Method, system and equipment for locating short-circuit fault of active power distribution network based on PMU state estimation

Abstract

The invention relates to the technical field of power grid fault detection, in particular to an active power distribution network short-circuit fault positioning method, system and device based on PMU state estimation, and the method comprises the steps: collecting the real-time measurement data of a PMU in a power distribution network and the pseudo measurement data before the fault; based on the real-time measurement data and the pseudo measurement data, dynamically updating a pseudo measurement value according to a state estimation algorithm, and generating a real-time state estimation result of the power distribution network; dividing the power distribution network into a plurality of candidate regions by combining topological structure features, performing state estimation on each candidate region, and calculating a measurement residual index of each candidate region; and determining a fault area according to the measurement residual index, performing state estimation on all buses in the fault area, calculating a local measurement residual index, and positioning a specific fault bus or line section. According to the method, the problems that traditional power distribution network fault positioning is low in precision under distributed energy access and complex topology, and bus and line section faults cannot be distinguished are effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid fault detection, and particularly to a method, system and device for short - circuit fault location in an active distribution network based on PMU state estimation. Background Art

[0002] The distribution network is an important part of the power system, and its reliability directly affects the power supply quality. With the wide application of distributed generation (DG), the topological structure of the distribution network has become increasingly complex, and traditional fault detection and location methods are difficult to meet the requirements of modern distribution networks. Currently, most distribution system fault location methods rely on breaker operation information or load flow measurement data. However, these traditional methods usually require a long fault diagnosis time and cannot reflect the complex changes in the distribution network in real - time.

[0003] In the prior art, fault detection methods based on phasor measurement units (PMUs) have gradually received attention. PMUs can provide real - time voltage and current phasor information, providing higher accuracy for the state estimation and fault location of the distribution network. However, existing PMU - based fault location methods often face problems such as insufficient fault location accuracy, poor real - time performance, and poor adaptability when dealing with factors such as network topology changes and distributed generation, and are difficult to effectively cope with the dynamic and complex distribution network environment.

[0004] The information disclosed in this background art section is only intended to deepen the understanding of the overall background art of the present disclosure, and should not be regarded as an admission or any form of suggestion that this information constitutes the prior art known to those skilled in the art. Summary of the Invention

[0005] The present invention provides a method, system and device for short - circuit fault location in an active distribution network based on PMU state estimation, which can effectively solve the problems in the background art.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is:

[0007] A method for short - circuit fault location in an active distribution network based on PMU state estimation, the method comprising:

[0008] Collecting real - time measurement data of PMUs in the distribution network and pseudo - measurement data before a fault;

[0009] Based on the real - time measurement data and pseudo - measurement data, dynamically updating the pseudo - measurement values according to a state estimation algorithm to generate a real - time state estimation result of the distribution network;

[0010] Dividing the distribution network into multiple candidate regions in combination with topological structure characteristics, performing state estimation on each candidate region, and calculating the measurement residual index of each candidate region;

[0011] Determine the fault area according to the measured residual index, perform the state estimation on all buses within the fault area, calculate the local measured residual index, and locate the specific faulty bus or line segment.

[0012] Further, dynamically update the pseudo-measurement values according to the state estimation algorithm, including:

[0013] In each iteration of the state estimation, update the pseudo-measurement values according to the voltage magnitude estimation result of the current iteration and the final voltage magnitude of the pre-fault state estimation;

[0014] If the voltage magnitude estimation result of the load bus is greater than or equal to 0.85 per unit value, update the pseudo-measurement values of the bus based on the ZIP model;

[0015] If the voltage magnitude estimation result of the load bus is less than 0.85 per unit value, update the pseudo-measurement values of the bus based on the constant impedance model in the ZIP model;

[0016] Calculate the injection current at the fault location according to the updated pseudo-measurement values, source injection current, and the real-time measurement data, and correct the iteration result of the state estimation.

