VMD-based low-current grounding power distribution network single-phase grounding fault line selection method

Abstract

The invention discloses a VMD-based low-current grounding power distribution network single-phase grounding fault line selection method, which comprises the following steps: firstly, monitoring three-phase voltage and current waveforms of each line in a power distribution network looped network system, when a single-phase grounding fault occurs in the line, recording the three-phase voltage and current waveforms of each line, and then performing fault feature extraction work; a subsequent transient fault line selection criterion is met; according to the transient fault line selection scheme, improved variational mode decomposition is carried out on zero-sequence fault features, phase comparison line selection is carried out in combination with Hilbert transformation, and a line selection result is obtained.

Description

Technical Field

[0001] The present invention belongs to the technical field of distribution network fault detection, and more specifically, relates to a single-phase grounding fault line selection method for a small current grounding distribution network based on an improved VMD algorithm. Background Art

[0002] In the current low-voltage distribution network, there are three ways to connect the neutral point, namely, the neutral point is not grounded, grounded through an arc suppression coil, and grounded through a high resistance, collectively referred to as a small current grounding system. According to the fault statistics of the power operation department, in the small current grounding system of the distribution network, single-phase grounding faults are the most common and have the highest incidence, exceeding 80% of the total faults. In addition, when a single-phase grounding fault occurs in a small current grounding system, the system can continue to operate for 1 to 2 hours under the fault condition due to the small fault current and fault voltage. Therefore, in order to prevent long-term overvoltage from causing insulation breakdown of other phases, causing the fault to develop into an interphase fault and affect power supply, it is very important to quickly and accurately select the fault line without interrupting the continuity of power supply and ensuring the safety of the distribution network.

[0003] At present, the line selection methods for small current grounding systems used in power systems are mainly divided into line selection methods based on steady-state components, line selection methods based on transient components and external injection methods. The line selection methods based on steady-state component analysis are all based on the fault signal of the single-phase grounding fault itself, and most of them are aimed at the zero-sequence component of the system, including group ratio amplitude ratio phase method, power frequency reactive direction method, negative sequence current method, fifth harmonic method, etc., which have certain stability and accuracy, but in some systems, the steady-state fault characteristics of the small current grounding system are weak, which will affect the sensitivity of line selection; although the route selection scheme based on transient components is more sensitive, it is difficult to accurately grasp the transient process of the single-phase grounding fault in the distribution network due to its short transient process; due to the need for additional equipment, the external injection method is not universal and practical.

[0004] The development of power system ring network is mainly to improve the reliability and economy of power system. In the traditional power system, the reliability of power system is guaranteed by unidirectional power supply and backup power supply, but this method has great defects, such as the backup power supply cannot guarantee 100% reliability and energy waste. The power supply ring network refers to the power supply mode in which the power supply and load point are connected into a ring by power lines. The ring network system refers to a power network composed of multiple power equipment, which are combined to form a ring structure. In this structure, the power equipment is responsible for connecting, regulating and managing voltage and current, and can provide energy through the power system. The ring network system is a network that consists of multiple connection contents on the power system, which can connect the electrical continuity of many devices in the power grid to meet the power generation, transmission, distribution and power consumption needs. The ring network system is a connected network system with high reliability wiring and a large number of nodes, which has a cyclic characteristic. It organically integrates the various parts of the power grid equipment to participate in the transmission of electric energy together, so that both the electric energy resources can be effectively utilized and the current can be guaranteed not to exceed the permitted standard. Ring power supply can improve power supply reliability. When a fault occurs in any section of the line in the ring, the faulty section can be cut off by the switch, so that the power supply to the load point will not be affected. It can also reduce voltage loss and power loss, improve power quality and economic efficiency of power supply. However, the relay protection and operation of ring power supply are relatively complicated.

[0005] At present, there are relatively few studies on line selection in distribution network rings, and the development and research of fault line selection schemes after single-phase grounding faults in distribution network rings is still an urgent problem to be solved. Therefore, accurate and rapid line selection in distribution network rings has important practical significance. Summary of the invention

[0006] The purpose of the present invention is to propose a single-phase grounding fault line selection method for a small current grounding distribution network based on an improved VMD algorithm in view of the current insufficient research on single-phase grounding faults in a small current grounding distribution network ring network system.

