A method, system, device and medium for voltage sag assessment

Abstract

The present application discloses a voltage sag assessment method, system, device and medium. Considering that the impact of the change in the distribution network topology structure on the nodal admittance matrix will reduce the applicability of the traditional voltage sag assessment method, the present application uses the XGBoost algorithm to perform topology identification based on the collected data such as the nodal voltage and nodal injection power of the distribution network, and then dynamically updates the nodal admittance matrix in the voltage sag assessment process. Further, after obtaining the real-time updated nodal admittance matrix of the distribution network, the short-circuit iterative calculation method in the sequence coordinates is combined with the Monte Carlo simulation method, and the voltage sag amplitude characteristics of each node in the unbalanced distribution network with inverter-type power sources are calculated by an iterative method to reduce the calculation error, thereby improving the accuracy of voltage sag assessment.

Description

Technical Field

[0001] This application relates to the technical field of distribution network operation evaluation, and particularly to a voltage sag evaluation method, system, device and medium. Background Art

[0002] Distribution network topology identification is an important task to ensure the safe and stable operation of the power system. It can provide topological data basis for the voltage sag evaluation of the distribution network and is the foundation for carrying out distribution network system analysis. Voltage sag refers to the phenomenon that the effective value of the power frequency voltage at a certain point in the power system is temporarily reduced to 10% - 90% of the rated voltage and lasts for 10ms - 1min, and then returns to the normal level.

[0003] Under the background of the construction of a new power system, the frequency of changes in the distribution network topology structure increases due to reasons such as power generation plans, load distribution, and economic operation. The access of various distributed generations (DGs) also makes the structure of the distribution network topology more complex. At the same time, the access of large-scale DGs on the one hand makes the power quality problems of the distribution network more prominent, and on the other hand, large-scale DGs are connected to the grid through power electronic devices, resulting in a large difference in their fault current characteristics compared with traditional AC synchronous generators.

[0004] However, on the one hand, the existing voltage sag evaluation methods do not consider the impact of real-time topology changes on system parameters, and on the other hand, they do not consider the impact of large-scale DG access on short-circuit current calculation, resulting in the inapplicability of their evaluation results. The evaluation results of voltage sags in the distribution network will be inaccurate due to the frequent changes in the topology structure and the deviation of short-circuit current calculation. In summary, it is urgent to study a voltage sag evaluation method that can adapt to DG access. Summary of the Invention

[0005] This application provides a voltage sag evaluation method, system, device and medium, which are used to solve the problem of inaccurate evaluation of voltage sags in the distribution network in the prior art.

[0006] In view of this, in the first aspect of this application, a voltage sag evaluation method is provided, and the method includes:

[0007] Establish a distribution network topology structure data set and a distribution network measurement data set, construct an initial data set for distribution network topology identification by combining the distribution network topology structure data set and the distribution network measurement data set, and after preprocessing, obtain a distribution network topology identification data set;

[0008] Train the XGBoost model with the distribution network topology identification data set to obtain a dynamic distribution network topology structure identification model;

[0009] Input the real-time monitoring data of the distribution network to be evaluated into the dynamic identification model of the distribution network topology structure to obtain the dynamically identified topology structure, and thus output the node association matrix corresponding to the real-time monitoring data;

[0010] Calculate the node admittance matrix and the node impedance matrix according to the node association matrix, and use the node admittance matrix and the node impedance matrix as the power network parameters for dynamic assessment of voltage sags;

[0011] Combine the short-circuit iterative calculation method in the sequence coordinate with the Monte Carlo simulation method, calculate the fault voltage sag amplitude of each node in the unbalanced distribution network with inverter-type power sources through iteration, and perform voltage sag assessment according to the fault voltage sag amplitude.

[0012] Optionally, after establishing the distribution network topology structure dataset and the distribution network measurement dataset, it further includes:

[0013] Evaluate the importance of the features in the distribution network measurement dataset, and screen the features in the distribution network measurement dataset according to the importance to obtain the screened distribution network measurement dataset.

