Leakage protection method suitable for low-voltage alternating current and direct current hybrid power distribution network and related equipment

Abstract

The invention belongs to the field of low-voltage power distribution networks, and discloses a leakage protection method and related equipment suitable for a low-voltage AC / DC hybrid power distribution network, and the method comprises the steps: obtaining residual current signals of the low-voltage power distribution network containing distributed photovoltaic access in different scenes; judging whether an electric leakage fault occurs or not based on the residual current signal, and after judging that the electric leakage fault occurs, calculating to obtain modal components of different frequency bands of the residual current signal; based on the modal components of different frequency bands of the residual current signal, electric leakage fault characteristics of the residual current signal during electric leakage fault are calculated and obtained; performing electric leakage scene identification on the fault features by using a pre-constructed electric leakage fault identification model to obtain an electric leakage fault type; and outputting a leakage protection execution instruction in combination with the obtained leakage fault type and a protection criterion corresponding to the leakage fault type, so as to execute a leakage protection action according to the leakage protection execution instruction. By adopting the method, the type and the source of the residual current can be effectively identified, and the safety performance is improved.

Description

Technical Field

[0001] The present invention belongs to the field of low-voltage distribution networks, specifically to the technical field of leakage protection actions of low-voltage AC / DC hybrid distribution networks, and in particular to a leakage protection method and related equipment suitable for low-voltage AC / DC hybrid distribution networks. Background Art

[0002] Traditional residual current devices (RCDs) use the residual current exceeding the threshold as the basis for determining whether the protection is activated. When the detection link detects that the residual current reaches or exceeds the threshold, the protection module can automatically disconnect the mechanical switch or combination of electrical appliances in the circuit to achieve fault isolation. RCDs are the most basic and important means of protection against leakage faults in low-voltage distribution networks. They can effectively avoid some personal and equipment safety accidents, significantly improving electricity safety.

[0003] At present, with the continuous transformation and upgrading of low-voltage distribution networks and the large-scale access of distributed photovoltaics, the causes of residual currents have become increasingly complex, and their values ​​have also increased accordingly. Therefore, traditional RCDs cannot effectively identify the type and source of residual currents, resulting in frequent misoperation, which in turn leads to a low commissioning rate or even inability to commission traditional RCDs in some areas. Specifically, distributed photovoltaic power generation devices mostly use non-isolated grid-connected inverters, and the non-isolated structure enables a direct electrical connection between photovoltaic components and the AC power grid. The high-frequency operation of the inverter switch tube causes a high-frequency common-mode voltage to be generated between the parasitic capacitance of the photovoltaic components to the ground, the inverter, the AC power grid, and the neutral point of the system, and then generates a high-frequency common-mode current, which is expressed as a residual current in the AC-side RCD. When a human body is electrocuted in a traditional low-voltage distribution network without distributed photovoltaics, the electric shock current is a power frequency current, and the traditional RCD operates according to the effective value of the residual current. When connected to distributed photovoltaics, the RCD will monitor the high-frequency common-mode current generated by the photovoltaic components. At this time, the waveform of the residual current is often very complex, and the increase in the medium and high frequency components will lead to different degrees of increase in the effective value of the monitored residual current. There are many types of existing photovoltaic inverters, with different manufacturers and powers. Some inverters have problems with high grid-connected voltage and large harmonics, which damage users' electrical equipment, and there is a risk of excessive leakage of photovoltaic battery components and inverters. There are many types of loads at users, and household appliances or electromechanical equipment may have load leakage due to aging lines, improper use, etc., which will cause power safety problems.

[0004] It can be seen that for low-voltage distribution networks connected to distributed photovoltaics, as the causes of residual current become increasingly complex and larger, the traditional residual current protection measures cannot effectively identify the type and source of residual current, resulting in false operation, which greatly affects the personal safety, equipment safety and power grid safety of low-voltage AC / DC hybrid distribution networks. Summary of the invention

[0005] The purpose of the present invention is to provide a leakage protection method and related equipment suitable for a low-voltage AC / DC hybrid distribution network, so as to solve the technical problem that for a low-voltage distribution network connected to a distributed photovoltaic system, the causes of residual current are becoming more and more complex and increasing, and the traditional residual current protection measures cannot effectively identify the type and source of residual current, resulting in false operation, which greatly affects the personal safety, equipment safety and grid safety of the low-voltage AC / DC hybrid distribution network.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: In a first aspect, the present invention provides a leakage protection method applicable to a low-voltage AC / DC hybrid distribution network, comprising: Obtain residual current signals in different scenarios of low-voltage distribution networks with distributed photovoltaic access; Based on the residual current signal, it is judged whether there is a leakage fault. After judging the leakage fault, the modal components of the residual current signal in different frequency bands are calculated. Based on the modal components of the residual current signal in different frequency bands, the leakage fault characteristics of the residual current signal during the leakage fault are calculated; Use the pre-built leakage fault identification model to identify the leakage scenario of the fault characteristics and obtain the leakage fault type; In combination with the obtained leakage fault type and the protection criterion corresponding to the leakage fault type, a leakage protection execution instruction is output to execute the leakage protection action according to the leakage protection execution instruction.

[0007] Furthermore, the specific steps of obtaining the residual current signal in different scenarios of the low-voltage distribution network with distributed photovoltaic access are as follows: Collect the user-side residual current data waveform corresponding to the normal operation state, abnormal photovoltaic leakage state and electric shock accident state of the low-voltage distribution network with distributed photovoltaic access.

