Power distribution network fault accurate positioning method and system based on multi-source data fusion

Abstract

The invention relates to the technical field of fault positioning, in particular to a power distribution network fault accurate positioning method and system based on multi-source data fusion, and the method comprises the steps: collecting the corona current data of each segment line in a power distribution network in a preset time period in real time, and the current data and voltage data of the two ends of each segment line; according to the decreasing condition of the pulse amplitude in the corona current data of each section line, obtaining the pulse decreasing coefficient of each section line; acquiring the wide kurtosis of each section line through the rising time and the duration of the pulse in the corona current data of each section line; obtaining a corona discharge coefficient of each section line; obtaining an electrical distortion characteristic value of each section line; adjusting control parameters of chaotic mapping; and performing fault positioning on the power distribution network by using the chaotic mapping after the control parameter adjustment and combining with a grey wolf optimization algorithm. The invention aims to improve the accuracy of fault positioning.

Description

Technical Field

[0001] This application relates to the field of fault location technology, specifically to a method and system for accurate fault location in distribution networks based on multi-source data fusion. Background Technology

[0002] The power distribution network is a crucial link in the power system for distributing electrical energy. With economic development, the scale of the distribution network has expanded rapidly, the number of nodes has increased, and the network topology has become more complex, making fault location more difficult. To improve the accuracy of fault location and its resistance to interference, multi-source data fusion methods are commonly used to enhance data fault tolerance by integrating multiple data sources.

[0003] The Grey Wolf Optimization Algorithm is one of the commonly used fault location algorithms in power distribution networks. However, its optimization effect is easily affected by the initial population. Currently, the initial population of the Grey Wolf Optimization Algorithm is generally initialized through the Tent mapping. The control parameters in the Tent mapping are usually taken as fixed values. However, in rainy weather, voltage and current are easily distorted by rainwater, increasing the difficulty of fault location. If the control parameters are too small, the population search range will be small, causing the fault interval to be concentrated in a certain range and unable to cover all candidate intervals. If the control parameters are too large, it will destroy the stability of the chaotic sequence, increase the randomness of the fault interval, and thus affect the accuracy of fault location. Summary of the Invention

[0004] In view of the above, it is necessary to provide a method and system for accurate fault location in distribution networks based on multi-source data fusion, which improves the accuracy of fault location compared to traditional methods based on multi-source data fusion. In a first aspect, embodiments of this application provide a method for accurate fault location in a distribution network based on multi-source data fusion, the method comprising the following steps: Real-time acquisition of corona current data of each section of the distribution network within a preset time period, as well as current and voltage data at both ends of each section of the line; The pulse decay coefficient of each line segment is obtained by analyzing the decrease in pulse amplitude in the corona current data of each line segment; the peak width of each line segment is obtained by analyzing the rise time and duration of pulses in the corona current data of each line segment; and the corona discharge coefficient of each line segment is obtained by analyzing the distribution of rise time of all pulses in the corona current data of each line segment, as well as the pulse decay coefficient and the peak width. By comparing the current and voltage data at both ends of each line segment, and combining the distortion of the voltage and current data at both ends of each line segment, the electrical distortion characteristic value of each line segment is obtained; the control parameters of the chaotic mapping are adjusted by the corona discharge coefficient and the electrical distortion characteristic value. The chaotic mapping after adjusting the control parameters is combined with the gray wolf optimization algorithm to locate faults in the distribution network.

[0005] In one embodiment, the process of obtaining the pulse decrement coefficient is as follows: Calculate the first-order difference in time sequence of all pulse peak values ​​in the corona current data of each section of the line; The pulse decay coefficient is the percentage of negative numbers in the first-order difference result.

[0006] In one embodiment, the process of obtaining the broad kurtosis is as follows: The pulse with the largest peak value in the corona current data of each section of the line is obtained. The product of the rise time and half-wave time of the obtained pulse is calculated. The ratio of the product to the square of the peak width of the obtained pulse is used as the peak width of each section of the line.

[0007] In one embodiment, the process of obtaining the corona discharge coefficient is as follows: The rise time and probability density of all pulse peaks in the corona current data of each section of the line are normally fitted to obtain a normal curve, and the goodness of fit of the normal curve is calculated. The corona discharge coefficient is inversely proportional to the peak width and the goodness of fit, and directly proportional to the pulse decrement coefficient.

[0008] In one embodiment, the process of obtaining the electrical distortion feature value is as follows: The electrical difference coefficient of each section of the line is obtained by measuring the degree of difference in current data at both ends of each section of the line. By combining the distortion of voltage data and current data at both ends of each section, the harmonic correction coefficient of each section of the line is obtained. The electrical distortion characteristic values ​​are positively correlated with the electrical difference coefficient and the harmonic correction coefficient, respectively.

