A low-voltage power distribution network fault detection method and device

Abstract

The application relates to the technical field of power distribution network fault detection, in particular to a low-voltage power distribution network fault detection method and device, which comprises the following steps: adopting a pulse signal injection method, injecting a pulse signal into a low-voltage power distribution network at different pulse widths each time, receiving the reflected signal, and recording the response signal after each pulse injection; performing modal decomposition on the response signal, obtaining a plurality of modal components, and calculating the relative difference of each modal component after each pulse injection; obtaining the modal component corresponding to the fault point; calculating the first interference value and the second interference value after each pulse injection; determining the pulse width adjustment amount and the pulse width of the next pulse injection, selecting the optimal response signal by continuously injecting pulses, and positioning a single-phase ground fault in the low-voltage power distribution network. The application improves the dynamic adaptability of the arc nonlinearity characteristics and the network topology in the power distribution network by adjusting the pulse width in real time, and improves the accuracy of the pulse signal injection method in fault detection.

Description

Technical Field

[0001] This application relates to the field of power distribution network fault detection technology, specifically to a method and device for detecting faults in low-voltage power distribution networks. Background Technology

[0002] Low-voltage distribution networks present significant challenges in fault detection due to their complex topology and the integration of various electrical devices. Single-phase grounding faults, in particular, occur frequently in low-current grounding systems and are difficult to detect, significantly impacting line repair and power supply reliability.

[0003] However, single-phase ground faults are often accompanied by intermittent arcing, and their impedance characteristics exhibit significant nonlinearity and time-varying properties, leading to severe distortion of the reflected signal waveform. Furthermore, interference from reflected signals from grid nodes and the nonlinear impedance characteristics of intermittent arcs further complicates the characteristics of ground faults, making accurate extraction difficult. Fixed-width pulse signals are ill-suited to complex signals with varying levels of interference and lack a dynamic adaptation mechanism to the nonlinear characteristics of arcs and the influence of network topology. The detection interference experienced by the pulse signal varies at different locations within the distribution network, resulting in decreased adaptability and accuracy of the pulse signal injection method for fault detection. Summary of the Invention

[0004] To address the aforementioned technical problems, a method and apparatus for detecting faults in low-voltage distribution networks are provided to resolve existing issues.

[0005] The solution to the technical problem of this application is to provide a method and device for detecting faults in low-voltage distribution networks, including the following steps:

[0006] In a first aspect, embodiments of this application provide a method for detecting faults in a low-voltage distribution network, the method comprising the following steps:

[0007] The pulse signal injection method is adopted. After each pulse signal is injected into the low-voltage distribution network with a different pulse width, the reflected signal is received and recorded as the response signal after each pulse injection.

[0008] Modal decomposition is performed on the response signal after each pulse injection to obtain multiple modal components, and the modal components are numbered. The differences in energy distribution in the frequency domain between each modal component after each pulse injection and the corresponding modal component after the previous pulse injection are analyzed, and the relative differences between each modal component after each pulse injection are calculated.

[0009] Based on the relative differences of all modal components after each pulse injection, the modal components corresponding to the fault point after each pulse injection are obtained.

[0010] After each pulse injection, the changes in the amplitudes of all signals surrounding the maximum signal amplitude in the modal component corresponding to the fault point are analyzed. Combined with the relative difference of the modal component corresponding to the fault point, the first interference value after each pulse injection is calculated. Combined with the energy proportion of the modal component corresponding to the fault point in the frequency domain, the second interference value after each pulse injection is determined.

[0011] Based on the second interference value after each pulse injection, the pulse width adjustment amount for the next pulse injection is determined. Combined with the pulse width at the time of each pulse injection, the pulse width at the time of the next pulse injection is obtained. By continuously injecting pulses, the optimal response signal is selected to locate single-phase grounding faults in low-voltage distribution networks.

[0012] Preferably, the step of numbering the modal components includes: numbering all modal components after each pulse injection in order from high frequency to low frequency.