[0014] Optionally, the combination of the short-circuit iterative calculation method in the sequence coordinate with the Monte Carlo simulation method, calculating the fault voltage sag amplitude of each node in the unbalanced distribution network with inverter-type power sources through iteration, and performing voltage sag assessment according to the fault voltage sag amplitude specifically includes:

[0015] S51. Set the sampling iteration times of Monte Carlo and establish a random probability model of short-circuit faults;

[0016] S52. Simulate the occurrence of short-circuit faults in the power grid based on the random probability model of short-circuit faults, construct a short-circuit calculation model for inverter-type distributed power sources to calculate the power grid operation parameters, and construct a sequence component model of the distribution network line according to the node impedance matrix;

[0017] S53. Establish the node voltage equation in the sequence coordinate for the normal operation network according to the node admittance matrix, and write the fault boundary conditions according to the fault type corresponding to the fault point when the power grid simulates a short-circuit fault;

[0018] S54. Combine the node voltage equation and the fault boundary conditions to establish the node equation of the fault component network;

[0019] S55. Based on the node equation of the fault component network, perform iteration according to the given initial value of the equivalent current source of the distributed power source, and use the theorem to obtain the voltage of each node after the power grid fails;

[0020] S56. Determine whether the difference between the node voltage values of two iterations meets the preset iteration accuracy. If not, return to step S53; if yes, calculate the amplitude of the temporary voltage drop of this fault according to the node voltage values of the two iterations.

[0021] S57. Determine whether the number of sampling iterations reaches the preset value. If yes, perform voltage sag assessment according to the amplitude of the temporary voltage drop of the fault; otherwise, return to step S52.

[0022] Optionally, the expression of the node voltage equation is:

[0023]

[0024] In the formula, is the node admittance matrix in the sequence coordinate system, is the positive, negative, and zero sequence voltage vectors of each node in the normal operation network, is the positive, negative, and zero sequence current vectors injected into each node.

[0025] Optionally, the fault boundary conditions specifically include: the boundary conditions for single-phase ground fault, the boundary conditions for two-phase interphase short-circuit fault, and the boundary conditions for two-phase ground short-circuit fault.

[0026] Optionally, the expression of the node equation of the fault component network is:

[0027]

[0028] In the formula, is the node admittance matrix in the sequence coordinate system, is the voltage of the fault component network, is the positive, negative, and zero sequence current vectors injected into node f.

[0029] The second aspect of this application provides a voltage sag assessment system, and the system includes:

[0030] A building unit, used to build a distribution network topology structure data set and a distribution network measurement data set, combine the distribution network topology structure data set and the distribution network measurement data set to construct an initial data set for distribution network topology identification, and after preprocessing, obtain a distribution network topology identification data set;

[0031] A training unit, used to train the XGBoost model through the distribution network topology identification data set to obtain a dynamic identification model of the distribution network topology structure;

[0032] An identification unit, which inputs the real-time monitoring data of the distribution network to be evaluated into the dynamic identification model of the distribution network topology structure together, obtains a dynamically identified topology structure, and thus outputs a node association matrix corresponding to the real-time monitoring data;

[0033] A calculation unit, configured to calculate a node admittance matrix and a node impedance matrix according to the node association matrix, and use the node admittance matrix and the node impedance matrix as power network parameters for dynamic assessment of voltage sags.

[0034] An evaluation unit, configured to combine a short-circuit iterative calculation method and a Monte Carlo simulation method in the sequence coordinate system, calculate the amplitude of the fault voltage sag at each node in an unbalanced distribution network with inverter-type power sources through an iterative method, and perform voltage sag assessment according to the amplitude of the fault voltage sag.

[0035] Optionally, the evaluation unit is specifically configured to:

[0036] S51. Set the sampling iteration times of Monte Carlo and establish a stochastic probability model of short-circuit faults;

[0037] S52. Simulate short-circuit faults occurring in the power grid based on the stochastic probability model of short-circuit faults, construct a short-circuit calculation model for inverter-type distributed power sources to calculate power grid operation parameters, and construct a sequence component model of the distribution network line according to the node impedance matrix;

[0038] S53. Establish a node voltage equation in the sequence coordinate system according to the node admittance matrix for the normal operation network, and write the fault boundary conditions according to the fault type corresponding to the fault point when the power grid simulates a short-circuit fault;

[0039] S54. Combine the node voltage equation and the fault boundary conditions to establish a node equation for the fault component network;

[0040] S55. Based on the node equation of the fault component network, perform iteration according to the initial value of the equivalent current source of the given distributed power source, and use the theorem to obtain the voltage at each node after the power grid fails;

[0041] S56. Determine whether the difference between the node voltage values of two iterations meets a preset iteration accuracy. If not, return to step S53. If so, calculate the amplitude of the fault voltage sag for this time according to the node voltage values of the two iterations;

[0042] S57. Determine whether the sampling iteration times reach a preset value. If so, perform voltage sag assessment according to the amplitude of the fault voltage sag. Otherwise, return to step S52.