[0007] To achieve the above-mentioned object of the invention, the present invention proposes a single-phase grounding fault line selection method for a small current grounding distribution network based on VMD, which is characterized by comprising the following steps:

[0008] (1) Establish a simulation model of a small current grounded distribution network ring network;

[0009] (1.1) Simulation model establishment;

[0010] A low-voltage, low-current grounded distribution network ring model is established in the electromagnetic transient simulation software; three lines consisting of two groups of overhead lines are respectively derived from two power sources with the same voltage level, the three lines have the same voltage level, and the lines are connected to different loads; the two power sources are respectively used as distributed power sources to supply power to one of the lines derived from the other power source to form a ring network;

[0011] (1.2), fault simulation;

[0012] In a line of the constructed low-current grounding distribution network ring network simulation model, a simulated single-phase grounding fault controlled by a logic time control model is added. The logic time control model can control the start and duration of the single-phase grounding fault.

[0013] (2) Record the waveforms of the three-phase current and voltage of each line when a single-phase grounding fault occurs;

[0014] (3) Extract the zero-sequence current and voltage of each line;

[0015] (4) Perform transient line selection in electromagnetic transient simulation software;

[0016] In the electromagnetic transient simulation software, the zero-sequence current and voltage of each line are decomposed by improved VMD to obtain a series of IMF components; the steady-state power frequency component and the noise component are eliminated to obtain the transient attenuation component; the transient attenuation component is subjected to Hilbert transformation to obtain the instantaneous phase;

[0017] Next, the line selection is completed using the instantaneous phase criterion, and the phase criterion for line selection is: the zero-sequence current phase angle of the fault line and the normal line differs by 90°.

[0018] The object of the invention of the present invention is achieved in this way:

[0019] The present invention discloses a single-phase grounding fault line selection method for a small current grounding distribution network based on VMD. The method first monitors the three-phase voltage and current waveforms of each line in a distribution network ring system. When a single-phase grounding fault occurs in the line, the three-phase voltage and current waveforms of each line are recorded, and then fault feature extraction is performed to meet the subsequent transient fault line selection criteria. The transient fault line selection scheme: an improved variational mode decomposition is performed on the zero-sequence fault feature, and then a phase comparison line selection is performed in combination with Hilbert transform to obtain a line selection result.

[0020] At the same time, the VMD-based single-phase grounding fault line selection method for a small current grounding distribution network of the present invention also has the following beneficial effects:

[0021] (1) The present invention adopts a combination of a steady-state line selection scheme and a transient line selection scheme after a single-phase grounding fault occurs in a small current distribution network system to perform comprehensive line selection, which has high accuracy, sensitivity and robustness;

[0022] (2) The present invention uses an improved VMD algorithm combined with Hilbert transform, which avoids the disadvantages of the common VMD algorithm and has certain advantages;

[0023] (3) The present invention has high accuracy and sensitivity, as well as high robustness, and is not only applicable to different grounding types of common small current grounding systems, but also to single-phase grounding fault line selection in distribution network ring network systems;

[0024] (4) The research on traditional single-phase grounding fault line selection scheme is more targeted at basic small current grounding distribution network systems; the development and research of fault line selection schemes after single-phase grounding faults in distribution network ring networks is still an urgent problem to be solved. Therefore, as in the present invention, in the distribution network ring network system, accurate and fast line selection has important practical significance; BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 This is a flow chart of the single-phase ground fault line selection method for small current grounding distribution network based on VMD;

[0026] Figure 2 The topological diagram of the simulation model that the present invention is expected to build;

[0027] Figure 3 It is the logic diagram for selecting the transient line selection scheme; Figure 4 is a flow chart of data processing of the present invention; DETAILED DESCRIPTION

[0028] The specific implementation of the present invention is described below in conjunction with the accompanying drawings so that those skilled in the art can better understand the present invention. It should be noted that in the following description, when the detailed description of known functions and designs may dilute the main content of the present invention, these descriptions will be omitted here.

