preliminary positions of each fault point determined based on the distance of each fault point relative to the measurement point.

[0038] S5. Use virtual simulation software to perform fault simulation according to the determined preliminary positions of each fault point, and collect the traveling wave signal generated by the simulation.

[0039] Specifically, since there are multiple situations in the initial positions of each fault point determined, in order to further accurately determine the position of the fault point, software simulation is used to generate a traveling wave signal to assist in locating the fault point.

[0040] Select appropriate virtual simulation software, such as MATLAB / Simulink or PSS / E, and build a virtual model of the overhead line based on the actual parameters of the overhead line (such as line length, conductor model, tower type, etc.), including key components such as conductors, poles, insulators, grounding devices, and line impedance, capacitance, inductance and other parameters in the virtual model of the overhead line.

[0041] By programming or using the built-in functions of the virtual simulation software (such as: a fault simulation function that can simulate various types of faults), for the traveling wave signal of each fault, a fault of the corresponding identified fault type is injected separately at the preliminary position of each corresponding fault point; for example: the preliminary position of the fault point K includes K1, K2 and K3, and a fault injection of the identified grounding fault type is performed at the K1 position alone, a fault injection of the identified grounding fault type is performed at the K2 position, and a fault injection of the identified grounding fault type is performed at the K3 position.

[0042] In an overhead line with a multi-branch structure, traveling waves will be refracted and reflected at the branch points. In order to ensure the accuracy of the simulation environment, the refraction and reflection coefficient calculations are introduced at each line branch point to simulate the actual propagation of traveling waves at the branch points. The introduced refraction and reflection coefficients can be used to obtain the corresponding refraction and reflection technology based on the experimental measurement data through data fitting technology.

[0043] The virtual simulation software is used to set the traveling wave signal collection points according to the preset measurement points of the overhead line, simulate the actual traveling wave sensor, and collect the simulated traveling wave signals accordingly. For example, obtain the simulated traveling wave signal X1 at the K1 position, the simulated traveling wave signal X2 at the K2 position, and the simulated traveling wave signal X3 at the K3 position.

[0044] S6. Compare the collected fault traveling wave signal with the simulated traveling wave signal to obtain the final position of the fault point and generate fault warning location prompt information.

[0045] Specifically, for each fault traveling wave signal, compare its similarity with the corresponding simulated traveling wave signal, select the initial position of the fault point corresponding to the simulated traveling wave signal with the greatest similarity as the final fault position of the fault point, and generate fault warning location prompt information; for example: the current traveling wave signal with the greatest similarity to the fault traveling wave signal is X3, then the corresponding final fault position of the fault point is K3, and the corresponding fault warning location prompt information of the K3 position fault is generated.

[0046] In addition, a visual interface can also be designed to display the location of the fault point of the overhead line in real time and receive fault feedback information from maintenance personnel to verify the final fault location; if the ratio of accurate feedback on the final fault location to all feedback in the fault feedback information is greater than the preset ratio value, it indicates that the final fault location is relatively accurate and no other operations are performed; otherwise, it indicates that the accuracy of the final fault location still needs to be improved, and the deep learning algorithm is continuously optimized to identify the faulty traveling wave signals of the subsequent collected traveling wave signals until the ratio of accurate feedback on the final fault location to all feedback in the fault feedback information is greater than the preset ratio value, and the optimization operation is stopped.

[0047] In a specific embodiment, considering that environmental factors may affect the transmission rate of the traveling wave signal, thereby causing errors in the calculation of the distance between the fault point and the measuring point, in order to ensure the accuracy of fault location, the method further includes: Using sensors to collect environmental data of the line where each measuring point is located; the environmental data includes: temperature, humidity and electromagnetic interference; The time when the traveling wave signal of the determined fault reaches the measuring point and the environmental data of the line where the measuring point is located are input into the distance calculation model to obtain the distance of the fault point relative to the measuring point.

[0048] Among them, the line where the measuring point is located is pre-set; the distance calculation model adopts a neural network model, recorded as the second neural network model, the input of the model is the time when the traveling wave signal of the fault reaches the measuring point, and the environmental data of the line where each measuring point is located, and the output is the distance of the fault point relative to the measuring point; the model is generated by training through historical environmental data, the time when the traveling wave signal of historical faults reaches the measuring point, and the distance of the corresponding historical actual fault point relative to the measuring point.

