A power distribution network single-end phase-to-phase short-circuit fault positioning method and system

Abstract

The application belongs to the technical field of power system relay protection and power distribution network automation, and discloses a power distribution network single-end phase-to-phase short-circuit fault positioning method and system. The application is based on a specific modal network construction of three-phase decoupling; endogenous wide-frequency harmonic of a medium-voltage direct-hanging multi-level inverter is used to extract wide-frequency input impedance of a power distribution network and convert the wide-frequency input impedance into a time-domain impulse response function; spatial distance function mapping based on a known traveling wave speed is performed; fault distance locking of a measurement point to an actual short-circuit fault point based on a sliding window waveform similarity difference comparison is performed; and specific fault branch determination based on fault shielding effect attenuation amplitude integration is performed. The application aims to effectively improve positioning accuracy and fault line selection efficiency of a power distribution network containing a high proportion of distributed power sources when a phase-to-phase short-circuit fault occurs, and significantly reduce dependence of the system on multi-end synchronous equipment and accurate line impedance parameters.

Description

Technical Field

[0001] This invention belongs to the field of power system relay protection and distribution network automation technology, and particularly relates to a method and system for locating single-phase-to-phase short-circuit faults in distribution networks. Background Technology

[0002] With the ongoing global energy transition, the large-scale integration of distributed generation (especially inverter-type power sources such as photovoltaics and wind power) into medium-voltage distribution networks has become an inevitable trend in power system development. Traditional distribution networks are gradually evolving into active distribution networks (ADNs) with multi-source integration and dynamic topology changes. However, the integration of inverter-type distributed generation alters the topology and short-circuit current characteristics of the distribution network, posing significant challenges to the reliable operation of traditional relay protection and fault location technologies.

[0003] In active distribution networks, inverters typically use isolation transformers for grid connection or actively suppress zero-sequence current output, thus having a relatively small impact on single-phase grounding faults. However, for phase-to-phase short-circuit faults with extremely high current and severe hazards, their electrical behavior differs significantly from that of traditional power grids. Existing methods for locating phase-to-phase short-circuit faults in distribution networks have obvious limitations, mainly manifested in the following ways: during a fault, the inverter often enters current-limiting or low-voltage ride-through (LVRT) mode, causing the output current to lose its natural response characteristics, and the equivalent impedance of the distribution network system to change dynamically, resulting in significant errors in traditional power frequency-based distance measurement methods.

[0004] To improve fault location accuracy, existing research often employs multi-terminal power frequency methods or multi-terminal traveling wave methods. However, these methods heavily rely on strict time synchronization of multiple nodes and reliable communication links. Furthermore, the traditional single-terminal power frequency method, when encountering high transition resistance during phase-to-phase short circuits, suffers significant deviations in its ranging results calculated from voltage and current phasors, rendering it completely unreliable. In addition, the traditional traveling wave method requires terminal equipment to maintain real-time, high-precision high-frequency sampling to effectively capture fault wavefronts, resulting in high equipment deployment costs and continuous energy consumption. Meanwhile, some active location methods based on control and protection coordination require actively altering the inverter's operating state, which not only introduces additional transient processes but may also interfere with the inverter's normal operation. Therefore, achieving accurate single-terminal fault location while overcoming high transition resistance interference, eliminating the need for multi-terminal synchronization, and ensuring uninterrupted operation of the distribution network system has become a hot topic and a challenge in current distribution network protection research.

[0005] Meanwhile, with the development of power electronics technology, medium-voltage direct-connected multilevel inverters (DCIs), characterized by transformerless structure and high efficiency, are being used more and more widely in power distribution networks. During PWM modulation and control, the physical transient processes of semiconductor switching devices naturally generate broadband harmonic components covering tens of kHz to several MHz. Current research mostly treats these broadband signals as interference that needs to be filtered out, with little in-depth exploration and application of the inverter's inherent broadband output as a natural high-frequency excitation source for passive fault location.

[0006] Therefore, it is particularly important to explore a passive single-ended ranging method based on the inherent wideband impedance characteristics of inverters to address the problems of phase-to-phase short-circuit fault location in distribution networks with inverter-type distributed power sources, such as limitations imposed by multi-terminal synchronization conditions, sensitivity to line parameters, and high energy consumption of high-frequency sampling.

[0007] Furthermore, with the large-scale integration of inverter-type distributed power sources (especially photovoltaic and wind power) into medium-voltage distribution networks, the topology and short-circuit current characteristics of active distribution networks have undergone fundamental changes. Particularly in the context of phase-to-phase short-circuit faults, relay protection and accurate fault location in distribution networks face significant challenges. Existing fault location methods for distribution networks fail to effectively adapt to the nonlinear transient control behavior of inverter-type power sources during faults (such as current limiting or low-voltage ride-through), especially in the case of high-resistance short-circuit faults or complex branch networks, where the traditional single-ended power frequency method often produces significant errors. Simultaneously, existing high-precision location methods heavily rely on high-precision time synchronization of multi-terminal equipment, and the real-time high-frequency sampling of the traditional traveling wave method incurs a substantial energy consumption burden. These issues not only affect the rapid isolation and power restoration efficiency of distribution network faults but also place higher demands on the deployment costs of distribution network terminal equipment and the overall reliability of the distribution network system. Summary of the Invention

[0008] To overcome the problems existing in related technologies, this invention discloses a method and system for locating single-phase-to-phase short-circuit faults in distribution networks, specifically involving a method for locating single-phase-to-phase short-circuit faults in distribution networks based on wideband impedance time-domain differences. This invention can fully utilize the inherent wideband output characteristics of medium-voltage direct-connected multilevel inverters, combined with decoupling analysis of specific mode networks, to achieve high-precision single-phase fault location without relying on multi-terminal communication and precise line parameters. This technology has significant practical implications for reducing energy consumption of distribution network terminal equipment, overcoming transition resistance interference, and comprehensively improving the rapid fault isolation and power restoration capabilities of active distribution networks. Furthermore, the technical problem this invention aims to solve is how to achieve the following objectives by comprehensively exploring the inherent wideband harmonic excitation characteristics of medium-voltage direct-connected multilevel inverters and optimizing the mode decoupling and wideband impedance analysis methods of the three-phase system of the distribution network: (1) Eliminate the dependence of fault location on multi-terminal synchronization and overcome the interference of transition resistance: By comprehensively considering the transient characteristics of active distribution network under different short-circuit faults, and making full use of a specific two-phase subtraction mode network (Mode3) to decouple the system and filter out the influence of complex impedance overlap, this invention proposes a single-terminal fault location method based on the broadband impedance characteristics of a specific mode network. This makes the ranging process not require external pulse injection and multi-terminal communication synchronization, and can adaptively overcome the interference of high transition resistance based on high frequency reflection characteristics, thereby improving the ranging accuracy under complex operating conditions.

