Distribution line tree line fault diagnosis method and isolation integrated switch device

Abstract

The invention discloses a distribution line tree line fault diagnosis method and a fault isolation integrated switch device, and relates to the technical field of distribution line fault diagnosis. The method comprises the steps of extracting a jump slope distribution sequence of a high-frequency traveling wave signal, performing feature extraction on the jump slope distribution sequence through an autoregression moving average model to obtain feature parameters of the autoregression moving average model, and learning the feature parameters through a support vector machine to obtain a fault type. According to the method, light-weight and high-precision tree line fault identification can be realized.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution line fault diagnosis, and in particular to a distribution line tree line fault diagnosis method and an isolation integrated switch device. Background Art

[0002] Power distribution lines in mountainous areas inevitably cross forests, grasslands and other areas, and tree line failures are difficult to avoid.

[0003] Tree line faults are caused by trees growing rapidly and touching the power lines, or by trees leaning against the power lines under wind. The transition resistance of the grounded trees during tree line faults can be as high as 10~100 It has the characteristics of small grounding current, weak electrical characteristics and difficult to identify. If the fault is not diagnosed and removed in time, it can easily cause forest fires.

[0004] Therefore, early warning and rapid handling of tree line faults are the key to preventing forest fires and ensuring power supply safety. However, the existing pole-mounted automatic switches can only effectively detect faults with a resistance of 5 Below ground fault, it is difficult to accurately detect tree line faults. Summary of the Invention

[0005] The present invention provides a method for diagnosing tree line faults in a distribution line and an integrated isolation switch device, so as to solve the problem that existing pole-mounted switches are difficult to accurately detect tree line faults.

[0006] The present invention is achieved through the following technical solutions: A first aspect of the present invention provides a method for diagnosing a tree line fault in a distribution line, comprising: Extracting a jump slope distribution sequence of a high-frequency traveling wave signal, wherein the high-frequency traveling wave signal is obtained from a signal output from a fault point of a power distribution line; Extracting features of the jump slope distribution sequence using an autoregressive moving average model to obtain characteristic parameters of the autoregressive moving average model; The characteristic parameters are learned by a support vector machine to obtain the fault type.

[0007] This method diagnoses tree-line faults based on high-frequency traveling wave signals. It first extracts the transition slope distribution sequence of the high-frequency traveling wave signal to obtain the transition characteristics of the high-frequency traveling wave signal. Then, it uses an autoregressive moving average model and a support vector machine to extract and identify the signal characteristics. The high-frequency traveling wave signal can be acquired using a high-frequency current sensor. Both the autoregressive moving average model and the support vector machine are lightweight models. The signal transition characteristics are used as input to the support vector machine for fault diagnosis. This method offers high diagnostic accuracy and is suitable for low-computing line switch applications. It can be installed in pole-mounted switches for accurate diagnosis of tree-line faults.

[0008] Furthermore, the extracting of the transition slope distribution sequence of the high-frequency traveling wave signal includes: Calculating extreme value points in a high-frequency traveling wave signal of a predetermined duration; The signal interval between two adjacent extreme value points is regarded as an extreme value interval. N The signal slope of the extreme value interval is calculated as:

[0009] in, Indicates the N The signal slope of the extreme value interval, 、 Respectively expressed in N +1 extreme value interval and in the N The value obtained by summing the sampled signals within the extreme value interval is 、 Respectively represent n +1 extreme point and n The sampling time between extreme points, It is n +1 extreme point and n The sampling time difference between the extreme points, that is, N The duration of the extreme value interval, N=n ; After calculating the signal slopes of all extreme value intervals, a jump slope distribution sequence is obtained, and the sequence points of the jump slope distribution sequence are the signal slopes of the corresponding extreme value intervals.

[0010] Furthermore, the step of calculating the extreme value points in the high-frequency traveling wave signal of the predetermined duration includes: Calculate the signal difference between each sampling point and the adjacent previous sampling point in the high-frequency traveling wave signal of a predetermined duration; If the polarity of the signal difference changes compared to the signal difference at the previous sampling point, the sampling point is marked as an extreme point.