[0017] Further, update the pseudo-measurement values of the bus based on the ZIP model, including:

[0018] Obtain the historical state estimation results of the bus before the fault, and extract the correlation data between the power consumption of the bus and the voltage magnitude estimation result;

[0019] Based on the correlation data, construct the optimization objective function of the ZIP model by weighted least squares method;

[0020] Solve the optimization objective function to determine the weight coefficients of the constant impedance, constant current, and constant power components in the ZIP model;

[0021] Dynamically calculate the pseudo-measurement values of the bus according to the weight coefficients and the voltage magnitude estimation result of the current iteration.

[0022] Further, construct the optimization objective function of the ZIP model by weighted least squares method, including:

[0023] Obtain the historical state estimation results of the bus under at least four voltage magnitude estimation results before the fault, and each historical state estimation result includes the voltage magnitude estimation result of the bus and the corresponding active power estimation value and reactive power estimation value;

[0024] Generate a theoretical power calculation value of the bus under the voltage amplitude estimation result according to the power expression of the ZIP model, where the theoretical power calculation value is the weighted sum of a constant impedance component, a constant current component, and a constant power component;

[0025] Construct the optimization objective function with the goal of minimizing the sum of squared errors between the theoretical power calculation value and the power estimation value in the historical state estimation result.

[0026] Further, locating a specific faulty bus or line segment includes:

[0027] Insert virtual buses on each line segment of the bus connected to the faulty area to simulate the fault location;

[0028] Perform the state estimation on each line segment containing the virtual bus and calculate the corresponding local measurement residual index;

[0029] Compare the local measurement residual index of the virtual bus with the local measurement residual index of the bus in the faulty area;

[0030] If the local measurement residual index of the virtual bus is low, it is determined that the fault is located in the corresponding line segment, otherwise it is determined that the fault is located in the bus in the faulty area.

[0031] Further, performing the state estimation on each line segment containing the virtual bus includes:

[0032] Modify the state vector of the distribution network, taking the voltage amplitude and injected current of the virtual bus as new state variables;

[0033] Based on the injected current of the virtual bus, the source injected current in the faulty area, and the real-time measurement data, calculate the current distribution of the line segment;

[0034] Iteratively update the state estimation result according to the current distribution and the voltage amplitude of the virtual bus until the convergence condition is met;

[0035] Among them, the injected current of the virtual bus is calculated through the conversion result of the phasor information and the pseudo-measurement value in the real-time measurement data.

[0036] Further, calculating the local measurement residual index includes:

[0037] Select the calculation range according to the topological structure type of the faulty area;

[0038] If the faulty area belongs to a meshed topology, calculate the local measurement residual index based on the residuals of all measurements of the distribution network;

[0039] If the fault area belongs to the radial topology, calculate the local measurement residual index based on the measurement residuals of the feeder where the bus of the fault area is located;

[0040] The local measurement residual index is the sum of the squares of all measurement residuals within the selected calculation range.

[0041] Furthermore, the determination conditions for the meshed topology and the radial topology include:

[0042] If there are two or more independent power supply paths in the fault area connecting to the same power supply node, determine that the fault area belongs to the meshed topology;

[0043] If all buses in the fault area are connected to the power supply node through a single feeder, determine that the fault area belongs to the radial topology.

[0044] An active distribution network short-circuit fault location system based on PMU state estimation, the system includes:

[0045] A data acquisition unit that acquires the real-time measurement data of PMUs in the distribution network and the pseudo-measurement data before the fault;

[0046] A state estimation unit that, based on the real-time measurement data and the pseudo-measurement data, dynamically updates the pseudo-measurement values according to the state estimation algorithm to generate the real-time state estimation result of the distribution network;

[0047] A residual calculation unit that divides the distribution network into multiple candidate areas in combination with the topological structure characteristics, performs state estimation on each candidate area, and calculates the measurement residual index of each candidate area;

[0048] A fault location unit that determines the fault area according to the measurement residual index, performs state estimation on all buses in the fault area, calculates the local measurement residual index, and locates the specific fault bus or line segment.

