Medium-voltage distribution network line fault positioning method and intelligent switch

Abstract

The invention provides a medium-voltage distribution network line fault positioning method and an intelligent switch, and belongs to the technical field of distribution networks. The method comprises the following steps: performing Hilbert-Huang transform on a distribution network fault current signal accessed to distributed energy to obtain an instantaneous frequency and an instantaneous amplitude of a key component signal; performing multi-dimensional signal phase space reconstruction based on the instantaneous frequency and the instantaneous amplitude of the key component signal to obtain a phase space reconstruction signal; and extracting the characteristic quantity of the time sequence of the distribution network fault current signal and the characteristic quantity of the phase-space reconstruction signal by using the convolutional neural network, mixing, and determining the fault occurrence position according to the mixed characteristic quantity. According to the method, firstly, fault current signals and time-frequency characteristics of the fault current signals in the medium-voltage distribution network containing the distributed resources are mined through Hilbert-Huang transform, then the internal law of the fault signals is captured through a phase-space reconstruction method, and fault line selection and accurate fault positioning of the medium-voltage distribution network containing the distributed resources are achieved through the convolutional neural network.

Description

Technical Field

[0001] The present invention relates to the technical field of power distribution networks, and in particular to a method for locating faults in a medium-voltage distribution network line and a smart switch. Background Art

[0002] The access of large-scale distributed resources (DG) has greatly increased the uncontrollability, volatility and randomness of the operation of medium-voltage distribution networks, changing the distribution network from a traditional radial network to a complex power grid with multiple power sources, and the power flow direction has changed from unidirectional to bidirectional. The dynamic switching of large-scale distributed resources and the randomness of output power make the flow direction of current in traditional power grids more complex and changeable. The fault location of distribution networks based on traditional traveling wave ranging methods and line selection methods will fail, resulting in damage to power equipment in the power grid system, affecting the normal power supply and distribution of the power grid and the stability of the power system, and even causing production safety accidents. Therefore, it is necessary to study a fast and accurate fault location method suitable for the access of large-scale distributed resources to distribution networks.

[0003] The fault location of distribution network is mainly divided into three categories: fault line selection, fault distance measurement, and section location. Fault line selection is to locate the fault line according to the reported fault information, quickly isolate the fault point, restore power supply to the non-fault area, and avoid further expansion of the accident. Fault distance measurement mainly determines the location of the fault based on the relationship between various electrical quantities when the fault occurs. Existing fault distance measurement methods mainly include impedance method, traveling wave method, and signal injection method. One traveling wave method combines frequency domain reflection method and inverse fast Fourier transform to estimate the exact location of the fault point in the cable. Another traveling wave method is to determine the fault interval based on the arrival time series of voltage traveling waves, and then use the double-end method to accurately locate the fault in the determined interval. Another method is based on the traveling wave characteristics and reference measurement points, calculate the distance from each fault point to the reference point, and determine the fault point location by selecting the maximum value. However, the above distribution network fault location methods do not consider the situation when the distribution network contains highly permeable distributed resources, and cannot achieve accurate positioning of large-scale distributed resources connected to the distribution network line fault.

[0004] As a core edge device of the distribution network, the smart switch integrates control, metering and communication functions, which can better complete the line breaking task, improve the reliability of breaking, and realize the intelligent operation of the circuit breaker. The smart switch can replace the traditional circuit breaker, contactor, thermal relay, fuse, etc., and can realize the isolation of the fault section of the distribution network and the grid connection operation after the fault recovery. Therefore, it is necessary to study the smart switch suitable for fast and accurate fault location, fault section isolation and fault recovery of large-scale distributed resources access to the distribution network. Summary of the invention

[0005] In order to solve the above technical defects, the present invention provides a medium voltage distribution network line fault location method and a smart switch.

[0006] One aspect of the present invention provides a method for locating a medium voltage distribution network line fault, comprising:

[0007] Performing Hilbert-Huang transform on the distribution network fault current signal connected to the distributed energy source to obtain the instantaneous frequency and instantaneous amplitude of multiple key component signals of the distribution network fault current signal;

[0008] Perform multi-dimensional signal phase space reconstruction based on the instantaneous frequencies and instantaneous amplitudes of multiple key component signals to obtain a phase space reconstructed signal; wherein the delay time of the signal phase space reconstruction is determined according to the cross-correlation function of the instantaneous frequencies and instantaneous amplitudes of the multiple key component signals;

[0009] A convolutional neural network is used to extract the characteristic quantities of the time series of the distribution network fault current signal and the characteristic quantities of the phase space reconstruction signal, the characteristic quantities of the time series of the distribution network fault current signal and the characteristic quantities of the phase space reconstruction signal are mixed, and the fault location is determined based on the mixed characteristic quantities.

[0010] In an embodiment of the present invention, the method of performing Hilbert-Huang transform on the distribution network fault current signal connected to the distributed energy to obtain the instantaneous frequency and instantaneous amplitude of multiple key component signals of the distribution network fault current signal includes: decomposing the distribution network fault current signal connected to the distributed energy into multiple different frequency components, and extracting multiple key component signals from the multiple different frequency components; performing Hilbert transform on the extracted multiple key component signals respectively to obtain the instantaneous frequency and instantaneous amplitude of the multiple key component signals.

[0011] In the embodiment of the present invention, the Hilbert transform is performed on the multiple key component signals respectively to obtain the instantaneous frequencies and instantaneous amplitudes of the multiple key component signals, including:

[0012] Performing Hilbert transform on each key component signal to obtain a complex signal, wherein the complex signal is composed of the key component signal and the Hilbert transform signal of the key component signal;

[0013] According to the key component signal and the Hilbert transform signal of the key component signal, the instantaneous amplitude and instantaneous phase of the key component signal are calculated, and according to the instantaneous phase of the key component signal, the instantaneous frequency of the key component signal is calculated.