[0011] Furthermore, the feature extraction of the jump slope distribution sequence by the autoregressive moving average model to obtain the feature parameters of the autoregressive moving average model includes: The jump slope distribution sequence is input into the autoregressive moving average model for regression fitting, and the obtained fitting parameters are the characteristic parameters of the autoregressive moving average model. The autoregressive moving average model is expressed as:

[0012] in, is the jump slope distribution sequence, x =1, 2, ...,N , N is the sequence length, and are characteristic parameters of the autoregressive moving average model, p and q are orders of the autoregressive moving average model, a previous value of the jump slope distribution sequence, is a previous random error term, is a prediction error of the autoregressive moving average model.

[0013] Further, before the jump slope distribution sequence is extracted by the autoregressive moving average model to obtain the characteristic parameters of the autoregressive moving average model, the method further comprises: determining the order of the autoregressive moving average model by minimizing a description length criterion, p and q the description length criterion is expressed as:

[0014] wherein, is the minimized description length, is a variance of the prediction error of the autoregressive moving average model, N is the length of the jump slope distribution sequence, is the order of the autoregressive moving average model, or , p and q is the order of the autoregressive moving average model.

[0015] Further, the method further comprises: evaluating the order of the autoregressive moving average model by a sensitivity function of the description length criterion, and the sensitivity function is expressed as:

[0016] wherein, represents the sensitivity function, is a change amount of under different orders, is the order change amount, is the order of the autoregressive moving average model, or .

[0017] Further, the learning of the characteristic parameters by the support vector machine to obtain the fault type comprises: inputting the characteristic parameters into the support vector machine model, and obtaining the output of the support vector machine model as the fault type; The support vector machine model is trained by inputting a plurality of high-frequency traveling wave signals output by a power distribution line with tree line faults and a high-frequency traveling wave signal output by a power distribution line without tree line faults, and the fault type includes a tree line fault type and a non-fault type.

[0018] In a second aspect of the present application, a power distribution line tree line fault isolation integrated switch device is provided, comprising a pole-mounted switch, a high-frequency current sensor, an intelligent diagnosis unit and an intelligent feeder terminal. The high-frequency current sensor is embedded in the pole-mounted switch and is used to collect high-frequency traveling wave signals output from a fault point of a power distribution line. The intelligent diagnosis unit is used to record the wave head arrival time of the high-frequency traveling wave signals and determine whether the fault point is within a preset range according to the wave head arrival time. If the fault point is within the preset range, the power distribution line tree line fault diagnosis method of any one of the first aspect of the present application is executed to obtain the fault type of the high-frequency traveling wave signals, and the fault type is sent to the intelligent feeder terminal. The intelligent feeder terminal is used to control a vacuum circuit breaker in the pole-mounted switch and cut off the vacuum circuit breaker when a fault occurs.

[0019] Further, the intelligent diagnosis unit is also clock-synchronized with other intelligent diagnosis units on the power distribution line and is also used to obtain the wave head arrival time of the high-frequency traveling wave signals recorded by other intelligent diagnosis units. The determination of whether the fault point is within the preset range of the intelligent diagnosis unit according to the wave head arrival time comprises: if the wave head arrival time of the high-frequency traveling wave signals recorded by the intelligent diagnosis unit itself is earlier than the wave head arrival time of the high-frequency traveling wave signals recorded by other intelligent diagnosis units, it is determined that the fault point is within the preset range.

[0020] Further, the intelligent diagnosis unit is also used to calculate the shortest time difference according to the wave head arrival time of the high-frequency traveling wave signals recorded by the intelligent diagnosis unit itself and the wave head arrival time of the high-frequency traveling wave signals recorded by other intelligent diagnosis units, and locate the position of the fault point according to the shortest time difference to inform the staff to timely eliminate the fault.

[0021] Compared with the prior art, the present application has the following advantages and beneficial effects: Based on high-frequency traveling wave signals, the jump slope distribution sequence of the high-frequency traveling wave signals is first extracted, and then signal feature extraction and recognition are realized through an autoregressive moving average model and a support vector machine, the whole process realizes lightweight and high precision, and is suitable for the pole-mounted switch scene.