[0043] The third aspect of the present application provides a voltage sag assessment device, which includes a processor and a memory:

[0044] The memory is used to store program codes and transmit the program codes to the processor;

[0045] The processor is configured to execute the steps of the voltage sag assessment method described in the first aspect above according to the instructions in the program code.

[0046] A fourth aspect of the present application provides a computer-readable storage medium for storing program code for executing the voltage sag assessment method described in the first aspect above.

[0047] From the above technical solutions, it can be seen that the present application has the following advantages:

[0048] The present application provides a voltage sag assessment method. Considering that the influence of the change in the distribution network topology structure on the nodal admittance matrix will reduce the applicability of the traditional voltage sag assessment method, the present application uses the XGBoost algorithm to perform topology identification by collecting data such as the nodal voltage and nodal injection power of the distribution network, and then dynamically updates the nodal admittance matrix in the voltage sag assessment process. Further, after obtaining the real-time updated nodal admittance matrix of the distribution network, the short-circuit iterative calculation method in the sequence coordinates is combined with the Monte Carlo simulation method, and the voltage sag amplitude characteristics of each node in the unbalanced distribution network with inverter-type power sources are calculated by an iterative method to reduce the calculation error, thereby improving the accuracy of the voltage sag assessment. Description of the Drawings

[0049] Figure 1 is a schematic flowchart of a voltage sag assessment method provided in an embodiment of the present application;

[0050] Figure 2 is a schematic flowchart of a voltage sag assessment process based on the Monte Carlo method provided in an embodiment of the present application;

[0051] Figure 3 is a schematic structural diagram of a voltage sag assessment system provided in an embodiment of the present application. Detailed Embodiments

[0052] In order to enable those skilled in the art to better understand the solution of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0053] Please refer to Figure 1 , a voltage sag assessment method provided in an embodiment of the present application includes:

[0054] Step 101: Establish a distribution network topology structure dataset and a distribution network measurement dataset. Combine the distribution network topology structure dataset and the distribution network measurement dataset to construct an initial dataset for distribution network topology identification. After preprocessing, a distribution network topology identification dataset is obtained.

[0055] It should be noted that with the development of smart distribution networks, Advanced Metering Infrastructure (AMI) and Phasor Measurement Unit (PMU) have become key measurement means for distribution network operation data. Therefore, this invention conducts topology dynamic identification through information such as distribution network node voltage and node injection power collected by the AMI system and PMU.

[0056] In specific implementation, Step 101 includes the following steps:

[0057] (1) Distribution network topology structure dataset:

[0058] Assume that the distribution network has m topology structures according to the states of different sectionalizing switches and tie switches, that is, corresponding to m distribution network topology node incidence matrices. Therefore, according to the lines in the distribution network that contain sectionalizing switches and tie switches, when the switches in the line are in different open / closed states, the corresponding distribution network topology structures can form the distribution network topology structure dataset y m .

[0059] (2) Distribution network measurement dataset:

[0060] Assume that the distribution network contains n nodes. According to the voltage amplitudes (V1~V n ) and injection powers (P1 + jQ1~P n + jQ n ) of each node obtained by the measurement devices in the distribution network, a distribution network measurement data feature dataset x can be constructed nm .

[0061] (3) Distribution network topology identification dataset:

[0062] In summary, the load data with the same topology structure in the distribution network and its corresponding topology are combined into the topology identification dataset of this topology structure. Furthermore, an initial dataset D for distribution network topology identification is established, and the Newton-Raphson method is further used for power flow calculation to preprocess the measurement data. The specific formula is as follows:

[0063]

[0064] The constraint conditions in power flow calculation include:

[0065]

[0066] Finally, the dataset D for distribution network topology recognition can be expressed as:

[0067]

[0068] where: x nm is the distribution network measurement dataset; y m is the distribution network topology structure dataset.

[0069] Furthermore, in one embodiment, after establishing the distribution network topology structure dataset and the distribution network measurement dataset in step 101, it further includes: evaluating the importance of the features in the distribution network measurement dataset, and screening the features in the distribution network measurement dataset according to the importance to obtain the screened distribution network measurement dataset.