[0013] Preferably, the calculation of the relative difference between each modal component after each pulse injection includes:

[0014] Frequency domain analysis is performed on each modal component to obtain the marginal spectrum of each modal component; after normalizing all energies in the marginal spectrum, an energy sequence is formed.

[0015] The relative difference is the distance between the energy sequence of each modal component after each pulse injection and the modal component with the same index after the previous pulse injection.

[0016] Preferably, the further acquisition process of the modal component corresponding to the fault point is as follows: the modal component corresponding to the maximum value of all the relative differences after each pulse injection is recorded as the modal component corresponding to the fault point.

[0017] Preferably, the calculation of the first interference value after each pulse injection includes:

[0018] After each pulse injection, all signal amplitudes between the maximum value of all signal amplitudes in the modal component corresponding to the fault point and the positions where the signal amplitudes on both sides drop to a preset multiple of the maximum value are formed into a signal sequence.

[0019] Calculate the energy of the signal sequence, denoted as the local energy, and perform a negative mapping on the local energy;

[0020] The first interference value is the product of the relative difference of the modal component corresponding to the fault point and the result of the negative mapping.

[0021] Preferably, determining the second interference value after each pulse injection includes:

[0022] The sum of all energies in the marginal spectrum of each modal component at each pulse injection is calculated and denoted as the total energy.

[0023] The ratio between the total energy corresponding to the modal component at the fault point during each pulse injection and the sum of the total energy of all modal components is denoted as the energy percentage.

[0024] The second interference value is the product of the energy percentage and the first interference value.

[0025] Preferred, the first Pulse width adjustment during sub-pulse injection The calculation formula is: ,in, For the first Pulse width adjustment during sub-pulse injection For the first The second interference value after the next pulse injection. For the first The second interference value after the next pulse injection. This is the normalization function.

[0026] Preferably, the pulse width at the next pulse injection is the sum of the pulse width at each pulse injection and the pulse width adjustment amount.

[0027] Preferably, the method for obtaining the optimal response signal is as follows: if the pulse width during the next pulse injection exceeds the preset adjustment range, the response signal after this pulse injection is recorded as the optimal response signal.

[0028] Secondly, embodiments of this application also provide a low-voltage distribution network fault detection device, 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 of the above-described low-voltage distribution network fault detection methods.

[0029] This application has at least the following beneficial effects:

[0030] This application obtains multiple modal components by performing modal decomposition on the response signal. Its advantage lies in decomposing the complex original signal, which involves superposition and interference, into multiple relatively simple modes, facilitating the subsequent extraction of signal features corresponding to each reflection point. By analyzing the energy distribution differences of the modal components in the frequency domain after two adjacent pulse injections, the relative difference is calculated. This considers the energy difference changes of the modal components corresponding to each reflection point after two pulse injections, allowing for the selection of modes with significant changes to obtain the modal components corresponding to the fault point. Furthermore, by analyzing the signal amplitude within the main pulse range of the modal components corresponding to the fault point, and the differences in the modal components corresponding to the fault point between two adjacent pulse injections, the first interference value after each pulse injection is calculated. This considers the degree of excitation of the intermittent arc at the single-phase grounding fault point by the pulse signal, reflecting the distortion of the response signal caused by the arc's nonlinear impedance characteristics, thereby assessing the detection interference generated by the nonlinear characteristics of the arc excited by the injected pulse. Further, it determines the relative difference in energy distribution of the modal components after two adjacent pulse injections. The second interference value after injection has the advantage of considering the dominance of the signal at the fault point in the overall response signal, reflecting the full excitation of the fault point by the pulse signal of this pulse width, and thus reflecting the degree of detection interference caused by the nonlinear characteristics of the pulse-excited arc. This allows for subsequent reduction of the pulse width to decrease the interference of the nonlinear characteristics of the arc. The pulse width adjustment amount for the next pulse injection is obtained, determining the pulse width for the next pulse injection. This has the advantage of adjusting the pulse width of the next pulse injection in real time based on the interference received after each pulse injection, thereby reducing the detection interference of the nonlinear characteristics of the pulse-excited arc. The optimal response signal is obtained for locating single-phase ground faults. This has the advantage of selecting the optimal response signal by adjusting the pulse width in real time, reducing the distortion of the signal waveform caused by the nonlinear characteristics of the arc, balancing the interference of reflected signals from grid nodes on the fault point, improving the dynamic adaptability to the nonlinear characteristics of the arc and the network topology in the distribution network, and improving the accuracy of the pulse signal injection method for fault detection. Attached Figure Description