[0049] An active distribution network short-circuit fault location device based on PMU state estimation, which is used to implement the active distribution network short-circuit fault location method based on PMU state estimation.

[0050] Through the technical solution of the present invention, the following technical effects can be achieved:

[0051] Through dynamic pseudo-measurement update and topology-adaptive measurement residual index calculation, improve the fault location accuracy and adaptability to complex distribution networks, combine network partition search and virtual bus modeling to achieve fast component-level location, shorten the fault troubleshooting time and improve the emergency repair efficiency.

[0052] The above description is only an overview of the technical solution of the present application. In order to be able to understand the technical means of the present application more clearly, it can be implemented in accordance with the content of the specification. And in order to make the above and other purposes, features and advantages of the present application more obvious and understandable, the specific embodiments of the present application are specifically exemplified below. Brief Description of the Drawings

[0053] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments recorded in the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0054] Figure 1 It is a schematic flowchart of a method for locating short - circuit faults in an active distribution network based on PMU state estimation;

[0055] Figure 2 It is a schematic flowchart of updating pseudo - measurement values;

[0056] Figure 3 It is a schematic flowchart of locating the fault location;

[0057] Figure 4 It is a schematic flowchart of calculating the local measurement residual index. Detailed Description of the Embodiments

[0058] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments.

[0059] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs. The terms used in the description of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0060] Embodiment 1;

[0061] As Figure 1 shown, the present application provides a method for locating short - circuit faults in an active distribution network based on PMU state estimation. The method includes:

[0062] S10: Collect the real - time measurement data of PMUs in the distribution network and the pseudo - measurement data before the fault;

[0063] S20: Dynamically update the pseudo-measurement values based on the real-time measurement data and pseudo-measurement data according to the state estimation algorithm, and generate the real-time state estimation result of the distribution network;

[0064] S30: Divide the distribution network into multiple candidate regions according to the topological structure characteristics, perform state estimation on each candidate region, and calculate the measurement residual index of each candidate region;

[0065] S40: Determine the fault region according to the measurement residual index, perform state estimation on all buses in the fault region, calculate the local measurement residual index, and locate the specific fault bus or line segment.

[0066] Specifically, first, the real-time voltage, current, and related measurement data of each bus in the distribution network are collected through PMUs. It is also necessary to obtain the pseudo-measurement data before the last fault to be used as the input for the state estimation process when a fault occurs. The real-time PMU data and historical pseudo-measurement data enter the state estimation process together. Based on the obtained real-time measurement data and pseudo-measurement data, the improved state estimation algorithm (RDSSE) is used to estimate the real-time state of the network. During the state estimation process, the pseudo-measurement values are dynamically updated according to the current voltage amplitude estimation result and the estimated voltage before the last fault; the distribution network is divided into multiple independent candidate regions according to geographical areas and topological structures. Each region is regarded as an independent part, and the state estimation algorithm is executed on the boundary buses of each region. By calculating the estimation results of the boundary buses of each region, the faulty region can be determined; after the state estimation algorithm is executed, the measurement residual index (MRI) is used to evaluate the fault region. MRI is an index defined based on the results of the state estimation algorithm and is used to represent the error between the assumed fault location and the actual fault location. By calculating the MRI value, the fault region and non-fault region can be distinguished, and finally, the specific location where the fault occurs can be determined; once the fault region is identified, further localization work will focus on this region. By performing state estimation on all buses and calculating the local measurement residual index (MRI), the specific fault bus or line segment can be determined. Two different versions of the MRI calculation method will be used for radial topology and meshed topology distribution networks respectively to consider the impact of faults on different topological structures; while fault location is being performed, according to the measurement data of PMUs on the substation and distributed generation (DG) buses, the fault type (such as three-phase fault, two-phase fault, or single-phase fault) is evaluated, which is achieved through sequence analysis of voltage and current, ensuring that the fault type can be accurately identified and further optimizing the fault location accuracy. Finally, through the optimized state estimation process, the accurate fault type and fault location are output. This process can provide fault information in a relatively short time, facilitating the rapid adoption of repair measures, thereby effectively reducing the power outage time and improving the stability and reliability of the distribution network.

