New energy station sending-out line single-ended protection method based on improved PCNN (Pulse Coupled Neural Network)

Abstract

The invention discloses a new energy station sending-out line single-ended protection method based on an improved PCNN, and the method comprises the steps: collecting fault current original data of a new energy station side, and carrying out the feature extraction of the fault current original data, and obtaining a fault current feature vector; taking the fault current feature vector and the fault current original data as input data of two input channels of the improved PCNN network model after offline training, fusing features extracted by the two channels through a feature fusion layer after a pooling layer of the improved PCNN network model, and outputting an identification result of a fault category; and performing corresponding circuit protection action according to the identification result of the fault types inside and outside the area. According to the new energy station sending-out line single-ended protection method, internal and external faults can be identified under various working conditions, and good protection performance is achieved.

Description

Technical Field

[0001] The invention relates to a circuit protection method, in particular to a single-ended quantity protection method for a new energy station transmission line based on an improved PCNN. Background Art

[0002] In order to solve the huge consumption problem of non-renewable energy, the use of renewable energy has developed rapidly. For example, in the power industry, environmentally friendly green power generation technologies such as wind power and photovoltaic power are widely used in power systems. When centralized new energy is connected to the power grid, a large number of power electronic equipment are included. The fault current waveform is significantly affected by the control strategy in the power electronic device. The fault current shows characteristics such as limited amplitude, phase change, and frequency offset. The threshold setting of traditional protection methods is complex and heavily dependent on manual experience.

[0003] At present, the protection of the transmission lines of new energy stations refers to the experience of conventional AC lines. The main protection adopts distance protection and longitudinal protection, and the backup protection adopts current protection. The overcurrent protection method is easily affected by the weak feedback of new energy, which reduces the protection range. The protection performance of distance protection is reduced under high transition resistance. These protection methods have an adverse impact on the large-scale application of new energy.

[0004] With the development of the new generation of artificial intelligence technology, protection methods based on deep learning algorithms are applied in the field of relay protection. This protection method can avoid complex threshold setting. In the existing technology, there is a method that uses voltage reverse wave as the input stream of the deep learning model to classify fault identification and pole selection. This protection method can achieve fault identification in a relatively short time; there is also a method that uses neural network for fault detection and location, but the training workload is very large; there is also a single-terminal protection scheme based on wavelet energy feature extraction and support vector machine, which can accurately identify faults under various working conditions, but the accuracy is low when training with small samples.

[0005] Deep learning algorithms can be used to process feature data, which is conducive to the fitting of nonlinear mapping relationships. Compared with threshold analysis and human judgment, they can improve computing performance and recognition accuracy. Summary of the invention

[0006] The purpose of the invention is to provide a single-ended quantity protection method for a new energy station transmission line based on an improved PCNN, which can identify internal and external faults under various working conditions and has good protection performance.

[0007] Technical solution: The single-ended quantity protection method for the new energy station transmission line based on the improved PCNN described in the present invention comprises the following steps: Step 1: collect the original data of fault current at the new energy station side, and extract the features of the original data of fault current to obtain the fault current feature vector; Step 2: The fault current feature vector and the original fault current data are used as the input data of the two input channels of the improved PCNN network model after offline training. After the pooling layer of the improved PCNN network model, the features extracted from the two channels are fused through the feature fusion layer to output the recognition result of the fault category. Step 3: Perform corresponding circuit protection actions according to the identification results of the fault types inside and outside the zone.

[0008] Furthermore, in step 1, the specific steps for collecting the original data of the fault current at the new energy station side are: The current on the new energy station side is collected in real time, and whether a fault occurs is determined based on the protection start criterion. The protection start criterion is: Δ I f >0.1 I n , where I n is the rated current of the station side, Δ I f is the sudden change of fault current at the station side; If the protection start criterion is established, it is determined that a fault has occurred, and the three-phase current data on the new energy station side within 10ms after the current moment is immediately intercepted.

[0009] Furthermore, in step 1, the calculation steps of the sudden change of fault current at the station side are as follows: Step 1.1, set the acquisition frequency of the station side current to P 0, and then set the sampling time width of the current mutation sampling to Lm ,and Lm ≥4 / P 0; Step 1.2, collect the current of each phase on the station side in real time according to the set collection frequency; Step 1.3, analyzing and judging the collected real-time current collection value, if the difference between the real-time current collection value collected the last time and the real-time current collection value collected the last time of a phase current is ≥ the amplitude growth threshold, then go to step 1.4, otherwise return to step 1.2; Step 1.4, start a timer and when the timer reaches Lm Get the number of times from the current collection time forward Lm The real-time current collection values ​​collected within the range are collected, and the current data sets corresponding to each timer are established in sequence according to the start time of each timer; Step 1.5: After the current data set is established, determine whether the number of all current data sets is greater than E , if the number is greater than E, then calculate the current average value of each real-time current acquisition value in the latest established current data set I 0, then go to step 1.6, otherwise return to step 1.2; Step 1.6, average current I 0 for analysis and judgment, if the current average value corresponding to each current data set I 0 gradually increases and the maximum difference of the start time of each timer is less than the set start time difference threshold, the first real-time current acquisition value in the first established current data set is selected as the first current value, and then the last real-time current acquisition value in the latest established current data set is selected as the second current value, and then go to step 1.7, otherwise delete each current data set and return to step 1.2; Step 1.7, after obtaining the second current value, calculate the difference between the second current value and the first current value as Δ I 0, for Δ I 0 for analysis and judgment, if Δ I 0>0.1 I n , then Δ I 0 is assigned to Δ I f , if Δ I 0≤0.1 I n , then return to step 1.2.