[0009] In one embodiment, the method for obtaining the electrical difference coefficient is as follows: Calculate the difference in voltage data at both ends of each section of the line, and calculate the difference in current data at both ends of each section of the line. The electrical difference coefficient is the average of the difference and the amount of difference.

[0010] In one embodiment, the method for obtaining the harmonic correction coefficient is as follows: Calculate the average value of the total voltage harmonic distortion rate of the voltage data at both ends of the line in each section; calculate the arithmetic mean of the total current harmonic distortion rate of the current data at both ends of the line in each section. The harmonic correction coefficient is the weighted sum of the average value and the arithmetic mean.

[0011] In one embodiment, the process of adjusting the control parameters of the chaotic mapping is as follows: Obtain the segmentation threshold of the corona discharge coefficient for all line segments, and adjust the control parameters of the chaotic mapping by using the average level of the electrical distortion characteristic values ​​of all line segments with corona discharge coefficients greater than the segmentation threshold.

[0012] In one embodiment, the method for adjusting the control parameters of the chaotic mapping is as follows: The mean value of the electrical distortion characteristics of all sections of the line with a corona discharge coefficient greater than the segmentation threshold is denoted as the mean distortion value. Calculate the product of the length of the preset range of the control parameter and the mean of the distortion, and use the sum of the product and the preset minimum value of the control parameter as the adjusted control parameter.

[0013] Secondly, embodiments of this application also provide a distribution network fault accurate location system based on multi-source data fusion, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above-described distribution network fault accurate location methods based on multi-source data fusion.

[0014] This application has at least the following beneficial effects: This application calculates the pulse decay coefficient to reflect the significant decrease in pulse amplitude in the corona current, indicating the change in discharge intensity caused by rain impacting the line; the kurtosis can reflect the narrowness of the corona current pulse, which is related to the degree of air ionization, further revealing the intensity of the discharge; the corona discharge coefficient comprehensively considers the distribution of pulse decay coefficient, kurtosis, and pulse rise time, and can comprehensively and accurately reflect the corona discharge characteristics of each section of the line under rainy weather. Furthermore, the electrical distortion characteristic value integrates the degree of difference between current and voltage data as well as harmonic distortion, which can accurately assess the degree of voltage and current distortion in each section of the line under complex conditions such as rainy weather. This assessment method not only considers the absolute difference of the data, but also combines the harmonic distortion, making the description of the line section status more comprehensive and accurate, and providing more reliable distortion characteristic information for subsequent fault location analysis. Furthermore, the control parameters of the chaotic mapping are adjusted according to the corona discharge coefficient and electrical distortion characteristic value, so that the control parameters can adaptively match the state changes of the distribution network under different weather conditions. This avoids problems such as an excessively small population search range or destruction of the stability of the chaotic sequence caused by inappropriate control parameters, thereby improving the accuracy and adaptability of the Grey Wolf optimization algorithm in fault location, thus enhancing the accuracy of fault location and helping to improve the efficiency of distribution network health management. Attached Figure Description

[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 A flowchart illustrating the steps of a method for accurate fault location in a distribution network based on multi-source data fusion, as provided in one embodiment of this application; Figure 2 A schematic diagram illustrating the steps for adjusting control parameters; Figure 3 This is a schematic diagram of the process for adjusting control parameters. Detailed Implementation

[0017] In the description of the embodiments in this application, the words "exemplary," "or," and "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplary," "or," and "for example" is intended to present the relevant concepts in a specific manner.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. It should be understood that, unless otherwise stated, " / " in this application means "or".

[0019] It should also be noted that the terms "first" and "second" in this application are used to distinguish similar objects, rather than to describe a specific order or sequence.

[0020] The following description, in conjunction with the accompanying drawings, details the specific scheme of the method and system for accurate fault location in distribution networks based on multi-source data fusion provided in this application.

[0021] Please see Figure 1 The diagram illustrates a flowchart of a method for accurate fault location in a distribution network based on multi-source data fusion, according to an embodiment of this application. The method includes the following steps: Step 1: Collect in real time the current overcurrent alarm information, voltage undervoltage alarm information, corona current data, and current and voltage data at both ends of each section of the distribution network within the preset time period.