[0014] In an embodiment of the present invention, the multi-dimensional signal phase space reconstruction based on the instantaneous frequencies and instantaneous amplitudes of multiple key component signals includes: performing three-dimensional signal phase space reconstruction based on the instantaneous frequency of each key component signal to obtain a three-dimensional space signal vector; performing three-dimensional signal phase space reconstruction based on the instantaneous amplitude of each key component signal to obtain a four-dimensional space signal vector.

[0015] In an embodiment of the present invention, the delay time of the signal phase space reconstruction is determined according to the cross-correlation function of the instantaneous frequency and instantaneous amplitude of multiple key component signals, including: based on the cross-correlation function of the instantaneous frequency and instantaneous amplitude of each key component signal, calculating the mutual information between the original key component signal and the delayed key component signal, and selecting the time point when the local minimum is first reached as the delay time of the phase space reconstructed signal.

[0016] In the embodiment of the present invention, the formula for calculating the mutual information between the original key component signal and the delayed key component signal is:

[0017]

[0018] Where I(τ) is the mutual information, x(t) is the original key component signal, x(t+τ) is the key component signal after a delay of τ, P(x(t)) is the probability distribution of the original key component signal x(t), P(x(t+τ)) is the probability distribution of the key component signal x(t+τ) after a delay of τ, and P(x(t), x(t+τ)) is the joint probability distribution of x(t) and x(t+τ).

[0019] In an embodiment of the present invention, the method of using a convolutional neural network to extract the characteristic quantities of the time series of the distribution network fault current signal and the characteristic quantities of the phase space reconstruction signal includes: using a one-dimensional convolutional neural network to mine the characteristic quantities of the time series of the distribution network fault current signal; and using a point voxel convolutional neural network to extract the characteristic quantities of the three-dimensional space signal vector and the characteristic quantities of the four-dimensional space signal vector.

[0020] In an embodiment of the present invention, the characteristic quantity of the time series of the distribution network fault current signal is mixed with the characteristic quantity of the phase space reconstruction signal, and the fault location is determined based on the mixed characteristic quantity, including: mixing the characteristic quantity of the time series of the distribution network fault current signal with the characteristic quantity of the three-dimensional space signal vector, performing fault line selection based on the mixed characteristic quantity to determine the line where the fault occurs; mixing the characteristic quantity of the time series of the distribution network fault current signal with the characteristic quantity of the four-dimensional space signal vector, and performing fault distance measurement based on the mixed characteristic quantity to determine the fault location.

[0021] Another aspect of the present invention provides a smart switch, comprising: a collection module and an edge computing module;

[0022] The edge computing module includes: a fault section positioning module and a fault ranging module;

[0023] The fault section positioning module is used to perform preliminary positioning of the fault section based on the distribution network fault current signal connected to the distributed energy source collected by the collection module;

[0024] The fault distance measurement module is used to perform fault distance measurement in the initially located fault section to locate the fault location, specifically including:

[0025] Performing Hilbert-Huang transform on the distribution network fault current signal connected to the distributed energy source to obtain the instantaneous frequency and instantaneous amplitude of multiple key component signals of the distribution network fault current signal;

[0026] Perform multi-dimensional signal phase space reconstruction based on the instantaneous frequencies and instantaneous amplitudes of multiple key component signals to obtain a phase space reconstructed signal; wherein the delay time of the signal phase space reconstruction is determined according to the cross-correlation function of the instantaneous frequencies and instantaneous amplitudes of the multiple key component signals;

[0027] A convolutional neural network is used to extract the characteristic quantities of the time series of the distribution network fault current signal and the characteristic quantities of the phase space reconstruction signal, the characteristic quantities of the time series of the distribution network fault current signal and the characteristic quantities of the phase space reconstruction signal are mixed, and the fault location is determined based on the mixed characteristic quantities.

[0028] In an embodiment of the present invention, signal phase space reconstruction is performed based on the instantaneous frequencies and instantaneous amplitudes of multiple key component signals, including: performing three-dimensional signal phase space reconstruction based on the instantaneous frequency of each key component signal to obtain a three-dimensional space signal vector; performing three-dimensional signal phase space reconstruction based on the instantaneous amplitude of each key component signal to obtain a four-dimensional space signal vector.

[0029] In the embodiment of the present invention, the fault distance measurement module is specifically used for:

[0030] Using a one-dimensional convolutional neural network to mine the characteristic quantity of the time series of the distribution network fault current signal;

[0031] Extracting the feature quantity of the three-dimensional space signal vector and the feature quantity of the four-dimensional space signal vector using a point voxel convolutional neural network;

[0032] Mixing the characteristic quantity of the time series of the distribution network fault current signal with the characteristic quantity of the three-dimensional space signal vector, and performing fault line selection according to the mixed characteristic quantity to determine the line where the fault occurs;

[0033] The characteristic quantity of the time series of the distribution network fault current signal is mixed with the characteristic quantity of the four-dimensional space signal vector, and fault distance measurement is performed according to the mixed characteristic quantity to determine the fault location.

[0034] In the embodiment of the present invention, the smart switch further comprises: a strategy library module and a control module;

[0035] The strategy library module is used to receive the fault location and fault information sent by the edge computing module, and search for a recovery strategy corresponding to the current fault in the control strategy library of the strategy library module;

[0036] The control module is used to receive the fault section location information sent by the fault section location module and the recovery strategy and fault location sent by the strategy library module, perform isolation operations according to the fault section location information, and restore the line at the fault location according to the recovery strategy.

[0037] In an embodiment of the present invention, the smart switch further comprises: a remote communication module;

[0038] The remote communication module is used to communicate with the remote master station, and when no recovery strategy corresponding to the current fault is searched in the strategy library module, the fault location and fault information are uploaded to the remote master station, the recovery strategy issued by the remote master station is received, and the recovery strategy issued by the remote master station is sent to the strategy library module;

[0039] The policy library module is also used to send the recovery policy issued by the remote master station to the control module;

[0040] The control module is also used to isolate the fault section corresponding to the fault location according to the recovery strategy sent by the remote master station.