[0009] (2) Reducing terminal ranging energy consumption and improving the engineering practicality of the positioning system: Taking into account the MHz-level broadband harmonic components naturally generated by inverter PWM modulation and the reflection patterns of impedance discontinuities in the distribution network (such as branch nodes and fault points), under the condition of known equivalent wave velocity of the line, this invention proposes a feature extraction and fault line selection method based on frequency domain-time domain impedance transformation and similarity matching of sliding windows before and after the fault. This method only requires short-term high-frequency sampling after topology change or fault occurrence, avoiding the huge energy consumption of all-weather high-frequency sampling in the traditional traveling wave method, and ensuring the safe, economical and stable operation of distribution networks with a high proportion of distributed power sources.

[0010] The technical solution is as follows: A method for locating single-phase-to-phase short-circuit faults in a distribution network, comprising the following steps: S1, Construct a specific mode network based on three-phase decoupling; S2, based on the established specific mode network, utilizes the endogenous wideband harmonics of the medium-voltage direct-connected multilevel inverter to extract the wideband input impedance of the distribution network and convert it into a time-domain impulse response function; S3, based on the extracted time-domain impulse response function, perform spatial distance function mapping based on the known traveling wave velocity; S4. Based on the spatial distance function mapping result, the fault distance from the measurement point to the actual short-circuit fault point is locked based on the comparison of waveform similarity difference in the sliding window. S5, based on the locked fault distance, performs specific fault branch determination based on the integral of the attenuation amplitude of the fault shielding effect.

[0011] In step S1, the specific mode network based on three-phase decoupling is defined as a Mode3 mode network, and the decoupling formula is: ; In the formula, These are the Mode 3 transient voltage and mode transient current obtained from decoupling, respectively. These are the transient voltages of phase B and phase C collected at the measurement terminal, respectively. These are the transient currents of phase B and phase C, respectively. For time.

[0012] In step S2, extracting the broadband input impedance of the distribution network and converting it into a time-domain impulse response function includes: Wideband input impedance calculation; the single-ended data acquisition and processing unit obtains the Mode 3 transient voltage. and modal transient current The signal is transmitted to the Fourier transform module for Fast Fourier Transform to obtain the frequency domain response. and Dividing the two yields the wideband input impedance frequency domain response of the distribution network from the measurement point to the fault point. The expression is: ; In the formula, Let be the frequency domain responses obtained by performing FFT on the measured signals, respectively. Angular frequency, subscript The impedance is the input impedance obtained at the measurement point; Time-domain characteristic signal transformation, the obtained frequency-domain impedance The signal is transmitted to the inverse Fourier transform module, where an inverse fast Fourier transform is performed, converting it into a time-domain impulse response function. : ; In the formula, This is an IFFT operation.

[0013] In step S3, the spatial distance function mapping based on the known traveling wave velocity includes: Utilizing the main line wave speed The processor performs spatiotemporal coordinate transformation, using the mapping relationship between time and distance. Or an equivalent conversion including transmission coefficients, where, Let distance be the distance variable measured along the propagation path from the measurement point; the time-domain impulse response function... The time coordinate system is transformed into the spatial coordinate system, thereby constructing a standard distance function measured by the spatial scale. : ; In the formula, This is the time-domain impulse response function; High-frequency sampling is performed when no fault occurs or for a period of time before the fault occurs, and the reference distance function before the fault is calculated using formulas (1), (2), (3) and (4) in steps S1, S2 and S3. .

[0014] In step S4, the fault distance locking from the measurement point to the actual short-circuit fault point based on the comparison of waveform similarity differences within a sliding window includes: When a phase-to-phase short-circuit fault occurs in the distribution network, the sampling equipment starts high-frequency sampling. The sampled waveform is processed for a period of time according to formulas (1), (2), (3) and (4) in steps S1, S2 and S3, and the spatial distance function after the fault is calculated and generated. ; Similarity sliding traversal, the digital signal processor (DSP) sets a length of... A sliding window is used to control the window to scroll gradually from the starting point to the endpoint, representing the maximum line distance from the measurement point to the farthest point in the distribution network. At each spatial translation feature point, the amplitude of the pre-fault reference distance function and the post-fault distance function is normalized, and the pre-fault reference distance function is calculated using equation (5). Distance function after fault Waveform similarity curves within this window range : ; In the formula, The length of the sliding window. The distance function before the fault is located at the end of the current sliding window. The function value at that point, The distance function after the fault is located at the end of the current sliding window. The function value at that point, For spatial integration, This represents the maximum line distance from the measurement point to the farthest end of the distribution network. The distance function before the fault is at the location The function value at that point, The distance function after the fault is at the location The function value at that location; Fault breakpoint identification: In the region in front of the fault point, the distance functions before and after the fault highly overlap, and the similarity curves... Approaching 1; once the fault point is crossed, the waveform similarity curve A fractured fall occurred at the spatial coordinates of the failure point; Distance determination is performed by setting a preset similarity mutation threshold. , The value is 0.5; when the waveform similarity curve The first sharp drop occurred and the price fell below the threshold. At that time, the location of the sudden change is locked, and the spatial abscissa corresponding to the peak value of the first impedance reflection pulse with large amplitude characteristics that appears immediately afterwards is determined as the fault distance from the measurement point to the actual short-circuit fault point. : ; In the formula, for The midpoint peak value corresponds to the set of distances to the measurement points. To be at the candidate peak position The local waveform similarity calculated at that location For set The first in Each candidate peak location is represented by a value indicating the distance between the candidate peak and the measurement point.