[0022] Feeder terminals (FTUs) are used to collect overcurrent alarm information from each section of the distribution network, and transformer monitoring terminals (TTUs) are used to collect undervoltage alarm information from each section of the distribution network. The FTUs are installed at the feeder pole switches, and the TTUs are installed at the distribution transformers. When a fault occurs in any section of the distribution network, the undervoltage alarm information for each section has two possible values: 0 for normal power supply and 1 for undervoltage alarm. The FTUs may collect overcurrent information in two directions. The positive direction is defined as the flow from the system power supply to the faulty section. The overcurrent alarm information for each section has three possible values: 1 for positive overcurrent, 0 for no overcurrent, and -1 for reverse overcurrent. Smart meters are installed at both ends of each section to collect voltage and current data. Corona current data for each section is collected using a high-speed data acquisition card to analyze the characteristics of corona current data during rainfall. The two types of alarm information are collected independently, complementing each other in terms of timeliness and having overlapping and redundant coverage. Information fusion can reduce the impact of alarm information distortion on fault location. Among them, the section line refers to a specific line segment between two nodes in the distribution network, such as the line between two switches, or the line between a switch and a transformer.

[0023] In this embodiment, overcurrent alarm information, voltage undervoltage alarm information, voltage data, current data, and corona current data are all collected in real time. The collection frequency of overcurrent alarm information, voltage undervoltage alarm information, voltage data, and current data is 10kHz, and the collection frequency of corona current data is 1MHz. The collection frequency is preset by the user and can be set by the implementer according to the actual situation. This application does not impose any special restrictions.

[0024] Furthermore, to avoid the impact of different dimensions on subsequent analysis, the collected voltage data, current data and corona current data were normalized respectively.

[0025] In this embodiment, the Min-Max normalization method is used to normalize the collected voltage data, current data, and corona current data respectively. The maximum and minimum values ​​in the Min-Max normalization method are obtained by statistical analysis of historical long-period data. The Min-Max normalization method is a well-known technology and will not be described in detail in this application. In this application, unless otherwise specified, all normalization operations use the Min-Max normalization method.

[0026] Step 2: Obtain the pulse decay coefficient, peak width, and corona discharge coefficient of each line segment; obtain the electrical distortion characteristic value of each line segment; adjust the control parameters of the chaotic mapping using the corona discharge coefficient and electrical distortion characteristic value.

[0027] Step 2.1: Obtain the pulse decay coefficient of each line segment by analyzing the decrease in pulse amplitude in the corona current data of each line segment; obtain the peak width of each line segment by analyzing the rise time and duration of pulses in the corona current data of each line segment; obtain the corona discharge coefficient of each line segment by analyzing the distribution of rise time of all pulses in the corona current data of each line segment, as well as the pulse decay coefficient and the peak width.

[0028] During rainy weather, rainwater coming into contact with power distribution lines usually triggers a discharge phenomenon. Because the process of rainwater hitting the lines is random, the discharge phenomenon usually has complex discharge characteristics. For example, when rainwater hits the lines, a tip pulse discharge usually occurs. As the raindrops slide down the lines, protrusions form on the surface of the conductors, and the amplitude of adjacent pulses in the corona current gradually decreases.

[0029] The preset fault location period for the distribution network is used as an example, taking the t-th period as an example. In order to measure the amplitude decrease of adjacent pulses, the pulse decrease coefficient of each line segment is obtained by analyzing the pulse amplitude decrease in the corona current data of each line segment. Specifically: The pulse peak values ​​in the corona current data of each line segment are obtained. The first-order difference of all pulse peak values ​​in the corona current data of each line segment in terms of time sequence is calculated. The proportion of negative numbers in the first-order difference results of each line segment is used as the pulse decay coefficient of each line segment. This coefficient reflects the significance of the decrease in pulse amplitude in the corona current data of each line segment in the distribution network under rainy weather. The larger the calculated pulse decay coefficient, the higher the significance of the decrease in pulse amplitude in the corona current data of each line segment. The calculation of the first-order difference is a well-known technique and will not be described in detail in this application.

[0030] In this embodiment, the length of the cycle is 10 seconds. The length of the cycle is preset by the user and can be set by the implementer according to the actual situation. This application does not impose any special restrictions.

[0031] In this embodiment, the Automatic Multiscale-based Peak Detection (AMPD) algorithm is used to obtain the pulse peak value in the corona current data. The AMPD algorithm is a well-known technology and will not be described in detail here. As other implementation methods, based on the ability to obtain the pulse peak value in the corona current data, the implementer may use other existing technologies, such as peak and trough detection algorithms, etc. This application does not impose any special restrictions.

[0032] Since the actual rainfall intensity is dynamic, if the rainfall suddenly increases, the pulse amplitude of the corona current will increase instead of decreasing. Therefore, the pulse decrease coefficient alone cannot accurately reflect the degree of corona discharge in the distribution network under the influence of rainfall. Further analysis is needed in conjunction with the core mechanism of corona discharge, namely the degree of air ionization.