[0049] In a specific embodiment, considering the existence of high-frequency fault points in overhead lines, accurate monitoring and early warning of high-frequency fault points are achieved, and the timeliness and effectiveness of fault early warning are improved. The method also includes: According to the final position of the acquired fault point, the position of the fault point in the overhead line within a period of time is counted, and the position of the fault point with a fault frequency greater than a preset frequency is calculated and recorded as a high-frequency fault point; the period of time of time can be set manually.

[0050] The overhead line is divided into several sections according to the length and type of the overhead line, so that the length difference between the overhead lines of every two sections is not greater than the preset length and different types of overhead lines (such as high-voltage overhead lines, medium-voltage overhead lines and low-voltage overhead lines) are not in the same section, and the two ends of each section are used as preset measuring points; the measuring points can also be adjusted according to the length of the overhead line, so that a preset number of measuring points are set in a certain length of overhead line.

[0051] Determine whether there are high-frequency fault points in the current section; if so, continue to divide the section to obtain several sub-sections so that the high-frequency fault points are evenly distributed in each sub-section, and use the two ends of each sub-section as the budgeted measurement points.

[0052] In a specific embodiment, considering that there may be a recognition error in identifying the fault traveling wave signal or there may be a problem with the collected traveling wave signal, resulting in a large difference from the fault traveling wave signal generated by simulation, in order to avoid such a situation and ensure the accuracy of fault point identification as much as possible, the method further includes: When selecting the initial position of the fault point corresponding to the simulated traveling wave signal with the largest similarity, the maximum similarity is compared with a preset similarity; the preset similarity can be set manually.

[0053] If the maximum similarity obtained is less than the preset similarity, the initial position of the fault point corresponding to the simulated traveling wave signal with the largest similarity is not directly selected as the final fault position of the fault point, but the deep learning algorithm is optimized and the corresponding collected traveling wave signal is re-identified; the optimization method can select incremental training or adjust parameters. If the recognition result is still a faulty traveling wave signal and the maximum similarity is still less than the preset similarity, it is determined that there is a problem with the collected traveling wave signal, and the spare sensor of the measuring point is used to collect the traveling wave signal in the next time period, and the fault point position at the current stage is not analyzed; wherein, each measuring point is provided with a spare traveling wave signal collection sensor.

[0054] In a specific embodiment, in order to improve the efficiency of fault location as much as possible, the method further includes: injecting the traveling wave signal for each fault into the preliminary position of each fault point obtained correspondingly, and the fault type corresponding to the identified fault type includes: According to the order of the collected traveling wave signals of each fault, the initial position of the corresponding fault point is injected in sequence; for the traveling wave signals of multiple faults with the same order, different servers are used to complete the fault injection simulation of the initial position of the fault point obtained corresponding to the traveling wave signal of one fault, and the different servers are all constructed with the same virtual model of the overhead line using virtual software; Among them, for the initial positions of multiple fault points corresponding to the traveling wave signal of each fault, a parallel processing method is used to inject a fault of the corresponding identified fault type into the initial position of each fault point obtained accordingly, that is, multiple parallel simulation tasks are set to synchronously complete the fault injection at the initial position of each fault point.

[0055] like Figure 3 As shown, the embodiment of the present application discloses a system for online monitoring and fault early warning location of overhead lines, specifically comprising: The traveling wave signal acquisition module 101 is used to collect traveling wave signals using sensors installed at preset measurement points on each overhead line; The fault traveling wave signal recognition module 102 is used to extract features of the collected traveling wave signals using a deep learning algorithm, and to identify the fault traveling wave signals and the fault type; The fault point preliminary location determination module 103 is used to use the traveling wave ranging technology to determine the time when the traveling wave signal of each fault reaches the measurement point, and obtain the distance of each fault point relative to the measurement point; using the GIS technology, the obtained distance of each fault point relative to the measurement point is matched with the geographic information and topological structure of the overhead line to obtain the preliminary location of each fault point; The simulated traveling wave signal acquisition module 104 is used to build a virtual model of the overhead line according to the parameters of the actual overhead line using the virtual simulation software; using the built-in function of the virtual simulation software, for each fault traveling wave signal, a fault of the corresponding identified fault type is injected separately at the preliminary position of each fault point obtained correspondingly, and the virtual simulation software is used to set the traveling wave signal acquisition points according to the measurement points of the overhead line to correspondingly acquire the simulated generated traveling wave signal; The final fault location determination module 105 is used to compare the similarity between the traveling wave signal of each fault and the corresponding simulated traveling wave signal, select the initial position of the fault point corresponding to the simulated traveling wave signal with the greatest similarity as the final fault location of the fault point, and generate fault warning location prompt information In a specific embodiment, the system further includes: an environmental data acquisition module 106 for acquiring environmental data of the line where each measuring point is located using a sensor; The module 103 for determining the preliminary position of the fault point is also used to input the time when the traveling wave signal of the determined fault reaches the measuring point and the environmental data of the line where the measuring point is located into a distance calculation model to obtain the distance of the fault point relative to the measuring point; the distance calculation model adopts a neural network model, and the input is the time when the traveling wave signal of the fault reaches the measuring point and the environmental data of the line where each measuring point is located, and the output is the distance of the fault point relative to the measuring point, which is generated through training of historical environmental data, the time when the traveling wave signal of historical faults reaches the measuring point, and the distance of the corresponding historical actual fault point relative to the measuring point.