[0031] The following description, in conjunction with the accompanying drawings, provides a more detailed explanation of a low-voltage distribution network fault detection method of this application.

[0032] Figure 1 A flowchart illustrating the steps of a low-voltage distribution network fault detection method provided in this application embodiment;

[0033] Figure 2 A flowchart illustrating the steps of a method for obtaining modal components corresponding to a fault point, as provided in an embodiment of this application.

[0034] Figure 3A flowchart illustrating the steps of a method for obtaining the first interference value after each pulse injection, as provided in an embodiment of this application. Detailed Implementation

[0035] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description of a low-voltage distribution network fault detection method and apparatus proposed in this application, in conjunction with the accompanying drawings and embodiments, is provided. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0036] 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 pertains.

[0037] Please see Figure 1 The diagram illustrates a flowchart of a low-voltage distribution network fault detection method according to an embodiment of this application. The method includes the following steps:

[0038] Step 1: Using the pulse signal injection method, after each pulse signal with a different pulse width is injected into the low-voltage distribution network, the reflected signal is received and recorded as the response signal after each pulse injection.

[0039] Because distribution network lines are complex with multiple branch points and are often mixed cable lines, accurate fault location is crucial after a fault occurs. Existing fault location methods can be divided into impedance methods, traveling wave methods, and waveform data analysis methods. Among them, the impedance method mainly analyzes the impedance characteristics before and after the fault, and is greatly affected by factors such as fault resistance and inaccurate line parameter settings. The waveform data analysis method uses a distance function constructed from voltage and current data after the fault to achieve fault distance measurement, and is greatly affected by high-frequency components in the collected signal. The traveling wave method mainly calculates the fault distance by detecting the propagation time of transient traveling waves on the grounding electrode line. Specifically, it can be divided into single-ended method, double-ended method, and pulse injection method. Among them, the pulse signal injection method, as an offline active fault location method, takes the faulty line out of operation after a fault occurs in the distribution network and injects a high-voltage pulse signal into the faulty line. By analyzing the reflected pulse response, the fault is detected and located, resulting in high detection accuracy.

[0040] Secondly, single-line grounding faults are more likely to occur in high-voltage transmission line faults, and the current in the faulted line is zero. All the unbalanced current in the system flows into the ground through the non-faulted lines and into the earth through the resistor at the grounding point, which causes the voltage of the line to rise to a certain extent. Therefore, the method of measuring the distance of the faulted grounding line by injecting multiple sets of pulse signals with variable pulse widths can effectively improve the accuracy of fault location.

[0041] Therefore, when a single-line grounding fault occurs in a low-voltage distribution network, a high-voltage pulse signal is injected into the single-phase line. The injected signal is a rectangular pulse, the pulse injection source is a DC voltage source, and the amplitude of the pulse signal is set to 5kV. As another implementation method, the implementer can set it according to the actual situation. The pulse signal injection process is controlled by a high-voltage circuit breaker. The pulse width of the high-voltage pulse source is set to an adjustment range of 1µs to 10µs. As another implementation method, the implementer can set it according to the actual situation.

[0042] An injection point is selected in the low-voltage distribution network. After the pulse injection source is started, a pulse signal with a different pulse width is injected into the low-voltage distribution network each time. The pulse signal propagates in the distribution network and will be reflected when it encounters the fault location point. The reflected signal is received by the signal receiving device, converted into an electrical signal, and preprocessed by filtering, amplification and other processes to obtain the response signal after each pulse injection.