[0022] An integrated pole-mounted switch device is proposed, which uses high-frequency current sensors to collect high-frequency traveling wave signals, performs fault location and fault diagnosis through an intelligent diagnostic unit, and cooperates with the intelligent feeder terminal to control the pole-mounted switch circuit breaker to achieve accurate positioning and phase selection of tree line faults and rapid isolation and power outage, solving the problems of difficult identification and positioning of tree line faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the following briefly introduces the drawings required for use in the examples. It should be understood that the following drawings only illustrate certain embodiments of the present invention and should not be considered as limiting the scope. A person of ordinary skill in the art can also derive other relevant drawings based on these drawings without inventive effort. In the drawings: Figure 1 This is a flow chart of a method for diagnosing tree line faults in a power distribution line according to an embodiment of the present invention; Figure 2 Schematic diagram of a model training method combining an autoregressive moving average model and a support vector machine according to an embodiment of the present invention; Figure 3 This is an application scenario diagram of a distribution line tree line fault isolation integrated switch device according to an embodiment of the present invention; Figure 4 This is a functional diagram of an intelligent diagnostic unit according to an embodiment of the present invention. DETAILED DESCRIPTION

[0024] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with examples and drawings. The exemplary embodiments of the present invention and their descriptions are only used to explain the present invention and are not intended to limit the present invention.

[0025] It should be noted that the terms "including" and "having" and any variations thereof in the specification and claims of the present invention and the above-mentioned drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to or inherent to other steps or units of the device.

[0026] The terms used in the various embodiments of the present invention are only used for the purpose of describing specific embodiments and are not intended to limit the various embodiments of the present invention. As used herein, the singular form is intended to also include the plural form, unless the context clearly indicates otherwise. Unless otherwise specified, all terms used herein (including technical terms and scientific terms) have the same meaning as those commonly understood by those skilled in the art to which the various embodiments of the present invention belong. The terms (such as those defined in generally used dictionaries) will be interpreted as having the same meaning as the contextual meaning in the relevant technical field and will not be interpreted as having an idealized meaning or an overly formal meaning unless clearly defined in the various embodiments of the present invention.

[0027] In a first aspect, an embodiment of the present invention provides a method for diagnosing tree line faults in a distribution line, so as to achieve accurate diagnosis of tree line faults and ensure power supply safety.

[0028] like Figure 1 This is a flow chart of a distribution line tree line fault diagnosis method proposed by the present invention, which includes the following steps.

[0029] S1, extract the jump slope distribution sequence of the high-frequency traveling wave signal; S2, extracting features from the jump slope distribution sequence through the autoregressive moving average model to obtain the characteristic parameters of the autoregressive moving average model; S3, takes the characteristic parameters of the autoregressive moving average model as input, learns the characteristic parameters through the support vector machine, and obtains the fault type.

[0030] The high-frequency traveling wave signal is obtained from the output of the distribution line fault point and can be collected by a high-frequency current sensor. This sensor can be embedded in the pole-mounted switch to receive the high-frequency traveling wave signal in real time. The sampling frequency is typically set between 2 and 10 MHz. Because the high-frequency traveling wave signal collected by the high-frequency current sensor contains a large amount of low-frequency interference signals such as the power frequency, the collected traveling wave signal is low-frequency filtered. The cutoff frequency is typically set between 500 and 1000 Hz to obtain a relatively pure high-frequency traveling wave signal.

[0031] In step S1, a signal transition feature is extracted from the high-frequency traveling wave signal, which can reflect the line fault status. In this embodiment, a transition slope distribution sequence of the high-frequency traveling wave signal is extracted. The transition slope is the slope change between signal transition points. The high-frequency traveling wave signal for a period of time is intercepted, and the signal transition point is first found among the sampling points. The sampling signal between adjacent previous and next transition points is used as an interval to calculate the slope. The slopes of multiple intervals constitute the transition slope distribution sequence.

[0032] In one embodiment, the method of extracting the transition slope distribution sequence of the high-frequency traveling wave signal by using the extreme value point in the high-frequency traveling wave signal as the transition point specifically includes: S2-1, calculating the extreme value points in the high-frequency traveling wave signal of a predetermined duration.