[0010] Furthermore, in step 1, the specific steps of extracting features from the original fault current data to obtain the fault current feature vector are: Firstly, the best IMF component of each group of fault current raw data is extracted according to the minimum information entropy criterion; Then, nine parameter indicators of each optimal IMF component are calculated respectively, and the nine parameter indicators are mean, variance, peak value, kurtosis, effective value, peak factor, impulse factor, waveform factor and margin factor; Finally, nine parameter indicators are used to construct the characteristic vector of each group of fault current raw data.

[0011] Furthermore, in step 1.1, the specific steps of extracting the best IMF component of each group of fault current raw data according to the minimum information entropy criterion are as follows: Firstly, the empirical mode decomposition model is used to decompose the various IMF components from the original data of the current group fault current; Then, the optimal IMF component of the original data of the current group fault current is found from the decomposed IMF components according to the minimum information entropy criterion; Repeat the above two steps until the best IMF components of each group of fault current raw data are extracted.

[0012] Furthermore, the specific steps of decomposing each IMF component from the original data of the current group fault current using the empirical mode decomposition model are as follows: First, the empirical mode decomposition model is constructed. The specific model formula is: , , , where As constraints, min {} means taking the minimum value, u k ={ u 1, u 2,…, u k} is the modal function, f （t） is the fault current signal, u k ( t ) is the original data of fault current, k is the number of modes, δ （t） is the Dirac distribution function, ∂ t is the partial derivative, * is the convolution operation, w k represents the center frequency of each mode, w k ={ w 1, w 2,…, w k}, using index Correction, so that the spectrum of each mode function is modulated to the corresponding baseband, α is the quadratic penalty factor, λ is the Lagrange multiplication operator, is the augmented Lagrangian function; Then the original data of the current group fault current is sent to the constructed empirical mode decomposition model, and the initial value of the number of modes of the empirical mode decomposition model is set; Finally, the empirical mode decomposition model decomposes the corresponding number of IMF components from the original data of the current group fault current according to the initially set number of modes.

[0013] Furthermore, the specific steps of finding the best IMF component of the current group fault current raw data from the decomposed IMF components according to the minimum information entropy criterion are as follows: First, record the kurtosis value of each group of IMF components that have been decomposed, and then select the mode number corresponding to the peak of the kurtosis value as the preferred mode number; Then, a parameter optimization model is constructed to optimize the number of modes. The parameter optimization model is: nThe population optimization model consists of canaries. The model formula of the population optimization model is: , where d represents the number of preferred modes, d ∈[0,1,2,…], n represents the number of canaries, x 1,1 ~ x n,d Represents each canary separately, and the matrix form of the fitness value of all canaries is: , where f is the fitness value, and the formula for updating the discoverer position is: , where t is the current iteration number, j =1,2,…, d , x max is the maximum number of iterations, X i,j Indicates i Canary j Location information in the dimension, α and r 2 are warning values. α∈ [0,1], r 2 ∈ [0,1], CT is a safe value, CT∈ [0.5,1], Q is a random number that follows a normal distribution, L Represents a 1× d The matrix, matrix L All elements in are 1; Construct the position update formula of the joiner: , where For the i The canary in the first stage, j The new location of the t is the current iteration number, SF i,j For the selected food, r i,j is a random number between [0,1], I i,j is a random number in the set {1,2}; When aware of danger, the canaries at the edge of the group will quickly move to a safe area to get a better position, while the canaries in the middle of the group will move randomly. The mathematical expression is: , where the canary individual with the lowest fitness value is taken as the global canary optimal position, X bestis the current global canary optimal position, β As the step size control parameter, it is a random number that follows a normal distribution with a mean of 0 and a variance of 1. K is a random number, K∈ [0,1], s i is the fitness value of the current canary individual, s g and s w are the current global best and worst fitness values, respectively. ε is the smallest constant that avoids zero in the denominator; Finally, the global optimal position X best Where d The value is used as the optimal component position value, and then the optimal IMF component at the corresponding position is found from the decomposed IMF components according to the optimal component position value.