[0041] In the embodiment of the present invention, the smart switch further comprises: a fault recording module;

[0042] The fault recording module is used to record fault information, fault location and fault handling method, and update them to the strategy library module;

[0043] The remote communication module is also used to upload the fault information, fault location and fault handling method updated by the policy library module to the remote master station to update the fault recovery strategy stored in the remote master station.

[0044] In the embodiment of the present invention, the smart switch further comprises: a self-checking and self-recovering module, and the self-checking and self-recovering module is used to perform fault self-checking and self-recovery on the smart switch.

[0045] The present invention firstly utilizes the Hilbert-Huang transform technology to fully explore the fault current signal and its time-frequency characteristics in the medium-voltage distribution network containing distributed resources, then uses the phase space reconstruction method to capture the inherent law of the fault signal, and further realizes the fault line selection and accurate fault location of the medium-voltage distribution network containing distributed resources through the convolutional neural network. The present invention also proposes a smart switch embedded with a fault location function module, which realizes the automatic recovery and grid connection of the fault line through the smart switch installed in the line, and can realize fault recovery without the need for other equipment, thus saving costs.

[0046] Other features and advantages of the technical solution of the present invention will be described in detail in the specific implementation section below. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0048] Figure 1 is a flow chart of a method for locating a medium voltage distribution network line fault provided by an embodiment of the present invention;

[0049] Figure 2 It is an architecture diagram of fault line selection and fault distance measurement based on convolutional neural network provided by an embodiment of the present invention;

[0050] Figure 3 is a block diagram of a smart switch provided by an embodiment of the present invention;

[0051] Figure 4 The present invention provides a flow chart of a distribution network line fault recovery process. DETAILED DESCRIPTION

[0052] In order to make the technical solutions and advantages of the embodiments of the present invention more clearly understood, the exemplary embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than an exhaustive list of all the embodiments. It should be noted that the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.

[0053] As introduced in the background technology, distribution network fault location is mainly divided into three categories: fault line selection, fault distance measurement, and section location. Fault line selection is to locate the fault line according to the reported fault information, quickly isolate the fault point, and restore power supply to the non-fault area. Fault distance measurement is mainly based on the relationship between various electrical quantities when the fault occurs to determine the location of the fault. The existing distribution network fault location methods do not consider the situation when the distribution network contains highly permeated distributed resources, and cannot achieve the precise location of large-scale distributed resources connected to the distribution network line fault. In the early days, the distribution network fault location relied on the fault indicator installed on the distribution network and the branch line. If the color of the fault indicator changes, it can be judged that a fault has occurred in the downstream area. With the continuous maturity of distribution network automation, distribution switch monitoring terminals (FTU) are installed in large numbers at feeder circuit breakers and automation switches. The distribution switch monitoring terminal and the fault indicator need to cooperate to achieve fault location, resulting in high installation costs.

[0054] In response to the above problems, the implementation of the present invention proposes a fault location method for a medium-voltage distribution network containing distributed energy. First, the Hilbert-Huang transform technology is used to fully explore the fault current signal and its time-frequency characteristics in the medium-voltage distribution network containing distributed resources. Secondly, the phase space reconstruction method is used to capture the inherent law of the fault signal, and then the convolutional neural network is further used to achieve the fault line selection and accurate location of the fault in the medium-voltage distribution network containing distributed resources. The implementation of the present invention also proposes a smart switch embedded with a fault location function module. The smart switch installed in the line realizes the automatic recovery and grid connection of the fault line, and the fault recovery can be achieved without the need for other equipment, saving costs.

[0055] Figure 1 FIG. 1 is a flow chart of a method for locating a medium voltage distribution network line fault provided by an embodiment of the present invention. Figure 1 As shown, the method for locating a medium voltage distribution network line fault provided in this embodiment includes the following steps:

[0056] S100, performing Hilbert-Huang transform on the distribution network fault current signal connected to the distributed energy source to obtain the instantaneous frequency and instantaneous amplitude of multiple key component signals of the distribution network fault current signal;

[0057] S200, performing multi-dimensional signal phase space reconstruction based on instantaneous frequencies and instantaneous amplitudes of multiple key component signals to obtain a phase space reconstructed signal;

[0058] S300, using a convolutional neural network to extract the characteristic quantities of the time series of the distribution network fault current signal and the characteristic quantities of the phase space reconstruction signal, mixing the characteristic quantities of the time series of the distribution network fault current signal with the characteristic quantities of the phase space reconstruction signal, and determining the fault location based on the mixed characteristic quantities.

[0059] Before the above step S100, for the distribution network connected to the distributed energy, the distribution network fault current signals under different topologies, different fault types, different lines, and different noise interference conditions are collected, the sampling frequency is 10kHz, and the distribution network fault current signals of two cycles after the fault are used as the original fault sample signals for fault line selection.

[0060] Hilbert-Huang Transhorm (HHT) is a time-frequency analysis method suitable for non-stationary and nonlinear signals, which mainly includes two steps: empirical mode decomposition (EMD) and Hilbert Spectral Analysis (HSA). Compared with traditional Fourier transform or wavelet transform, Hilbert-Huang transform has better adaptability and higher accuracy in processing complex nonlinear signals.

[0061] When a fault occurs in the traditional distribution network, such as a single-phase ground fault, a phase-to-phase short circuit fault, a two-phase ground fault, etc., different fault types will generate different fault currents in the distribution line. However, after a high proportion of new energy is added to the distribution network, distributed power sources with different capacities, different states, and different grid-connected locations have an unknown impact on the waveform characteristics of the traditional fault current. Therefore, it is necessary to first extract the key component signals from the original fault current signal through empirical mode decomposition (EMD) of the distribution network fault current containing distributed energy, and then perform Hilbert spectrum analysis on each key component signal.