[0033] In order to balance the calculation accuracy and calculation efficiency, the predetermined time length can be selected as a high-frequency traveling wave signal of the first 5s-10s. S2-2, the signal interval between two adjacent extreme value points is regarded as an extreme value interval, N The signal slope of the extreme value interval is calculated as:

[0034] in, Indicates the N The signal slope of the extreme value interval, 、 Respectively expressed in N +1 extreme value interval and in the N The value obtained by summing the sampled signals within the extreme value interval is 、 Respectively represent n +1 extreme point and n The sampling time between extreme points, It is n +1 extreme point and n The sampling time difference between the extreme points, that is, N The duration of the extreme value interval, N=n ; S2-3, after calculating the signal slopes of all extreme value intervals, we get the jump slope distribution sequence { }.

[0035] The sequence points of the transition slope distribution sequence are the signal slopes corresponding to the extreme value intervals, and the transition slope distribution sequence can be visualized as a two-dimensional signal transition slope distribution probability graph, that is, the horizontal axis is the extreme value interval and the vertical axis is the transition slope of the extreme value interval.

[0036] Furthermore, the method for calculating the extreme value points in the high-frequency traveling wave signal of a predetermined duration in step S2-1 specifically includes: S2-1-1, calculate each sampling point in the high-frequency traveling wave signal of a predetermined duration i The adjacent previous sampling point i -1 signal difference , expressed as:

[0037] in, Indicates the ithe signal difference of the i-th sampling point i , , the signal amplitude of the i-th sampling point and the i-1-th sampling point i , i -1, respectively i if -1≤0, then .

[0038] S2-1-2, if the signal difference of the i-th sampling point i is changed in polarity compared with the signal difference of the i-1-th sampling point , then the i-th sampling point is marked as an extreme point. i

[0039] In step S3, the jump slope distribution sequence is extracted by an Auto-Regression and Moving Average Model (ARMA), i.e., the jump slope distribution sequence is fitted as an input of the ARMA model. The ARMA model is an important method for studying time series, which is "mixed" based on an Auto-Regression model (AR) and a Moving Average model (MA), has the characteristics of wide application range and small prediction error, and is expressed by the formula:

[0040] wherein, is the jump slope distribution sequence, x =1, 2, …, N , N is the sequence length, and are characteristic parameters of the Auto-Regression and Moving Average Model, p and q are orders of the Auto-Regression and Moving Average Model, is a previous value of the jump slope distribution sequence, is a previous random error term, is a prediction error of the Auto-Regression and Moving Average Model.

[0041] Specifically, the modeling steps of the ARMA model are as follows: (1) The input data is judged to determine whether it is a stationary non-random sequence, if it is stationary, then it directly enters step (2); if it is not stationary, it is processed, and then enters step (2).

[0042] (2) The established model is identified and ordered by autocorrelation and partial autocorrelation functions, and combined with AIC or BIC criteria. ​​

[0043] (3) After completing model identification and order determination, enter the model parameter estimation stage.

[0044] (4) After completing parameter estimation, the fitted model is tested for adaptability. If the fitted model passes the test, the prediction phase begins. If the model fails the test, the model identification and test are repeated, i.e., step (2) is repeated to reselect the model.

[0045] (5) Finally, a highly adaptable fitting model is used to predict the future trend of the sequence.

[0046] In this example, an autoregressive moving average model is established using M series of transition slope distributions of high-frequency traveling wave signals when a tree-line fault occurs and Z series of transition slope distributions of signals when no tree-line fault occurs. M ranges from 10 to 30, Z ranges from 10 to 20, and the signal duration is 5 seconds. Furthermore, the order of the autoregressive moving average model is determined by minimizing the description length (MDL) criterion. p and q , the criterion of minimizing description length is expressed as:

[0047] in, To minimize description length, is the variance of the forecast error of the autoregressive moving average model, N is the length of the jump slope distribution sequence, is the order of the autoregressive moving average model, or , p and q is the order of the autoregressive moving average model.