[0015] In step S5, the specific fault branch determination based on the integral of the attenuation amplitude of the fault shielding effect includes: If the fault distance is determined in step S4 If multiple solutions exist in the spatial topology, meaning the location result is not unique, then the branch selection determination procedure is invoked: The digital signal processor (DSP) extracts the set of all inherent feature points downstream of each possible fault branch before the fault occurs. By calculating the overall topology matching index of each branch before and after the failure, : ; In the formula, This is the set of distances between all inherent feature points downstream of the faulty branch and the measurement point. For collection The sub-elements in the equation represent the distances between all inherent feature points downstream of the fault branch and the measurement point. The baseline distance function before the fault is in The characteristic amplitude at that location; Calculate the pre-fault state using formula (7) After the fault Compare the difference in matching degree before and after the fault. ; In the actual short-circuit fault branch, the broadband harmonic energy received by the inverter at downstream characteristic points is attenuated to the greatest extent due to the truncation caused by the short-circuit fault, resulting in... The indicator shows the maximum value among all alternative branches; The candidate line segment with the largest descent is uniquely identified as the final actual fault branch, and the output includes the specific fault branch name and the mileage.

[0016] Another objective of this invention is to provide a single-phase-to-phase short-circuit fault location system for a distribution network. This system implements the aforementioned single-phase-to-phase short-circuit fault location method for a distribution network. The system includes: The upstream power supply side structure includes a 110kV upstream system power supply, which reduces the voltage to 10kV through the main transformer and connects to the 10kV bus of the distribution network. The power transmission and distribution line topology consists of a complex tree-like network structure extending downwards from the 10kV busbar, which comprises the main line, multiple branch lines, and sub-branch lines extending further below the branch lines. The grid-connected source-side structure and measurement configuration system includes a high-capacity medium-voltage direct-connected multilevel inverter (DCI) directly connected to a specific node on the 10kV bus or main line. A high-frequency voltage acquisition module is connected in parallel to the three-phase output ports of the DCI, and a high-frequency current acquisition module is connected in series to form a single-ended data acquisition and processing unit. This single-ended data acquisition and processing unit integrates a high-speed analog-to-digital converter (ADC), a digital signal processor (DSP), and a data storage device for storing the coordinates of characteristic points of the transmission and distribution line topology.

[0017] Furthermore, in the transmission and distribution line topology, the line types include overhead lines and underground cable lines. Between the main line and branch lines, between lines of different diameters, and between the load connection points at the end of the line, due to the discontinuity of physical medium and electrical impedance, several inherent impedance discontinuity characteristic points with fixed spatial coordinates are naturally formed.

[0018] Furthermore, the aforementioned single-phase-to-phase short-circuit fault location system for distribution networks is applied to medium-voltage distribution networks with direct-connected inverters and other new energy sources.

[0019] Furthermore, the single-phase-to-phase short-circuit fault location system for the distribution network is mounted on a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it can realize the functions of the single-phase-to-phase short-circuit fault location system for the distribution network.

[0020] Combining all the above technical solutions, the beneficial effects of this invention are as follows: First, this invention combines the inherent wideband harmonic excitation characteristics of medium-voltage direct-connected multilevel inverters with the decoupling technology of specific mode networks. It comprehensively considers the reflection patterns and similarity comparison indices of time-domain characteristic signals before and after a fault, establishing a fault location model based on sliding window waveform comparison and its multi-branch fuzzy location identification method. This forms a single-ended fault location strategy based on the shielding effect of impedance discontinuities. This invention aims to effectively improve the location accuracy and fault line selection efficiency of distribution networks with a high proportion of distributed power sources when phase-to-phase short-circuit faults occur, significantly reducing the system's dependence on multi-terminal synchronization equipment and precise line impedance parameters.

[0021] Secondly, this invention can reduce the equipment configuration cost, communication synchronization cost, and on-site modification difficulty of the distribution network fault location system, improve the fault section identification efficiency and power supply restoration speed after a phase-to-phase short-circuit fault, and is suitable for the protection, fault location, and distribution automation system upgrade of medium-voltage distribution networks under the background of high proportion of distributed renewable energy access. Existing methods for locating phase-to-phase short-circuit faults in distribution networks mostly rely on power frequency voltage and current phasors, multi-terminal synchronous traveling wave measurement, or active harmonic / pulse injection. The above methods are easily affected by inverter current limiting control, low voltage ride-through control, line parameter uncertainty, multi-branch topology, and communication synchronization conditions in active distribution networks containing inverter-type distributed power sources. Unlike existing solutions, this invention uses the inherent broadband harmonics of medium-voltage direct-connected multilevel inverters as a usable natural broadband excitation source, and combines it with the Mode3 specific mode network formed by subtracting the two fault phases to extract the broadband input impedance time-domain reflection characteristics before and after the fault, and achieves single-terminal fault location through distance function difference and sliding window similarity. This technical approach differs from traditional power frequency ranging, traditional traveling wave ranging, and active injection positioning methods, providing a new technical implementation path for locating phase-to-phase short-circuit faults in medium-voltage distribution networks with direct-connected inverters.

[0022] Third, in distribution networks with a high proportion of inverter-type distributed power sources, the inverter output characteristics are significantly affected by the control strategy during phase-to-phase short-circuit faults. Traditional power frequency distance measurement is prone to large errors due to short-circuit current limitations, dynamic changes in equivalent impedance, and the influence of transition resistance. Traditional traveling wave methods typically require continuous high-frequency sampling and strict multi-terminal time synchronization, leading to high deployment costs and complex operation and maintenance. How to achieve reliable single-terminal phase-to-phase short-circuit fault location without relying on multi-terminal synchronization, precise line impedance parameters, or interfering with normal inverter operation has long been a key technical challenge in the field of active distribution network protection and automation. This invention constructs a specific modal network, extracts the time-domain response of wideband input impedance, and compares the differences in impedance reflection characteristics before and after the fault. This enables the identification of fault distance under single-terminal measurement conditions. When the same fault distance corresponds to multiple possible branches, the shielding attenuation effect of the short-circuit fault point on the downstream impedance discontinuity characteristic points is further utilized to determine the actual fault branch, thus providing an effective solution to the above problems.