[0008] Furthermore, the specific steps of determining whether a leakage fault occurs based on the residual current signal are as follows: From the residual current signal collected in real time Select the first i The sampling points are used as reference signals, and based on the reference signals, the residual current signal of the previous cycle with a length of N is selected to calculate its mean value. Estimation sequence:

[0009] Perform a mean operation on the sampled value data sequence to obtain a new data sequence:

[0010] Calculate the corresponding i The absolute value of the skewness estimate for the sample points , then:

[0011] According to the calculated absolute value of the skewness estimate, the leakage fault is judged, and the leakage fault time judgment criterion is defined as:

[0012] In the formula, Indicates the preset threshold; Indicates the time judgment criteria for leakage fault; when When , it indicates that there is no leakage fault; when When , it indicates that there is a leakage fault; If the judgment result is that no leakage fault occurs, the residual current signal of the next sampling point is extracted, and the calculation and fault judgment of the latest collected residual current signal are repeated until it is judged that a leakage fault occurs.

[0013] Furthermore, after determining that a leakage fault has occurred, the steps of calculating and obtaining the modal components of the residual current signal in different frequency bands are as follows: The HOA-CEEMDAN decomposition algorithm is used to decompose the current residual current signal to obtain the modal components of the residual current signal in different frequency bands. The specific steps and formulas are as follows: For the original residual current signal Add different noise instances to generate a new set of signals:

[0014] in, is the original residual current signal, is the noise signal, is a random noise instance; For each noisy signal Perform empirical mode decomposition to obtain a set of intrinsic mode functions:

[0015] in, It is i The signal k The intrinsic mode functions, Nis the number of eigenmode functions; It is i The residual of a signal;

[0016] in, It is the final k Intrinsic mode functions; M is the number of noise instances; Select the most stable mode from the set of intrinsic mode functions as the modal components of the residual current signal in different frequency bands; In the decomposition process, the HOA algorithm is used to optimize the noise standard deviation and the number of noise repetitions in the CEEMDAN algorithm.

[0017] Furthermore, based on the modal components of the residual current signal in different frequency bands, the steps of calculating the leakage fault characteristics of the residual current signal during the leakage fault are as follows: Based on the modal components of the residual current signal in different frequency bands, the fuzzy entropy corresponding to each modal component is calculated. The specific calculation steps are as follows: According to the modal components of the residual current signal in different frequency bands, the time series data is divided into m The sequence is denoted as ,in , and subtract its mean from , get the sequence after removing the mean; Define two sequences and The distance between is the Chebyshev distance; Introducing fuzzy membership To measure and The similarity is defined using an exponential function:

[0018] in, n is the exponential of the fuzzy function, r is the similarity tolerance threshold; Calculate the average similarity of each sequence , and its calculation formula is:

[0019] in, N is the total length of the sequence; Increase the dimension to m +1, double counting , and then calculate the fuzzy entropy :

[0020] The fuzzy entropy of each IMF component is used as an element to construct the feature vector T Characteristics of leakage fault as residual current signal: .

[0021] Furthermore, the leakage fault identification model is constructed based on the NRBO algorithm combined with the random forest model, and the specific construction process is as follows: The NRBO algorithm is used to optimize the parameters of the random forest model; the Newton-Raphson search rule expression of the NRBO algorithm is as follows:

[0022] In the formula, is the first-order derivative of the objective function, is the second-order derivative; x represents the position; n represents the number of iterations; The Newton-Raphson search rule is calculated as follows:

[0023] In the formula, Indicates the worst position, Indicates the best position, is the disturbance amount; The TAO strategy is used to calculate and obtain the optimized parameters. The TAO strategy is used to prevent the algorithm from falling into the local optimum. The solution update rule is as follows:

[0024] in, is the solution update under the condition that DF It is the determining factor in avoiding pitfalls; is the new solution after applying the TAO strategy; It is a solution update that does not satisfy the conditions; The optimized parameters are applied to the random forest model, and the training set data is used to train the random forest model to obtain the leakage fault identification model.

[0025] Furthermore, the protection criteria corresponding to the leakage fault type include distributed photovoltaic leakage protection criteria, user-side electric shock protection criteria and grid-connected line-side leakage protection criteria; If the protection criterion corresponding to the leakage fault type is the distributed photovoltaic leakage protection criterion: when the residual current effective value reaches 30mA, an alarm signal is issued and displayed as abnormal distributed photovoltaic leakage; when the residual current effective value reaches 10mA, the protection execution device at the output end of the photovoltaic inverter will immediately trip; If the protection criterion corresponding to the leakage fault type is the user-side electric shock protection criterion: when the effective value of the residual current reaches 20mA, the protection execution device at the user's incoming line end will immediately trip; If the protection criterion corresponding to the leakage fault type is the grid-connected line side leakage protection criterion: when the effective value of the residual current reaches 50mA, an alarm signal is issued and displayed as abnormal line leakage; when the effective value of the residual current reaches 100mA, the protection execution device at the grid-connected end trips after a delay of 2s.

[0026] In a second aspect, the present invention provides a leakage protection system applicable to a low voltage AC / DC hybrid distribution network, comprising: The current signal acquisition module is used to obtain the residual current signal in different scenarios of the low-voltage distribution network with distributed photovoltaic access; A modal component calculation module is used to determine whether a leakage fault occurs based on the residual current signal, and after determining that a leakage fault occurs, calculate the modal components of different frequency bands of the residual current signal; A leakage fault feature acquisition module is used to calculate the leakage fault feature of the residual current signal when a leakage fault occurs based on the modal components of the residual current signal in different frequency bands; A leakage fault identification module is used to identify leakage scenarios based on fault features using a pre-built leakage fault identification model to obtain leakage fault types; The leakage protection execution module is used to output a leakage protection execution instruction based on the obtained leakage fault type and the protection criterion corresponding to the leakage fault type, so as to execute the leakage protection action according to the leakage protection execution instruction.

[0027] In a third aspect, the present invention provides a device, comprising: Memory for storing computer programs; The processor is used to implement the steps of the above-mentioned leakage protection method applicable to a low-voltage AC / DC hybrid distribution network when executing the computer program.

[0028] In a fourth aspect, the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the computer program is used to implement the steps of the above-mentioned leakage protection method applicable to a low-voltage AC / DC hybrid distribution network.