[0048] Furthermore, the order of the autoregressive moving average model is evaluated by minimizing the sensitivity function describing the length criterion. The sensitivity function is expressed as:

[0049] in, represents the sensitivity function, At different levels The amount of change, is the order variation, is the order of the autoregressive moving average model, or .

[0050] The order of the autoregressive moving average model is determined to be between 2 and 4 by minimizing the description length criterion. The selected order is verified by the residual test. If the residual is a random normal distribution and is not autocorrelated, it means that all useful signals have been extracted into the autoregressive moving average model. The designed order is then used. p 、 q Build an autoregressive moving average model, otherwise re-order and evaluate again.

[0051] The autoregressive moving average model is used to perform regression fitting on the high-frequency traveling wave signal to be diagnosed, and the characteristic parameters of the autoregressive moving average model are obtained. and , input it into the Support Vector Machine (SVM) model for diagnosis and obtain the fault type.

[0052] In one embodiment, the support vector machine model uses M The jump slope distribution sequence and Z The signal jump slope distribution sequence when no treeline fault occurs is trained. The features of the two jump slope distribution sequences are first extracted by the autoregressive moving average model. The labeled fault type is used as the model output to obtain the optimal model parameters. The characteristic parameters to be diagnosed are predicted based on the model parameters, and the model output is the diagnosed fault type, where the fault type includes treeline fault type and non-fault type.

[0053] like Figure 2 The figure shows a schematic diagram of a model training method combining an autoregressive moving average model and a support vector machine, which includes the following steps.

[0054] Stpe1, obtain M high-frequency traveling wave signals S1 with tree-line faults and Z high-frequency traveling wave signals S2 without tree-line faults as training samples; Step 2, low-pass filter the high-frequency traveling wave signals S1 and S2 to obtain pure high-frequency traveling wave signals S1 and S2; Step 3: Extract the transition slope distribution sequences S1' and S2' from the pure high-frequency traveling wave signals S1 and S2; Step 4, determine the order of the autoregressive moving average model by minimizing the description length criterion; Step 5: Use the pure high-frequency traveling wave signals S1' and S2' as the input of the autoregressive moving average model for regression fitting to obtain the autoregressive moving average model fitting parameters; Step 6: Evaluate the order of the autoregressive moving average model using the sensitivity function. If the requirements are met (residuals are randomly normally distributed and not autocorrelated), proceed to the next step; otherwise, return to Step 4 and re-determine the order. In step 7, the autoregressive moving average model fitting parameters obtained in step 5 are used as feature parameter inputs, and the fault type marked by the feature parameters is used as the prediction output to train the support vector machine model to obtain the optimal model parameters.

[0055] The present invention adopts a lightweight autoregressive moving average model to extract signal features, combined with a lightweight support vector machine as a diagnostic model, to achieve lightweight deployment and accurate identification of the tree line diagnostic method of the present invention on the pole-mounted switch, and the diagnostic method can be carried out in parallel with the grounding fault diagnosis of the pole-mounted switch without conflict.

[0056] In a second aspect, an embodiment of the present invention further provides a distribution line tree line fault isolation integrated switch device, see Figure 3 As shown, the integrated switchgear includes a pole-mounted switch, a high-frequency current sensor, an intelligent diagnostic unit and an intelligent feeder terminal.

[0057] Pole-mounted switches are installed on distribution lines. When used, multiple integrated switch devices are spaced apart on the distribution lines, preferably within a spacing of 3-5 kilometers. Pole-mounted switches utilize standardized interfaces and are designed to comply with the grid company's standardized design plan for integrated primary and secondary circuit breakers (deeply integrated, pillar-type). The switch itself must meet IP67 protection, effectively isolating the operating environment from dust, dirt, rain, and humidity, ensuring reliable operation. Inlet and outlet terminals are equipped with insulating caps for full insulation, and the switch tripping time must be ≤45ms.