[0062] In the above step S100, the distribution network fault current signal (i.e., the original fault sample signal) connected to the distributed energy is first decomposed into multiple different frequency components (IMF) and a residual component by the empirical mode decomposition (EMD) algorithm, and each IMF component represents a different frequency component of the original fault sample signal. Then, multiple key component signals are extracted from the multiple different frequency components, and the extracted multiple key component signals are Hilbert transformed respectively to obtain the instantaneous frequency and instantaneous amplitude of the multiple key component signals. Among them, each key component signal is Hilbert transformed to obtain a complex signal, which is composed of the key component signal and the Hilbert transform signal of the key component signal; according to the key component signal and the Hilbert transform signal of the key component signal, the instantaneous amplitude and instantaneous phase of the key component signal are calculated, and the instantaneous frequency of the key component signal is calculated according to the instantaneous phase of the key component signal.

[0063] In a specific example, the EMD algorithm process is as follows:

[0064] Step 1: Assume that the original signal of the fault current received by the smart switch is x(t), use the spline interpolation method to fit the maximum and minimum points of the original signal to form the upper and lower envelopes, and calculate the average envelope x o (t), and then subtract the mean from the original signal to get a new signal h1(t), expressed as: h1(t) = x(t)-x o (t);

[0065] Step 2: determine whether the signal h1(t) satisfies the characteristic condition of the intrinsic mode function (IMF). If so, the signal h1(t) is the first IMF component IMF1. If not, continue to loop step 1 until h1(t) satisfies the characteristic condition of the intrinsic mode function (IMF).

[0066] Step 3, separate the IMF1 component in the original signal x(t) to obtain a new signal r1(t), the expression of the signal r1(t) is: r1(t) = x(t) - IMF1, and repeat until the second IMF component IMF2 is obtained;

[0067] Step 4: Repeat the above steps until the Kth IMF component is obtained. Finally, the result of the empirical mode decomposition can be expressed as:

[0068]

[0069] Among them, r k is the remainder, representing the central trend of the signal.

[0070] Finally, the original signal is decomposed into a series of IMF components and a residual component, and each IMF component represents a different frequency component of the signal. For example, after the original distribution network fault current signal is processed by EMD decomposition, the intrinsic mode function of the original distribution network fault current signal is obtained, which is the time-frequency data matrix of the original signal, including the time-frequency characteristics of the original signal.

[0071] In a specific example, during Hilbert spectrum analysis (HSA), a complex signal z(t) is obtained after Hilbert transforming the IMF component signal x(t):

[0072]

[0073] in, is the Hilbert transform of x(t);

[0074] The calculation expression of the instantaneous amplitude A(t) of the component signal x(t) is:

[0075]

[0076] The instantaneous phase of the component signal x(t) The calculation expression is:

[0077]

[0078] The calculation expression of the instantaneous frequency f(t) of the component signal x(t) is:

[0079]

[0080] In the above step S200, a multi-dimensional signal phase space reconstruction is performed based on the instantaneous frequency and instantaneous amplitude of each key component (IMF) signal. For the signal phase space reconstruction, it is necessary to determine the reconstruction dimension and delay time. The reconstruction dimension determines the dimension of the phase space, and an automatic information dimension algorithm can be used to select a suitable dimension to capture the dynamic characteristics of the signal. The delay time determines how to extract the value of the time series from the signal. The selection of the delay time is very important. If the delay is too small, the trajectory of the phase space reconstruction may be redundant; if the delay is too large, the dynamic characteristics of the signal may not be accurately represented. In this embodiment, the delay time of the signal phase space reconstruction is determined according to the cross-correlation function of the instantaneous frequency and instantaneous amplitude of multiple key component signals. According to the dimension of the signal phase space reconstruction and the delay time of the phase space reconstruction signal, the delay embedding method is used to convert multiple key component signals into trajectories in the high-dimensional phase space.

[0081] Since the instantaneous frequency can reflect the frequency change caused by the fault event (such as short circuit, ground fault). The present invention extracts the instantaneous frequency of the intrinsic mode component through the Hilbert-Huang transform, constructs the three-dimensional space signal vector of the instantaneous frequency of each IMF component, and constructs the four-dimensional space signal vector of the energy of the IMF component signal. For fault line selection, the cross-correlation function (CCF) of the instantaneous frequency and instantaneous amplitude of the IMF component is calculated, and for fault distance measurement, the autocorrelation function (ACF) of the energy of the IMF component is calculated to determine the delay time.

[0082] In one embodiment, three-dimensional images of the instantaneous frequency and instantaneous amplitude of the IMF component signal are respectively constructed for the fault line selection process, and a four-dimensional image of the energy of the IMF component signal is constructed. Specifically, a three-dimensional space signal vector is constructed based on the instantaneous frequency of each key component signal, and a three-dimensional space signal vector is constructed based on the instantaneous amplitude of each key component signal; at the same time, a four-dimensional space signal vector is constructed based on the instantaneous amplitude (energy) of each key component signal (the signal amplitude is used as the fourth dimension signal parameter). Based on the cross-correlation function of the instantaneous frequency and instantaneous amplitude of each key component signal, the mutual information of the original key component signal and the delayed key component signal is calculated, and the time point at which the local minimum is first reached is selected as the delay time of the phase space reconstruction signal.

[0083] Assume that the instantaneous frequency signal of the fault current signal after Hilbert-Huang transform (HHT) is f(t), and the process of signal phase space reconstruction is:

[0084] F(t)=[f(t), f(t+τ), f(t+2τ),..., f(t+(m-1)τ)];

[0085] Wherein, F(t) is the reconstructed vector, which contains the delayed coordinates of the signal; m is the embedding dimension, and when the reconstructed signal is three-dimensional, m=2; τ is the delay time.

[0086] At the moment of fault, the change of amplitude is usually significant. Capturing the change of signal strength over time helps to identify the peak value in the current. The instantaneous amplitude of each IMF is calculated by Hilbert transform, and the three-dimensional spatial signal of the instantaneous amplitude of each IMF is reconstructed. Assuming that the instantaneous amplitude of the fault current signal after HHT transformation is A(t), the process of signal phase space reconstruction is:

[0087] A'(t)=[A(t), A(t+τ), A(t+2τ),...,A(t+(m-1)τ)];

[0088] Among them, A'(t) is the reconstructed vector, m is the embedding dimension, and τ is the delay time.