[0067] Through the technical solution of the present invention, by means of dynamic pseudo-measurement update and topology-adaptive measurement residual index calculation, the fault location accuracy and the adaptability to complex distribution networks are improved. Combining network partition search and virtual bus modeling, fast component-level location is achieved, the fault troubleshooting time is shortened, and the emergency repair efficiency is improved.

[0068] Furthermore, the pseudo-measurement values are dynamically updated according to the state estimation algorithm, including:

[0069] In each iteration of the state estimation, the pseudo-measurement values are updated according to the voltage magnitude estimation result of the current iteration and the final voltage magnitude of the pre-fault state estimation;

[0070] If the voltage magnitude estimation result of the load bus is greater than or equal to 0.85 per unit value, the pseudo-measurement value of the bus is updated based on the ZIP model;

[0071] If the voltage magnitude estimation result of the load bus is less than 0.85 per unit value, the pseudo-measurement value of the bus is updated based on the constant impedance model in the ZIP model;

[0072] According to the updated pseudo-measurement values, source injection current, and real-time measurement data, the injection current at the fault location is calculated, and the iteration result of the state estimation is corrected.

[0073] As a preference of the above embodiments, when implementing the short-circuit fault location method based on PMU state estimation, the dynamic update of the pseudo-measurement value is a key step to ensure the fault location accuracy. During each state estimation (SE) iteration process, the pseudo-measurement value is updated according to the current voltage amplitude estimation result and the final voltage amplitude of the pre-fault state estimation. During each state estimation iteration process, first obtain the current voltage amplitude estimation result, then compare it with the voltage amplitude estimation before the last fault, and update the pseudo-measurement value by calculating the difference between the two to make it reflect the network state change under fault conditions; in the distribution network, according to the different load bus voltage amplitude estimation results, the strategy for updating the pseudo-measurement value is also different. If the load bus voltage amplitude estimation result is greater than or equal to 0.85 per unit value, the entire ZIP model is used to update the pseudo-measurement value. The ZIP model includes three parts: constant impedance, constant current, and constant power. At this time, the power change of the load reflects the voltage change, so all components in the ZIP model need to be considered to comprehensively reflect the impact of voltage on the load; if the load bus voltage amplitude estimation result is less than 0.85 per unit value, the pseudo-measurement value is updated according to the constant impedance model in the ZIP model. The constant impedance model is applicable when the voltage amplitude is low, and the power change of the load is proportional to the square of the voltage amplitude. At this time, the updated pseudo-measurement value should reflect this characteristic; to determine the state of the distribution network under fault conditions, the PMU can be directly used during the state estimation process. In the event of a fault, the pseudo-measurement value cannot represent the approximate load values of all nodes in the network, especially near the fault point. In fact, as mentioned above, when a distribution network fails, important parts of the network, especially the buses of the fault feeder, will face the collapse of the voltage distribution, and the power consumed by it will change. In this case, eliminating the pseudo-measurement from the state estimation process will lead to the loss of network observability. Therefore, the pseudo-measurement value can be set to zero or its pre-fault value can be used during the state estimation process under fault conditions. This assumption may cause significant calculation errors in the calculation of high impedance or low impedance. Before using the pseudo-measurement in each iteration of the state estimation, their values are updated according to the estimated voltage in the current iteration and the final estimated voltage in the last pre-fault state estimation. For this reason, for network buses with a voltage amplitude greater than 0.85 per unit value, the constant impedance-constant current-constant power (ZIP) model should be considered, as shown in the following formula:

[0074]