[0058] A high-frequency current sensor (high-frequency CT) and vacuum circuit breaker are embedded within the pole-mounted switch. The high-frequency current sensor uses a self-integrating Rogowski coil sensor with excellent high-frequency response. It collects high-frequency traveling wave signals from the distribution line fault point in real time. The sampling frequency is set between 2 and 10 MHz, and a low-pass filter is connected to filter the collected traveling wave signals (with a cutoff frequency set between 500 and 1000 Hz). The filtered signal is then transmitted to the intelligent diagnostic unit 303 via the aviation socket on the pole-mounted switch.

[0059] The intelligent diagnosis unit uses a microcontroller chip, which has two functions: one is to locate the fault point, and the other is to automatically extract and identify the fault characteristics. Figure 4 The functional diagram of the intelligent diagnostic unit is shown.

[0060] Function 1 includes: recording the arrival time of the wave head of the high-frequency traveling wave signal, and judging whether the fault point is within the preset range based on the wave head arrival time.

[0061] The fault traveling wave current signal propagates from the fault point to both ends of the conductor at a constant speed. v=0.29m / ns. Multiple intelligent diagnostic units installed on the line use nanosecond-level clock synchronization to achieve wide-area clock synchronization of less than 200ns. By recording the arrival time of the high-frequency traveling wave's front end and synchronizing the signals, the two intelligent diagnostic units closest to the fault point are identified based on the different arrival times of the front end.

[0062] Taking the intelligent diagnosis unit 1 as an example, the method of judging whether the fault point is within the preset range based on the wave head arrival time is specifically implemented in the following way: By synchronizing the clocks with other intelligent diagnostic units (intelligent diagnostic unit 2, intelligent diagnostic unit 3, etc.) on the distribution line, and obtaining the wave head arrival time of the high-frequency traveling wave signal recorded by other intelligent diagnostic units through information synchronization; By comparing the wave head arrival time of the high-frequency traveling wave signal recorded by itself with the wave head arrival time of the high-frequency traveling wave signal recorded by other intelligent diagnostic units, if the wave head arrival time of the high-frequency traveling wave signal recorded by itself is earlier than the wave head arrival time of the high-frequency traveling wave signal recorded by other intelligent diagnostic units, it can be judged that it is the intelligent diagnostic unit closest to the fault point; otherwise, it is not the intelligent diagnostic unit closest to the fault point, and its own position can be judged based on the comparison sorting.

[0063] When the wave head arrival time of the high-frequency traveling wave signal recorded by itself is earlier than the wave head arrival time of the high-frequency traveling wave signal recorded by other intelligent diagnostic units, the trigger condition for judging that the fault point is within its own preset range is met, and function 2 is executed by itself, otherwise it is not executed.

[0064] Function 2 includes: feature extraction and fault type diagnosis of high-frequency traveling wave signals.

[0065] A lightweight data processing module and treeline fault diagnosis are deployed on the intelligent diagnosis unit, which is similar to feature extraction and fault type diagnosis. The data processing module pre-processes the collected high-frequency traveling wave signal, including (1): intercepting the high-frequency traveling wave signal of a predetermined time length starting from the wave head arrival time, and extracting the jump slope distribution sequence of the high-frequency traveling wave signal of the predetermined time length; the treeline fault diagnosis includes (2): extracting the features of the jump slope distribution sequence through the autoregressive moving average model to obtain the feature parameters of the autoregressive moving average model; and (3): learning the feature parameters through the support vector machine, outputting the fault type, and transmitting the treeline fault diagnosis results to the intelligent feeder terminal through the aviation socket.

[0066] The above data processing and tree line fault diagnosis process refers to the distribution line tree line fault diagnosis method of the aforementioned embodiment of the present invention. The autoregressive moving average model and support vector machine are deployed in the intelligent diagnosis unit after being trained using the aforementioned training method.

[0067] The intelligent feeder terminal receives the fault diagnosis results from the intelligent diagnostic unit. When the fault diagnosis result is a tree line fault, the vacuum circuit breaker in the switch on the control column is opened, cutting off the line to prevent the tree line fault from causing a large-scale forest fire.

[0068] Furthermore, the intelligent diagnosis unit can also accurately locate the fault point through function 1, specifically: When the fault point is judged to be within the preset range based on the wave front arrival time, the shortest time difference is calculated based on the wave front arrival time of the high-frequency traveling wave signal recorded by itself and the wave front arrival time of the high-frequency traveling wave signal recorded by other intelligent diagnosis units; The position of the fault point is located according to the shortest time difference, so as to notify the staff to eliminate the fault in time.