[0014] Furthermore, in step 2, the improved PCNN network model includes a feature fusion layer, a fully connected layer, a multi-head attention mechanism, a Softmax classifier, an output layer and two convolutional pooling branches; the convolutional pooling branch includes a sequentially connected convolutional layer, a pooling layer and an acceleration layer; the two convolutional pooling branches receive the fault current feature vector and the fault current original data respectively, wherein the convolutional layer is used to perform convolution calculations on the fault current feature vector and the fault current original data, and output a group of one-dimensional vectors respectively, the pooling layer is used to perform average pooling operations on the output vectors, the acceleration layer is used to improve the calculation efficiency, the feature fusion layer is used to concatenate and fuse the two groups of one-dimensional vectors and send them to the fully connected layer, and then the fully connected layer performs probability calculation processing and then sends them to the multi-head attention mechanism, and then the multi-head attention mechanism performs feature enhancement processing and then sends them to the Softmax classifier, and the Softmax classifier performs probability classification processing and finally outputs them by the output layer.

[0015] Furthermore, in step 2, the identification results of the output fault categories include an A-phase grounding fault within the zone, a B-phase grounding fault within the zone, a C-phase grounding fault within the zone, a multi-phase fault within the zone, and an out-of-zone fault.

[0016] Furthermore, in step 3, when the corresponding circuit protection action is performed according to the identification result of the fault type in the zone, if there is a phase A grounding fault in the zone, the phase A fault protection device will trip, if there is a phase B grounding fault in the zone, the phase B fault protection device will trip, if there is a phase C grounding fault in the zone, the phase C fault protection device will trip, if there is a multi-phase fault in the zone, all three-phase fault protection devices will trip, and if there is a fault outside the zone, the fault protection device in the zone will not operate.

[0017] Compared with the prior art, the present invention has the following beneficial effects: by collecting the fault current on the new energy station side and improving the variational mode decomposition (VMD) to extract the features of the fault current, the extracted feature vector and the fault current data are used as the input of the two channels of the model, a feature fusion layer is added after the pooling layer, the features extracted from the two channels are fused and a multi-head self-attention mechanism is introduced for training, so that the differentiation of various fault characteristics is improved, and the fault category identification inside and outside the zone and the corresponding protection actions are realized. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is a flow chart of the method of the present invention; Figure 2 The improved PCNN network model structure diagram of the present invention; Figure 3 It is a route map for sending out a new energy station according to an embodiment of the present invention; Figure 4 This is a curve diagram of the change in loss values ​​of the training set and the test set of an embodiment of the present invention; Figure 5 This is a curve diagram of the accuracy change of the training set and the test set of an embodiment of the present invention; Figure 6 This is a test set classification result diagram of an embodiment of the present invention; Figure 7 This is a classification result diagram of different new energy station capacities according to an embodiment of the present invention; Figure 8 This is a test result diagram of different fault locations of an embodiment of the present invention; Fig. 9 This is a diagram showing classification results of different transition resistances according to an embodiment of the present invention; Fig.10 Graph showing test results of different noise intensities according to an embodiment of the present invention; Fig.11 Different network result diagrams of embodiments of the present invention; Fig.12 This is a PCNN output result diagram of an embodiment of the present invention; Fig.13 This is a single-channel CNN output result diagram of an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The technical solution of the present invention is described in detail below with reference to the accompanying drawings, but the protection scope of the present invention is not limited to the embodiments.

[0020] like Figure 1 As shown, the single-ended quantity protection method for the new energy station transmission line based on the improved PCNN disclosed in the present invention comprises the following steps: Step 1: collect the original data of fault current at the new energy station side, and extract the features of the original data of fault current to obtain the fault current feature vector; Step 2: The fault current feature vector and the original fault current data are used as the input data of the two input channels of the improved PCNN network model after offline training. After the pooling layer of the improved PCNN network model, the features extracted from the two channels are fused through the feature fusion layer to output the recognition results of the fault categories inside and outside the area. Step 3: Perform corresponding circuit protection actions according to the identification results of the fault types inside and outside the zone.

[0021] Furthermore, in step 1, the specific steps for collecting the original data of the fault current at the new energy station side are: The current on the new energy station side is collected in real time, and whether a fault occurs is determined based on the protection start criterion. The protection start criterion is: Δ I f >0.1 I n , where I n is the rated current of the station side, Δ I f is the sudden change of fault current at the station side; If the protection start criterion is established, it is determined that a fault has occurred, and the three-phase current data on the new energy station side within 10ms after the current moment is immediately intercepted.