[0089] The delay time τ is selected as follows:

[0090] Calculate the mutual information between the key component signal and the delayed key component signal:

[0091]

[0092] Where I(τ) is the mutual information, x(t) is the original key component signal, x(t+τ) is the key component signal after a delay of τ, P(x(t)) is the probability distribution of the original key component signal x(t), P(x(t+τ)) is the probability distribution of the key component signal x(t+τ) after a delay of τ, and P(x(t), x(t+τ)) is the joint probability distribution of x(t) and x(t+τ).

[0093] Calculate the values ​​of I(τ1), I(τ2), I(τ3), etc. corresponding to P(x(t)) and P(x(t+2τ)), P(x(t+2τ)) and P(x(t+3τ))… respectively, where the delay time τ corresponding to the minimum value I(τ) is the time point when the local minimum value is first reached.

[0094] In the above step S300, a convolutional neural network is used to extract the time series feature of the distribution network fault current signal and the feature of the phase space reconstruction signal. Figure 2 , the one-dimensional convolutional neural network (1D-CNN) is used to mine the time series feature of the distribution network fault current signal, and the point voxel convolutional neural network (PVCNN) is used to extract the feature of the three-dimensional space signal vector and the feature of the four-dimensional space signal vector. The feature of the time series of the distribution network fault current signal is mixed with the feature of the three-dimensional space signal vector, and the fault line selection is performed based on the mixed feature to determine the faulty line; the feature of the time series of the distribution network fault current signal is mixed with the feature of the four-dimensional space signal vector, and the fault distance is performed based on the mixed feature to determine the fault location.

[0095] In a specific embodiment, the one-dimensional convolutional neural network (1D-CNN) includes: a one-dimensional convolutional layer, a normalization layer, an activation function layer, an exhibit layer, and a fully connected layer. The process of using the point voxel convolutional neural network (PVCNN) to extract three-dimensional features includes: inputting a three-dimensional spatial signal vector into the PVCNN network, the data size is n×3, n represents n points, and each point is represented by a three-dimensional coordinate (x, y, z); the feature extraction layer (PV Conv Layers) extracts local features of the data, and each layer extracts the local features of the point through the Pointwise Convolution (PV Conv) in PointNet. The feature extraction layer (PV ConvLayers) forms a high-dimensional feature space by increasing the input dimension. The first layer: n×3→n×64, extracts preliminary features through Pointwise Convolution of PointNet; the second layer: n×64→n×128, extracts higher-dimensional features; the third layer: n×128→n×128, continues to enrich the local features of the points; the fourth layer: n×128→n×512, extracts high-order point features. Then, the fully connected layer (FC) is used to map the point features: n×512→n×2048, and the features of all points are aggregated into a global feature: 1×2048 through the Max Pooling operation. The global feature can represent the semantic information of the entire point cloud.

[0096] For fault line selection, the three-dimensional features extracted by the one-dimensional convolutional neural network (1D-CNN) and the point voxel convolutional neural network (PVCNN) are mixed, and the mixed features are used as the input of the classifier Sigmoid, and the probability of the faulty line is output in the form of probability to realize the selection of the faulty line.

[0097] For fault location, by calculating the energy of the IMF component (for example, by calculating the square amplitude), the energy distribution of each frequency band in the signal can be understood. When a fault occurs, the energy may be concentrated in certain frequency bands, and the characteristics of the fault mode can be identified through energy analysis. By analyzing the change in instantaneous amplitude or energy, the time and duration of the fault can be determined, which can be used for fault line location.

[0098] Assume that the energy of the fault current after HHT transformation is E(t), and the signal reconstruction process is: E'(t) = [E(t), E(t+τ), E(t+2τ), ..., E(t+(m-1)τ)], constructing the four-dimensional spatial signal of the signal energy, and selecting the energy signal amplitude as the fourth-dimensional signal parameter.

[0099] Extraction of four-dimensional features: The four-dimensional signal features can be expressed as n×(x,y,z,e), where n represents the number of signal sampling points; x,y,z represent the spatial coordinates of the first three dimensions; e is the fourth dimension, representing the signal energy. The energy e in n×4 is directly used as one of the initial input dimensions of the convolution or fully connected layer. Feature fusion: Energy information and geometric information (x,y,z) are jointly processed to calculate the joint features of the energy distribution. The shape of the input data is n×4, and each point contains four-dimensional information. Local feature extraction (PV Conv Layers) is performed on the data. Each layer extracts the local features of the points through Pointwise Convolution (PVConv) in PointNet. The feature extraction layer (PV Conv Layers) forms a high-dimensional feature space by increasing the input dimension. The first layer: n×4→n×64, extracts preliminary features through Pointwise Convolution of PointNet; the second layer: n×64→n×128, extracts higher-dimensional features; the third layer: n×128→n×128, continues to enrich the local features of the points; the fourth layer: n×128→n×512, extracts high-order point features. The fully connected layer (FC) is used to map the point features: n×512→n×2048, and the features of all points are aggregated into a global feature: 1×2048 through the Max Pooling operation.

[0100] The four-dimensional features extracted by the one-dimensional convolutional neural network (1D-CNN) and the point voxel convolutional neural network (PVCNN) are mixed. The mixed features are mapped from the features to the distance of the fault point through CNN. The mapping result is used as the input of the classifier Sigmoid. Its accuracy is calculated and the distance of the fault position with the corresponding probability is mapped to the distance of the smart switch. The distance with the highest accuracy is output separately. In order to ensure that no possible fault location information is missed, the fault point location with an accuracy greater than a certain threshold is output as a reference to achieve fault distance measurement.

[0101] An embodiment of the present invention also provides a smart switch, which has off-grid and edge computing functions, as well as a fault detection function, and can perform preliminary section positioning of medium-voltage line faults connected to the smart switch, and based on this, realize fault line selection and accurate fault distance measurement.

[0102] like Figure 3 As shown, the smart switch includes: a collection module, an edge computing module, a policy library module, a control module and a remote communication module.