[0023] Fourth, in traditional distribution network protection and power quality analysis, broadband harmonics generated during inverter switching are typically considered interference signals that need to be suppressed or filtered out. In engineering practice, their impact is often reduced through filtering, shielding, or control optimization. However, this invention reverses this by utilizing the broadband harmonics naturally generated during the operation of a medium-voltage direct-connected multilevel inverter as an endogenous excitation source for extracting broadband impedance characteristics of the distribution network, eliminating the need for additional detection signals. Furthermore, traditional techniques generally consider high-precision fault location to rely on multi-terminal synchronous measurement and external high-frequency pulse injection. This invention, however, utilizes the difference in the time-domain reflection response of broadband impedance before and after a fault, combined with a specific mode network to reduce the modal coupling effect under phase-to-phase short circuits, achieving fault distance locking and branch determination under single-terminal conditions. Therefore, this invention breaks through the traditional understanding that inverter broadband harmonics are simply considered interference and that high-precision location necessarily relies on multi-terminal synchronization or active injection. Attached Figure Description

[0024] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the disclosure of this invention and, together with the description, serve to explain the principles of this disclosure; Figure 1 This is a diagram of the single-phase-to-phase short-circuit fault location system for power distribution networks provided by the present invention; Figure 2 This is a schematic diagram of the single-phase-to-phase short-circuit fault location method for power distribution networks provided by the present invention. Figure 3 This is a flowchart of the method for locating single-phase-to-phase short-circuit faults in a power distribution network provided by the present invention; Figure 4 This is a waveform diagram of the inverter output three-mode voltage and current provided by the present invention; Figure 5 This is a fault location result diagram provided by the present invention. Detailed Implementation

[0025] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0026] The innovation of this invention lies in the following: This invention utilizes the naturally generated wideband harmonics during the operation of a medium-voltage direct-connected multilevel inverter as the excitation source for wideband impedance identification of the system, eliminating the need for additional high-voltage pulse generators or active harmonic injection devices; by constructing a Mode 3 specific mode network that subtracts the fault phases, the impact of mode coupling on wideband impedance feature extraction under phase-to-phase short-circuit faults is reduced; by comparing the difference in distance functions obtained from the time-domain impedance reflection response mapping before and after the fault, single-end fault distance locking is achieved; and when multiple topological solutions exist, the shielding attenuation effect of the fault point on downstream feature points is used to complete the specific fault branch determination.

[0027] Example 1.

[0028] The distribution network hardware and physical structure upon which this invention is based are as follows: The distribution network single-phase-to-phase short-circuit fault location method provided in this invention is applied to the distribution network single-phase-to-phase short-circuit fault location system (e.g., Figure 1 The system's physical and electrical topology includes: Upstream power supply side structure: Includes 110kV upstream system power supply, which is reduced to 10kV through the main transformer and connected to the 10kV bus of the distribution network.

[0029] Transmission and distribution line topology: The 10kV busbar extends downwards into a complex tree-like network structure consisting of the main line, multiple branch lines (such as branch line 1, branch line 2, branch line 3, etc.), and sub-branch lines extending further below the branch lines. Line types include overhead lines (such as double-circuit or single-circuit main lines on tall poles) and buried cable lines. Between the main line and branch lines, between lines of different wire diameters, and between the load connection points at the end of the line, due to the discontinuity of physical medium and electrical impedance, several inherent impedance discontinuity characteristic points (referred to as characteristic points) with fixed spatial coordinates are naturally formed.

[0030] Grid-connected source-side structure and measurement configuration system: A high-capacity medium-voltage direct-connected multilevel inverter (DCI) is directly connected at a specific node on the 10kV bus or main line. This inverter does not require a step-up transformer for isolation. A high-frequency voltage acquisition module is connected in parallel to the three-phase output ports of the DCI, and a high-frequency current acquisition module is connected in series to form a single-ended data acquisition and processing unit. This single-ended data acquisition and processing unit integrates a high-speed analog-to-digital converter (ADC), a digital signal processor (DSP), and a data storage device for storing the coordinates of characteristic points of the transmission and distribution line topology.

[0031] For example, Figure 2This is a schematic diagram of the single-phase-to-phase short-circuit fault location method for distribution networks provided by the present invention. After acquiring normal waveforms through high-frequency sampling and forming a reference distance function before the fault, the system continuously uses low-frequency sampling for fault detection. When a phase-to-phase short-circuit fault is detected, high-frequency sampling is initiated to acquire electrical quantities after the fault and form a distance function after the fault. Then, the fault distance is determined by the similarity change of the distance functions before and after the fault. When there are multiple possible locations of the determined fault distance in the distribution network topology, the actual fault branch is further determined based on the matching degree change of the candidate line before and after the fault. Specifically, such as Figure 3 As shown, the method for locating single-phase-to-phase short-circuit faults in a distribution network provided by this invention mainly includes the following five core steps: S1, Construct a specific mode network based on three-phase decoupling; For the three-phase transient voltage and current signals acquired by the single-ended data acquisition and processing unit, in order to eliminate the interference of electromagnetic coupling between the three-phase lines of the active distribution network on high-frequency waveform analysis, the digital signal processor (DSP) calls the built-in Karrenbauer linear transformation matrix to linearly decouple the acquired three-phase transient electrical quantities. This invention innovatively proposes and preferably uses a specific mode network (defined as a Mode3 mode network) independently composed of two fault phases, and its decoupling formula is as follows: ; In the formula, These are the Mode 3 transient voltage and mode transient current obtained from decoupling, respectively. These are the transient voltages of phase B and phase C collected at the measurement terminal, respectively. These are the transient currents of phase B and phase C, respectively. For time.

[0032] Physical Principle: During a fault, the equivalent impedance of a conventional line-mode network (such as Mode1 or Mode2) is subject to crosstalk from other phase sequence networks, resulting in an extremely complex impedance waveform. In contrast, the Mode3 network preferred in this invention has its input impedance at the measurement point determined entirely by the line distribution parameters and the transition resistance at the fault point under that specific mode. This shields against electromagnetic crosstalk from other healthy phases, significantly simplifying the frequency domain and the converted time domain impedance reflection characteristic waveforms. Furthermore, it possesses a completely uniform applicability to both symmetrical (three-phase) and asymmetrical (two-phase) short-circuit faults.