[0070] It should be noted that in order to further improve the speed and accuracy of topology recognition, it is necessary to evaluate the importance of the features in the distribution network measurement dataset, and based on this, eliminate the features with smaller contributions. The feature screening process is as follows:

[0071] (1) Use six features as inputs for XGBoost prediction, and select the Root Mean Square Error (RMSE) R RMSE as the key parameter for evaluating feature importance. Its expression is as follows:

[0072]

[0073] where n is the number of calculation results, R op,true,i represents the true value of the i-th operation resilience index, and R op,pre,i represents the predicted result of the i-th operation resilience index.

[0074] (2) Eliminate each feature quantity in turn, that is, input five features for each regression prediction, so as to obtain the root mean square error R of the evaluation result after eliminating the i-th feature RMSE,i (i = 1, 2, ···, 6).

[0075] (3) The importance of each feature is represented by the increment E of the evaluation result error after eliminating this feature i . According to the sorting of the size of E i , the contribution ranking of each feature to the accuracy of operation resilience evaluation can be obtained, and eliminating the feature with the smallest contribution can achieve feature importance screening. The expression of E i is:

[0076]

[0077] Taking the screened distribution network topology recognition dataset D as the input and training it with the XGBoost algorithm, a topology recognition model with better accuracy and algorithm speed can be obtained. By inputting the real-time monitoring data of the distribution network, the topology structure can be dynamically recognized, and then the corresponding node association matrix A=(a ij ) can be output, where:

[0078]

[0079] In the formula, +1 indicates that the specified direction of the branch current is flowing out of the node; -1 indicates that the specified direction of the branch current is flowing into the node.

[0080] Step 102: Train the XGBoost model with the distribution network topology recognition dataset to obtain a dynamic recognition model of the distribution network topology structure.

[0081] It should be noted that XGBoost is a Boosting integrated learning algorithm, which can effectively learn the relationship between the characteristics of the distribution network monitoring data and the topology structure from the sample data. As a serial decision tree model, the XGBoost model can be understood as a summation model of multiple decision trees and can be written as:

[0082]

[0083] In the formula: K is the number of trees; is the set space of the trees; f k is a function in the set space, used to characterize the correlation between the structure q and the leaf weight w of the kth tree.

[0084] The objective function of XGBoost is as follows:

[0085]

[0086] In the formula: The objective function mainly consists of two parts. The first half is the loss function, usually represented by the squared error between y i and , which is used to describe the degree of model fitting to the data; the second half is the regularization term, used to control the complexity of the model. Further processing the formula by means of Taylor expansion, etc., we can get:

[0087]

[0088] Solving, we can get that when , the objective function is minimized to:

[0089]

[0090] Randomly select 80% of the samples from the distribution network topology recognition dataset D as the training set, and the remaining 20% of the samples as the validation set. After training with the XGBoost algorithm, a model for dynamic recognition of the distribution network topology structure can be obtained.

[0091] Step 103: Input the real-time monitoring data of the distribution network to be evaluated into the dynamic recognition model of the distribution network topology structure to obtain the dynamically recognized topology structure, and thus output the node association matrix corresponding to the real-time monitoring data.

[0092] It should be noted that the voltage sag assessment based on the Monte Carlo method requires the topology information of the distribution network. Inputting the distribution network data for which voltage sag assessment is required into the trained distribution network topology identification and prediction model based on XGBoost can obtain the topology structure of the distribution network and be used for dynamic assessment of voltage sag.

[0093] Step 104: Calculate the node admittance matrix and the node impedance matrix according to the node association matrix, and use the node admittance matrix and the node impedance matrix as the power network parameters for dynamic assessment of voltage sag.

[0094] It should be noted that the dynamic recognition of the node admittance matrix in Step 104 specifically includes:

[0095] Assume that the power network has n nodes and L branches. For each branch, the equation can be written as:

[0096]

[0097] In the formula: I lk is the current of branch k; V Lk is the voltage drop of branch k; y Lk is the admittance of branch k.

[0098] Writing the above formula in matrix form gives:

[0099]

[0100] In the formula: I l is the branch current column vector; V L is the branch voltage drop column vector; Y L is the branch admittance matrix.

[0101] Using the node association matrix and the branch admittance matrix, the node admittance matrix Y and the node impedance matrix Z can be obtained through the following formula:

[0102]

[0103] In summary, the dynamic recognition of the power network parameters is realized, and thus data support is provided for the subsequent dynamic assessment of voltage sag.

[0104] Step 105: Combine the short-circuit iterative calculation method in the sequence coordinates with the Monte Carlo simulation method, calculate the fault voltage sag values of each node in the unbalanced distribution network with inverter power supplies through iteration, and conduct voltage sag assessment based on the fault voltage sag values.