[0029] Compared with the prior art, the present invention has the following beneficial effects: The present invention provides a leakage protection method suitable for a low-voltage AC / DC hybrid distribution network. The method obtains residual current signals in different scenarios, determines whether there is a leakage fault, and then calculates the modal components of the residual current signal in different frequency bands, thereby accurately identifying the characteristics of the leakage fault; the pre-built leakage fault identification model can accurately identify the type of leakage fault, and finally outputs a leakage protection execution instruction in combination with the fault type and the corresponding protection criterion to achieve accurate leakage protection action. The use of this method not only effectively avoids the personal safety, equipment safety and power grid safety problems that may be caused by the misoperation of traditional residual current protection measures, but also improves the safety and reliability of the low-voltage AC / DC hybrid distribution network.

[0030] Preferably, in the present invention, by collecting residual current data waveforms under different states, rich data support is provided for subsequent leakage fault judgment, which helps to more accurately identify the type and source of the leakage fault and improve the accuracy of leakage protection.

[0031] Preferably, in the present invention, this method uses the absolute value of the skewness estimate to judge the leakage fault, which can effectively distinguish between the normal state and the leakage fault state, and improve the sensitivity and accuracy of leakage fault detection; at the same time, through real-time calculation and judgment, timely response to the leakage fault can be achieved.

[0032] Preferably, in the present invention, the HOA-CEEMDAN decomposition algorithm is used to decompose the residual current signal, so that modal components of different frequency bands can be obtained, which is helpful for a more in-depth analysis of the characteristics of the leakage fault; at the same time, the accuracy and stability of the decomposition are improved by the optimization algorithm, which provides strong support for the subsequent leakage fault feature extraction.

[0033] Preferably, in the present invention, by calculating the fuzzy entropy corresponding to each modal component, a leakage fault feature vector is constructed, which provides an effective input feature for the leakage fault identification model, thereby making full use of the information in the residual current signal and improving the accuracy and reliability of leakage fault identification.

[0034] Preferably, in the present invention, the NRBO algorithm is used to optimize the parameters of the random forest model, thereby improving the recognition ability and generalization performance of the model; at the same time, the TAO strategy is introduced to avoid the algorithm from falling into the local optimum, thereby further improving the optimization effect of the model; the leakage fault identification model constructed by this method can more accurately identify the type and source of the leakage fault, providing strong support for leakage protection.

[0035] Preferably, in the present invention, different protection criteria and action thresholds are set according to the type of leakage fault, so that targeted protection can be performed for different types of leakage faults; at the same time, by setting reasonable alarm signals and tripping actions, the timeliness and effectiveness of leakage protection are ensured, and the impact of leakage faults on the power grid and users is reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 The field application principle of leakage protection provided by the embodiment of the present invention; Figure 2 The user-side residual current waveforms under three leakage scenarios provided by the embodiments of the present invention, where (a), (b), and (c) are respectively the normal operating state, the photovoltaic abnormal leakage state, and the state of electric shock accident; Figure 3 The IMF component after the AC leakage signal is optimized according to the embodiment of the present invention; Figure 4 The IMF component after the photovoltaic leakage signal is optimized according to the embodiment of the present invention; Figure 5 The IMF component after optimization of the biological body electric shock leakage signal provided by the embodiment of the present invention; Figure 6 The user-photovoltaic leakage protection integrated action criterion process provided by the embodiment of the present invention; Figure 7 A flow chart of a leakage protection method applicable to a low-voltage AC / DC hybrid distribution network provided by the present invention; Figure 8 A structural schematic diagram of a leakage protection system suitable for a low-voltage AC / DC hybrid distribution network provided by the present invention. DETAILED DESCRIPTION

[0037] Example 1 As described in the background technology, traditional RCDs cannot effectively identify the type and source of residual current, resulting in frequent misoperation, which in turn leads to a low commissioning rate or even inability to operate traditional RCDs in some areas; when connected to distributed photovoltaics, RCDs will monitor the high-frequency common-mode current generated by photovoltaic modules. At this time, the waveform of the residual current is often very complex, and the increase in medium and high-frequency components will lead to different degrees of increase in the effective value of the monitored residual current. There are many existing photovoltaic inverters, with different manufacturers and powers. Some inverters have problems with high grid-connected voltage and large harmonics, which cause damage to users' electrical equipment, and there is a risk of excessive leakage of photovoltaic battery modules and inverters. There are many types of loads at users, and household appliances or electromechanical equipment have load leakage during use due to aging of lines, improper use, etc., which will cause power safety problems. .

[0038] In order to solve the above problems, the present invention provides a leakage protection method suitable for low-voltage AC / DC hybrid distribution network. The method can effectively identify the type and source of residual current and avoid false operation, which is of great significance to ensuring the reliable operation of the power grid and personal safety.

[0039] This embodiment provides a leakage protection method applicable to a low-voltage AC / DC hybrid distribution network, comprising the following steps: Step 1: Residual current signal data acquisition.

[0040] Collect the residual current signal of the user connection point of the low-voltage distribution network with distributed photovoltaic access.

[0041] Step 2: Obtain modal components of different frequency bands of the residual current signal according to the residual current signal.

[0042] After obtaining the modal components of the residual current signal in different frequency bands under different scenarios based on the HOA-CEEMDAN decomposition algorithm, proceed to step 3.

[0043] Step 3: Obtain the leakage fault feature set based on the modal components of the residual current signal in different frequency bands.

[0044] The fuzzy entropy of the modal components in each frequency band is calculated and used as the leakage fault feature set.

[0045] Step 4: Identify the type of leakage fault.

[0046] The fault characteristics obtained in step 3 are input into the NRBO-RF-based leakage fault identification model in the microprocessor to identify the leakage scenario. If it is identified as distributed photovoltaic leakage, go to step 5; if it is identified as electric shock on the user side, go to step 6; if it is not the above situation, go to step 7.

[0047] Step 5: Execute the distributed photovoltaic leakage protection criteria.

[0048] When the effective value of the residual current reaches 30mA, an alarm signal is issued and displayed as abnormal leakage of distributed photovoltaics; when the effective value of the residual current reaches 10mA, the protection execution device at the output end of the photovoltaic inverter will immediately trip.