[0033] S2, based on the established specific mode network, utilizes the endogenous wideband harmonics of the medium-voltage direct-connected multilevel inverter to extract the wideband input impedance of the distribution network and convert it into a time-domain impulse response function; This invention does not require the high-voltage pulse generator or harmonic injection device found in conventional technologies. Because medium-voltage direct-connected multilevel inverters (DCIs) employ multilevel pulse width modulation (PWM) control, their internal power semiconductor devices (such as IGBTs) switch on and off at high frequencies within an extremely short time (typically less than 100 nanoseconds), naturally generating continuous broadband harmonic components ranging from tens of kilohertz (kHz) to several megahertz (MHz) at the output.

[0034] Wideband input impedance calculation: The single-ended data acquisition and processing unit obtains the Mode 3 transient voltage. and modal transient current The signal is transmitted to the Fourier Transform module for Fast Fourier Transform (FFT) to obtain the frequency domain response. and Dividing the two yields the wideband input impedance frequency domain response of the distribution network from the measurement point to the fault point. The expression is: ; In the formula, Let be the frequency domain responses obtained by performing FFT on the measured signals, respectively. Angular frequency, subscript The impedance is the input impedance obtained at the measurement point; Time-domain characteristic signal transformation: further transforming the obtained frequency-domain impedance The data is transmitted to the inverse Fourier transform module, where an inverse fast Fourier transform (IFFT) is performed to convert it into a time-domain impulse response function. : ; In the formula, This is an IFFT operation.

[0035] This function is physically equivalent to the reflected waveform of the system under a unit electromagnetic pulse excitation. When a high-frequency electromagnetic wave encounters an impedance discontinuity while propagating in a line, a reflection will occur. The series of pulse peaks appearing in the time-domain waveform represent the high-frequency traveling wave two-way transmission delay time from the measurement point to the impedance discontinuity point and back.

[0036] S3, based on the extracted time-domain impulse response function, perform spatial distance function mapping based on the known traveling wave velocity; Before a fault occurs or during the topology initialization phase, the single-ended data acquisition and processing unit reads the pre-measured and stored solid-state wave propagation velocity (wave velocity) of the main distribution network line and its branches. and , (This indicates the line type of this branch line). Wave velocity is a known parameter of the system, determined by the physical materials and geometric constants of the transmission line. It can be obtained by consulting the cable manufacturer's manual or pre-entered using existing conventional offline measurement methods for double-ended traveling waves.

[0037] Retrieve the lengths of each line segment in the distribution network from the built-in Geographic Information System (GIS) or asset database, and categorize them according to the wave speed of each line segment. With the main line wave speed The ratio is reduced to the equivalent length under the main line wave velocity condition, thus obtaining the equivalent line length of the distribution network under a unified wave velocity scale.

[0038] Utilizing the main line wave speed The processor performs spatiotemporal coordinate transformation, using the mapping relationship between time and distance. Or an equivalent conversion including transmission coefficients, where, Let distance be the distance variable measured along the propagation path from the measurement point; the time-domain impulse response function... The time coordinate system is transformed into the spatial coordinate system, thereby constructing a standard distance function measured by the spatial scale. : ; In the formula, This is the time-domain impulse response function; High-frequency sampling (sampling frequency between 500kHz and 5MHz, preferably 1MHz) is performed for a period of time before a fault occurs (preferably 0.04s, i.e., two power frequency cycles, or the range can be set to 0.02-0.08s) to obtain the normal waveform before the fault. The reference distance function before the fault is calculated using formulas (1), (2), (3) and (4) in steps S1, S2 and S3. .

[0039] S4. Based on the spatial distance function mapping result, the fault distance from the measurement point to the actual short-circuit fault point is locked based on the comparison of waveform similarity difference in the sliding window. When a phase-to-phase short-circuit fault occurs in the distribution network, the sampling equipment starts high-frequency sampling (sampling frequency between 500kHz and 5MHz, preferably 1MHz). The sampled waveform is processed for a period of time (preferably 0.04s, i.e., two power frequency cycles, or the range can be set to 0.02-0.08s) according to formulas (1), (2), (3) and (4) in steps S1, S2 and S3, to calculate and generate the spatial distance function after the fault. .

[0040] Similarity sliding traversal, the digital signal processor (DSP) sets a length of... A sliding window is used to control the window to scroll gradually from the starting point to the endpoint, representing the maximum line distance from the measurement point to the farthest point in the distribution network. At each spatial translation feature point, the amplitude of the pre-fault reference distance function and the post-fault distance function is normalized, and the pre-fault reference distance function is calculated using equation (5). Distance function after fault Waveform similarity curves within this window range : ; In the formula, The length of the sliding window. The distance function before the fault is located at the end of the current sliding window. The function value at that point, The distance function after the fault is located at the end of the current sliding window. The function value at that point, For spatial integration, This represents the maximum line distance from the measurement point to the farthest end of the distribution network. The distance function before the fault is at the location The function value at that point, The distance function after the fault is at the location The function value at that location; Fault Point Identification: Based on the laws of electromagnetic wave propagation, since the physical structure of the line ahead of the short-circuit fault point remains unchanged, the high-frequency harmonic diffuse reflection echoes are completely consistent. Therefore, in the region ahead of the fault point, the distance functions before and after the fault highly overlap, and the similarity curves are identical. Approaching 1. Once past the fault point, due to the metallic short circuit or resistive short circuit forming a very strong new impedance discontinuity at the fault location, the high-frequency signal undergoes near total reflection at this point, causing the inherent reflected waveform behind the fault point to be significantly distorted or shielded, resulting in a waveform similarity curve. It will break apart and fall at the spatial coordinates of the failure point.

[0041] Distance determination: The fault location method sets a preset similarity abrupt change threshold. (Preferably 0.5). When the waveform similarity curve (sliding similarity curve) The first sharp drop occurred and the price fell below the threshold. At that time, the location of the sudden change is locked, and the spatial abscissa corresponding to the peak value of the first impedance reflection pulse with large amplitude characteristics that appears immediately afterwards is innovatively proposed in this invention to be determined as the fault distance from the measurement point to the actual short-circuit fault point. : ; In the formula, for The midpoint peak value corresponds to the set of distances to the measurement points. To be at the candidate peak position The local waveform similarity calculated at that location For set The first in Each candidate peak location is represented by a value indicating the distance between the candidate peak and the measurement point.