[0033] Rainwater increases the ionization of the air surrounding the conductor, rapidly forming a corona current, characterized by a steep rise edge and short half-wave time. To analyze the rise time and half-wave time characteristics of the corona current pulse under the influence of rain discharge, the peak width of each line segment is obtained by analyzing the rise time and duration of the pulses in the corona current data. Specifically: Taking the i-th segment of the distribution network as an example, the pulse with the largest peak value in the corona current data of the i-th segment is obtained. The product of the pulse's rise time and half-wave time is calculated. The ratio of this product to the square of the pulse's peak width is used as the peak width of the i-th segment. This peak width reflects the narrowness of the pulse in the corona current data of the i-th segment of the distribution network under rainy weather. The smaller the calculated peak width, the greater the ionization of the air around the i-th segment due to rainwater, the faster the corona current forms, and the narrower the pulse. The rise time refers to the time interval from 10% to 90% of the peak value, the half-wave time refers to the time interval from 50% of the peak value at the rising edge to 50% of the peak value at the falling edge, and the peak width refers to the time interval from 10% to 10% of the peak value at the rising edge to 10% of the peak value at the falling edge.

[0034] In practical applications, corona discharge in power distribution networks is not solely caused by rainwater. Factors such as conductor burrs and mud on insulator surfaces can also trigger corona discharge. Furthermore, the pulse width in the corona discharge current caused by these factors is relatively small. Therefore, it is difficult to accurately determine whether the corona discharge is caused by rainwater. Typically, the rise time of the pulse in the corona discharge current generated by rainwater impact on conductors exhibits a normal distribution.

[0035] Based on the above analysis, in order to analyze the normal distribution of the rise time of the pulses in the corona current, the normal distribution coefficient of each line segment is obtained by analyzing the distribution of the rise time of all pulses in the corona current data of each segment. Specifically: The rise time of each pulse in the corona current data of the i-th segment of the line is obtained. A class interval for the rise time is set, and the probability density of the rise time within each class interval is calculated. The rise time and probability density of all pulses in the corona current data of the i-th segment of the line are statistically analyzed to obtain a probability density histogram of the rise time. A normal distribution curve is obtained by fitting the probability density histogram to a normal distribution. The goodness of fit of the normal distribution curve is used as the normal distribution coefficient of the i-th segment of the line, reflecting the significance of the normal distribution of the rise time of the pulses in the corona current data of the i-th segment of the line. The smaller the calculated normal distribution coefficient, the more the rise time of the pulses conforms to a normal distribution, and the higher the significance of the corona discharge phenomenon of the i-th segment of the line being affected by rainfall. The probability density calculation, probability density histogram, and normal fitting are all known techniques and will not be elaborated upon in this application.

[0036] In this embodiment, the class interval is 100 μs, which was calculated from experimental data.

[0037] In this embodiment, the goodness of fit of the normal curve is obtained by the chi-square test method. The goodness of fit is specifically the chi-square coefficient. The chi-square test method is a well-known technique and will not be described in detail in this application. As other implementation methods, based on the ability to obtain the goodness of fit of the normal curve, the implementer may use other existing techniques, such as the Shapiro-Wilk test, etc. This application does not impose any special restrictions.

[0038] Based on the above analysis, the corona discharge coefficient of each section of the line is obtained by using the pulse decay coefficient, kurtosis, and normal distribution coefficient. Specifically: The corona discharge coefficient of each section of the line is inversely proportional to the peak width and normal distribution coefficient of each section of the line, and directly proportional to the pulse decay coefficient of each section of the line.

[0039] In this embodiment, the expression for the corona discharge coefficient of each section of the line is: In the formula, This represents the corona discharge coefficient of the i-th section of the line in the distribution network; This represents the pulse decay coefficient of the i-th section of the distribution network; This represents the peak width of the i-th segment of the distribution network. This represents the normal distribution coefficient of the i-th section of the distribution network; This indicates a preset value greater than 0, used to avoid a denominator of 0, and also to avoid... The value of affects the calculation results of the corona discharge coefficient. The value of should be extremely small. In this embodiment, The value is 0.005.

[0040] In another embodiment, the expression for the corona discharge coefficient of each section of the line is: In the formula, This represents the corona discharge coefficient of the i-th section of the line in the distribution network; This represents the pulse decay coefficient of the i-th section of the distribution network; This represents the peak width of the i-th segment of the distribution network. This represents the normal distribution coefficient of the i-th section of the distribution network.