[0105] In one embodiment, as Figure 2 shown, Step 105 specifically includes:

[0106] S1051: Set the sampling iteration times of Monte Carlo and establish a random probability model for short-circuit faults.

[0107] It should be noted that when establishing a random probability model for short-circuit faults in Step 1051:

[0108] Short-circuit faults in the network occur randomly. To better reflect the randomness of fault occurrence, a random probability model for short-circuit faults can be established, and then fault information can be generated to simulate a certain short-circuit fault that occurs randomly in the power grid.

[0109] (1) Selection of fault lines:

[0110] In an actual power system, the incidence of multiple faults is relatively low. Therefore, only two states of the line are considered, namely: normal state and single-fault state. To ensure that the number of times each line is selected is proportional to its fault probability, a random method is also used to generate the starting line number for each sampling in the algorithm, that is:

[0111]

[0112] where: L is the number of lines; INT represents taking the integer of the result; μ1 follows a uniform distribution in the interval [0, 1].

[0113] (2) Determine whether the line has a fault

[0114] Generate a random number μ2 that follows a uniform distribution in [0, 1], then there is:

[0115]

[0116] where: P(l i ) is the probability of line i having a fault.

[0117] (3) Determine the fault type

[0118] There are usually 4 types of short-circuit faults: single-phase ground short-circuit (LG), two-phase ground short-circuit (2LG), two-phase interphase short-circuit (2L), and three-phase ground short-circuit (3LG). Assume that the random number μ3 follows a uniform distribution in [0, 1], and its corresponding fault type probability model is expressed as

[0119]

[0120] Wherein, P 3LG , P 2L , P 2LG , P LG are respectively the occurrence probabilities of corresponding fault types. Generally, P 3LG is considered as 5%, P 2L is 5%, P 2LG is 10%, and P LG is 80%.

[0121] (4)Judgment of fault duration

[0122] Generate a random number μ4 that follows a uniform distribution on [0,1], then there is:

[0123]

[0124] Wherein: p t is the failure rate of instantaneous faults, and p p is the failure rate of permanent faults; According to the judgment results, the instantaneous failure rate or the permanent failure rate is respectively called and substituted into the calculation.

[0125] (5)Determine the location of the line fault

[0126] Generate a random number μ5 that follows a uniform distribution on [0,1], then there is:

[0127]

[0128] Wherein: L ij is the line distance, k is the fault point, and L ik is the distance from the line head to the fault point; i and j respectively represent the node labels at the head and end of the line.

[0129] S1052. Based on the random probability model of short - circuit faults, simulate the occurrence of short - circuit faults in the power grid, construct a short - circuit calculation model for inverter - type distributed power sources to calculate the power grid operation parameters, and construct a sequence component model of the distribution network line according to the node impedance matrix.

[0130] Regarding the construction of the short - circuit calculation model for inverter - type distributed power sources in step 1052 to calculate the power grid operation parameters, it should be noted that:

[0131] When the IIDG adopts the low - voltage ride - through control strategy of suppressing negative - sequence current, its short - circuit calculation model is equivalent to a controlled current source, and only a positive - sequence network exists, as follows:

[0132]

[0133] Wherein, I IIDGis the positive-sequence component amplitude of the IIDG fault current, φ is its phase angle, and α1 is the phase angle of the positive-sequence voltage at the connection point of the IIDG; U IIDG（1） is the amplitude of the positive-sequence voltage at the connection point of the IIDG after the fault, and i d * and i q * are respectively the active current command and the reactive current command of the IIDG during the fault, and i d0 is the active command value of the IIDG before the fault.

[0134] Regarding the construction of the sequence component model of the distribution network line based on the node impedance matrix in step 1052, it should be noted that:

[0135] The multi-phase distribution network line sequence component model is established for unbalanced grid voltage sag assessment, and the establishment steps are as follows:

[0136] If a certain line is a three-phase line, its sequence component impedance model can be established as:

[0137]

[0138] In the formula, the superscripts 1, 2, and 0 represent positive, negative, and zero sequences respectively, and Z ij 120 represents the impedance matrix between nodes i and j in the sequence coordinate system, and the value of a is e j120° , T is the transformation formula for transforming the positive, negative, and zero sequence coordinates to the three-phase coordinates, and Z ij ABC represents the impedance matrix between nodes i and j in the phase coordinate system, and can be expressed as:

[0139]

[0140] In the formula, Z ij kk is the virtual impedance, and its value can be calculated by the following formula:

[0141]

[0142] For other cases of single-phase lines and two-phase lines, the establishment of the virtual impedance of the missing phase is similar to the above two cases.