[0042] S5, based on the locked fault distance, perform specific fault branch determination based on the integral of the attenuation amplitude of the fault shielding effect; Medium-voltage distribution networks often exhibit a multi-branch radial structure. If the fault distance determined in step S4... If multiple solutions exist in the spatial topology, meaning the location result is not unique (for example, a location 2.56 kilometers behind the measuring point corresponds to both a position on the main line and a position on branch line 1), then the branch selection determination program should be invoked: The digital signal processor (DSP) extracts the set of all inherent feature points downstream of each possible faulty branch (such as the main line or branch line 1) before the fault occurs (the echoes behind these feature points will be severely attenuated or disappear after the fault due to the electromagnetic shielding effect of the short circuit point).

[0043] By calculating the overall topology matching index of each possible branch before and after the failure, : ; In the formula, This is the set of distances between all inherent characteristic points and measurement points downstream of the faulty branch (such as the main line or branch line 1). For collection The sub-elements in the equation represent the distances between all inherent feature points downstream of the fault branch and the measurement point. The baseline distance function before the fault is in The characteristic amplitude at that location; Calculate the pre-fault state using formula (7) After the fault Compare the difference in matching degree before and after the fault. ; In the case of a line branch where a short-circuit fault actually occurs, the inverter's broadband harmonic energy received at downstream characteristic points is attenuated to the greatest extent due to the truncation caused by the short-circuit fault, resulting in its... The indicator represents the maximum value among all candidate branches. The system will... The candidate line segment with the largest descent is uniquely identified as the final actual fault branch, thus outputting the final location result containing "the specific name of the fault branch + the precise mileage".

[0044] As can be seen from the above embodiments, the present invention does not require multi-terminal synchronization and communication: it breaks through the rigid dependence of existing traveling wave method or multi-terminal power frequency protection on multi-terminal communication and high-precision clock synchronization, and greatly reduces the system implementation cost and management complexity.

[0045] Superior resistance to transition resistance interference: This invention overcomes the shortcomings of traditional power frequency impedance methods, which are highly sensitive to transition resistance. Based on the theory of high-frequency reflection coefficient, even in the event of a high-impedance short circuit (e.g., 30 ohms), the short-circuit transition resistance remains much smaller than the characteristic impedance of the line at high frequencies, the reflection characteristics are still significant, and the positioning accuracy is almost unaffected by the transition resistance.

[0046] Eliminating external signal injection source devices: This invention directly utilizes the inherent wideband (MHz level) harmonic components of the high-capacity medium-voltage direct-connected multilevel inverter as a natural high-frequency excitation source, without the need for additional high-frequency pulse injection devices, and 100% meets the requirements for grid-connected safe operation.

[0047] Significantly reduces the energy consumption burden of high-frequency sampling: Compared with the traditional traveling wave method, which requires high-frequency sampling to be turned on around the clock to capture randomly occurring fault wavefronts, this invention adopts a passive impedance measurement mechanism, which only needs to start short-term high-frequency sampling when a topology change occurs (preparation stage) or after a fault occurs, which greatly reduces the operating energy consumption of terminal equipment.

[0048] Mode 3 analysis reduces computational complexity: This invention uses a specific two-phase subtraction mode (Mode 3) network, which, compared to the traditional mode 1 network, filters out the complex impedance overlap effects of other modes, greatly reducing the characteristic clutter of the frequency domain-time domain transformation, improving the accuracy of the algorithm, and is uniformly applicable to symmetrical three-phase and asymmetrical two-phase short-circuit faults.

[0049] Furthermore, the single-ended fault location method proposed in this invention, based on wideband impedance characteristics, combines wideband harmonic excitation of the inverter with mode 3 decoupling and frequency-domain to time-domain impedance conversion. It can flexibly cope with the complex topology of the distribution network and the changes in electrical characteristics under different fault conditions, ensuring accurate distance measurement even when facing complex situations such as high-impedance short circuits, and has strong adaptability.

[0050] An innovative decoupling strategy based on two-phase subtraction of a specific mode (Mode3) is proposed, which can better filter out the interference caused by the overlap of impedances of other mode networks in a three-phase system. It greatly smooths and simplifies the impedance characteristic curves in the frequency and time domains, reduces the complexity of signal processing and feature extraction, and is uniformly applicable to symmetrical three-phase and asymmetrical two-phase short-circuit faults.

[0051] Using Fourier transform and its inverse transform, and under the condition of known equivalent wave velocity of the line, the frequency domain broadband impedance of the measurement point viewed into the system is converted into a distance function containing physical space information. It can accurately map abstract high-frequency input impedance into intuitive spatial location signals of distribution network impedance discontinuities (characteristic points).

[0052] In practical applications, the inherent shielding effect of the fault point is utilized to greatly overcome the shortcomings of the traditional power frequency method, which is highly sensitive to the transition resistance. Tests show that even in harsh test environments such as high transition resistance of up to 30 ohms or low signal-to-noise ratio (SNR), the differences in waveform characteristics are still clearly visible, significantly reducing the impact of environmental interference on ranging and improving the practical value of engineering.

[0053] An innovative sliding window-based similarity matching strategy for fault pre- and post-fault events is introduced. By filtering out useless small signal fluctuations, the distance function before and after the fault is accurately compared using a set difference threshold. and The local similarity is used to maximize the accuracy of single-end localization when locking feature peak mutations.

[0054] In complex networks with multiple branch lines, the local matching degree of the main feeder and each branch line is calculated, constrained by the amplitude attenuation characteristics of the fault point on the lines behind the fault. By adaptively selecting lines based on the decrease in matching degree, the traditional method effectively solves the problem of accurately determining fault sections in multi-branch structures.

[0055] By comprehensively considering the inherent high-frequency harmonic passive measurement and sliding window similarity matching strategy of medium-voltage direct-connected multilevel inverters, this patent can effectively improve the single-end positioning accuracy of active distribution networks under phase-to-phase short-circuit faults, eliminating the need for all-weather high-frequency sampling energy consumption and multi-end time synchronization equipment in the traditional traveling wave method, and providing important technical support for the safe, rapid self-healing and efficient operation of distribution networks.