[0043] In this embodiment, a pulse signal with a pulse width of 5µs is injected for the first time, and a pulse signal with a pulse width of 6µs is injected for the second time. The time period of each pulse injection is 2.5s, and the frequency of the received signal is 1MHz. As other implementation methods, the implementer can set the frequency according to the actual situation. Next, a bandpass filter is used to filter the signal to remove noise, and an amplifier is used to amplify the signal. The preprocessing processes such as filtering and amplification are known technologies and will not be described in detail here.

[0044] Thus, the response signal after each pulse injection is obtained.

[0045] Step 2: Perform modal decomposition on the response signal after each pulse injection to obtain multiple modal components and number them; analyze the difference in energy distribution in the frequency domain between each modal component after each pulse injection and the corresponding modal component after the previous pulse injection, and calculate the relative difference of each modal component after each pulse injection; based on the relative difference of all modal components after each pulse injection, obtain the modal component corresponding to the fault point after each pulse injection.

[0046] When a pulse signal propagates in a distribution network, it is reflected upon reaching low-voltage distribution network nodes and line fault points due to changes in wave impedance. Secondly, single-phase ground faults in low-voltage distribution networks easily generate intermittent arcs. Because the distances between the grid node, the line fault point, and the injection point vary, the excitation degree of the intermittent arc by the high-voltage pulse differs. The nonlinear impedance characteristics of the arc cause varying degrees of waveform distortion in the fault signal, increasing its complexity and affecting the detection of singularities. Therefore, for single-phase ground faults in low-voltage distribution networks, pulse signals with fixed pulse widths experience different levels of detection interference at different reflection locations, leading to a decrease in the adaptability and accuracy of fault detection using the pulse signal injection method. Thus, the choice of pulse width has a significant impact on the ranging accuracy of the pulse injection method: narrower pulse widths reduce the ranging dead zone but contain more high-frequency components, resulting in a shorter measurable fault distance; conversely, wider pulse widths are easier to identify but increase the ranging dead zone. Choosing an appropriate pulse width can improve the accuracy of fault location.

[0047] Secondly, when a pulse signal propagates in a distribution network, it is reflected due to changes in wave impedance at different locations such as fault points and grid nodes. This results in the response signal actually containing comprehensive information from multiple reflected signals. These reflected signals superimpose and interfere with each other, making the original response signal quite complex and difficult to directly distinguish the characteristics corresponding to each reflection point. By performing mode decomposition on the response signal, the complex response signal is broken down into multiple relatively simple modes. Each mode reflects the specific modal characteristics of the reflected signals generated at different locations in the distribution network, which facilitates the subsequent extraction of signal features corresponding to each reflection point and helps identify the location of the fault point. Specifically:

[0048] Modal decomposition is performed on the response signal after each pulse injection to obtain multiple modal components;

[0049] In this embodiment, the Hilbert transform algorithm is used for decomposition to obtain multiple modal components. The Hilbert transform algorithm is a well-known technology and will not be described in detail here. As other implementation methods, implementers can use other methods of the prior art, such as the LMD algorithm, etc. This embodiment does not impose any special restrictions on this.

[0050] Furthermore, when the pulse width changes slightly, the operating state of the distribution network remains relatively stable during the short time interval between two adjacent injected pulse signals, except for the dynamic characteristics of the fault point, such as intermittent arcs, which may change. The reflection characteristics of other normal nodes change little. Secondly, since the number of grid nodes and fault points around the injection point is fixed, the number of reflected signals generated by the wave impedance characteristics is also fixed. Therefore, the number of modal components decomposed from the response signal after two adjacent pulse injections is the same, and each modal component has specific frequency and time characteristics, corresponding to a specific reflection source of the signal in the distribution network. The position of the reflection source does not change between two adjacent injections, so the order of the decomposed modal components is also the same.