[0056] In a specific embodiment, the method further includes: a traveling wave signal measurement point preset module 107, which is used to count the positions of fault points in the overhead line within a period of time, calculate and obtain the positions of fault points with a fault frequency greater than a preset frequency, and record them as high-frequency fault points; divide the overhead line into several sections according to the length and type of the overhead line, and use the two ends of each section as preset measurement points to determine whether there is a high-frequency fault point in the current section; if so, divide the section, obtain several sub-sections so that the high-frequency fault points are evenly distributed in each sub-section, and use the two ends of each sub-section as the budgeted measurement points.

[0057] In a specific embodiment, the module 105 for determining the final position of the fault point in the system is also used to compare the maximum similarity with the preset similarity when selecting the initial position of the fault point corresponding to the simulated traveling wave signal with the greatest similarity; if the obtained maximum similarity is less than the preset similarity, the initial position of the fault point corresponding to the simulated traveling wave signal with the greatest similarity is not directly selected as the final fault position of the fault point, but an optimized deep learning algorithm is selected and the corresponding collected traveling wave signal is re-identified; if the identification result is still a faulty traveling wave signal and the maximum similarity is still less than the preset similarity, it is determined that there is a problem with the collected traveling wave signal, and the traveling wave signal is collected using a spare sensor in the next time period.

[0058] In a specific embodiment, the fault point preliminary location determination module 103 in the system is also used to obtain the geographic information and topological structure of the overhead line in real time; the geographic information and topological structure of the overhead line at the time of collecting the faulty traveling wave signal are obtained in advance, and the distance of the acquired fault point relative to the measuring point is matched with the geographic information and topological structure of the overhead line at the time of collecting the faulty traveling wave signal.

[0059] The embodiment of the present application also discloses a computer-readable storage medium.

[0060] Specifically, the computer-readable storage medium stores a computer program that can be loaded by a processor and execute the method for online monitoring and fault warning and locating of overhead lines as described above. 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.

[0061] The embodiment of the present application also discloses a computer device.

[0062] 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 early warning and locating of overhead lines.

[0063] 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 location of overhead lines, characterized in that: include: The traveling wave signals are collected by using sensors installed at preset measuring points on each line of the overhead line; Use deep learning algorithms to extract features from the collected traveling wave signals and identify the traveling wave signals and fault types of the faults; Using the traveling wave ranging technology, determine the time when the traveling wave signal of each fault reaches the measurement point, and obtain the distance of each fault point relative to the measurement point; Using GIS technology, the distances of each fault point relative to the measurement point are matched with the geographic information and topological structure of the overhead line to obtain the preliminary location of each fault point; Using virtual simulation software, a virtual model of the overhead line is built according to the parameters of the actual overhead line; using the built-in function of the virtual simulation software, for each fault traveling wave signal, a fault of the corresponding identified fault type is injected separately at the preliminary position of each fault point obtained correspondingly, and using the virtual simulation software to set the traveling wave signal collection point according to the measurement point of the overhead line, the simulated generated traveling wave signal is correspondingly collected; For each fault traveling wave signal, compare its similarity with the corresponding simulated traveling wave signal, select the initial position of the fault point corresponding to the simulated traveling wave signal with the greatest similarity as the final fault position of the fault point, and generate fault warning location prompt information.

2. The method for online monitoring and fault early warning location of overhead lines according to claim 1, characterized in that: Also includes: Use sensors to collect environmental data of the lines where each measuring point is located; The time when the traveling wave signal of the determined fault reaches the measuring point and the environmental data of the line where the measuring point is located are input into the distance calculation model to obtain the distance of the fault point relative to the measuring point; the distance calculation model adopts a neural network model, and the input is the time when the traveling wave signal of the fault reaches the measuring point and the environmental data of the line where each measuring point is located, and the output is the distance of the fault point relative to the measuring point, which is generated through training of historical environmental data, the time when the traveling wave signal of historical faults reaches the measuring point, and the distance of the corresponding historical actual fault point relative to the measuring point.