[0029] Example

[0030] Figure 1 It is a flow chart of the single-phase grounding fault line selection method of a small current grounding distribution network based on VMD of the present invention.

[0031] In this embodiment, the simulation is performed in electromagnetic transient simulation software (such as PSCAD), such as Figure 1 As shown, a line selection scheme for a single-phase grounding fault in a small current grounding system of a distribution network ring network in the present invention comprises the following steps:

[0032] S1. Build a small current grounding distribution network ring network;

[0033] In the electromagnetic transient simulation software simulation, a low-voltage, low-current grounded distribution network is established (the present invention adopts a 10KV voltage level); three lines of different lengths are respectively drawn from the power supply (20KM, 30KM, and 41KM in the embodiment of the present invention), the three lines have the same voltage level, and different loads are connected; another power supply of the same voltage level is connected to the transformer on the load side of one of the lines, three lines with the same parameters are drawn from the power supply, and the main power supply supplies power to one of the lines to form a ring network, such as Figure 2 As shown;

[0034] S2. Simulate the occurrence of a single-phase grounding fault on a certain line of the system;

[0035] In a line of a built distribution network with a small current grounding ring, a simulated single-phase grounding fault controlled by a logic time control model is added. The logic time control model can control the start and duration of the single-phase grounding fault; the topological model of the system, the relevant modeling parameters and the expected location of the single-phase grounding fault are as follows: Figure 2 As shown;

[0036] S3. Record the current and voltage of each line in the system when a fault occurs;

[0037] The appearance of positive sequence, negative sequence and zero sequence is to analyze when the system voltage and current are asymmetrical, and to decompose the asymmetrical components of the three phases into symmetrical components and zero-sequence components of the same direction; as long as it is a three-phase system, these three components can be decomposed. For an ideal power system, due to the symmetry of the three phases, the values ​​of the negative sequence and zero sequence components are both zero. The conditions for generating zero sequence current: whether it is a longitudinal fault or a transverse fault, or the asymmetry between normal and abnormal times, as long as there is a zero sequence voltage and a zero sequence current path, a zero sequence current will be generated; the zero sequence component of the system fault is an important basis for the line selection of the present invention, so the zero sequence component is first extracted from the system;

[0038] The adder and multiplier can calculate the zero-sequence current and voltage of each line; the output element and waveform display element in the electromagnetic transient simulation software can display the current and voltage of each phase when the system is running in real time, and record the output data;

[0039] S4. Calculate the zero-sequence component when a fault occurs in each line;

[0040] The adder and multiplier can calculate the zero-sequence current and voltage of each line; the output element and waveform display can display the zero-sequence current and voltage of each phase in real time when the system is running;

[0041] S5. Single-phase grounding fault in ring network has the characteristics of diverse fault current flow direction, easy interference of fault characteristics, unstable phase and polarity of fault characteristics. The VMD algorithm can automatically adapt to the local characteristics and changes of the signal, and make corresponding adjustments according to the scale and frequency characteristics of the signal. This makes the VMD algorithm better suitable for complex and changeable signals; and the VMD algorithm also has orthogonality and energy accumulation. Compared with algorithms such as empirical mode decomposition (EMD) and local mean decomposition (LMD), it avoids the problems of mutual interference and overlap that may exist in traditional modal decomposition methods; and the VMD algorithm has strong scalability and flexibility, can be applied to different types of signal analysis, and can also be combined with other signal processing methods, with better analysis results and a larger scope of use.

[0042] Although the VMD algorithm has many advantages in signal analysis and modal decomposition, it also has some disadvantages: there are some parameters in the VMD algorithm that need to be selected, such as regularization parameters and number of iterations. The selection of these parameters is crucial to the accuracy and stability of the decomposition results. However, if the parameters are not selected appropriately, it may lead to unsatisfactory results or failure to converge; the VMD algorithm is sensitive to the noise present in the signal. Noise introduces random components into the signal, which may interfere with the modal decomposition of the signal. Therefore, an improved transient line selection scheme is used for line selection in the present invention.