[0049] Step 6: Execute the user-side electric shock protection criteria.

[0050] When the effective value of the residual current reaches 20mA, the protection execution device at the user's incoming line end will immediately trip.

[0051] Step 7: Execute the grid-connected line side leakage protection criteria.

[0052] When the effective value of the residual current reaches 50mA, an alarm signal is issued and it is displayed as abnormal line leakage; when the effective value of the residual current reaches 100mA, the protection execution device at the grid-connected end should trip after a delay of 2s.

[0053] The leakage protection method applicable to the low voltage AC / DC hybrid distribution network provided in this embodiment is further explained in conjunction with the accompanying drawings: like Figure 1 As shown in the figure, the design principle of its leakage protection field application is as follows: The residual current detection module collects the residual current signal in real time and performs data analysis and interpretation; the action execution module responds quickly and can quickly disconnect the faulty load to ensure personal safety.

[0054] First, residual current detection and identification module The residual current detection and identification module collects residual current signals through high-precision transformers, determines the type of residual current through artificial intelligence algorithms, performs fault diagnosis, displays and executes load switching actions to ensure power safety.

[0055] The fault diagnosis module is based on the HOA-CEEMDAN (fully adaptive noise ensemble empirical mode based on pedestrian optimization algorithm) decomposition algorithm and the NRBO-RF identification algorithm, which constitutes a combined model for leakage identification of AC / DC hybrid distribution networks containing distributed power sources based on HCEEMDAN-NRBO-RF. The process is described in detail as follows: Step 1: Obtain the residual current signal in different scenarios of the low-voltage distribution network with distributed photovoltaic access.

[0056] Collect the residual current data waveform of the user side in different states of the low-voltage distribution network with distributed photovoltaic access, including normal operation state, photovoltaic abnormal leakage state and electric shock accident state; as shown in the attached Figure 2 As shown in (a), (b) and (c).

[0057] Step 2: Leakage fault detection method The residual current skewness calculation method is adopted, and the sudden change characteristics of the absolute value of the skewness of the residual current signal transient process are used as the basis for leakage fault detection to detect the leakage fault moment of the low-voltage distribution network. The principle is as follows: Assumptions is a time sampling sequence of residual current in low-voltage distribution lines, considering the following binary assumptions: (1) In the formula, T 0 and T 1 corresponds to the steady-state operation state and fault state of the low-voltage power distribution system respectively; n ( i) is the sum of the fundamental component and the harmonic component in the residual current signal under steady-state operation; f ( i ) is the sum of the fundamental component, harmonic component and transient component in the residual current signal under fault condition; k ( i ) is the noise in the residual current signal.

[0058] Under normal operating conditions, the residual current signal contains fundamental components, harmonic components and noise, and the skewness of these components is theoretically zero. When a leakage fault occurs, the fundamental amplitude and phase of the residual current signal change. The residual current signal contains not only steady-state components and noise, but also exponentially decaying DC components and other high-frequency transient components. Since the skewness of the transient component is not zero, the skewness of the residual current signal is no longer equal to zero. Therefore, the problem of detecting the fault moment can be transformed into the problem of monitoring the skewness of the residual current. Formula (1) can be transformed into: (2) However, in practice, the length of the measured data is finite, and the calculation of skewness cannot be accurately obtained through the expectation calculation involving the sampling sequence. Therefore, the expectation operation is replaced by the sample mean to obtain the estimate of skewness. For a finite length data sequence, The estimate of its skewness is: (3) In the formula, N is the number of sampling points in one power frequency cycle.

[0059] The skewness estimate calculated by formula (3) has a small offset to zero, so in order to avoid making wrong judgments, it is necessary to set a threshold, namely: (4) In the formula, S set The setting of is related to the sampling rate and noise intensity of the signal, because the accuracy of skewness estimation is affected by the number of sampling points and noise intensity of the signal.

[0060] By real-time monitoring of the absolute value of the residual current deviation and comparing it with the setting threshold, it is possible to detect whether an electric shock fault occurs in the low-voltage power distribution system. Figure 6 The specific steps are as follows: 1) Real-time acquisition of residual current signal ; 2) Select the first i sampling points as the reference signal, and select the previous cycle (length N) residual current sampling signal and calculate its mean value Estimation sequence: (5) Perform a mean operation on the sampled value data sequence to obtain a new data sequence: (6) Calculate the corresponding i The absolute value of the skewness estimate of the sampling points is: (7) 3) According to the absolute value of the skewness estimate calculated in 2), the electric shock fault is determined. The electric shock fault time judgment criterion is defined as: (8) In the formula, when When , it indicates that there is no leakage fault (normal operation state); when , it indicates that an electric shock fault has occurred.

[0061] 4) If the result of step 3) is that no leakage fault occurs, return to extract the residual current signal of the next sampling point, repeat 1) to 3) to calculate and determine the fault of the latest collected residual current sampling value.

[0062] Step 3: Obtain the fault characteristics of the residual current signal during leakage fault.

[0063] When step 2 indicates that a leakage fault has occurred in the low-voltage distribution network, the fault characteristics are obtained through step 3. The residual current signal is decomposed using the HOA-CEEMDAN decomposition algorithm. The HOA-CEEMDAN algorithm decomposes the complex residual current signal into K intrinsic mode functions (IMFs) by introducing adaptive noise and multiple iterations. The energy characteristics are extracted by using the residual current waveforms and data obtained when the low-voltage distribution network system is operating abnormally, photovoltaic leakage, and electric shock to organisms. The specific steps of CEEMDAN decomposition are as follows: 1) Adding noise For the original signal Add different noise instances to generate a new set of signals: (9) in, is the original signal, is a signal with noise added, is an instance of random noise.

[0064] 2) EMD decomposition For each noisy signal Perform empirical mode decomposition (EMD) to obtain a set of intrinsic mode functions (IMFs).