[0022] Furthermore, in step 1, the calculation steps of the sudden change of fault current at the station side are as follows: Step 1.1, set the acquisition frequency of the station side current to P 0, and then set the sampling time width of the current mutation sampling to Lm ,and Lm ≥4 / P 0; Step 1.2, collect the current of each phase on the station side in real time according to the set collection frequency; Step 1.3: Analyze and judge the collected real-time current values. If the difference between the real-time current value collected last time and the real-time current value collected last time of a phase current is greater than or equal to the amplitude growth threshold, the amplitude growth threshold is 0.02. I n , then go to step 1.4, otherwise return to step 1.2; Step 1.4, start a timer and when the timer reaches Lm Get the number of times from the current collection time forward LmThe real-time current collection values ​​collected within the range are collected, and the current data sets corresponding to each timer are established in sequence according to the start time of each timer; Step 1.5: After the current data set is established, determine whether the number of all current data sets is greater than E , E Preferably 2, if the number is greater than E , then calculate the current average value of each real-time current acquisition value in the latest established current data set I 0, then go to step 1.6, otherwise return to step 1.2; Step 1.6, average current I 0 for analysis and judgment, if the current average value corresponding to each current data set I 0 gradually increases and the maximum difference of the start time of each timer is less than the set start time difference threshold, the start time difference threshold is 0.15s, and the judgment of the mean value increment and the start time difference threshold can avoid the misjudgment of the abnormal mutation of the current waveform in segmented form. The first real-time current acquisition value in the first established current data set is selected as the first current value, and the last real-time current acquisition value in the latest established current data set is selected as the second current value, and then enter step 1.7, otherwise delete each current data set and return to step 1.2; Step 1.7, after obtaining the second current value, calculate the difference between the second current value and the first current value as Δ I 0, for Δ I 0 for analysis and judgment, if Δ I 0>0.1 I n , then Δ I 0 is assigned to Δ I f , if Δ I 0≤0.1 I n , then return to step 1.2.

[0023] Furthermore, in step 1, the specific steps of extracting features from the original fault current data to obtain the fault current feature vector are: Firstly, the best IMF component of each group of fault current raw data is extracted according to the minimum information entropy criterion; Then, nine parameter indicators of each optimal IMF component are calculated respectively, and the nine parameter indicators are mean, variance, peak value, kurtosis, effective value, peak factor, impulse factor, waveform factor and margin factor; Finally, nine parameter indicators are used to construct the characteristic vector of each group of fault current raw data.

[0024] Furthermore, in step 1.1, the specific steps of extracting the best IMF component of each group of fault current raw data according to the minimum information entropy criterion are as follows: Firstly, the empirical mode decomposition model is used to decompose the various IMF components from the original data of the current group fault current; Then, the optimal IMF component of the original data of the current group fault current is found from the decomposed IMF components according to the minimum information entropy criterion; Repeat the above two steps until the best IMF components of each group of fault current raw data are extracted.

[0025] Furthermore, the specific steps of decomposing each IMF component from the original data of the current group fault current using the empirical mode decomposition model are as follows: First, the empirical mode decomposition model is constructed. The specific model formula is: , , , where As constraints, min {} means taking the minimum value, u k ={ u 1, u 2,…, u k} is the modal function, f （t） is the fault current signal, u k ( t ) is the original data of fault current, k is the number of modes, δ （t） is the Dirac distribution function, ∂ t is the partial derivative, * is the convolution operation, w k represents the center frequency of each mode, w k ={ w 1, w 2,…, w k}, using index Correction, so that the spectrum of each mode function is modulated to the corresponding baseband, α is the quadratic penalty factor, λ is the Lagrange multiplication operator, is the augmented Lagrangian function; Then the original data of the current group fault current is sent to the constructed empirical mode decomposition model, and the initial value of the number of modes of the empirical mode decomposition model is set; Finally, the empirical mode decomposition model decomposes the corresponding number of IMF components from the original data of the current group fault current according to the initially set number of modes.