[0056] The present invention discloses a method for locating single-ended short-circuit faults between phases in a distribution network based on wideband impedance characteristics. Specifically, this method significantly improves the accuracy of locating short-circuit faults between phases in the distribution network and the engineering practicality of the ranging system.

[0057] By employing a decoupling and frequency-to-time impedance transformation strategy based on a specific mode (Mode3), the extreme value characteristics of the impedance curve become clearer and simpler, reducing the severe overlapping interference caused by multi-mode coupling in three-phase systems. This optimized extraction of characteristic signals not only reduces the analysis difficulty under complex topologies but also improves the overall accuracy and adaptability of the ranging algorithm. Through a sliding window similarity matching strategy based on impedance reflection feature points before and after a fault, the system's ability to resist high transition resistance is effectively improved. The optimized single-end ranging method overcomes the defect of the traditional power frequency impedance method being highly sensitive to transition resistance, enabling the system to accurately locate the fault under different fault conditions (such as high-resistance short circuits up to 30 ohms), thereby better supporting the rapid isolation of distribution network faults. By directly utilizing the inherent wideband harmonics of medium-voltage direct-connected multilevel inverters as a natural excitation source, and only performing short-term high-frequency sampling after topology changes or faults, the energy consumption of the ranging terminal is significantly reduced. This eliminates the hardware losses and operational pressure caused by the all-weather high-frequency sampling of the traditional traveling wave method, while also eliminating the equipment cost of external signal injection sources. The positioning strategy of this invention has strong robustness and adaptability, maintaining high-precision ranging capabilities under different inverter connection locations, multi-inverter grid-connected environments, and low signal-to-noise ratio (SNR) conditions. This enhances the system's ability to respond to sudden short-circuit events in complex active distribution networks, ensuring the safety and stability of the power grid. In the simulation and verification of a 10kV medium-voltage distribution network with a high proportion of inverter-type distributed power sources, the implementation of the optimized strategy significantly improved the reliability and anti-interference capability of phase-to-phase short-circuit fault ranging, and the ranging error was strictly controlled within a very small range. In summary, this invention provides effective technical support for the intelligent, automated, and highly reliable development of active distribution networks.

[0058] This invention can be applied to medium-voltage distribution networks with direct-connected inverters and other renewable energy sources.

[0059] A typical 10kV distribution network model is built using Matlab / Simulink, and its structure is as follows: Figure 1 The diagram shows a single-phase-to-phase short-circuit fault location system for a distribution network (10kV distribution network model structure). Here, Source is the 110kV system power source, and T is the system main transformer with a transformation ratio of 110 / 10.5.

[0060] The system includes one main line and four branch lines. Some branch lines have subordinate branch lines. The lengths of each line segment are as follows: Figure 3 As indicated by the markings. The overhead line diameters are 240mm (Type 1), 150mm (Type 2), and 95mm (Type 3), with overhead line height and pole type set according to typical structures. The cable type is copper core cross-linked polyethylene insulated power cable with a diameter of 95mm. The line simulation model uses a distributed parameter model. The following describes the inverter connection point to the DCI and the fault location settings. Taking time as an example, we will provide a detailed introduction.

[0061] After the fault occurs, record the 3-mode voltage and current at the inverter's DCI input, such as... Figure 4 The inverter output 3-mode voltage and current waveforms are shown in the figure, where a fault occurs at 60ms.

[0062] After obtaining the selected fault ( Figure 4 Distance function between 80-120ms and when there is no fault ( Figure 4 Recorded in 10-50ms Calculate the similarity between waveforms to obtain, as follows Figure 5 The results of the fault location are shown in the figure.

[0063] First, by setting a minimum threshold... Screening is performed to avoid interference from similarity calculated using excessively small and unstable values. Furthermore, based on testing experience, setting a similarity threshold of 0.5 allows us to obtain the position of the first pulse peak after a rapid decrease in similarity. Figure 5 As shown in the figure, the peak value represents the preliminary location of the fault location from the inverter, which is 2.4642 km.

[0064] Observe the structure of the distribution network lines Figure 3 It can be seen that the fault point at this distance may exist on the main line or branch line 1, and further calculation of the main line matching degree is needed. Matching degree with branch line The changes for each are shown in Table 1.

[0065] Table 1 Before and after the fault and

[0066]

[0067] As can be seen from the results in Table 1, The change is greater than Therefore, the fault occurred on branch line 1.

[0068] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by those skilled in the art within the scope of the technology disclosed in the present invention and within the spirit and principles of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A method for locating single-phase-to-phase short-circuit faults in a distribution network, characterized in that, The method includes the following steps: S1, Construct a specific mode network based on three-phase decoupling; S2, based on the established specific mode network, utilizes the endogenous wideband harmonics of the medium-voltage direct-connected multilevel inverter to extract the wideband input impedance of the distribution network and convert it into a time-domain impulse response function; S3, based on the extracted time-domain impulse response function, perform spatial distance function mapping based on the known traveling wave velocity; S4. Based on the spatial distance function mapping result, the fault distance from the measurement point to the actual short-circuit fault point is locked based on the comparison of waveform similarity difference in the sliding window. S5, based on the locked fault distance, performs specific fault branch determination based on the integral of the attenuation amplitude of the fault shielding effect.

2. The method for locating single-phase-to-phase short-circuit faults in a distribution network according to claim 1, characterized in that, In step S1, the specific mode network based on three-phase decoupling is defined as a Mode3 mode network, and the decoupling formula is: ； In the formula, These are the Mode 3 transient voltage and mode transient current obtained from decoupling, respectively. These are the transient voltages of phase B and phase C collected at the measurement terminal, respectively. These are the transient currents of phase B and phase C, respectively. For time.

3. The method for locating single-phase-to-phase short-circuit faults in a distribution network according to claim 2, characterized in that, In step S2, extracting the broadband input impedance of the distribution network and converting it into a time-domain impulse response function includes: Wideband input impedance calculation; the single-ended data acquisition and processing unit obtains the Mode 3 transient voltage. and modal transient current The signal is transmitted to the Fourier transform module for Fast Fourier Transform to obtain the frequency domain response. and Dividing the two yields the wideband input impedance frequency domain response of the distribution network from the measurement point to the fault point. The expression is: ； In the formula, Let be the frequency domain responses obtained by performing FFT on the measured signals, respectively. Angular frequency, subscript The impedance is the input impedance obtained at the measurement point; Time-domain characteristic signal transformation, the obtained frequency-domain impedance The signal is transmitted to the inverse Fourier transform module, where an inverse fast Fourier transform is performed, converting it into a time-domain impulse response function. : ； In the formula, This is an IFFT operation.