[0065] (10) in, It is i The signal k The intrinsic mode functions, N is the number of eigenmode functions. It is i The residual of a signal.

[0066] 3) Ensemble average (11) in, It is the final k The intrinsic mode functions. M is the number of noise instances.

[0067] 4) Mode selection The most stable modes are selected from the set, which have high consistency across different noise instances. The performance and accuracy of CEEMDAN can be improved by optimizing the noise standard deviation (noise level) and the number of noise repetitions in the CEEMDAN algorithm using the HOA (Hike Optimization Algorithm).

[0068] At this time, the residual current signals in the three scenarios are decomposed by optimized CEEMDAN and the results are as shown in the attached figure. Figure 3 , Figure 4 and Figure 5 shown.

[0069] The fuzzy entropy (FE) is calculated for each IMF component obtained by decomposing the residual current signal. Fuzzy entropy is an indicator to measure the complexity and uncertainty of the signal, which can effectively characterize the randomness and regularity of the signal.

[0070] The principle of fuzzy entropy is similar to that of sample entropy and approximate entropy, both of which are used to measure the complexity or regularity of time series data. Unlike sample entropy and approximate entropy, which use unit step functions to judge the similarity between sequences, fuzzy entropy introduces fuzzy set theory and measures the similarity between sequences through fuzzy membership, which makes fuzzy entropy more robust when processing noisy data. Its calculation steps include: 1) Sequence definition The time series data is divided into m The sequence is denoted as ,in , and subtract its mean from , and get the mean-removed sequence.

[0071] 2) Distance definition Define two sequences and The distance between is the Chebyshev distance, that is, the absolute value of the maximum difference between the corresponding elements of the two sequences.

[0072] 3) Similarity definition Introducing fuzzy membership To measure and The similarity is defined using an exponential function: (12) in, n is the exponential of the fuzzy function, r It is the similarity tolerance threshold, which is usually taken as a certain proportion of the sequence standard deviation (such as 0.2 times the standard deviation).

[0073] 4) Average similarity calculation Calculate the average similarity of each sequence , and its calculation formula is: (13) in, N is the total length of the sequence.

[0074] 5) Fuzzy entropy calculation Increase the dimension to m +1, repeat the above steps to calculate , and then calculate the fuzzy entropy: (14) This formula shows that when the sequence dimension changes from m Increase to m +1, the change in sequence complexity.

[0075] The fuzzy entropy of each IMF component is combined into a column vector to form the feature vector of the leakage signal. The fuzzy entropy of each IMF component is used as an element to construct the feature vector T As characteristics of residual current signal: (15) Step 4: Identify the type of leakage fault.

[0076] After obtaining the fault characteristics of different leakage scenarios with distributed photovoltaic access, they are used as input to establish a leakage fault type recognition model based on NRBO-RF to identify different leakage scenarios.

[0077] The feature vectors of all leakage signals are divided into 70% as training set and 30% as test set. The training set is used to train the NRBO-RF classification model, while the test set is used to evaluate the performance of the model.

[0078] Use the NRBO (Newton-Raphson-Based Optimizer) algorithm to optimize the parameters of the RandomForest (RF) model. The NRBO algorithm simulates the Newton-Raphson method to find the optimal combination of random forest parameters, such as the maximum number of iterations, the maximum depth of the tree, and the learning rate. The key to the NRBO algorithm is its Newton-Raphson search rule (NRSR), which is based on the Newton-Raphson method and iteratively finds the root of the function. In optimization problems, this means that the algorithm will continuously adjust the solution in order to find a solution that minimizes or maximizes the objective function. NRSR calculates the first and second order derivatives of the function at the current point, and then uses this derivative information to determine the search direction and step size.

[0079] It uses the first few terms of the Taylor series to estimate the value of the function near the assumed root. For a polynomial function, NRM finds the next point closer to the root by starting from the initial point and then using the Taylor series evaluated at that point. In NRBO, the mathematical expression of the Newton-Raphson search rule (NRSR) is as follows: (16) in, is the first-order derivative of the objective function, is the second-order derivative x represents the position; n represents the number of iterations. This formula shows that the solution can be updated by calculating the derivative of the function at the current point to get closer to the root of the function.

[0080] To implement NRSR in NRBO, the algorithm needs to estimate the first and second order derivatives of the function. This can be done using the finite difference method: (17) (18) Through these estimation calculations, the algorithm can update the solution and move in the direction of reducing the objective function value. At the beginning of the algorithm, an initial random population is generated within the boundary of the candidate solution. Each population consists of fuzzy decision variables and vectors. The population generation rules are as follows: (19) in, Indicates the population n The individual in j The position of the dimension, rand Represents a random number between (0,1), lband ub Respectively represent the upper and lower bounds of the variable.

[0081] The NRSR control vector allows more accurate exploration of the feasible area and better positions. NRSR is based on the concept of the Newton-Raphson method, which is designed to promote exploration trends and accelerate convergence. The calculation formula of NRSR is as follows: (20) in, Indicates the worst position, Indicates the best position, is the disturbance amount.

[0082] The TAO strategy is used to prevent the algorithm from falling into a local optimum. The solution update rule is as follows: (twenty one) in, is a solution update under certain conditions. DF It is the determining factor in avoiding pitfalls; is the new solution after applying the TAO strategy; It is a solution update that does not satisfy the conditions.

[0083] The optimized parameters are applied to the random forest model, and the model is trained using the training set data. It is implemented on the basis of decision trees and is a supervised ensemble learning method of the bagging type. It improves the prediction accuracy by integrating multiple decision trees together. The random forest generates different training sample sets through the bagging algorithm. During training, a part of the samples and a part of the features are randomly sampled from the original data with replacement, and these samples and features are used to train multiple decision trees. During classification, each decision tree classifies the samples, and the random forest selects the classification result by majority voting. This randomization method can reduce the problem of overfitting and improve generalization ability. It has high accuracy and generalization ability, is suitable for high-dimensional data and complex classification problems, can automatically handle missing values ​​and outliers, has good robustness to noisy data, can evaluate the importance of each feature, and provide a basis for feature selection. It has fast training and prediction speeds and is suitable for large-scale data processing.