[0051] Secondly, because the coupling amplification of signals of different frequencies at low-voltage distribution network nodes exhibits good linearity, their response to different frequencies changes relatively little. However, the intermittent arc effect at a single-phase ground fault point has strong nonlinear characteristics. Therefore, different reflection points—the fault point and the grid node—have different energy response characteristics at different frequencies in the frequency domain. Specifically, due to the nonlinear characteristics of the arc, the energy of the reflected signal at the fault point may vary significantly within certain frequency ranges, while the energy distribution of the reflected signal at a normal grid node is relatively stable. By analyzing the marginal spectrum, these differences in energy distribution can be observed intuitively, thus effectively distinguishing the modes corresponding to the fault point and the normal grid node.

[0052] Based on the above analysis, frequency domain analysis is performed on the modal components. By analyzing the differences in frequency domain between modal components of the same sequence between two adjacent pulse injections, the modal components corresponding to the fault point are selected. The flowchart of the method for obtaining the modal components corresponding to the fault point provided in this application embodiment is as follows: Figure 2 As shown, it specifically includes:

[0053] All modal components after each pulse injection are numbered in order from high frequency to low frequency;

[0054] Frequency domain analysis is performed on each modal component to obtain the marginal spectrum of each modal component;

[0055] In this embodiment, the Hilbert transform algorithm is used to obtain the marginal spectrum. The Hilbert transform algorithm is a well-known technique and will not be described in detail here.

[0056] After normalizing all energies in the marginal spectrum, an energy sequence is formed;

[0057] In this embodiment, the maximum-minimum normalization method is used for processing. The maximum-minimum normalization method is a well-known technique and will not be described in detail here. As other implementation methods, implementers can use other methods of the prior art, such as the Z-score normalization method. This embodiment does not impose any special restrictions on this.

[0058] It should be noted that, due to the characteristics of different modal components corresponding to specific frequency components, the range of energy in the marginal spectrum of different modal components varies greatly. By normalizing, the comparison results can be avoided due to the large difference in the absolute value of energy when comparing and analyzing different modal components.

[0059] Calculate the distance between the energy sequence of each modal component after each pulse injection and the modal component with the same index after the previous pulse injection, and record it as the relative difference.

[0060] In this embodiment, the distance is measured by calculating the DTW distance between each modal component after each pulse injection and the modal component with the same index after the previous pulse injection. The calculation of the DTW distance is a well-known technique and will not be described in detail here. As other implementation methods, implementers may use other methods of the prior art, such as Euclidean distance, etc. This embodiment does not impose any special restrictions on this.

[0061] The modal component corresponding to the maximum value of all the relative differences after each pulse injection is denoted as the modal component corresponding to the fault point.

[0062] It should be noted that the larger the relative difference, the more significant the change in the energy distribution of the corresponding modal component in the frequency domain when the pulse width changes. This reflects a significant change in the signal characteristics of the reflection point corresponding to the modal component, indicating a greater degree of excitation of the arc effect at the fault point. The greater the detection interference generated by the nonlinear characteristics of the arc, the more likely the modal component is to be the fault point.

[0063] Thus, the modal components corresponding to the fault point after each pulse injection are obtained.

[0064] Step 3: After each pulse injection, analyze the changes in the amplitudes of all signals surrounding the maximum signal amplitude in the modal component corresponding to the fault point, and calculate the first interference value after each pulse injection by combining the relative difference of the modal components corresponding to the fault point.

[0065] Furthermore, the flowchart of the method for obtaining the first interference value after each pulse injection provided in this application embodiment is as follows: Figure 3 As shown.

[0066] First, analyze the changes in the signal amplitude within the main pulse range of the modal component corresponding to the fault point, and calculate the local energy, specifically:

[0067] From the maximum value of all signal amplitudes in the modal component corresponding to the fault point after each pulse injection, extend to both sides until the signal amplitudes on both sides drop to a preset multiple of the maximum value. All signal amplitudes between these positions are used to form a signal sequence.

[0068] In this embodiment, the preset multiple is set to 0.707. In signal processing, decibels (dB) are used to describe the relative change of a signal. 3dB corresponds to 0.707 times the amplitude change. When the signal amplitude drops to 0.707 times the maximum value, the corresponding power is halved. Therefore, the range of the main pulse is usually defined by a certain proportion of the signal amplitude dropping to the maximum value. The 3dB point, that is, 0.707 times the maximum value, is a widely used reference standard in signal processing.