[0103] The edge computing module includes: a fault section location module and a fault distance measurement module. The fault section location module is used to perform preliminary location of the fault section based on the fault current signal of the distribution network connected to the distributed energy source collected by the acquisition module; the fault distance measurement module is used to perform fault distance measurement within the initially located fault section to locate the location of the fault.

[0104] The strategy library module is used to receive the fault location and fault information sent by the edge computing module, and search for the recovery strategy corresponding to the current fault in the control policy library of the strategy library module; the control policy library stores the historical control strategies issued by the remote master station, and when a fault occurs, the control policy library is first searched for a recovery strategy matching the current fault.

[0105] The control module is used to receive the fault section location information sent by the fault section location module and the recovery strategy and fault location sent by the strategy library module, perform isolation operations (opening and closing operations of the fault section) according to the fault section location information, and restore the line at the fault location according to the recovery strategy.

[0106] The remote communication module is used to communicate with a remote master station (such as a remote intelligent fusion terminal). When the recovery strategy corresponding to the current fault is not found in the policy library module, the fault location and fault information are uploaded to the remote master station, the recovery strategy issued by the remote master station is received, and the recovery strategy issued by the remote master station is sent to the policy library module.

[0107] The strategy library module is also used to send the recovery strategy sent by the remote master station to the control module. The control module performs an isolation operation on the fault section corresponding to the fault location according to the recovery strategy sent by the remote master station.

[0108] The edge computing module of the smart switch of the embodiment of the present invention adopts the medium-voltage distribution network line fault location method provided in the above embodiment to perform fault line selection and fault distance measurement. Specifically, the fault distance measurement module performs Hilbert-Huang transform on the distribution network fault current signal connected to the distributed energy source to obtain the instantaneous frequency and instantaneous amplitude of multiple key component signals of the distribution network fault current signal; performs multi-dimensional signal phase space reconstruction based on the instantaneous frequency and instantaneous amplitude of multiple key component signals to obtain a phase space reconstructed signal; uses a convolutional neural network to extract the characteristic quantity of the time series of the distribution network fault current signal and the characteristic quantity of the phase space reconstructed signal, mixes the characteristic quantity of the time series of the distribution network fault current signal with the characteristic quantity of the phase space reconstructed signal, and determines the location of the fault according to the mixed characteristic quantity.

[0109] In a specific embodiment, the fault distance measurement module is preset with a Hilbert-Huang transform unit, a signal reconstruction unit, and a convolutional neural network model. First, the connected distribution network fault current signal is decomposed into multiple different frequency components by the Hilbert-Huang transform unit, and multiple key component signals are extracted from the multiple different frequency components. The extracted multiple key component signals are subjected to Hilbert transform respectively to obtain the instantaneous frequency and instantaneous amplitude of the multiple key component signals. Then, the three-dimensional signal phase space is reconstructed based on the instantaneous frequency of each key component signal by the signal reconstruction unit to obtain a three-dimensional space signal vector, and the three-dimensional signal phase space is reconstructed based on the instantaneous amplitude of each key component signal to obtain a four-dimensional space signal vector; according to the dimension of the signal phase space reconstruction and the delay time of the phase space reconstruction signal, the delay embedding method is used to convert the multiple key component signals into trajectories in the high-dimensional phase space.

[0110] The delay time of signal phase space reconstruction can be determined based on the cross-correlation function of the instantaneous frequency and instantaneous amplitude of multiple key component signals. Specifically, the mutual information between the key component signal and the delayed key component signal is calculated:

[0111]

[0112] Where I(τ) is the mutual information, x(t) is the original key component signal, x(t+τ) is the key component signal after a delay of τ, P(x(t)) is the probability distribution of the original key component signal x(t), P(x(t+τ)) is the probability distribution of the key component signal x(t+τ) after a delay of τ, and P(x(t), x(t+τ)) is the joint probability distribution of x(t) and x(t+τ).

[0113] Calculate the values ​​of I(τ1), I(τ2), I(τ3), etc. corresponding to P(x(t)) and P(x(t+2τ)), P(x(t+2τ)) and P(x(t+3τ))… respectively, where the delay time τ corresponding to the minimum value I(τ) is the time point when the local minimum value is first reached.

[0114] In a specific embodiment, the fault distance measurement module uses a convolutional neural network model to extract the characteristic quantities of the time series of the distribution network fault current signal and the characteristic quantities of the phase space reconstruction signal. Specifically, a one-dimensional convolutional neural network is used to mine the characteristic quantities of the time series of the distribution network fault current signal, and a point voxel convolutional neural network is used to extract the characteristic quantities of the three-dimensional space signal vector and the characteristic quantities of the four-dimensional space signal vector. Then, the characteristic quantities of the time series of the distribution network fault current signal are mixed with the characteristic quantities of the three-dimensional space signal vector, and the fault line is selected based on the mixed characteristic quantities to determine the line where the fault occurs; the characteristic quantities of the time series of the distribution network fault current signal are mixed with the characteristic quantities of the four-dimensional space signal vector, and the fault distance is measured based on the mixed characteristic quantities to determine the location of the fault.

[0115] In one embodiment, the smart switch further includes a fault recording module. The fault recording module is used to record fault information, fault location, and fault handling method, and package and update them together to the policy library module. The remote communication module then uploads the fault information, fault location, and fault handling method updated by the policy library module to the remote master station to update the fault recovery strategy stored in the remote master station.

[0116] In one embodiment, the smart switch further comprises: a self-checking and self-recovering module and a power supply module. The self-checking and self-recovering module is used to detect and handle faults of the smart switch, and realize self-checking and self-recovering of faults. The power supply module is used to supply power to the smart switch as a whole.

[0117] The present invention is directly applied to smart switches installed in the lines. It does not need to understand the distribution network topology structure and only focuses on the line information. First, the Hilbert-Huang transform technology is used to fully explore the fault current signal and its time-frequency characteristics in the medium-voltage distribution network containing distributed resources. Secondly, the phase space reconstruction method is used to capture the inherent law of the fault signal. Then, the convolutional neural network is used to further realize the fault line selection and accurate positioning of the fault in the medium-voltage distribution network containing distributed resources.