[0084] Random forests are highly random, and the selection of features for each sampling and splitting is different. They can be effectively applied to large-scale samples with high accuracy. Random forests have become a powerful tool for solving various machine learning tasks through their efficient computing performance, powerful regularization mechanism, flexible missing value processing, and wide range of application scenarios.

[0085] Use the trained random forest model to predict the test set, obtain the classification results of leakage identification, and evaluate the model performance with reference to the accuracy rate.

[0086] 2. Action execution module The action execution module uses a fast solid-state relay and contactor to disconnect the load switch. The solid-state relay is used as a high-speed response switch, and the action response time can reach The level is much better than the current common 0.1s response speed. At the same time, the contactor is used as a safe disconnection point to improve power safety.

[0087] 3. Data display, residual current waveform recording, data storage A color LCD touch screen is used as the real-time display and control interface to display the residual current value and curve in real time. The contactor opening and closing actions can be manually performed on the screen.

[0088] The USB interface is used to connect the U disk as real-time data storage. The storage duration depends on the capacity of the U disk. The stored data can be analyzed on the computer later. The memory card is used to store the scheduled operation data and instantaneous fault data.

[0089] 4. Realizable functions (1) Residual current detection and protection at the grid-connected end When the residual current detection module identifies abnormal leakage of the grid-connected line, the grid-connected leakage protection is executed. Two thresholds are set for this point. The first threshold is 50mA. When it reaches 50mA, the display screen displays the abnormal warning signal of the line residual current and issues a prompt; the second threshold is 100mA. When it reaches 100mA, the protection should be delayed for 2s before tripping. When the action value disappears, the operation state is automatically restored.

[0090] (2) Residual current detection and protection at the user end When the residual current detection module identifies an electric shock fault at the user end, the leakage protection at the user end is executed. A threshold of 20mA is set for this point. If it reaches 20mA, it will act instantly, and if it does not reach 20mA, it will continue to detect. When the action value disappears, the operation state will be automatically restored.

[0091] (3) Residual current detection and protection at the output of photovoltaic inverter When the residual current detection module identifies a leakage fault at the output of the photovoltaic inverter, the leakage protection at the output of the photovoltaic inverter is executed. Two thresholds are set for this point. The first threshold is 30mA. If it reaches 30mA, a warning signal is displayed on the display and a prompt is issued; the second threshold is 10mA / kVA. If it reaches 10mA / kVA (5~8kVA for general users, the specific situation is adjusted according to the site), it will act immediately. When the action value disappears, the operation state is automatically restored.

[0092] Example 2 like Figure 7 As shown, this embodiment provides a leakage protection method applicable to a low-voltage AC / DC hybrid distribution network, comprising the following steps: Obtain residual current signals in different scenarios of low-voltage distribution networks with distributed photovoltaic access; Based on the residual current signal, it is judged whether there is a leakage fault. After judging the leakage fault, the modal components of the residual current signal in different frequency bands are calculated. Based on the modal components of the residual current signal in different frequency bands, the leakage fault characteristics of the residual current signal during the leakage fault are calculated; Use the pre-built leakage fault identification model to identify the leakage scenario of the fault characteristics and obtain the leakage fault type; In combination with the obtained leakage fault type and the protection criterion corresponding to the leakage fault type, a leakage protection execution instruction is output to execute the leakage protection action according to the leakage protection execution instruction.

[0093] like Figure 8 As shown, this embodiment also provides a leakage protection system suitable for a low-voltage AC / DC hybrid distribution network, including: a current signal acquisition module, used to obtain residual current signals in different scenarios of a low-voltage distribution network with distributed photovoltaic access; a modal component calculation module, used to determine whether a leakage fault occurs based on the residual current signal, and after determining that a leakage fault occurs, calculate the modal components of the residual current signal in different frequency bands; a leakage fault feature acquisition module, used to calculate the leakage fault feature of the residual current signal during a leakage fault based on the modal components of the residual current signal in different frequency bands; a leakage fault identification module, used to use a pre-built leakage fault identification model to identify the fault feature for the leakage scenario, and obtain the leakage fault type; a leakage protection execution module, used to combine the obtained leakage fault type and the protection criterion corresponding to the leakage fault type, and output a leakage protection execution instruction to execute the leakage protection action according to the leakage protection execution instruction.

[0094] The present invention also provides a device, comprising: a memory for storing a computer program; a processor for implementing the steps of the leakage protection method applicable to a low-voltage AC / DC hybrid distribution network when executing the computer program.

[0095] When the processor executes the computer program, the above-mentioned steps of leakage protection applicable to the low-voltage AC / DC hybrid distribution network are implemented, for example: obtaining the residual current signal in different scenarios of the low-voltage distribution network with distributed photovoltaic access; judging whether a leakage fault occurs based on the residual current signal, and after judging that a leakage fault occurs, calculating the modal components of the residual current signal in different frequency bands; calculating the leakage fault characteristics of the residual current signal in the case of a leakage fault based on the modal components of the residual current signal in different frequency bands; using a pre-built leakage fault identification model to identify the fault characteristics for leakage scenarios, and obtaining the leakage fault type; combining the obtained leakage fault type and the protection criterion corresponding to the leakage fault type, outputting a leakage protection execution instruction, so as to execute the leakage protection action according to the leakage protection execution instruction.

[0096] Alternatively, when the processor executes the computer program, it implements the functions of each module in the above-mentioned system, for example: a current signal acquisition module, which is used to obtain the residual current signal in different scenarios of the low-voltage distribution network with distributed photovoltaic access; a modal component calculation module, which is used to determine whether a leakage fault occurs based on the residual current signal, and after determining that a leakage fault occurs, calculate the modal components of the residual current signal in different frequency bands; a leakage fault feature acquisition module, which is used to calculate the leakage fault feature of the residual current signal during a leakage fault based on the modal components of the residual current signal in different frequency bands; a leakage fault identification module, which is used to use a pre-built leakage fault identification model to identify the fault feature according to the leakage scenario, and obtain the leakage fault type; a leakage protection execution module, which is used to output a leakage protection execution instruction in combination with the obtained leakage fault type and the protection criterion corresponding to the leakage fault type, so as to execute the leakage protection action according to the leakage protection execution instruction.