3. The method for online monitoring and fault early warning location of overhead lines according to claim 1, characterized in that: Also includes: Count the locations of fault points in the overhead line within a certain period of time, calculate and obtain the locations of fault points with a fault frequency greater than a preset frequency, and record them as high-frequency fault points; The overhead line is divided into several sections according to its length and type. The two ends of each section are used as preset measurement points to determine whether there are high-frequency fault points in the current section. If so, the section is divided to obtain several sub-sections so that the high-frequency fault points are evenly distributed in each sub-section, and the two ends of each sub-section are used as budgeted measurement points.

4. The method for online monitoring and fault early warning location of overhead lines according to claim 1, characterized in that: Also includes: When selecting the initial position of the fault point corresponding to the simulated traveling wave signal with the greatest similarity, the maximum similarity is compared with a preset similarity; If the maximum similarity obtained is less than the preset similarity, the initial position of the fault point corresponding to the simulated traveling wave signal with the largest similarity is not directly selected as the final fault position of the fault point. Instead, the optimized deep learning algorithm is selected and the corresponding collected traveling wave signal is re-identified. If the identification result is still a faulty traveling wave signal and the maximum similarity is still less than the preset similarity, it is determined that there is a problem with the collected traveling wave signal, and the spare sensor is used to collect the traveling wave signal in the next time period.

5. The method for online monitoring and fault early warning location of overhead lines according to claim 1, characterized in that: Also includes: Obtain geographic information and topological structure of overhead lines in real time; The geographic information and topological structure of the overhead line at the time of collecting the traveling wave signal of the fault are obtained in advance, and the distance of the acquired fault point relative to the measuring point is matched with the geographic information and topological structure of the overhead line at the time of collecting the traveling wave signal of the fault.

6. The method for online monitoring and fault early warning location of overhead lines according to claim 1, characterized in that: The traveling wave signal for each fault, separately injecting a fault type corresponding to the identified fault type at the initial position of each fault point obtained corresponding to the fault signal, includes: According to the order of the collected traveling wave signals of each fault, the fault injection is performed at the initial position of the corresponding fault point, and the fault injection is completed at the initial position of each fault point obtained corresponding to the traveling wave signal of the fault by parallel processing; For the traveling wave signals of multiple faults with the same sequence, multiple servers with virtual models of the same overhead line constructed based on virtual software are used to complete the fault injection of the preliminary position of the fault point obtained corresponding to the traveling wave signal of each fault.

7. The method for online monitoring and fault early warning location of overhead lines according to claim 1, characterized in that: Also includes: Receive fault feedback information for the final fault location. If the ratio of accurate feedback for the final fault location to all feedback in the fault feedback information is greater than the preset ratio value, no operation is performed. Otherwise, the deep learning algorithm is continuously optimized to identify the traveling wave signals of the fault on the subsequently collected traveling wave signals until the ratio of accurate feedback for the final fault location to all feedback in the fault feedback information is greater than the preset ratio value.

8. A system for online monitoring and fault early warning location of overhead lines, characterized in that: include: A traveling wave signal acquisition module, used to collect traveling wave signals using sensors installed at preset measurement points on each overhead line; The fault traveling wave signal recognition module is used to extract features of the collected traveling wave signals using a deep learning algorithm, and to identify the fault traveling wave signals and the fault type; The module for determining the preliminary position of the fault point is used to determine the time when the traveling wave signal of each fault reaches the measuring point by using the traveling wave ranging technology, and obtain the distance of each fault point relative to the measuring point; using the GIS technology, the distance of each fault point relative to the measuring point is matched with the geographic information and topological structure of the overhead line to obtain the preliminary position of each fault point; The simulated traveling wave signal acquisition module is used to build a virtual model of the overhead line according to the parameters of the actual overhead line using the virtual simulation software; using the built-in function of the virtual simulation software, for each fault traveling wave signal, a fault of the corresponding identified fault type is injected separately at the preliminary position of each fault point obtained correspondingly, and the virtual simulation software is used to set the traveling wave signal acquisition points according to the measurement points of the overhead line to correspondingly acquire the simulated generated traveling wave signal; The module for determining the final position of the fault point is used to compare the similarity of the traveling wave signal of each fault with the corresponding simulated traveling wave signal, select the initial position of the fault point corresponding to the simulated traveling wave signal with the greatest similarity as the final fault position of the fault point, and generate fault warning location 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