[0069] Calculate the energy of the signal sequence and denote it as the local energy;

[0070] It should be noted that the calculation of the energy of the signal sequence is a well-known technique. The specific calculation process is the sum of the squares of all signal amplitudes within the signal sequence, which is denoted as the local energy.

[0071] It should be noted that the smaller the local energy, the lower the signal amplitude and the weaker the intensity of the signal sequence. When performing abrupt change detection on the signal reflected from the fault point, the weaker signal is more susceptible to noise and other interference factors. Furthermore, when the signal is weak, the detection algorithm has difficulty accurately identifying abrupt changes in the signal, which can easily lead to false detections or missed detections, thus introducing a large error. This will affect the accurate location of the fault point and the accurate assessment of the fault severity.

[0072] Furthermore, based on the relative difference and the local energy, a first interference value is determined, specifically as follows:

[0073] Negative mapping is applied to the local energy;

[0074] In this embodiment, the specific process of negative mapping is as follows: negative mapping is performed on the local energy using an exponential function. Let the local energy be denoted as E. The result is used as the result of the negative mapping, where, It is an exponential function with the natural constant as the base.

[0075] The product of the relative difference of the modal component corresponding to the fault point after each pulse injection and the result of the negative mapping is used as the first interference value after each pulse injection.

[0076] It should be noted that the relative difference of the modal components corresponding to the fault point represents the maximum value of all relative differences after each pulse injection. Secondly, the larger the relative difference of the modal components corresponding to the fault point, the more obvious the nonlinear impedance characteristics of the intermittent arc at the fault point, the greater the degree of arc excitation and dynamic changes, resulting in significant changes in the frequency and energy distribution of the reflected signal. The smaller the local energy, the larger the negative mapping result, indicating that the intensity of the reflected signal corresponding to the fault point is weak and the reflection characteristics are not obvious. The larger the first interference value, the more significant the distortion of the response signal caused by the nonlinear impedance characteristics of the arc at the single-phase grounding fault point, and the weaker the signal. This may be due to the high degree of excitation of the arc nonlinear characteristics by the pulse and the excessively low local energy, resulting in significant signal distortion. Therefore, the detection interference generated by the nonlinear characteristics of the arc excited by the injected pulse is large. The pulse width can be appropriately reduced to reduce the degree of arc excitation and thus reduce interference.

[0077] Thus, the first interference value after each pulse injection is obtained.

[0078] Step 4: Based on the energy ratio of the modal component corresponding to the fault point in the frequency domain after each pulse injection, and in combination with the first interference value, determine the second interference value after each pulse injection.

[0079] Furthermore, if the energy proportion of the reflected signal at the fault point is larger, the interference from the reflected signals of surrounding power grid nodes to the fault point will be smaller. By comparing the signal strengths corresponding to different modal components after each pulse injection, and combining them with the first interference value, the second interference value is calculated, specifically:

[0080] Calculate the sum of all energies in the marginal spectrum of each modal component after each pulse injection, and denote it as the total energy;

[0081] The ratio between the total energy of the modal component corresponding to the fault point after each pulse injection and the sum of the total energy of all modal components is denoted as the energy percentage.

[0082] The product of the energy percentage and the first interference value is used as the second interference value after each pulse injection;

[0083] It should be noted that the larger the energy ratio, the more dominant the signal of the modal component corresponding to the fault point is in the overall response signal, indicating that the pulse width of this pulse is sufficient to excite the fault point. In this case, the pulse width can be appropriately reduced to avoid signal distortion or other interference. The larger the second interference value, the larger the detection interference generated by the nonlinear characteristics of the arc excited by the injected pulse is, and the pulse signal of this pulse width is sufficient to excite the fault point. In order to reduce the interference of the nonlinear characteristics of the arc, the pulse width can be reduced in the future.

[0084] Thus, the second interference value after each pulse injection is obtained.