[0143] S1053. Establish the node voltage equation in the sequence coordinate system according to the node admittance matrix for the normal operation network, and write the fault boundary conditions according to the fault type corresponding to the fault point when the power grid simulates a short-circuit fault.

[0144] It should be noted that when writing the nodal voltage equation in the sequence coordinates under normal network operation, the initial value of the injection current at the node where the inverter-type power supply is connected can be set to the sequence current value before the fault, and the written equation is as follows:

[0145]

[0146] Where, Y 120 is the nodal admittance matrix in the sequence coordinates, U th 120 is the positive, negative, and zero-sequence voltage vectors of each node in the normal operating network, and I 120 is the positive, negative, and zero-sequence current vectors injected into each node. The voltage vector corresponding to the i-th node can be expressed as: U th 120 （i） =[U i 1 , U i 2 , U i 0 T . Similarly, the current vector injected into the i-th node can be expressed as I （i） 120 =[I i 1 , I i 2 , I i 0 T .

[0147] Assume that a fault occurs at node f, and then write the fault boundary conditions according to the fault type to solve the short-circuit current and voltage at the fault point.

[0148] The boundary condition equation for a single-phase ground fault is

[0149]

[0150] The boundary condition equation for a two-phase short circuit between phases is

[0151]

[0152] The boundary condition equation for a two-phase ground short circuit is

[0153]

[0154] In the formula, z f is the transition resistance, , , are the positive, negative, and zero-sequence voltages at the fault point f respectively, , ,​​ They are the positive, negative, and zero-sequence currents at the fault point f, respectively.

[0155] S1054. Establish the nodal equations of the fault component network by combining the nodal voltage equations and the fault boundary conditions. S1055. Based on the nodal equations of the fault component network, perform iterations according to the initial values of the equivalent current sources of the given distributed power sources, and use the theorem to obtain the voltages of each node after the power grid fails.

[0156] In the fault component network, only the negative value of the fault current is injected at the fault point f, and the current injection amounts at other nodes are 0. Then, the nodal equations of the fault component network are as follows:

[0157]

[0158] After obtaining the voltages of the fault component network, the superposition theorem can be used to obtain the voltages of each node after the power grid fails, as shown in the following formula:

[0159]

[0160] S1056. Determine whether the difference between the nodal voltage values of two consecutive iterations meets the preset iteration accuracy. If not, return to step S1053. If so, calculate the amplitude value of the temporary voltage drop of this fault according to the nodal voltage values of the two consecutive iterations.

[0161] It should be noted that those skilled in the art can preset the iteration accuracy according to actual needs, which is not limited here.

[0162] S1057. Determine whether the number of sampling iterations has reached the preset value. If so, perform a voltage sag assessment according to the amplitude value of the temporary voltage drop of the fault. Otherwise, return to step S1052.

[0163] It should be noted that when the number of sampling iterations reaches the number of sampling iterations set in step 1051, a voltage sag assessment is performed according to the amplitude value of the temporary voltage drop of the fault. Those skilled in the art can perform a voltage sag assessment based on historical data or experience accumulated in work according to the amplitude value of the temporary voltage drop of the fault, which will not be elaborated here.

[0164] In the embodiment of the present application, a voltage sag assessment method is provided. Considering that the impact of the change in the distribution network topology structure on the nodal admittance matrix will reduce the applicability of the traditional voltage sag assessment method, the present application uses the XGBoost algorithm to perform topology identification by collecting data such as the nodal voltages and nodal injection powers of the distribution network, and then dynamically updates the nodal admittance matrix in the voltage sag assessment process. Further, after obtaining the real-time updated nodal admittance matrix of the distribution network, the short-circuit iterative calculation method in the sequence coordinates is combined with the Monte Carlo simulation method, and the voltage sag amplitude characteristics of each node in the unbalanced distribution network with inverter-type power sources are calculated by an iterative method to reduce the calculation error, thereby improving the accuracy of the voltage sag assessment.

[0165] The above is a voltage sag assessment method provided in the embodiment of the present application. The following is a voltage sag assessment system provided in the embodiment of the present application.

[0166] Please refer to Figure 3 , a voltage sag assessment system provided in the embodiment of the present application includes:

[0167] A building unit 201, configured to build a distribution network topology structure data set and a distribution network measurement data set, construct an initial data set for distribution network topology identification by combining the distribution network topology structure data set and the distribution network measurement data set, and obtain a distribution network topology identification data set after preprocessing.