[0041] It should be noted that before calculating the corona discharge coefficient, the pulse decline coefficient, kurtosis, and normal distribution coefficient are normalized to normalize them to the interval (0,1], ensuring that their contributions to the corona discharge coefficient are more balanced.

[0042] It should be noted that the corona discharge coefficient is used to characterize the degree of corona discharge caused by rainfall in each section of the distribution network. The larger the calculated corona discharge coefficient, the higher the degree of corona discharge caused by rainfall in each section of the distribution network.

[0043] Step 2.2: By comparing the current data and voltage data at both ends of each section of the line, and combining the distortion of the voltage data and the distortion of the current data at both ends of each section of the line, the electrical distortion characteristic value of each section of the line is obtained.

[0044] In power distribution networks, sections with severe corona discharge often accelerate the aging process of the lines, leading to more significant voltage and current distortions. Under such circumstances, the difficulty of accurately locating faults will also increase accordingly.

[0045] Based on the above analysis, to analyze the degree of voltage and current distortion under rainfall conditions, the threshold value of the corona discharge coefficient of all sections of the distribution network is used as the discharge threshold. Under the same rainfall conditions, the more significant the corona discharge, the higher the degree of aging of the section of the line, and the more significant the voltage and current distortion. Therefore, sections of the distribution network with corona discharge coefficients greater than the threshold value are selected, and each selected section is analyzed. By comparing the current and voltage data at both ends of each section of the line, and combining the distortion of the voltage and current data at both ends of each section of the line, the electrical distortion characteristic value of each section of the line is obtained. It should be noted that: the electrical distortion characteristic value of each section of the line with a corona discharge coefficient greater than the threshold value is calculated only. The specific process of obtaining the electrical distortion characteristic value is as follows: Due to the impact of rainfall, unstable voltage discharge exists in each section of the line, resulting in different voltage and current variations at both ends of the line. The difference in voltage data at both ends of each section of the line is calculated, as well as the difference in current data at both ends of each section of the line. The average of the difference and the difference in the difference between each section of the line is used as the electrical difference coefficient of each section of the line. This coefficient reflects the degree of voltage and current distortion in each section of the distribution network under the influence of rainfall. The larger the calculated electrical difference coefficient, the greater the degree of voltage and current distortion in each section of the line. Furthermore, due to the influence of the transmission distance of the line section itself, the voltage and current differences at both ends of the line section are relatively large. Therefore, it is necessary to further analyze the harmonic characteristics in the voltage and current data under rainfall. Rainfall corona discharge has nonlinear discharge characteristics, which interferes with the electromagnetic environment around the line and generates a large number of harmonics, leading to an increase in voltage and current harmonic distortion rates. The harmonic distortion degree of the voltage and current data is analyzed, and the electrical difference coefficient is corrected by the obtained harmonic distortion degree so that the electrical difference coefficient can more accurately reflect the distortion degree of voltage and current of the line section. The average value of the total voltage harmonic distortion rate of the voltage data at both ends of each line section is calculated. The arithmetic mean of the total current harmonic distortion rate of the current data at both ends of each line section is calculated. The weighted sum of the average value and the arithmetic mean is used as the harmonic correction coefficient of each line section. The electrical distortion characteristic value of each line section is positively correlated with the electrical difference coefficient and the harmonic correction coefficient of each line section. The calculation of the total harmonic distortion rate of voltage and the total harmonic distortion rate of current are well known techniques and will not be described in detail in this application.

[0046] In this embodiment, the fundamental frequency is 50Hz.

[0047] It should be noted that positive correlation means that the variables change in the same direction; when one variable increases, the other variable also increases, and when one variable decreases, the other variable also decreases.

[0048] In this embodiment, when calculating the weighted sum, the weights of the average and the arithmetic mean are 0.55 and 0.45, respectively, and are derived from experimental data.

[0049] In this embodiment, the Otsu threshold segmentation algorithm is used to obtain the segmentation threshold of the corona discharge coefficient. The Otsu threshold segmentation algorithm is a well-known technology and will not be described in detail here. As other implementation methods, based on the ability to obtain the segmentation threshold of the corona discharge coefficient, implementers may use other existing technologies, such as iterative threshold segmentation, global threshold segmentation, etc. This application does not impose any special restrictions.

[0050] In this embodiment, the calculation process for the voltage data difference is as follows: the voltage data of each end of each section of the line are arranged in time sequence to form the voltage sequence of each end of each section of the line, and the DTW (Dynamic Time Warping) distance between the voltage sequences at both ends of each section of the line is calculated. The calculation of the DTW distance is a well-known technique and will not be described in detail in this application. As other implementation methods, based on the ability to measure the degree of difference between voltage sequences, the implementer may use other existing techniques, such as Euclidean distance, etc. This application does not impose any special restrictions. The calculation process for the difference in current data is the same as the calculation process for the voltage data difference, except that the voltage data is replaced with current data.