[0085] Step 5: Based on the second interference value after each pulse injection, determine the pulse width adjustment amount for the next pulse injection. Combined with the pulse width during each pulse injection, obtain the pulse width for the next pulse injection. By continuously injecting pulses, select the optimal response signal to locate single-phase grounding faults in the low-voltage distribution network.

[0086] Furthermore, a narrower pulse width contains more high-frequency components. While this provides high time resolution, the pulse is prone to distortion over long distances, leading to poorer fault location. Conversely, a wider pulse width increases the proportion of low-frequency components in the injected signal, making it more likely to overlap with the frequency components of reflected signals from power grid nodes. Simultaneously, the low-frequency reflection characteristics of longer low-voltage distribution network transmission lines exacerbate the node reflection effects of wider pulses, causing the reflected signal characteristics of the fault point and power grid nodes to become more consistent, increasing node reflection interference and reducing the accuracy of fault location. The increased influence of node reflections in the low-voltage distribution network further reduces the accuracy of fault location. To balance these two effects, the pulse width is adjusted based on the second interference value, specifically as follows:

[0087]

[0088] in, For the first Pulse width adjustment during sub-pulse injection For the first Pulse width adjustment during sub-pulse injection For the first The second interference value after the next injection pulse. For the first The second interference value after the next pulse injection. As the normalization function, in this embodiment, the arctangent function is used for normalization, so that the value range of the normalization result is within... The arctangent function is a well-known technique and will not be elaborated upon here.

[0089] The sum of the pulse width during each pulse injection and the pulse width adjustment amount is used as the pulse width during the next pulse injection.

[0090] It should be noted that if the first The second pulse injection was compared to the first... If the second interference value is large during the next pulse injection, then the If it is negative, it means that the first... During the second pulse injection, the pulse width should be appropriately reduced. Therefore, the first pulse injection... Pulse width adjustment during sub-pulse injection If the value is negative, the result is... The pulse width during the second pulse injection is greater than that of the first pulse injection. The pulse width of the next pulse injection is relatively small.

[0091] It should be noted that, in this embodiment, the pulse signal is injected with a pulse width of 5µs for the first time and a pulse width of 6µs for the second time. Therefore, starting from the third time, the pulse width at the time of the third pulse injection is obtained according to the pulse width adjustment amount. This process can be repeated to obtain the pulse width at the time of multiple pulse injections.

[0092] By continuously performing multiple pulse injections, if the pulse width of the next pulse injection exceeds the preset adjustment range, the response signal of this pulse injection is selected and recorded as the optimal response signal.

[0093] In this embodiment, the preset adjustment range is 1µs to 10µs. As for other implementation methods, the implementer can set it according to the actual situation.

[0094] It should be noted that, in order to avoid infinite pulse injection, when the number of pulse injections to the low-voltage distribution network reaches the preset maximum number, the response signal at the last injection is recorded as the optimal response signal. The preset maximum number is 30. As another implementation method, the implementer can set it according to the actual situation.

[0095] Based on the optimal response signal, the location of a single-phase ground fault in the low-voltage distribution network is specifically as follows:

[0096] The arrival time of the fault signal in the optimal response signal is extracted using wavelet analysis and denoted as the time delay.

[0097] It should be noted that the process of extracting the arrival time of the fault signal from the optimal response signal using wavelet analysis is a well-known technique and will not be elaborated here.

[0098] Based on the aforementioned time delay and the propagation speed of the pulse signal in the line, the distance to a single-phase ground fault point is determined as follows:

[0099] Distance between the fault point and the pulse injection point for: ,in, This refers to the propagation speed of a pulse signal in a line. In this embodiment, the propagation speed is the time delay. It is 298 m / μs. Based on the calculated distance... To pinpoint the exact location of the fault and perform maintenance on the power distribution network.

[0100] Based on the same inventive concept as the above method, this application embodiment also provides a low-voltage distribution network fault detection device, 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 low-voltage distribution network fault detection methods.

[0101] It should be understood that, although Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0102] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0103] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application. Therefore, any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of this application, without departing from the content of the technical solution of this application, shall fall within the protection scope of the technical solution of this application.