[0026] Furthermore, the specific steps of finding the best IMF component of the current group fault current raw data from the decomposed IMF components according to the minimum information entropy criterion are as follows: First, record the kurtosis value of each group of IMF components that have been decomposed, and then select the mode number corresponding to the peak of the kurtosis value as the preferred mode number; Then, a parameter optimization model is constructed to optimize the number of modes. The parameter optimization model is: n The population optimization model consists of canaries. The model formula of the population optimization model is: , where d represents the number of preferred modes, d ∈[0,1,2,…], n represents the number of canaries, x 1,1 ~ x n,d Represents each canary separately, and the matrix form of the fitness value of all canaries is: , where f is the fitness value, and the formula for updating the discoverer position is: , where t is the current iteration number, j =1,2,…, d , x max is the maximum number of iterations, X i,j Indicates i Canary j Location information in the dimension, α and r 2 are warning values. α∈ [0,1], r 2 ∈ [0,1], CT is a safe value, CT∈ [0.5,1], Q is a random number that follows a normal distribution, L Represents a 1× d The matrix, matrix L All elements in are 1; Construct the position update formula of the joiner: , where is i The canary in the first stage, j The new location of the t is the current iteration number, SFi,j For the selected food, r i,j is a random number between [0,1], I i,j is a random number in the set {1,2}; When aware of danger, the canaries at the edge of the group will quickly move to a safe area to get a better position, while the canaries in the middle of the group will move randomly. The mathematical expression is: , where the canary individual with the lowest fitness value is taken as the global canary optimal position, X best is the current global canary optimal position, β As the step size control parameter, it is a random number that follows a normal distribution with a mean of 0 and a variance of 1. K is a random number, K∈ [0,1], s i is the fitness value of the current canary individual, s g and s w are the current global best and worst fitness values, respectively. ε is the smallest constant that avoids zero in the denominator; Finally, the global optimal position X best Where d The value is used as the optimal component position value, and then the optimal IMF component at the corresponding position is found from the decomposed IMF components according to the optimal component position value.

[0027] Furthermore, in step 2, the improved PCNN network model includes a feature fusion layer, a fully connected layer, a multi-head attention mechanism, a Softmax classifier, an output layer, and two convolutional pooling branches, such as Figure 2 As shown in the figure; the convolutional pooling branch includes a convolutional layer, a pooling layer and an acceleration layer connected in sequence; the two convolutional pooling branches receive the fault current feature vector and the fault current original data respectively, wherein the convolutional layer is used to perform convolution calculation on the fault current feature vector and the fault current original data, and output a group of one-dimensional vectors respectively, the pooling layer is used to perform average pooling operation on the output vectors, the acceleration layer is used to improve the calculation efficiency, the feature fusion layer is used to concatenate and fuse the two groups of one-dimensional vectors and send them to the fully connected layer, and then the fully connected layer performs probability calculation processing and then sends them to the multi-head attention mechanism, and then the multi-head attention mechanism performs feature enhancement processing and then sends them to the Softmax classifier, and the Softmax classifier performs probability classification processing and finally outputs them by the output layer.

[0028] Furthermore, in step 2, the identification results of the output fault categories include an A-phase grounding fault within the zone, a B-phase grounding fault within the zone, a C-phase grounding fault within the zone, a multi-phase fault within the zone, and an out-of-zone fault.

[0029] Furthermore, in step 3, when the corresponding circuit protection action is performed according to the identification result of the fault type in the zone, if there is a phase A grounding fault in the zone, the phase A fault protection device will trip, if there is a phase B grounding fault in the zone, the phase B fault protection device will trip, if there is a phase C grounding fault in the zone, the phase C fault protection device will trip, if there is a multi-phase fault in the zone, all three-phase fault protection devices will trip, and if there is a fault outside the zone, the fault protection device in the zone will not operate.

[0030] In order to verify the reliability of the single-end quantity protection method of the new energy station transmission line of the present invention, the following simulation experiment is carried out: First, establish the new energy transmission line structure, such as Figure 3 As shown in the figure, a new energy transmission line model is built in PSCAD / EMTDC. The main parameters are as follows: the voltage level of the transmission line is 220kV, and the rated capacity of the new energy station is 100MW. The external system capacity is 400MVA, the positive sequence impedance is 0.5Ω, the positive sequence impedance phase angle is 80°, the zero sequence impedance is 1Ω, and the frequency is 50Hz. The total length of the transmission line is 100km. The transmission line is an overhead transmission line with a positive sequence resistance of 6.76×10 -8 pu / m, positive sequence inductive reactance 9.6×10 -7 pu / m, positive sequence capacitive reactance 5.78×10 5 pu·m, zero sequence resistance 6.86×10 -7 pu / m, zero sequence inductance 2.5×10 -6 pu / m, zero sequence capacitance 8.14×10 5pu·m. The capacity of the main transformer is 500MVA, and the rated capacity of the box transformer is 300MVA. A large number of simulation experiments were carried out under different fault conditions. The sampling frequency was set to 4kHz, the fault time was 8s, the original data of the fault current within 10ms after the fault was intercepted, the features of the original data of the fault current were extracted, and the fault current feature vector and the original data of the fault current were input into the deep learning model for waveform feature mining and training. According to the protection action principle of 220kV transmission line, it is divided into single-phase faults in the area (A phase grounding, B phase grounding, C phase grounding), which are recorded as AG, BG, CG, multi-phase faults in the area and faults outside the area, a total of 5 fault types. The transition resistance range is 1~100Ω, considering a total of 10 different situations; the fault distance is set to 10~100km, considering a total of 10 different situations; the capacity of the new energy station is 75~150MW, considering a total of 5 different situations, so the original data of the fault current is 5×10×10×5=2500 groups. The improved VMD is used for feature extraction to obtain the fault current feature vector, and the original data of the fault current is divided according to the ratio of 7:3, as shown in Table 1: Table 1 is the data sample division table

[0031] The Adam optimizer was used for optimization, and the initial learning rate was set to 0.001. The loss function was the cross entropy function. The network training was completed after 90 iterations, and the learning rate was adjusted 60 times. The accuracy curve and loss value curve were obtained as follows: Figure 4 and Figure 5 shown.