[0069] The shortest time difference mentioned above is the difference in wave head arrival time recorded by the two intelligent diagnosis units at both ends closest to the fault point. Taking the earliest arrival time of the wave head received by the intelligent diagnosis unit 2 as an example, when the installation distance between the two intelligent diagnosis units is known, L In the case of The shortest distance between the fault point and the intelligent diagnosis unit 2 can be calculated to accurately locate the fault and send the location to the control center, so that the operation and maintenance personnel can troubleshoot the fault and restore power supply in time.

[0070] The present invention utilizes high-frequency current sensors, intelligent feeder terminals, and intelligent diagnostic systems, and adopts a technology for accurately locating and quickly isolating distribution line wildfire faults. First, a signal jump slope distribution probability graph is drawn, and then the features of the signal jump slope distribution probability graph are extracted and identified through an autoregressive moving average model and a support vector machine. The entire process is lightweight and highly precise, suitable for line switching situations with low computing power, and can achieve accurate positioning and phase selection of tree line faults and rapid isolation and power outage.

[0071] In a third aspect, embodiments of the present invention further provide an electronic device comprising a processor and a memory, where the number of processors may be one or more. The memory, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules. The processor executes the software programs, instructions, and modules stored in the memory to perform various functional applications and data processing of the electronic device, thereby implementing the distribution line tree fault diagnosis method of any of the above-mentioned embodiments of the present invention.

[0072] The memory may primarily include a program storage area and a data storage area. The program storage area may store an operating system and at least one application required for a function; the data storage area may store data generated based on the use of the terminal, etc. Furthermore, the memory may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state memory device. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories may be connected to the electronic device via a network. Examples of the aforementioned networks include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0073] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the distribution line tree line fault diagnosis method of any of the above embodiments of the present invention.

[0074] The computer storage media of the embodiments of the present invention may employ any combination of one or more computer-readable media. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or component, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device, or component.

[0075] A computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0076] In a fifth aspect, an embodiment of the present invention further provides a computer program product, which, when executed on a computer, enables the computer to execute the distribution line tree line fault diagnosis method of any of the above embodiments of the present invention.

[0077] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for diagnosing tree line faults in a power distribution line, characterized in that: include: Extracting a jump slope distribution sequence of a high-frequency traveling wave signal, wherein the high-frequency traveling wave signal is obtained from a signal output from a fault point of a power distribution line; Extracting features of the jump slope distribution sequence using an autoregressive moving average model to obtain characteristic parameters of the autoregressive moving average model; The characteristic parameters are learned by a support vector machine to obtain the fault type.

2. The method for diagnosing a tree line fault in a power distribution line according to claim 1, wherein: The step of extracting a transition slope distribution sequence of a high-frequency traveling wave signal comprises: Calculating extreme value points in a high-frequency traveling wave signal of a predetermined duration; The signal interval between two adjacent extreme value points is regarded as an extreme value interval. N The signal slope of the extreme value interval is calculated as: in, Indicates the N The signal slope of the extreme value interval, 、 Respectively expressed in N +1 extreme value interval and in the N The value obtained by summing the sampled signals within the extreme value interval is 、 Respectively represent n +1 extreme point and n The sampling time of the extreme point, It is n +1 extreme point and n The sampling time difference between the extreme points, that is, N The duration of the extreme value interval, N=n ; After calculating the signal slopes of all extreme value intervals, a jump slope distribution sequence is obtained, and the sequence points of the jump slope distribution sequence are the signal slopes of the corresponding extreme value intervals.

3. The method for diagnosing a tree line fault in a power distribution line according to claim 2, characterized in that: The step of calculating the extreme value points in the high-frequency traveling wave signal of a predetermined duration includes: Calculate the signal difference between each sampling point and the adjacent previous sampling point in the high-frequency traveling wave signal of a predetermined duration; If the polarity of the signal difference changes compared to the signal difference at the previous sampling point, the sampling point is marked as an extreme point.