[0075] In the formula, and are the active and reactive powers of the i-th load bus, which can be calculated in the last pre-fault state estimation; and are the updated active and reactive powers of the i-th load bus during the k-th iteration during the state estimation fault, respectively; and are respectively the estimated value of the voltage amplitude of the i-th load bus of the last pre-fault state estimation and the estimated value of the k-th iteration during the state estimation fault; in addition, α, β, γ are constant impedance, constant current and constant power factor respectively; on the other hand, if the voltage amplitude of the network bus is less than 0.85 times the per-unit value, the constant impedance model is used, as shown in the following formula:

[0076]

[0077] The most common method to obtain the coefficients for each load bus is the least squares method. In this method, the goal is to minimize the square of the error between the estimated power consumption and the measured power consumption according to at least four different voltage values. In distribution networks, state estimation leads to different time snapshots and their estimated measurement functions related to the power consumption on the load bus. Equivalent measurements are used to determine the parameters of the ZIP model. In order to take into account the impact of possible faults on the state estimation process, the fault current must be calculated and used to update the injected current at the fault location according to the following formula

[0078]

[0079] In each iteration of the state estimation, the source injected current and the converted injected current are used The fault current IFault,k is calculated based on the pseudo-measurement value, where NB is the total number of medium-voltage buses in the network; the updated pseudo-measurement value is combined with the source injection current and the real-time measurement data to correct the injection current at the fault location. In this process, by calculating the source injection current and the current data under the fault condition, the current change in the fault area can be accurately estimated, and the state estimation result can be corrected. The calculation process is to calculate the injection current of the fault point based on the updated pseudo-measurement value and the source injection current, combined with the real-time measurement data, and correct the iterative results in the state estimation (SE) process. Through repeated iterative optimization calculations, the accurate fault location is finally obtained. In each iteration, the state estimation is updated according to the corrected injection current, and the positioning accuracy of the fault location is gradually improved. By optimizing the update process of the pseudo-measurement value, the state estimation result of the distribution network can better reflect the actual situation after the fault occurs, and improve the real-time and accuracy of fault detection and positioning.

[0080] Specifically, the pseudo-measurement values ​​of the bus are updated based on the ZIP model, including:

[0081] Obtain the historical state estimation result of the bus before the fault, and extract the correlation data of the power consumption and voltage amplitude estimation result of the bus;

[0082] Based on the associated data, the optimization objective function of the ZIP model is constructed by weighted square method;

[0083] Solve the optimization objective function to determine the weight coefficients of the constant impedance, constant current, and constant power components in the ZIP model;

[0084] Dynamically calculate the pseudo-measurement value of the bus according to the weight coefficients and the estimated result of the voltage amplitude in the current iteration.

[0085] As an optimization of the above embodiment, in the process of updating the pseudo-measurement value based on the ZIP model, it is first necessary to obtain the historical state estimation result of the bus before the fault. The historical state estimation result includes the estimated voltage amplitude and power consumption data before the fault. Obtain the historical voltage amplitude and power consumption data of each load bus, record its change trend under normal working conditions, and form the associated data between voltage and power consumption. By extracting the associated data between voltage amplitude and power consumption from these historical state estimation results, it can provide a basis for the subsequent optimization of the ZIP model parameters; after obtaining the associated data between voltage and power consumption, the next step is to use the weighted least squares method to construct the optimization objective function of the ZIP model. The weighted least squares method (WLS) is usually used as the main criterion for state estimation to minimize the weighted sum of squares of the deviation between the estimated measurement value and the actual measurement value. Consider the actual measurement value given by the vector z:

[0086] z = h(x) + e;

[0087] In the formula, h T = [h1(x), h2(i), …, hm(x)]; hi(x) (estimated measurement value) is a function associated with the measurement value i and the vector x; x T = [x1, x2, …, xn] is the state vector, and e T = [e1, e2, …, en] is the measurement error vector; in linear state estimation, hi(x) is a linear function of the state variables, as shown in the following formula:

[0088]