[0032] from Figure 5 It can be concluded that the prediction accuracy of the test set, which accounts for 30% of the sample data, reaches 100%. Table 2 shows the specific classification results of different fault categories in the sample test set. Figure 6 As shown in the figure is the test set classification result.

[0033] Table 2 is the protection action results of different fault categories in the sample test

[0034] Large-scale renewable energy stations are connected to the power grid. In order to test the impact of renewable energy station capacity on protection performance, test samples with renewable energy station capacity of 120MW, 160MW, 180MW, 200MW, 220MW and 260MW are input into the deep learning network for testing experiments. Figure 7 Table 3 is the classification result of different renewable energy station capacities, and Table 4 is the protection action performance analysis result under different renewable energy station capacities.

[0035] Table 3 shows the protection action results of different new energy station capacities

[0036] from Figure 7 It can be concluded from Table 3 that under different new energy station capacities, faults can be correctly identified and protection can be reliably operated.

[0037] In the transmission lines of new energy stations, the protection can still operate reliably at different fault locations. In order to test the protection action performance of the entire length of the transmission line, 10 groups of test samples with different fault locations were randomly input into the deep learning network for testing experiments. Figure 8 Table 4 is the classification result of different fault locations and Table 5 is the analysis result of protection action performance under different fault locations.

[0038] Table 4 shows the protection action results for different fault locations

[0039] from Figure 8 It can be concluded from Table 4 that, under different fault locations, the fault can still be correctly identified and the protection can be reliably operated.

[0040] When the transmission line of the new energy station fails to ground, ground resistance appears. The maximum transition resistance value of the transmission line of the new energy station is 120Ω. In order to test the ability of the proposed protection method to resist transition resistance, test samples with ground resistance of 100Ω, 105Ω, 110Ω, 115Ω, and 120Ω are input into the deep learning network for testing experiments. Fig. 9 Table 5 is the classification result of different transition resistances, and Table 6 is the analysis result of protection action performance under different transition resistances.

[0041] Table 5 shows the protection action results of different transition resistances

[0042] from Fig. 9 It can be concluded from Table 5 that under different fault resistance conditions, the fault can be correctly identified and the protection can be reliably operated.

[0043] A large amount of noise often occurs in power transmission lines. In order to test the anti-noise ability of the proposed protection method, test samples with noise intensities of 40db, 30db, 20db, and 10db were input into the deep learning network for testing experiments. Fig.10 Table 6 is the noise test classification result, and Table 7 is the protection action performance analysis result under noise interference. Fig.10 It can be concluded from Table 6 that when noise is added to the data, the fault can be correctly identified and the protection can be reliably operated.

[0044] Table 6 shows the protection action results for different noises

[0045] In order to verify the superiority of the single-ended quantity protection based on the improved PCNN network, 10 ms of original current data was superimposed with 10 dB noise. The data set was divided into a training set and a test set at a ratio of 7:3. The noise data was brought into the CNN single-channel network and the PCNN network for anti-noise interference testing. The test comparison results are shown in the figure. Fig.11 shown. Fig.12 and Fig.13 The output results of PCNN and single-channel CNN.

[0046] from Fig.11 and Fig.12 , 13 It can be seen that the classification accuracy of the single-channel CNN and the original PCNN is not as high as the accuracy of the method proposed in this paper. Therefore, the new energy station transmission line protection scheme based on the improved PCNN can still reliably identify faults inside and outside the area under strong noise interference and protect the correct action.

[0047] In summary, the present invention aims at the problem that the four characteristics of the transmission line protection of the new energy station are difficult to coordinate, and proposes a single-ended quantity protection method for the transmission line of the new energy station based on the improved PCNN. First, the fault transient current on the new energy station side is collected, and data features are extracted. The original data samples and feature vectors are input into the improved PCNN to deeply mine the waveform features of the fault transient current, and finally realize the identification of the fault categories inside and outside the area. The results show that the proposed protection method has the following advantages: (1) Adopting the deep learning algorithm with adaptive learning to avoid complex calculations and manual threshold setting; (2) Using a dual-channel neural network, compared with a single data input, the two data types complement each other. Even if one of the data characteristics is weak, the fault characteristics can be fully explored; (3) As backup protection, single-ended data is used for offline training with a detection time of 10 ms, meeting the protection speed requirements.