[0043] The variational mode decomposition (VMD) method is a new nonlinear and non-stationary signal analysis method proposed by DRAGOMIRETSKIY and ZOSSO. VMD is used to decompose the fault components of each line of the system when single-phase grounding occurs, and the sample entropy of each inherent mode function (IMF) is obtained together with the time domain characteristics of the original signal to form a feature matrix. In the VMD algorithm, the preset scale number K and the penalty parameter α are the most important setting parameters in the signal decomposition process. If the K value is not selected properly, it is very easy to produce under-decomposition or over-decomposition. At present, the parameter setting method commonly used by researchers is the default value method, which will not ensure whether the parameter selection is suitable for the signal being analyzed. In addition, although intelligent algorithms can be used to optimize relevant parameters, the process is relatively complicated and inefficient, and the influence of the selected parameters on the signal reconstruction characteristics is not considered.

[0044] When using the VMD algorithm, its parameters need to be selected. Generally, the selection is based on experience or experiments. The present invention uses a method for selecting a penalty parameter α based on a preset scale number K of a time-frequency entropy stability criterion and a multi-scale sample entropy minimum criterion.

[0045] In order to effectively extract the morphological spectrum characteristics of the signal, the background components and environmental noise in the signal need to be filtered out. To this end, the collected fault feature signal is subjected to VMD decomposition. The preset scale number K and the penalty parameter α are the most important parameters that affect the decomposition accuracy during the VMD decomposition process. Among them, the K value determines the number of IMF components decomposed. If it is improperly selected, it is very easy to produce under-decomposition or over-decomposition, resulting in modal aliasing problems; the selection of α determines the bandwidth of each IMF component. The smaller α is, the larger the bandwidth of the IMF component is, and vice versa, the smaller the bandwidth of each IMF component is. At present, the default parameter setting value method is a common method, but this will not ensure whether the selection of parameters is suitable for the analyzed signal. In addition, although intelligent algorithms can be used to optimize relevant parameters, the process is relatively complicated and inefficient, and the influence of the selected parameters on the signal reconstruction characteristics is not considered.

[0046] To this end, the present invention proposes to apply a method for selecting a preset scale number K based on the time-frequency entropy stability criterion and a penalty parameter α of the multi-scale sample entropy minimum criterion to an improved VMD algorithm. The time-frequency distribution of a signal is an effective characterization of the energy change of the signal at different frequencies within the sampling time, and is less affected by the noise factors in the signal. The time-frequency entropy is a quantitative description of the degree of this change. The IMF components obtained by VMD decomposition are usually distributed from low frequency to high frequency. If the optimal preset scale number K is obtained, the time-frequency distribution of each IMF component is more reasonable; as the K value increases, the time-frequency distribution characteristics should have better stability. Therefore, the fault characteristics are analyzed by improved variational mode decomposition: the IMF components obtained by VMD decomposition are usually distributed from low frequency to high frequency. If the optimal preset scale number K is obtained, the time-frequency distribution of each IMF component is more reasonable; as the K value increases, the time-frequency distribution characteristics should have better stability. Therefore, the authors superimpose the amplitudes of the same instantaneous frequency points obtained by Hilbert transform of each IMF component under different K values, construct a sparse matrix, and divide the matrix into equal-area regions to obtain the time-frequency entropy value, and determine the value of K based on the time-frequency entropy value.

[0047] The expression of VMD decomposition current signal i(t) is:

[0048]

[0049] Where: K is the number of modal components (IMF); are the impulse function and its partial derivative with respect to t respectively; {Hk}, {ωk} are the Kth IMF after the current signal is decomposed and its corresponding center frequency.

[0050] Introducing the quadratic penalty factor α and the Lagrangian multiplication operator λ(t), the extended Lagrangian expression is:

[0051]

[0052] The alternating direction multiplier iteration algorithm is used to iteratively update the solution. Combined with the principle of VMD center frequency adaptive selection, VMD is used to decompose zero-sequence current and zero-sequence voltage to obtain the corresponding transient quantity.