[0118] The smart switch of the embodiment of the present invention has off-grid and edge computing functions, as well as fault detection functions, which reduces the cost of matching feeder equipment. The smart switch can accurately locate medium-voltage distribution network faults of high-penetration distributed energy, and ensures the accuracy and efficiency of positioning through intelligent algorithms (specifically refer to the above-mentioned medium-voltage distribution network line fault location method). It can also realize automatic fault recovery and grid connection of faulty lines through edge-end collaboration, without the need to use other equipment, saving costs and reducing losses caused by line faults in the power grid.

[0119] Reference Figure 4 , the fault recovery process of the medium-voltage line based on the above-mentioned smart switch to achieve edge-end coordination is as follows:

[0120] Initial stage: The acquisition module sends the collected fault current signal of the distribution network connected to the distributed energy to the edge computing module, and the edge computing module extracts the line information characteristics;

[0121] Fault section location: The fault section location module performs preliminary location of the fault section according to the distribution network fault current signal, and the control module isolates the fault section according to the fault section location result;

[0122] Processing of the edge computing module: The fault distance measurement module performs fault distance measurement in the initially located fault section, locates the fault location (i.e., the fault line distance), and sends the fault location and fault information (fault type, etc.) to the strategy library module;

[0123] Execution of fault recovery strategy: The fault distance measurement module uses reinforcement learning to measure the fault line based on the fault section; the control policy library of the policy library module searches for a recovery strategy corresponding to the current fault (checks whether there is a corresponding strategy for the reported fault): If there is a recovery strategy corresponding to the current fault in the control policy library, the recovery strategy is sent to the control module, and the control module restores the line at the fault location according to the recovery strategy; If there is no recovery strategy corresponding to the current fault in the control policy library, it is uploaded to the remote master station to trigger the remote master station to calculate the recovery strategy and issue the recovery strategy. The control module implements automatic recovery of the fault line according to the strategy issued by the policy library.

[0124] Check whether the fault is recovered: If the fault is recovered successfully, record the fault information and recovery strategy, and update the recovery strategy in the strategy library; if the fault is not recovered, check whether the request time is greater than the set threshold X S seconds (the selection of the threshold time is determined according to actual needs); if so, request manual intervention for recovery; if not, request a new recovery strategy from the terminal.

[0125] Trigger-end recovery strategy: The remote master station sends the recovery strategy to the strategy library module's strategy library, and the control module performs automatic recovery according to the strategy sent by the strategy library.

[0126] The embodiment of the present invention further provides a computer device, including: a memory and a processor, the memory stores a computer program, and the processor is used to execute the computer program to implement the above-mentioned medium voltage distribution network line fault locating method.

[0127] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The schemes in the embodiments of the present invention may be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.

[0128] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0129] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0130] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0131] Although preferred embodiments of the present invention have been described, additional changes and modifications may be made to these embodiments by those skilled in the art once the basic inventive concepts are known. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention. Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for locating a medium voltage distribution network line fault, characterized in that: include: Performing Hilbert-Huang transform on the distribution network fault current signal connected to the distributed energy source to obtain the instantaneous frequency and instantaneous amplitude of multiple key component signals of the distribution network fault current signal; Perform multi-dimensional signal phase space reconstruction based on the instantaneous frequencies and instantaneous amplitudes of multiple key component signals to obtain a phase space reconstructed signal; wherein the delay time of the signal phase space reconstruction is determined according to the cross-correlation function of the instantaneous frequencies and instantaneous amplitudes of the multiple key component signals; A convolutional neural network is used to extract the characteristic quantities of the time series of the distribution network fault current signal and the characteristic quantities of the phase space reconstruction signal, the characteristic quantities of the time series of the distribution network fault current signal and the characteristic quantities of the phase space reconstruction signal are mixed, and the fault location is determined based on the mixed characteristic quantities.

2. The method for locating a medium voltage distribution network line fault according to claim 1, characterized in that: The Hilbert-Huang transform is performed on the distribution network fault current signal connected to the distributed energy to obtain the instantaneous frequency and instantaneous amplitude of multiple key component signals of the distribution network fault current signal, including: Decomposing the fault current signal of the distribution network connected to the distributed energy into multiple different frequency components, and extracting multiple key component signals from the multiple different frequency components; The extracted multiple key component signals are respectively subjected to Hilbert transform to obtain instantaneous frequencies and instantaneous amplitudes of the multiple key component signals.

3. The method for locating a medium voltage distribution network line fault according to claim 2, characterized in that: The Hilbert transform is performed on the multiple key component signals respectively to obtain the instantaneous frequencies and instantaneous amplitudes of the multiple key component signals, including: Performing Hilbert transform on each key component signal to obtain a complex signal, wherein the complex signal is composed of the key component signal and the Hilbert transform signal of the key component signal; According to the key component signal and the Hilbert transform signal of the key component signal, the instantaneous amplitude and instantaneous phase of the key component signal are calculated, and according to the instantaneous phase of the key component signal, the instantaneous frequency of the key component signal is calculated.

4. The method for locating a fault in a medium voltage distribution network according to claim 1, characterized in that: The multi-dimensional signal phase space reconstruction based on the instantaneous frequencies and instantaneous amplitudes of the multiple key component signals includes: Based on the instantaneous frequency of each key component signal, a three-dimensional signal phase space is reconstructed to obtain a three-dimensional space signal vector; The three-dimensional signal phase space is reconstructed based on the instantaneous amplitude of each key component signal to obtain a four-dimensional space signal vector.

5. The method for locating a fault in a medium voltage distribution network according to claim 4, characterized in that: The delay time of the signal phase space reconstruction is determined according to the cross-correlation function of the instantaneous frequency and instantaneous amplitude of multiple key component signals, including: Based on the cross-correlation function of the instantaneous frequency and instantaneous amplitude of each key component signal, the mutual information between the original key component signal and the delayed key component signal is calculated, and the time point when the local minimum is first reached is selected as the delay time of the phase space reconstructed signal.