[0048] As described above, although the present invention has been shown and described with reference to specific preferred embodiments, it should not be construed as limiting the present invention itself. Various changes in form and details may be made without departing from the spirit and scope of the present invention as defined in the appended claims.

Claims

1. A single-ended quantity protection method for a new energy station transmission line based on an improved PCNN, characterized in that: The steps include: Step 1: collect the original data of fault current at the new energy station side, and extract the features of the original data of fault current to obtain the fault current feature vector; Step 2: The fault current feature vector and the original fault current data are used as the input data of the two input channels of the improved PCNN network model after offline training. After the pooling layer of the improved PCNN network model, the features extracted from the two channels are fused through the feature fusion layer to output the recognition result of the fault category. Step 3: Perform corresponding circuit protection actions according to the identification results of the fault types inside and outside the zone.

2. According to the improved PCNN-based single-ended quantity protection method for new energy station transmission lines according to claim 1, it is characterized in that: In step 1, the specific steps for collecting the original data of fault current at the new energy station side are: The current on the new energy station side is collected in real time, and whether a fault occurs is determined based on the protection start criterion. The protection start criterion is: Δ I f >0.1 I n , where I n is the rated current of the station side, Δ I f is the sudden change of fault current at the station side; If the protection start criterion is established, it is determined that a fault has occurred, and the three-phase current data on the new energy station side within 10ms after the current moment is immediately intercepted.

3. The single-ended quantity protection method for the new energy station transmission line based on the improved PCNN according to claim 2 is characterized in that: In step 1, the calculation steps of the fault current mutation amount at the station side are: Step 1.1, set the acquisition frequency of the station side current to P 0, and then set the sampling time width of the current mutation sampling to Lm ,and Lm ≥4 / P 0; Step 1.2, collect the current of each phase on the station side in real time according to the set collection frequency; Step 1.3, analyzing and judging the collected real-time current collection value, if the difference between the real-time current collection value collected the last time and the real-time current collection value collected the last time of a phase current is ≥ the amplitude growth threshold, then go to step 1.4, otherwise return to step 1.2; Step 1.4, start a timer and when the timer reaches Lm Get the number of times from the current collection time forward Lm The real-time current collection values ​​collected within the range are collected, and the current data sets corresponding to each timer are established in sequence according to the start time of each timer; Step 1.5: After the current data set is established, determine whether the number of all current data sets is greater than E , if the number is greater than E , then calculate the current average value of each real-time current acquisition value in the latest established current data set I 0, then go to step 1.6, otherwise return to step 1.2; Step 1.6, average current I 0 for analysis and judgment, if the current average value corresponding to each current data set I 0 gradually increases and the maximum difference of the start time of each timer is less than the set start time difference threshold, the first real-time current acquisition value in the first established current data set is selected as the first current value, and then the last real-time current acquisition value in the latest established current data set is selected as the second current value, and then go to step 1.7, otherwise delete each current data set and return to step 1.2; Step 1.7, after obtaining the second current value, calculate the difference between the second current value and the first current value as Δ I 0, for Δ I 0 for analysis and judgment, if Δ I 0>0.1 I n , then Δ I Assign 0 to Δ I f , if Δ I 0≤0.1 I n , then return to step 1.

2.

4. The single-ended quantity protection method for the new energy station transmission line based on the improved PCNN according to claim 1 is characterized in that: In step 1, the specific steps of extracting features from the original fault current data to obtain the fault current feature vector are as follows: Firstly, the best IMF component of each group of fault current raw data is extracted according to the minimum information entropy criterion; Then, nine parameter indicators of each optimal IMF component are calculated respectively, and the nine parameter indicators are mean, variance, peak value, kurtosis, effective value, peak factor, impulse factor, waveform factor and margin factor; Finally, nine parameter indicators are used to construct the characteristic vector of each group of fault current raw data.

5. The single-ended quantity protection method for the new energy station transmission line based on the improved PCNN according to claim 4 is characterized in that: In step 1.1, the specific steps of extracting the best IMF component of each group of fault current raw data according to the minimum information entropy criterion are as follows: Firstly, the empirical mode decomposition model is used to decompose the various IMF components from the original data of the current group fault current; Then, the optimal IMF component of the original data of the current group fault current is found from the decomposed IMF components according to the minimum information entropy criterion; Repeat the above two steps until the best IMF components of each group of fault current raw data are extracted.