[0089] In the formula, H is the correlation coefficient matrix, and the elements in the H matrix represent the degree of association between the state variables and the measurement values. In addition, the following assumptions are also used for the measurement error statistics in the SE study. For example, ei is a Gaussian random variable, the average value of the measurement error is zero, and the measurement errors are independent. The covariance measurement error matrix (R) can be expressed as follows:

[0090]

[0091] In the formula, σ i is the standard deviation of the i-th measurement, which can be obtained from its measurement accuracy; in the linear WLS state estimator, the following function should be minimized:

[0092] J(x) = [z - Hx] T R -1 [z - Hx];

[0093] Finally, the estimated vector can be calculated as:

[0094]

[0095] Through the solution process of weighted least squares, the weight coefficients of the constant impedance (Z), constant current (I), and constant power (P) components in the ZIP model are determined. These weight coefficients reflect the response characteristics of the load bus under different voltage conditions, especially the power response of the load when the voltage changes; by solving the objective function, the optimal coefficient values are obtained to describe the influence of voltage magnitude on load power consumption. For example, the coefficient of the constant impedance component represents the response of the load impedance to voltage changes, the coefficient of the constant current component represents the current response, and the coefficient of the constant power component represents the stability of the load power to voltage changes; after determining the weight coefficients of each component in the ZIP model, during each iteration of the state estimation, the pseudo-measurement values of the load bus are dynamically calculated using the voltage magnitude estimation results of the current iteration and these weight coefficients; according to the current voltage magnitude estimation results, substitute them into the calculation formula in the ZIP model to calculate the pseudo-measurement values of each load bus at the current voltage magnitude; by combining the dynamically calculated pseudo-measurement values with real-time measurement data, the state estimation results of the distribution network are updated. This process helps to correct the changes in current and voltage distributions caused by faults, and then accurately determine the fault location.

[0096] Furthermore, by constructing the optimization objective function of the ZIP model through the least squares method, it includes:

[0097] Obtain the historical state estimation results of the bus under at least four voltage magnitude estimation results before the fault. Each historical state estimation result includes the voltage magnitude estimation result of the bus and the corresponding active power estimation value and reactive power estimation value;

[0098] According to the power expression of the ZIP model, generate the theoretical power calculation value of the bus under the voltage magnitude estimation result. The theoretical power calculation value is the weighted sum of the constant impedance component, constant current component, and constant power component;

[0099] Taking the minimization of the sum of the squares of the errors between the theoretical power calculation value and the power estimation value in the historical state estimation result as the optimization objective, construct the optimization objective function.

[0100] As a preference of the above embodiments, when constructing the optimization objective function of the ZIP model, it is first necessary to obtain the historical state estimation results of the load bus before the fault. Specifically, at least four sets of historical state estimation data under different voltage amplitude estimation results are required. Each historical state estimation result should include the voltage amplitude estimation value of the bus, as well as the corresponding active power and reactive power estimation values. Extract the state estimation results of the load bus at different voltage amplitudes from the historical measurement data to ensure that these data cover the load power responses under various voltage amplitude changes. Each historical estimation result should include the voltage amplitude estimation value, the corresponding active power estimation value, and the corresponding reactive power estimation value; based on the power expression of the ZIP model, generate the theoretical power calculation values of the bus under different voltage amplitude estimation results. The theoretical power is calculated by the weighted sum of three parts (constant impedance, constant current, and constant power) in the ZIP model; based on the weighted least squares method, the optimization objective function is constructed to minimize the sum of the squares of the errors between the theoretical power calculation values and the power estimation values in the historical state estimation results. Specifically, this objective function is used to adjust the component coefficients (constant impedance, constant current, and constant power) in the ZIP model to match the error between the actual measured power and the theoretical power. By minimizing the sum of the squares of the power errors, the optimization objective function will adjust the coefficients in the ZIP model to ensure that the difference between the calculated theoretical power and the historical estimated power is minimized. Solve the optimization objective function using the weighted least squares method to obtain the weight coefficients of each component in the ZIP model. These coefficients represent the power response characteristics of the load under different voltage amplitudes. By optimizing these coefficients, the model can more accurately describe the response of the load to voltage changes; after obtaining the optimal weight coefficients, combined with the current voltage amplitude estimation results, the pseudo-measurement values can be calculated according to the ZIP model. This process ensures that the update of the pseudo-measurement values can accurately reflect the impact of voltage changes on the load, thereby improving the accuracy of state estimation.