Claims

1. A method for detecting faults in a low-voltage distribution network, characterized in that, The method includes the following steps: The pulse signal injection method is adopted. After each pulse signal is injected into the low-voltage distribution network with a different pulse width, the reflected signal is received and recorded as the response signal after each pulse injection. Modal decomposition is performed on the response signal after each pulse injection to obtain multiple modal components, and the modal components are numbered. The differences in energy distribution in the frequency domain between each modal component after each pulse injection and the corresponding modal component after the previous pulse injection are analyzed, and the relative differences between each modal component after each pulse injection are calculated. Based on the relative differences of all modal components after each pulse injection, the modal components corresponding to the fault point after each pulse injection are obtained. After each pulse injection, the changes in the amplitudes of all signals surrounding the maximum signal amplitude in the modal component corresponding to the fault point are analyzed. Combined with the relative difference of the modal component corresponding to the fault point, the first interference value after each pulse injection is calculated. Combined with the energy proportion of the modal component corresponding to the fault point in the frequency domain, the second interference value after each pulse injection is determined. Based on the second interference value after each pulse injection, the pulse width adjustment amount for the next pulse injection is determined. Combined with the pulse width for each pulse injection, the pulse width for the next pulse injection is obtained. By continuously injecting pulses, the optimal response signal is selected to locate single-phase grounding faults in low-voltage distribution networks. The calculation of the first interference value after each pulse injection includes: After each pulse injection, all signal amplitudes between the maximum value of all signal amplitudes in the modal component corresponding to the fault point and the positions where the signal amplitudes on both sides drop to a preset multiple of the maximum value are formed into a signal sequence. Calculate the energy of the signal sequence, denoted as the local energy, and perform a negative mapping on the local energy; The first interference value is the product of the relative difference of the modal component corresponding to the fault point and the result of the negative mapping; Determining the second interference value after each pulse injection includes: The sum of all energies in the marginal spectrum of each modal component at each pulse injection is calculated and denoted as the total energy. The ratio between the total energy corresponding to the modal component at the fault point during each pulse injection and the sum of the total energy of all modal components is denoted as the energy percentage. The second interference value is the product of the energy percentage and the first interference value.

2. The low-voltage distribution network fault detection method as described in claim 1, characterized in that, The process of numbering the modal components includes: numbering all modal components after each pulse injection in order from high frequency to low frequency.

3. The low-voltage distribution network fault detection method as described in claim 1, characterized in that, The calculation of the relative difference of each modal component after each pulse injection includes: Frequency domain analysis is performed on each modal component to obtain the marginal spectrum of each modal component; after normalizing all energies in the marginal spectrum, an energy sequence is formed. The relative difference is the distance between the energy sequence of each modal component after each pulse injection and the modal component with the same index after the previous pulse injection.

4. The low-voltage distribution network fault detection method as described in claim 1, characterized in that, The further acquisition process of the modal component corresponding to the fault point is as follows: the modal component corresponding to the maximum value of all the relative differences after each pulse injection is recorded as the modal component corresponding to the fault point.

5. The low-voltage distribution network fault detection method as described in claim 1, characterized in that, No. Pulse width adjustment during sub-pulse injection The calculation formula is: ,in, For the first Pulse width adjustment during sub-pulse injection For the first The second interference value after the next pulse injection. For the first The second interference value after the next pulse injection. This is the normalization function.

6. The low-voltage distribution network fault detection method as described in claim 1, characterized in that, The pulse width for the next pulse injection is the sum of the pulse width for each pulse injection and the pulse width adjustment amount.

7. The low-voltage distribution network fault detection method as described in claim 1, characterized in that, The method for obtaining the optimal response signal is as follows: if the pulse width during the next pulse injection exceeds the preset adjustment range, the response signal after this pulse injection is recorded as the optimal response signal.

8. A low-voltage distribution network fault detection device, 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 low-voltage distribution network fault detection method as described in any one of claims 1-7.

Images