[0097] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of completing preset functions, and the instruction segments are used to describe the execution process of the computer program in the leakage protection device suitable for a low-voltage AC / DC hybrid distribution network. For example, the computer program can be divided into a current signal acquisition module, a modal component calculation module, a leakage fault feature acquisition module, a leakage fault identification module and a leakage protection execution module; the specific functions of each module are as follows: a current signal acquisition module, used to obtain the residual current signal in different scenarios of a low-voltage distribution network with distributed photovoltaic access; a modal component calculation module, used to determine whether a leakage fault has occurred based on the residual current signal, and after determining that a leakage fault has occurred, calculate the modal components of the residual current signal in different frequency bands; a leakage fault feature acquisition module, used to calculate the leakage fault feature of the residual current signal during a leakage fault based on the modal components of the residual current signal in different frequency bands; a leakage fault identification module, used to use a pre-built leakage fault identification model to perform leakage scenario identification on the fault feature to obtain the leakage fault type; a leakage protection execution module, used to output a leakage protection execution instruction in combination with the obtained leakage fault type and the protection criterion corresponding to the leakage fault type, so as to execute the leakage protection action according to the leakage protection execution instruction.

[0098] The leakage protection device applicable to the low-voltage AC / DC hybrid distribution network may be a computing device such as a desktop computer, a notebook, a PDA, and a cloud server. The leakage protection device applicable to the low-voltage AC / DC hybrid distribution network may include, but is not limited to, a processor and a memory. Those skilled in the art may understand that the above is an example of a leakage protection device applicable to a low-voltage AC / DC hybrid distribution network, and does not constitute a limitation on the leakage protection device applicable to the low-voltage AC / DC hybrid distribution network. It may include more components than the above, or a combination of certain components, or different components. For example, the leakage protection device applicable to the low-voltage AC / DC hybrid distribution network may also include input and output devices, network access devices, buses, etc.

[0099] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc. The processor is the control center of the leakage protection for low-voltage AC / DC hybrid distribution network, and uses various interfaces and lines to connect various parts of the leakage protection device for low-voltage AC / DC hybrid distribution network.

[0100] The memory can be used to store the computer program and / or module, and the processor implements various functions of the leakage protection device suitable for a low-voltage AC / DC hybrid distribution network by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory.

[0101] The memory may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0102] The present invention also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the leakage protection method applicable to a low-voltage AC / DC hybrid distribution network are implemented.

[0103] If the integrated module / unit of the leakage protection system suitable for a low-voltage AC / DC hybrid distribution network is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium.

[0104] Based on such understanding, the present invention implements all or part of the processes in the leakage protection method applicable to the low-voltage AC / DC hybrid distribution network, and can also be completed by instructing related hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, and when the computer program is executed by the processor, the steps of the leakage protection method applicable to the low-voltage AC / DC hybrid distribution network can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or preset intermediate form, etc.

[0105] The computer-readable storage medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium, etc.

[0106] It should be noted that the content contained in the computer-readable storage medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable storage media do not include electrical carrier signals and telecommunication signals.

[0107] In summary, the present invention provides a leakage protection method and related equipment suitable for low-voltage AC / DC hybrid distribution networks. The method combines the new leakage protection principle and setting strategy to develop a new leakage protection device suitable for low-voltage AC / DC hybrid distribution networks, and provides an effective solution to the problem of malfunction of leakage protection devices caused by AC / DC distribution networks with large-scale photovoltaic access, thereby achieving personal safety, equipment safety and power grid safety in low-voltage AC / DC hybrid distribution networks.

[0108] The above embodiment is only one of the implementation methods that can realize the technical solution of the present invention. The scope of protection claimed by the present invention is not limited only to this embodiment, but also includes changes, replacements and other implementation methods that can be easily thought of by any technician familiar with the technical field within the technical scope disclosed by the present invention.

Claims

1. A leakage protection method suitable for low voltage AC / DC hybrid distribution network, characterized in that: include: Obtain residual current signals in different scenarios of low-voltage distribution networks with distributed photovoltaic access; Based on the residual current signal, it is judged whether there is a leakage fault. After judging the leakage fault, the modal components of the residual current signal in different frequency bands are calculated. Based on the modal components of the residual current signal in different frequency bands, the leakage fault characteristics of the residual current signal during the leakage fault are calculated; Use the pre-built leakage fault identification model to identify the leakage scenario of the fault characteristics and obtain the leakage fault type; In combination with the obtained leakage fault type and the protection criterion corresponding to the leakage fault type, a leakage protection execution instruction is output to execute the leakage protection action according to the leakage protection execution instruction.

2. The leakage protection method applicable to a low voltage AC / DC hybrid distribution network according to claim 1, characterized in that: The specific steps of obtaining the residual current signal in different scenarios of the low-voltage distribution network with distributed photovoltaic access are as follows: Collect the user-side residual current data waveform corresponding to the normal operation state, abnormal photovoltaic leakage state and electric shock accident state of the low-voltage distribution network with distributed photovoltaic access.