[0168] A training unit 202, configured to train an XGBoost model through the distribution network topology identification data set to obtain a dynamic distribution network topology structure identification model.

[0169] An identification unit 203, together with inputting the real-time monitoring data of the distribution network to be evaluated into the dynamic distribution network topology structure identification model to obtain a dynamically identified topology structure, and thus outputting a node association matrix corresponding to the real-time monitoring data.

[0170] A calculation unit 204, configured to calculate a nodal admittance matrix and a nodal impedance matrix according to the node association matrix, and use the nodal admittance matrix and the nodal impedance matrix as the power network parameters for dynamic voltage sag assessment.

[0171] An evaluation unit 205, configured to combine the short-circuit iterative calculation method in the sequence coordinates with the Monte Carlo simulation method, calculate the fault voltage sag amplitude of each node in the unbalanced distribution network with inverter-type power sources by an iterative method, and perform voltage sag assessment according to the fault voltage sag amplitude.

[0172] Furthermore, in the embodiment of the present application, a voltage sag assessment device is also provided. The device includes a processor and a memory:

[0173] The memory is used to store program code and transmit the program code to the processor;

[0174] The processor is used to execute the steps of the voltage sag evaluation method described in the above method embodiment according to the instructions in the program code.

[0175] Furthermore, in the embodiments of the present application, a computer-readable storage medium is also provided. The computer-readable storage medium is used to store program code, and the program code is used to execute the method described in the above method embodiment.

[0176] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be described herein again.

[0177] It should be understood that in the present application, the terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not have to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0178] It should be understood that in the present application, "at least one (item)" means one or more, and "a plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships can exist. For example, "A and / or B" can mean: only A exists, only B exists, and both A and B exist at the same time. Among them, A and B can be singular or plural. The character " / " generally means that the associated objects before and after are in an "or" relationship. "At least one (one) of the following" or a similar expression means any combination of these items, including any combination of single item (one) or plural items (ones). For example, at least one (one) of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0179] In several embodiments provided in the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of devices or units can be in electrical, mechanical, or other forms.

[0180] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0181] In addition, in each embodiment of the present application, the functional units can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

[0182] If the above-mentioned integrated unit 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. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (English full name: Read-Only Memory, English abbreviation: ROM), random access memories (English full name: Random Access Memory, English abbreviation: RAM), magnetic disks, or optical discs that can store program codes.

[0183] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than limiting them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A voltage sag assessment method, characterized in that: include: Establishing a distribution network topology data set and a distribution network measurement data set, combining the distribution network topology data set and the distribution network measurement data set to construct a distribution network topology identification initial data set, and after preprocessing, obtaining a distribution network topology identification data set; The XGBoost model is trained by using the distribution network topology identification data set to obtain a distribution network topology structure dynamic identification model; Inputting the real-time monitoring data of the distribution network to be evaluated into the dynamic identification model of the distribution network topology structure to obtain a dynamic identification topology structure, thereby outputting a node association matrix corresponding to the real-time monitoring data; Calculate a node admittance matrix and a node impedance matrix according to the node association matrix, and use the node admittance matrix and the node impedance matrix as power network parameters for dynamic evaluation of voltage sag; The short-circuit iterative calculation method under ordinal coordinates is combined with the Monte Carlo simulation method, and the fault voltage sag amplitude of each node in the unbalanced distribution network containing inverter power sources is calculated in an iterative manner, and the voltage sag is evaluated according to the fault voltage sag amplitude; The method combines the short-circuit iterative calculation method under the sequence coordinate with the Monte Carlo simulation method, iteratively calculates the fault voltage sag amplitude of each node in the unbalanced distribution network containing the inverter power supply, and performs voltage sag evaluation according to the fault voltage sag amplitude, including: S51, setting the number of Monte Carlo sampling iterations, and establishing a random probability model of short circuit fault; S52, simulating a short-circuit fault in the power grid based on a random probability model of the short-circuit fault, and constructing an inverter-type distributed power supply short-circuit calculation model for calculating power grid operation parameters, and constructing a distribution network line sequence component model according to the node impedance matrix; S53, establishing a normal operating network according to the node admittance matrix, writing node voltage equations under ordinal coordinates, and writing fault boundary conditions according to the fault type corresponding to the fault point when a short circuit fault occurs in the power grid simulation; S54, establishing a fault component network node equation in combination with the node voltage equation and the fault boundary condition; S55, based on the fault component network node equation, iterate according to the given initial value of the distributed power source equivalent current source, and use the theorem to obtain the voltage of each node after the fault occurs in the power grid; S56, determining whether the difference between the node voltage values ​​of the two iterations meets the preset iteration accuracy, if not, returning to step S53, if yes, calculating the voltage sag amplitude of the fault according to the node voltage values ​​of the two iterations; S57, determining whether the number of sampling iterations reaches a preset value, if so, performing voltage sag evaluation according to the fault voltage sag amplitude, otherwise returning to step S52.