[0053] S6. Line selection based on Hilbert transform:

[0054] Due to the scalability and flexibility of the VMD algorithm, it can be combined with other analysis methods for line selection; and due to the shortcomings of the VMD algorithm such as sensitivity in parameter selection, possible modal aliasing, and sensitivity to noise, it is also necessary to combine other algorithms for the final transient line selection. In the present invention, the line selection is combined with Hilbert transform:

[0055] Hilbert transform is a signal processing technique, usually used to analyze the phase information of a signal. Suppose there is a real-valued function x(t), whose Hilbert transform is denoted by Or H[x(t)]:

[0056]

[0057] The inverse transform is:

[0058]

[0059] Compared with the concept of convolution, we can find that the above Hilbert transform expression is actually the result of convolving an original signal with another signal. Therefore, Hilbert transform can be regarded as passing the original signal through a filter or a system.

[0060] Hilbert transform has the advantages of analytical signal representation, energy preservation, comprehensive analysis in time domain and frequency domain, and high computational efficiency. When used in combination with VMD algorithm, it can significantly improve the stability of line selection and overcome the disadvantage that VMD algorithm is susceptible to interference. Therefore, the present invention uses Hilbert transform for the final transient line selection:

[0061] Collect signal data: First, you need to collect the signal data for line selection. This can be sound, electrical signals, or other types of waveform data.

[0062] Apply Hilbert Transform: Use Hilbert Transform to transform the signal data. Hilbert Transform generates a complex signal which includes the phase information of the original signal. This can be done by using Hilbert Transform function in signal processing tools or programming languages.

[0063] Extract phase information: Extract phase information from the result of Hilbert transform. This can be done by calculating the magnitude and phase angle of the complex signal.

[0064] Select phase information: Filter or select the extracted phase information based on selection criteria. This may involve setting thresholds, applying filters, or performing other signal processing operations to select specific phase angles or frequency components.

[0065] The transient current and transient voltage of a single-phase grounding fault have opposite phases and polarities in the first half-wave; the transient current and transient voltage of a normal operating circuit have the same first half-wave and polarity. Comparing the more obvious modal functions extracted by the VMD algorithm, the line with the opposite polarity of transient current to other lines is the fault line.

[0066] In summary, we perform improved VMD decomposition on the original signal to obtain a series of IMF components; remove the steady-state power frequency component and the noise component to obtain the transient attenuation component; perform Hilbert transform on the transient attenuation component to obtain the instantaneous phase; and use the phase criterion to complete the line selection. The phase criterion for line selection is: the zero-sequence current phase angle of the fault line and the normal line differs by 90°.

[0067] In this embodiment, the logic diagram of the transient line selection scheme provided by the present invention is as follows: Figure 3 shown.

[0068] Although the above describes the illustrative specific embodiments of the present invention to facilitate those skilled in the art to understand the present invention, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations using the concept of the present invention are protected.

Claims

1. A VMD-based single-phase grounding fault line selection method for a small current grounding distribution network, characterized in that: The following steps are involved: (1) Establish a simulation model of a small current grounded distribution network ring network; (1.1) Simulation model establishment; A low-voltage, low-current grounded distribution network ring model is established in the electromagnetic transient simulation software; three lines consisting of two groups of overhead lines are respectively derived from two power sources with the same voltage level, the three lines have the same voltage level, and the lines are connected to different loads; the two power sources are respectively used as distributed power sources to supply power to one of the lines derived from the other power source to form a ring network; (1.2), fault simulation; In a line of the constructed low-current grounding distribution network ring network simulation model, a simulated single-phase grounding fault controlled by a logic time control model is added. The logic time control model can control the start and duration of the single-phase grounding fault. (2) Record the waveforms of the three-phase current and voltage of each line when a single-phase grounding fault occurs; (3) Extract the zero-sequence current and voltage of each line; (4) Perform transient line selection in electromagnetic transient simulation software; In the electromagnetic transient simulation software, the zero-sequence current and voltage of each line are decomposed by improved VMD to obtain a series of IMF components; the steady-state power frequency component and noise component are eliminated to obtain the transient attenuation component; Perform Hilbert transform on the transient attenuation component to obtain the instantaneous phase; Next, the line selection is completed using the instantaneous phase criterion, and the phase criterion for line selection is: the zero-sequence current phase angle of the fault line and the normal line differs by 90°.

Images