[0051] In this embodiment, the expression for the electrical distortion characteristic value of each section of the line is: In the formula, This represents the electrical distortion characteristic value of the j-th section of the line; This indicates a normalization operation; This represents the electrical difference coefficient of the j-th section of the line; Represents an exponential function with the natural constant as its base, used to... The inverse proportional mapping is applied to the interval (0,1]. This represents the harmonic correction coefficient for the j-th section of the line.

[0052] In another embodiment, the expression for the electrical distortion characteristic value of each section of the line is: In the formula, This represents the electrical distortion characteristic value of the j-th section of the line; This indicates a normalization operation; This represents the electrical difference coefficient of the j-th section of the line; This represents the harmonic correction coefficient for the j-th section of the line.

[0053] It should be noted that the harmonic correction coefficient reflects the degree of voltage and current harmonic distortion in a section of the distribution network due to rainfall. The greater the impact of rainfall on the distribution network, the larger the electrical difference coefficient and the larger the harmonic correction coefficient. A larger harmonic correction coefficient indicates a higher likelihood of voltage and current distortion in the section of the network. Therefore, the harmonic correction coefficient exhibits an inverse proportional mapping effect. and The smaller the value, the less the correction for the electrical difference coefficient, and the larger the final calculated electrical distortion characteristic value, indicating that the voltage and current distortion of the distribution network section under the influence of rainy weather is more significant.

[0054] Step 2.3: Adjust the control parameters of the chaotic mapping using the corona discharge coefficient and the electrical distortion characteristic value.

[0055] The Grey Wolf Optimization Algorithm based on the Tent chaotic map is one of the commonly used fault location algorithms for power distribution networks. However, the control parameters of the Tent chaotic map are usually taken as fixed values. Under the influence of rainy weather, voltage and current data are usually highly distorted. If the control parameters are too small, the population search range will be small, and the fault interval will be concentrated in a certain range, failing to cover all candidate intervals. If the control parameters are too large, it will destroy the stability of the chaotic sequence, making the fault interval more random, thus affecting the accuracy of using the Grey Wolf Optimization Algorithm for power distribution network fault location. Therefore, it is necessary to adaptively adjust the size of the control parameters according to the significant distortion of the voltage and current data of the sections and lines in the power distribution network. The smaller the control parameters, the more suitable they are for periodic data. The larger the control parameters, the higher the chaos and the more suitable they are for data with significant distortion.

[0056] Based on the above analysis, the control parameters of the Tent chaotic mapping are adjusted by using the corona discharge coefficient and electrical distortion characteristic value of the line segment. The expression is as follows: In the formula, This represents the adjusted control parameters; B represents the average electrical distortion characteristic value of all sections of the distribution network whose corona discharge coefficient is greater than the segmentation threshold. Indicates the length of the preset value range of the control parameter; This indicates the preset minimum value of the control parameter. A diagram illustrating the control parameter adjustment steps is shown below. Figure 2 As shown in the diagram. The flowchart for adjusting the control parameters is as follows. Figure 3 As shown in the figure. Here, B is denoted as the mean distortion.

[0057] In this embodiment, the preset value range of the control parameter is [0,2], and the preset minimum value of the control parameter is 0. Both the preset value range and the preset minimum value of the control parameter are obtained through historical experiments.

[0058] Step 3: Use the chaotic mapping after adjusting the control parameters combined with the gray wolf optimization algorithm to locate faults in the distribution network.

[0059] Based on the network topology of the distribution network, construct interval functions respectively. and switching functions Where K represents the total number of distributed power sources; N represents the total number of sections of the distribution network. This indicates matrix multiplication by its positions, i.e., calculating the Hadamard product of the matrices; , Let X and Y represent the voltage fault state matrix and current fault state matrix of all line segments only when the k-th distributed power source is connected, respectively. Taking voltage fault as an example, if any line segment fails, the downstream line experiences voltage loss fault, meaning that the voltage state of each downstream line segment is 1, and the voltage state of each upstream line segment is 0. The size of the voltage fault state matrix is ​​N×1. X represents the state column vector of the distribution network line segments. T denotes matrix transpose. This indicates the status value of the first section of the line. When the first section of the line is faulty, When the first section of the line is normal, , This indicates the status value of the Nth section of the line. When the Nth section of the line experiences a fault, When the Nth section of the line is normal, ; The sign function is used to determine the direction of the overcurrent alarm, but does not determine the magnitude of the overcurrent. A positive overcurrent is 1, no overcurrent is 0, and reverse overcurrent is -1. Indicates calculation The Hadamard product of X, Indicates calculation The Hadamard product of X. Both the distribution network topology construction and the Hadamard product are well-known technologies and will not be elaborated upon in this application.