6. The method for locating a fault in a medium voltage distribution network according to claim 5, characterized in that: The formula for calculating the mutual information between the original key component signal and the delayed key component signal is: Where I(τ) is the mutual information, x(t) is the original key component signal, x(t+τ) is the key component signal after a delay of τ, P(x(t)) is the probability distribution of the original key component signal x(t), P(x(t+τ)) is the probability distribution of the key component signal x(t+τ) after a delay of τ, and P(x(t), x(t+τ)) is the joint probability distribution of x(t) and x(t+τ).

7. The method for locating a fault in a medium voltage distribution network according to claim 4, characterized in that: The method of extracting the characteristic quantity of the time series of the distribution network fault current signal and the characteristic quantity of the phase space reconstruction signal by using a convolutional neural network includes: Using a one-dimensional convolutional neural network to mine the characteristic quantity of the time series of the distribution network fault current signal; A point voxel convolutional neural network is used to extract the feature quantity of the three-dimensional space signal vector and the feature quantity of the four-dimensional space signal vector.

8. The method for locating a fault in a medium voltage distribution network according to claim 7, characterized in that: Mixing the characteristic quantity of the time series of the distribution network fault current signal with the characteristic quantity of the phase space reconstruction signal, and determining the fault occurrence location according to the mixed characteristic quantity, including: Mixing the characteristic quantity of the time series of the distribution network fault current signal with the characteristic quantity of the three-dimensional space signal vector, and performing fault line selection according to the mixed characteristic quantity to determine the line where the fault occurs; The characteristic quantity of the time series of the distribution network fault current signal is mixed with the characteristic quantity of the four-dimensional space signal vector, and fault distance measurement is performed according to the mixed characteristic quantity to determine the fault location.

9. A smart switch, characterized in that: include: Acquisition module and edge computing module; The edge computing module includes: a fault section positioning module and a fault ranging module; The fault section positioning module is used to perform preliminary positioning of the fault section based on the distribution network fault current signal connected to the distributed energy source collected by the collection module; The fault distance measurement module is used to perform fault distance measurement in the initially located fault section to locate the fault location, specifically including: Performing Hilbert-Huang transform on the distribution network fault current signal connected to the distributed energy source to obtain the instantaneous frequency and instantaneous amplitude of multiple key component signals of the distribution network fault current signal; Perform multi-dimensional signal phase space reconstruction based on the instantaneous frequencies and instantaneous amplitudes of multiple key component signals to obtain a phase space reconstructed signal; wherein the delay time of the signal phase space reconstruction is determined according to the cross-correlation function of the instantaneous frequencies and instantaneous amplitudes of the multiple key component signals; A convolutional neural network is used to extract the characteristic quantities of the time series of the distribution network fault current signal and the characteristic quantities of the phase space reconstruction signal, the characteristic quantities of the time series of the distribution network fault current signal and the characteristic quantities of the phase space reconstruction signal are mixed, and the fault location is determined based on the mixed characteristic quantities.

10. The smart switch according to claim 9, characterized in that: The signal phase space reconstruction based on the instantaneous frequencies and instantaneous amplitudes of the multiple key component signals comprises: Based on the instantaneous frequency of each key component signal, a three-dimensional signal phase space is reconstructed to obtain a three-dimensional space signal vector; The three-dimensional signal phase space is reconstructed based on the instantaneous amplitude of each key component signal to obtain a four-dimensional space signal vector.

11. The smart switch according to claim 10, characterized in that: The fault distance measurement module is specifically used for: Using a one-dimensional convolutional neural network to mine the characteristic quantity of the time series of the distribution network fault current signal; Extracting the feature quantity of the three-dimensional space signal vector and the feature quantity of the four-dimensional space signal vector using a point voxel convolutional neural network; Mixing the characteristic quantity of the time series of the distribution network fault current signal with the characteristic quantity of the three-dimensional space signal vector, and performing fault line selection according to the mixed characteristic quantity to determine the line where the fault occurs; The characteristic quantity of the time series of the distribution network fault current signal is mixed with the characteristic quantity of the four-dimensional space signal vector, and fault distance measurement is performed according to the mixed characteristic quantity to determine the fault location.

12. The smart switch according to claim 9, characterized in that: Also includes: Strategy library module and control module; The strategy library module is used to receive the fault location and fault information sent by the edge computing module, and search for a recovery strategy corresponding to the current fault in the control strategy library of the strategy library module; The control module is used to receive the fault section location information sent by the fault section location module and the recovery strategy and fault location sent by the strategy library module, perform isolation operations according to the fault section location information, and restore the line at the fault location according to the recovery strategy.

13. The smart switch according to claim 12, characterized in that: Also includes: Telecommunications module; The remote communication module is used to communicate with the remote master station, and when no recovery strategy corresponding to the current fault is searched in the strategy library module, the fault location and fault information are uploaded to the remote master station, the recovery strategy issued by the remote master station is received, and the recovery strategy issued by the remote master station is sent to the strategy library module; The policy library module is also used to send the recovery policy issued by the remote master station to the control module; The control module is also used to isolate the fault section corresponding to the fault location according to the recovery strategy sent by the remote master station.

14. The smart switch according to claim 13, characterized in that: Also includes: Fault recording module; The fault recording module is used to record fault information, fault location and fault handling method, and update them to the strategy library module; The remote communication module is also used to upload the fault information, fault location and fault handling method updated by the policy library module to the remote master station to update the fault recovery strategy stored in the remote master station.

15. The smart switch according to claim 14, characterized in that: Also includes: The self-checking and self-recovering module is used to perform fault self-checking and self-recovery on the smart switch.

16. A computer device, characterized in that: include: a memory storing a computer program; A processor is used to execute the computer program to implement the medium voltage distribution network line fault locating method as described in any one of claims 1-8.

Images