6. The single-ended quantity protection method for the new energy station transmission line based on the improved PCNN according to claim 5 is characterized in that: The specific steps of decomposing each IMF component from the original data of the current group fault current using the empirical mode decomposition model are as follows: First, the empirical mode decomposition model is constructed. The specific model formula is: , , , where As constraints, min {} means taking the minimum value, u k ={ u 1, u 2,…, u k } is the modal function, f （t） is the fault current signal, u k ( t ) is the original data of fault current, k is the number of modes, δ （t） is the Dirac distribution function, ∂ t is the partial derivative, * is the convolution operation, w k represents the center frequency of each mode, w k ={ w 1, w 2,…, w k }, using index Correction, so that the spectrum of each mode function is modulated to the corresponding baseband, α is the quadratic penalty factor, λ is the Lagrange multiplication operator, is the augmented Lagrangian function; Then the original data of the current group fault current is sent to the constructed empirical mode decomposition model, and the initial value of the number of modes of the empirical mode decomposition model is set; Finally, the empirical mode decomposition model decomposes the corresponding number of IMF components from the original data of the current group fault current according to the initially set number of modes.

7. The single-ended quantity protection method for the new energy station transmission line based on the improved PCNN according to claim 5 is characterized in that: The specific steps of finding the best IMF component of the current group fault current raw data from the decomposed IMF components according to the minimum information entropy criterion are as follows: First, record the kurtosis value of each group of IMF components that have been decomposed, and then select the mode number corresponding to the peak of the kurtosis value as the preferred mode number; Then, a parameter optimization model is constructed to optimize the number of modes. The parameter optimization model is: n The population optimization model consists of canaries. The model formula of the population optimization model is: , where d represents the number of preferred modes, d ∈[0,1,2,…], n represents the number of canaries, x 1,1 ~ x n,d Represents each canary separately, and the matrix form of the fitness value of all canaries is: , where f is the fitness value, and the formula for updating the discoverer position is: , where t is the current iteration number, j =1,2,…, d , x max is the maximum number of iterations, X i,j Indicates i Canary j Location information in the dimension, α and r 2 are warning values. α∈ [0,1], r 2 ∈ [0,1], CT is a safe value, CT∈ [0.5,1], Q is a random number that follows a normal distribution, L Represents a 1× d The matrix, matrix L All elements in are 1; Construct the position update formula of the joiner: , where For the i The canary in the first stage, j The new location of the t is the current iteration number, SF i,j For the selected food, r i,j is a random number between [0,1], I i,j is a random number in the set {1,2}; When aware of danger, the canaries at the edge of the group will quickly move to a safe area to get a better position, while the canaries in the middle of the group will move randomly. The mathematical expression is: , where the canary individual with the lowest fitness value is taken as the global canary optimal position, X best is the current global canary optimal position, β As the step size control parameter, it is a random number that follows a normal distribution with a mean of 0 and a variance of 1. K is a random number, K∈ [0,1], s i is the fitness value of the current canary individual, s g and s w are the current global best and worst fitness values, respectively. ε is the smallest constant that avoids zero in the denominator; Finally, the global optimal position X best Where d The value is used as the optimal component position value, and then the optimal IMF component at the corresponding position is found from the decomposed IMF components according to the optimal component position value.

8. The single-ended quantity protection method for the new energy station transmission line based on the improved PCNN according to claim 1 is characterized in that: In step 2, the improved PCNN network model includes a feature fusion layer, a fully connected layer, a multi-head attention mechanism, a Softmax classifier, an output layer and two convolutional pooling branches; the convolutional pooling branch includes sequentially connected convolutional layers, pooling layers and acceleration layers; the two convolutional pooling branches receive the fault current feature vector and the fault current original data respectively, wherein the convolutional layer is used to perform convolution calculations on the fault current feature vector and the fault current original data, and output a group of one-dimensional vectors respectively, the pooling layer is used to perform average pooling operations on the output vectors, the acceleration layer is used to improve the calculation efficiency, the feature fusion layer is used to concatenate and fuse the two groups of one-dimensional vectors and send them to the fully connected layer, and then the fully connected layer performs probability calculation processing and then sends them to the multi-head attention mechanism, and then the multi-head attention mechanism performs feature enhancement processing and then sends them to the Softmax classifier, and the Softmax classifier performs probability classification processing and finally outputs them from the output layer.

9. The single-ended quantity protection method for the new energy station transmission line based on the improved PCNN according to claim 1 is characterized in that: In step 2, the identification results of the output fault categories include an A-phase grounding fault within the zone, a B-phase grounding fault within the zone, a C-phase grounding fault within the zone, a multi-phase fault within the zone, and an out-of-zone fault.

10. The single-ended quantity protection method for the new energy station transmission line based on the improved PCNN according to claim 9 is characterized in that: In step 3, when the corresponding circuit protection action is performed according to the identification result of the fault type in the zone, if there is a phase A grounding fault in the zone, the phase A fault protection device will trip, if there is a phase B grounding fault in the zone, the phase B fault protection device will trip, if there is a phase C grounding fault in the zone, the phase C fault protection device will trip, if there is a multi-phase fault in the zone, all three-phase fault protection devices will trip, and if there is a fault outside the zone, the fault protection device in the zone will not operate.

Images