2. The voltage sag evaluation method according to claim 1, characterized in that: The step of establishing a distribution network topology data set and a distribution network measurement data set further includes: Importance evaluation is performed on features in the distribution network measurement data set, and features in the distribution network measurement data set are screened according to the importance to obtain a screened distribution network measurement data set.

3. The voltage sag evaluation method according to claim 1, characterized in that: The expression of the node voltage equation is: ； In the formula, is the node admittance matrix in ordinal coordinates, is the positive and negative zero-sequence voltage vector of each node in the normal operation network, are the positive and negative zero-sequence current vectors injected into each node.

4. The voltage sag evaluation method according to claim 1, characterized in that: The fault boundary conditions specifically include: boundary conditions for a single-phase grounding fault, boundary conditions for a two-phase phase-to-phase short-circuit fault, and boundary conditions for a two-phase grounding short-circuit fault.

5. The voltage sag evaluation method according to claim 1, characterized in that: The expression of the fault component network node equation is: ； In the formula, is the node admittance matrix in ordinal coordinates, is the voltage of the fault component network, are the positive and negative zero-sequence current vectors injected into node f.

6. A voltage sag assessment system, characterized in that: include: An establishing unit is used to establish a distribution network topology data set and a distribution network measurement data set, and to construct a distribution network topology identification initial data set by combining the distribution network topology data set and the distribution network measurement data set, and to obtain a distribution network topology identification data set after preprocessing; A training unit, used to train the XGBoost model using the distribution network topology identification data set to obtain a distribution network topology structure dynamic identification model; The identification unit inputs the real-time monitoring data of the distribution network to be evaluated into the dynamic identification model of the distribution network topology structure to obtain a dynamic identification topology structure, thereby outputting a node association matrix corresponding to the real-time monitoring data; A calculation unit, used for calculating a node admittance matrix and a node impedance matrix according to the node association matrix, and using the node admittance matrix and the node impedance matrix as power network parameters for dynamic evaluation of voltage sag; An evaluation unit, used for combining a short-circuit iterative calculation method under ordinal coordinates with a Monte Carlo simulation method, calculating the fault voltage sag amplitude of each node in an unbalanced distribution network containing an inverter power supply in an iterative manner, and performing voltage sag evaluation according to the fault voltage sag amplitude; Wherein, the evaluation unit is specifically used for: S51, setting the number of Monte Carlo sampling iterations, and establishing a random probability model of short circuit fault; S52, simulating a short-circuit fault in the power grid based on a random probability model of the short-circuit fault, and constructing an inverter-type distributed power supply short-circuit calculation model for calculating power grid operation parameters, and constructing a distribution network line sequence component model according to the node impedance matrix; S53, establishing a normal operating network according to the node admittance matrix, writing node voltage equations under ordinal coordinates, and writing fault boundary conditions according to the fault type corresponding to the fault point when a short circuit fault occurs in the power grid simulation; S54, establishing a fault component network node equation in combination with the node voltage equation and the fault boundary condition; S55, based on the fault component network node equation, iterate according to the given initial value of the distributed power source equivalent current source, and use the theorem to obtain the voltage of each node after the fault occurs in the power grid; S56, determining whether the difference between the node voltage values ​​of the two iterations meets the preset iteration accuracy, if not, returning to step S53, if yes, calculating the voltage sag amplitude of the fault according to the node voltage values ​​of the two iterations; S57, determining whether the number of sampling iterations reaches a preset value, if so, performing voltage sag evaluation according to the fault voltage sag amplitude, otherwise returning to step S52.

7. A voltage sag assessment device, characterized in that: The device comprises a processor and a memory: The memory is used to store program code and transmit the program code to the processor; The processor is used to execute the voltage sag assessment method according to any one of claims 1 to 5 according to the instructions in the program code.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium is used to store program codes, and the program codes are used to execute the voltage sag assessment method according to any one of claims 1 to 5.

Images