[0101] Furthermore, locating the specific fault bus or line segment includes:

[0102] Insert virtual buses on each line segment of the bus connected to the fault area to simulate the fault location;

[0103] Perform state estimation on each line segment containing virtual buses and calculate the corresponding local measurement residual index;

[0104] Compare the local measurement residual index of the virtual bus with the local measurement residual index of the bus in the fault area;

[0105] If the local measurement residual index of the virtual bus is low, it is determined that the fault is located in the corresponding line segment. Otherwise, it is determined that the fault is located in the bus in the fault area.

[0106] As an optimization of the above embodiments, during the fault location process of the distribution network, when the fault area is determined, it is first necessary to simulate possible fault locations by inserting virtual buses. Specifically, for each line segment connected to the fault area, a virtual bus is inserted into the line segment. The virtual bus is used to simulate the fault occurrence point and help calculate the electrical characteristics of the fault point; a virtual bus is inserted at an appropriate position in each line segment, and the bus is located at the assumed fault point position on the line segment. By inserting virtual buses in different line segments, tests can be carried out at multiple assumed fault positions, thereby improving the accuracy of fault location; after the virtual bus is inserted, state estimation is performed on each line segment containing the virtual bus. The purpose is to calculate the voltage, current and other electrical parameters of each line segment where the virtual bus is located through the state estimation algorithm, so as to further evaluate the characteristics of the fault point represented by the virtual bus. For each line segment into which a virtual bus is inserted, the same state estimation steps as those of the actual line segment are performed. By analyzing measurement values such as voltage and current, the state estimation algorithm will calculate the electrical state of the virtual bus and its connected line segments; after performing state estimation, for each line segment where the virtual bus is located, calculate the corresponding local measurement residual index (MRI). The local measurement residual index is used to evaluate the difference between the measurement data and the estimated value at the assumed fault position, and further evaluate the rationality of the assumed fault position; compare the local measurement residual index of each virtual bus with the local measurement residual index of the bus in the fault area. By comparing the MRIs calculated at different positions (virtual bus and fault area bus), the fault location can be judged; by comparing the MRI of the virtual bus and the MRI of the bus in the fault area, the specific fault location can be determined. If the MRI value of the virtual bus is low, it means that the measurement result at the assumed fault position is more matched with the estimated result. Therefore, it can be considered that the fault is located in the line segment where the virtual bus is located. On the contrary, if the MRI value of the virtual bus is high, it means that the fault may not be located in this line segment, and the fault location is more likely to be the bus in the fault area; make a fault location decision based on the comparison result. If the MRI of the virtual bus is low, it is determined that the fault is located in this line segment; on the contrary, if the MRI of the virtual bus is high, it is determined that the fault is located in the bus in the fault area. This decision-making process can accurately narrow down the fault range and finally accurately locate the fault point in the distribution network; based on the comparison result, automatically select the most likely fault position, output this position and provide fault type information, for example, short circuit fault type.

[0107] Furthermore, performing state estimation on each line segment containing a virtual bus includes:

[0108] Modify the state vector of the distribution network, and use the voltage amplitude and injected current of the virtual bus as new state variables;

[0109] Calculate the current distribution of the line segment based on the injected current of the virtual bus, the source injected current of the fault area, and the real-time measurement data;

[0110] Iteratively update the state estimation result according to the current distribution and the voltage amplitude of the virtual bus until the convergence condition is met;

[0111] Among them, the injected current of the virtual bus is calculated from the conversion result of the phasor information and the pseudo-measurement value in the real-time measurement data.