4. The method for diagnosing tree line faults in a power distribution line according to claim 1, characterized in that: The feature extraction of the jump slope distribution sequence by the autoregressive moving average model to obtain the feature parameters of the autoregressive moving average model includes: The jump slope distribution sequence is input into the autoregressive moving average model for regression fitting, and the obtained fitting parameters are the characteristic parameters of the autoregressive moving average model. The autoregressive moving average model is expressed as: in, is the jump slope distribution sequence, x =1, 2, ..., N , N is the sequence length, and is the characteristic parameter of the autoregressive moving average model, p and q is the order of the autoregressive moving average model, is the previous value of the jump slope distribution sequence, is the random error term in the early stage, is the forecast error of the autoregressive moving average model.

5. The method for diagnosing tree line faults in a power distribution line according to claim 4, characterized in that: Before extracting features from the jump slope distribution sequence using an autoregressive moving average model to obtain feature parameters of the autoregressive moving average model, the method further includes: The order of the autoregressive moving average model is determined by minimizing the description length criterion p and q , the minimization description length criterion is expressed as: in, To minimize description length, is the variance of the forecast error of the autoregressive moving average model, N is the length of the jump slope distribution sequence, is the order of the autoregressive moving average model, or , p and q is the order of the autoregressive moving average model.

6. The method for diagnosing a tree line fault in a power distribution line according to claim 5, characterized in that: The method further includes evaluating the order of the autoregressive moving average model by minimizing a sensitivity function describing a length criterion, wherein the sensitivity function is expressed as: in, represents the sensitivity function, At different levels The amount of change, is the order variation, is the order of the autoregressive moving average model, or .

7. The method for diagnosing a tree line fault in a power distribution line according to claim 1, characterized in that: The method of learning the characteristic parameters by a support vector machine to obtain the fault type includes: Inputting the characteristic parameters into a support vector machine model, obtaining an output of the support vector machine model as a fault type; The support vector machine model is trained by inputting high-frequency traveling wave signals output by multiple distribution lines with tree-line faults and high-frequency traveling wave signals output by distribution lines without tree-line faults. The fault types include tree-line fault types and non-fault types.

8. The integrated switchgear for isolating tree line faults of distribution lines is characterized by: It includes pole-mounted switches, high-frequency current sensors, intelligent diagnostic units and intelligent feeder terminals; The high-frequency current sensor is embedded in the pole switch and is used to collect the high-frequency traveling wave signal output from the fault point of the distribution line; The intelligent diagnosis unit is used to record the wave front arrival time of the high-frequency traveling wave signal, and determine whether the fault point is within a preset range based on the wave front arrival time; and If the fault point is within a preset range, executing the distribution line tree line fault diagnosis method according to any one of claims 1 to 7 to obtain the fault type of the high-frequency traveling wave signal, and sending the fault type to the smart feeder terminal; The intelligent feeder terminal is used to control the vacuum circuit breaker in the pole switch according to the fault type, and cut off the vacuum circuit breaker when a fault occurs.

9. The integrated switchgear for isolating tree-line faults of distribution lines according to claim 8, characterized in that: The intelligent diagnostic unit is also synchronized with the clocks of other intelligent diagnostic units on the power distribution line, and is also used to obtain the wave head arrival time of the high-frequency traveling wave signal recorded by other intelligent diagnostic units; The method of judging whether the fault point is within the preset range of the intelligent diagnostic unit based on the wave head arrival time includes: if the wave head arrival time of the high-frequency traveling wave signal recorded by itself is earlier than the wave head arrival time of the high-frequency traveling wave signal recorded by other intelligent diagnostic units, then judging that the fault point is within the preset range.

10. The integrated switchgear for isolating tree-line faults of distribution lines according to claim 9, characterized in that: The intelligent diagnostic unit is also used to calculate the shortest time difference based on the wave head arrival time of the high-frequency traveling wave signal recorded by itself and the wave head arrival time of the high-frequency traveling wave signal recorded by other intelligent diagnostic units, and locate the position of the fault point according to the shortest time difference to notify the staff to eliminate the fault in time.

Images