3. The leakage protection method applicable to a low voltage AC / DC hybrid distribution network according to claim 1, characterized in that: The specific steps of judging whether a leakage fault occurs based on the residual current signal are as follows: From the residual current signal collected in real time Select the first i The sampling points are used as reference signals, and based on the reference signals, the residual current signal of the previous cycle with a length of N is selected to calculate its mean value. Estimation sequence: Perform a mean operation on the sampled value data sequence to obtain a new data sequence: Calculate the corresponding i The absolute value of the skewness estimate for the sample points , then: According to the calculated absolute value of the skewness estimate, the leakage fault is judged, and the leakage fault time judgment criterion is defined as: In the formula, Indicates the preset threshold; Indicates the time judgment criteria for leakage fault; when When , it indicates that there is no leakage fault; when When , it indicates that a leakage fault occurs; If the judgment result is that no leakage fault occurs, the residual current signal of the next sampling point is extracted, and the calculation and fault judgment of the latest collected residual current signal are repeated until it is judged that a leakage fault occurs.

4. The leakage protection method applicable to a low voltage AC / DC hybrid distribution network according to claim 1, characterized in that: After determining that a leakage fault has occurred, the steps for calculating the modal components of the residual current signal in different frequency bands are as follows: The HOA-CEEMDAN decomposition algorithm is used to decompose the current residual current signal to obtain the modal components of the residual current signal in different frequency bands. The specific steps and formulas are as follows: For the original residual current signal Add different noise instances to generate a new set of signals: in, is the original residual current signal, is the noise signal, is a random noise instance; For each noisy signal Perform empirical mode decomposition to obtain a set of intrinsic mode functions: in, It is i The signal k The intrinsic mode functions, N is the number of eigenmode functions; It is i The residual of a signal; in, It is the final k Intrinsic mode functions; M is the number of noise instances; Select the most stable mode from the set of intrinsic mode functions as the modal components of the residual current signal in different frequency bands; In the decomposition process, the HOA algorithm is used to optimize the noise standard deviation and the number of noise repetitions in the CEEMDAN algorithm.

5. The leakage protection method applicable to a low voltage AC / DC hybrid distribution network according to claim 1, characterized in that: Based on the modal components of the residual current signal in different frequency bands, the steps for calculating the leakage fault characteristics of the residual current signal during the leakage fault are as follows: Based on the modal components of the residual current signal in different frequency bands, the fuzzy entropy corresponding to each modal component is calculated. The specific calculation steps are as follows: According to the modal components of the residual current signal in different frequency bands, the time series data is divided into m The sequence is denoted as ,in , and subtract its mean from , get the sequence after removing the mean; Define two sequences and The distance between is the Chebyshev distance; Introducing fuzzy membership To measure and The similarity is defined using an exponential function: in, n is the exponential of the fuzzy function, r is the similarity tolerance threshold; Calculate the average similarity of each sequence , and its calculation formula is: in, N is the total length of the sequence; Increase the dimension to m +1, double counting , and then calculate the fuzzy entropy : The fuzzy entropy of each IMF component is used as an element to construct the feature vector T Characteristics of leakage fault as residual current signal: 。 6. The leakage protection method applicable to a low voltage AC / DC hybrid distribution network according to claim 1, characterized in that: The leakage fault identification model is constructed based on the NRBO algorithm combined with the random forest model. The specific construction process is as follows: The NRBO algorithm is used to optimize the parameters of the random forest model; the Newton-Raphson search rule expression of the NRBO algorithm is as follows: In the formula, is the first-order derivative of the objective function, is the second-order derivative; x represents the position; n represents the number of iterations; The Newton-Raphson search rule is calculated as follows: In the formula, Indicates the worst position, Indicates the best position, is the disturbance amount; The TAO strategy is used to calculate and obtain the optimized parameters. The TAO strategy is used to prevent the algorithm from falling into the local optimum. The solution update rule is as follows: in, is the solution update under the condition that DF It is the determining factor in avoiding pitfalls; is the new solution after applying the TAO strategy; It is a solution update that does not satisfy the conditions; The optimized parameters are applied to the random forest model, and the training set data is used to train the random forest model to obtain the leakage fault identification model.

7. The leakage protection method applicable to a low voltage AC / DC hybrid distribution network according to claim 1, characterized in that: The protection criteria corresponding to the leakage fault type include distributed photovoltaic leakage protection criteria, user-side electric shock protection criteria and grid-connected line-side leakage protection criteria; If the protection criterion corresponding to the leakage fault type is the distributed photovoltaic leakage protection criterion: when the residual current effective value reaches 30mA, an alarm signal is issued and displayed as abnormal distributed photovoltaic leakage; when the residual current effective value reaches 10mA, the protection execution device at the output end of the photovoltaic inverter will immediately trip; If the protection criterion corresponding to the leakage fault type is the user-side electric shock protection criterion: when the effective value of the residual current reaches 20mA, the protection execution device at the user's incoming line end will immediately trip; If the protection criterion corresponding to the leakage fault type is the grid-connected line side leakage protection criterion: when the effective value of the residual current reaches 50mA, an alarm signal is issued and displayed as abnormal line leakage; when the effective value of the residual current reaches 100mA, the protection execution device at the grid-connected end trips after a delay of 2s.

8. A leakage protection system suitable for low voltage AC / DC hybrid distribution network, characterized in that: include: The current signal acquisition module is used to obtain the residual current signal in different scenarios of the low-voltage distribution network with distributed photovoltaic access; A modal component calculation module is used to determine whether a leakage fault occurs based on the residual current signal, and after determining that a leakage fault occurs, calculate the modal components of different frequency bands of the residual current signal; A leakage fault feature acquisition module is used to calculate the leakage fault feature of the residual current signal when a leakage fault occurs based on the modal components of the residual current signal in different frequency bands; A leakage fault identification module is used to identify leakage scenarios based on fault features using a pre-built leakage fault identification model to obtain leakage fault types; The leakage protection execution module is used to output a leakage protection execution instruction based on the obtained leakage fault type and the protection criterion corresponding to the leakage fault type, so as to execute the leakage protection action according to the leakage protection execution instruction.

9. A device, characterized in that: include: Memory for storing computer programs; A processor is used to implement the steps of the leakage protection method applicable to a low-voltage AC / DC hybrid distribution network as described in any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, it is used to implement the steps of the leakage protection method applicable to a low-voltage AC / DC hybrid distribution network as described in any one of claims 1 to 7.

Images