[0060] Among them, the interval function and the switching function are used to determine the expected alarm information for TTUs and FTUs when a distribution network fault occurs. By finding the expected alarm information that minimizes the difference from the actual information, fault interval location can be achieved. Objective functions for voltage undervoltage alarm information and current overcurrent alarm information are constructed respectively, with the following expressions: ; In the formula, , These represent the objective functions for voltage undervoltage alarm information and current overcurrent alarm information, respectively; N represents the total number of lines in the distribution network. These represent the actual voltage undervoltage alarm information and the actual current overcurrent alarm information for the nth line segment, respectively. This indicates the absolute value operation; Representing an interval function The nth element in; Represents the switching function The nth element in the equation; since there are three possible values ​​for the overcurrent alarm information: 1, 0, and -1, with 1 and -1 having an error of 2, their impact on the objective function differs from that of 1 to 0 and 0 to -1. Therefore, an XOR operation is used. Represents the XOR symbol.

[0061] Furthermore, M column vectors of length N are initialized using the Tent chaotic map as the initial gray wolf population, where the chaotic map uses adjusted control parameters. The probability of undervoltage faults in each section of the line is obtained by iterative optimization using the gray wolf optimization algorithm. ,in, The probability of a voltage failure in the nth section of the line is represented by ; M represents the number of individual gray wolves in a single iteration. This represents the state value of the nth segment of the line in the mth gray wolf individual; the overcurrent fault probability of each segment of the line is obtained using the method for obtaining the undervoltage fault probability; among them, the Tent chaotic mapping and the gray wolf optimization algorithm are well-known technologies, and will not be described in detail in this application.

[0062] Furthermore, based on the DS (Dempster-Shafer) synthesis rules, the undervoltage fault probability and overcurrent probability data of the distribution network are fused to calculate the fault probability of each section of the line. ,in, This represents the probability of failure in the nth section of the line; This represents the probability of a voltage failure in the nth section of the line. Let represent the overcurrent fault probability of the nth section of the line; N represents the total number of sections in the distribution network. The fault probabilities of all sections are sorted in descending order, and the first section is identified as the faulty line, thus enabling fault location in the distribution network.

[0063] In this embodiment, the value of M is 30, and the maximum number of iterations of the Grey Wolf optimization algorithm is 100. The values ​​of M and the maximum number of iterations are preset by the user, and the implementer can set them according to the actual situation. This application does not impose any special restrictions.

[0064] Based on the same inventive concept as the above method, this application embodiment also provides a distribution network fault accurate location system based on multi-source data fusion, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above-described distribution network fault accurate location methods based on multi-source data fusion.

[0065] In summary, this application uses the pulse decay coefficient to reflect the significant decrease in pulse amplitude in the corona current, indicating the change in discharge intensity caused by rain impacting the line; the kurtosis can reflect the narrowness of the corona current pulse, which is related to the degree of air ionization, further revealing the intensity of the discharge; the corona discharge coefficient comprehensively considers the distribution of pulse decay coefficient, kurtosis, and pulse rise time, and can comprehensively and accurately reflect the corona discharge characteristics of different sections of the line under rainy weather. Furthermore, the electrical distortion characteristic value integrates the degree of difference between current and voltage data as well as harmonic distortion, which can accurately assess the degree of voltage and current distortion in each section of the line under complex conditions such as rainy weather. This assessment method not only considers the absolute difference of the data, but also combines the harmonic distortion, making the description of the line section status more comprehensive and accurate, and providing more reliable distortion characteristic information for subsequent fault location analysis. Furthermore, the control parameters of the chaotic mapping are adjusted according to the corona discharge coefficient and electrical distortion characteristic value, so that the control parameters can adaptively match the state changes of the distribution network under different weather conditions. This avoids problems such as an excessively small population search range or destruction of the stability of the chaotic sequence caused by inappropriate control parameters, thereby improving the accuracy and adaptability of the Grey Wolf optimization algorithm in fault location, thus enhancing the accuracy of fault location and helping to improve the efficiency of distribution network health management.

[0066] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than that shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. Each block in a block diagram and / or flowchart, and combinations of blocks in a block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0067] It will be apparent to those skilled in the art that this application is not limited to the details of the exemplary embodiments described above, and that this application can be implemented in other specific forms without departing from its essential characteristics. Therefore, the embodiments described above should be considered exemplary and non-limiting in all respects.