4. The method for locating single-phase-to-phase short-circuit faults in a distribution network according to claim 3, characterized in that, In step S3, the spatial distance function mapping based on the known traveling wave velocity includes: Utilizing the main line wave speed The processor performs spatiotemporal coordinate transformation, using the mapping relationship between time and distance. Or an equivalent conversion including transmission coefficients, where, Let distance be the distance variable measured along the propagation path from the measurement point; the time-domain impulse response function... The time coordinate system is transformed into the spatial coordinate system, and a standard distance function measured by the spatial scale is constructed. : ； In the formula, This is the time-domain impulse response function; High-frequency sampling is performed when no fault occurs or for a period of time before the fault occurs, and the reference distance function before the fault is calculated using formulas (1)-(4). .

5. The method for locating single-phase-to-phase short-circuit faults in a distribution network according to claim 4, characterized in that, In step S4, the fault distance locking from the measurement point to the actual short-circuit fault point based on the comparison of waveform similarity differences within a sliding window includes: When a phase-to-phase short-circuit fault occurs in the distribution network, the sampling equipment starts high-frequency sampling. The sampled waveform is processed according to formulas (1)-(4) for a period of time to calculate and generate the spatial distance function after the fault. ; Similarity sliding traversal, the digital signal processor (DSP) sets a length of... A sliding window is used to control the window to scroll gradually from the starting point to the endpoint, representing the maximum line distance from the measurement point to the farthest point in the distribution network. At each spatial translation feature point, the amplitude of the pre-fault reference distance function and the post-fault distance function is normalized, and the pre-fault reference distance function is calculated using equation (5). Distance function after fault Waveform similarity curves within this window range : ； In the formula, The length of the sliding window. The distance function before the fault is located at the end of the current sliding window. The function value at that point, The distance function after the fault is located at the end of the current sliding window. The function value at that point, For spatial integration, This represents the maximum line distance from the measurement point to the farthest end of the distribution network. The distance function before the fault is at the location The function value at that point, The distance function after the fault is at the location The function value at that location; Fault breakpoint identification: In the region in front of the fault point, the distance functions before and after the fault highly overlap, and the similarity curves... Approaching 1; once the fault point is crossed, the waveform similarity curve A fractured fall occurred at the spatial coordinates of the failure point; Distance determination is performed by setting a preset similarity mutation threshold. , The value is 0.5; when the waveform similarity curve The first sharp drop occurred and the price fell below the threshold. At that time, the location of the sudden change is locked, and the spatial abscissa corresponding to the peak value of the first impedance reflection pulse with large amplitude characteristics that appears immediately afterwards is determined as the fault distance from the measurement point to the actual short-circuit fault point. : ； In the formula, for The midpoint peak value corresponds to the set of distances to the measurement points. To be at the candidate peak position The local waveform similarity calculated at that location For set The first in Candidate peak positions.

6. The method for locating single-phase-to-phase short-circuit faults in a distribution network according to claim 5, characterized in that, In step S5, the specific fault branch determination based on the attenuation amplitude integral of the fault shielding effect includes: if the fault distance determined in step S4... If there are multiple solutions in the spatial topology, that is, the location result is not unique, then the branch selection determination program is called: the digital signal processor (DSP) extracts the set of all inherent feature points downstream of each faulty branch before the fault occurs. Calculate the overall topology matching index of each branch before and after the failure. : ； In the formula, This is the set of distances between all inherent feature points downstream of the faulty branch and the measurement point. For collection Sub-elements in; The baseline distance function before the fault is in The characteristic amplitude at that location; Calculate the pre-fault state using formula (7) After the fault Compare the difference in matching degree before and after the fault. ; In the actual short-circuit fault branch, the broadband harmonic energy received by the inverter at downstream characteristic points is attenuated to the greatest extent due to the truncation caused by the short-circuit fault, resulting in... The indicator shows the maximum value among all alternative branches; The candidate line segment with the largest descent is uniquely identified as the final actual fault branch, and the output includes the specific fault branch name and the mileage.

7. A single-phase-to-phase short-circuit fault location system for a distribution network, characterized in that, The system implements the single-phase-to-phase short-circuit fault location method for distribution networks as described in any one of claims 1-6, and the system includes: The upstream power supply side structure includes a 110kV upstream system power supply, which reduces the voltage to 10kV through the main transformer and connects to the 10kV bus of the distribution network. The power transmission and distribution line topology consists of a complex tree-like network structure extending downwards from the 10kV busbar, which comprises the main line, multiple branch lines, and sub-branch lines extending below the branch lines. The grid-connected source-side structure and measurement configuration system includes a high-capacity medium-voltage direct-connected multilevel inverter (DCI) directly connected to a specific node on the 10kV bus or main line. A high-frequency voltage acquisition module is connected in parallel to the three-phase output ports of the DCI, and a high-frequency current acquisition module is connected in series to form a single-ended data acquisition and processing unit. This single-ended data acquisition and processing unit integrates a high-speed analog-to-digital converter (ADC), a digital signal processor (DSP), and a data storage device for storing the coordinates of characteristic points of the transmission and distribution line topology.

8. The distribution network single-phase-to-phase short-circuit fault location system according to claim 7, characterized in that, In the transmission and distribution line topology, the line types include overhead lines and underground cable lines. Between the main line and branch lines, between lines of different wire diameters, and between the load connection points at the end of the line, due to the discontinuity of physical medium and electrical impedance, several inherent impedance discontinuity characteristic points with fixed spatial coordinates are naturally formed.

9. The distribution network single-phase-to-phase short-circuit fault location system according to claim 7, characterized in that, This system is applied to medium-voltage distribution networks with direct-connected inverters and other renewable energy sources.

10. The distribution network single-phase-to-phase short-circuit fault location system according to claim 7, characterized in that, The system is mounted on a computer-readable storage medium that stores a computer program. When the computer program is executed by a processor, it can implement the method for locating single-phase-to-phase short-circuit faults in the power distribution network.

Images