Claims

1. A method for accurate fault location in distribution networks based on multi-source data fusion, characterized in that, The method includes the following steps: Real-time acquisition of corona current data of each section of the distribution network within a preset time period, as well as current and voltage data at both ends of each section of the line; The pulse decay coefficient of each line segment is obtained by analyzing the decrease in pulse amplitude in the corona current data of each line segment; the peak width of each line segment is obtained by analyzing the rise time and duration of pulses in the corona current data of each line segment; and the corona discharge coefficient of each line segment is obtained by analyzing the distribution of rise time of all pulses in the corona current data of each line segment, as well as the pulse decay coefficient and the peak width. By comparing the current and voltage data at both ends of each line segment, and combining the distortion of the voltage and current data at both ends of each line segment, the electrical distortion characteristic value of each line segment is obtained; the control parameters of the chaotic mapping are adjusted by the corona discharge coefficient and the electrical distortion characteristic value. The chaotic mapping after adjusting the control parameters is combined with the gray wolf optimization algorithm to locate faults in the distribution network.

2. The method for accurate fault location in a distribution network based on multi-source data fusion as described in claim 1, characterized in that, The process of obtaining the pulse decrement coefficient is as follows: Calculate the first-order difference in time sequence of all pulse peak values ​​in the corona current data of each section of the line; The pulse decay coefficient is the percentage of negative numbers in the first-order difference result.

3. The method for accurate fault location in distribution networks based on multi-source data fusion as described in claim 1, characterized in that, The process of obtaining the broad peak width is as follows: The pulse with the largest peak value in the corona current data of each section of the line is obtained. The product of the rise time and half-wave time of the obtained pulse is calculated. The ratio of the product to the square of the peak width of the obtained pulse is used as the peak width of each section of the line.

4. The method for accurate fault location in distribution networks based on multi-source data fusion as described in claim 1, characterized in that, The process of obtaining the corona discharge coefficient is as follows: The rise time and probability density of all pulse peaks in the corona current data of each section of the line are normally fitted to obtain a normal curve, and the goodness of fit of the normal curve is calculated. The corona discharge coefficient is inversely proportional to the peak width and the goodness of fit, and directly proportional to the pulse decrement coefficient.

5. The method for accurate fault location in distribution networks based on multi-source data fusion as described in claim 1, characterized in that, The process for obtaining the electrical distortion feature value is as follows: The electrical difference coefficient of each section of the line is obtained by measuring the degree of difference in current data at both ends of each section of the line. By combining the distortion of voltage data and current data at both ends of each section, the harmonic correction coefficient of each section of the line is obtained. The electrical distortion characteristic values ​​are positively correlated with the electrical difference coefficient and the harmonic correction coefficient, respectively.

6. The method for accurate fault location in a distribution network based on multi-source data fusion as described in claim 5, characterized in that, The method for obtaining the electrical difference coefficient is as follows: Calculate the difference in voltage data at both ends of each section of the line, and calculate the difference in current data at both ends of each section of the line. The electrical difference coefficient is the average of the difference and the amount of difference.

7. The method for accurate fault location in a distribution network based on multi-source data fusion as described in claim 5, characterized in that, The method for obtaining the harmonic correction coefficient is as follows: Calculate the average value of the total voltage harmonic distortion rate of the voltage data at both ends of the line in each section; calculate the arithmetic mean of the total current harmonic distortion rate of the current data at both ends of the line in each section. The harmonic correction coefficient is the weighted sum of the average value and the arithmetic mean.

8. The method for accurate fault location in a distribution network based on multi-source data fusion as described in claim 1, characterized in that, The process of adjusting the control parameters of the chaotic mapping is as follows: Obtain the segmentation threshold of the corona discharge coefficient for all line segments, and adjust the control parameters of the chaotic mapping by using the average level of the electrical distortion characteristic values ​​of all line segments with corona discharge coefficients greater than the segmentation threshold.

9. The method for accurate fault location in a distribution network based on multi-source data fusion as described in claim 8, characterized in that, The method for adjusting the control parameters of the chaotic mapping is as follows: The mean value of the electrical distortion characteristics of all sections of the line with a corona discharge coefficient greater than the segmentation threshold is denoted as the mean distortion value. Calculate the product of the length of the preset range of the control parameter and the mean of the distortion, and use the sum of the product and the preset minimum value of the control parameter as the adjusted control parameter.

10. A distribution network fault precise location system based on multi-source data fusion, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method for accurate fault location in distribution networks based on multi-source data fusion as described in any one of claims 1-9.

Images