[0112] As an optimization of the above embodiment, when performing state estimation on each line segment containing a virtual bus, it is first necessary to modify the state vector of the distribution network. Specifically, the voltage amplitude and injected current of the virtual bus will be added as new state variables to the state vector of the distribution network. The modified state vector includes the original distribution network state variables (such as voltage, current, etc.), and at the same time, the voltage amplitude and injected current of the virtual bus are used as new state variables. This modification enables the state estimation algorithm to simultaneously estimate the electrical state of the virtual bus and further optimize the fault location process; Next, it is necessary to calculate the current distribution of the line segment based on the injected current of the virtual bus, the source injected current of the fault area, and the real-time measurement data. The calculation of the current distribution helps to further understand the electrical characteristics near the fault point and assist in determining the fault location. During the calculation of the current distribution, first consider the injected current of the virtual bus and the source injected current from the fault area, and combine the real-time measurement data provided by the PMU to calculate the current distribution on each line segment; After calculating the current distribution, perform iterative update of the state estimation according to the current distribution and the voltage amplitude of the virtual bus. In each iteration, the voltage amplitude and injected current of the virtual bus will be gradually updated according to the current voltage and current estimation results until the convergence condition is reached; The convergence condition means that after multiple iterations, the estimation errors of the voltage and current are lower than the set threshold, or the update amount approaches zero, indicating that the state estimation result has reached an accurate level, thus completing the iteration; The injected current of the virtual bus is calculated from the conversion result of the phasor information and the pseudo-measurement value in the real-time measurement data. The injected current of the virtual bus is calculated through the real-time measurement data and the pseudo-measurement value to ensure that the injected current of the virtual bus is consistent with other real-time measurement data of the distribution network, thereby providing a more accurate state estimation result; Using the above calculation results, iteratively update the voltage amplitude and injected current of the virtual bus, and gradually optimize the state estimation result until the convergence condition is met. In each iteration process, the new injected current and voltage estimation values will be fed back into the state estimation process to optimize the accuracy of the fault location.

[0113] Furthermore, calculate the local measurement residual index, including:

[0114] Select the calculation range according to the topological structure type of the fault area;

[0115] If the fault area belongs to a mesh topology, calculate the local measurement residual index based on the residuals of all measurements in the distribution network;

[0116] If the fault area belongs to a radial topology, calculate the local measurement residual index based on the measurement residuals of the feeder where the bus of the fault area is located;

[0117] The local measurement residual index is the sum of the squares of all measurement residuals within the selected calculation range.

[0118] As a preference of the above embodiments, before calculating the local measurement residual index, it is first necessary to select a suitable calculation range according to the topological structure type of the fault area. The topological structure type determines the range of the distribution network that needs to be considered when calculating the residual. For different topological structures of the distribution network, the range of calculating the residual is different. If the fault area is a meshed topological structure, the measurement data of the entire distribution network needs to be covered during the calculation. If the fault area belongs to a radial topological structure, only the measurement data of the feeder where the fault area is located is considered during the calculation. If the fault area belongs to a meshed topological structure, the local measurement residual index will be calculated based on the residuals of all measurement values of the distribution network. The meshed topological structure has multiple electrical loops, and the measurement values in all loops will affect the overall state estimation result. Therefore, it is necessary to calculate the residuals of the entire network. Under the meshed topological structure, first collect the real-time measurement data provided by all PMU (phasor measurement unit) devices in the distribution network. According to the difference (i.e., the residual) between these data and the state estimation result, calculate the local measurement residual index. The specific method is to sum up the squared values of all measurement residuals to obtain the local measurement residual index. If the fault area belongs to a radial topological structure, the local measurement residual index is calculated only based on the measurement residuals of the feeder where the fault area is located. The radial topological structure usually has only a single path connecting each node. Therefore, only the measurement data of the feeder related to the fault area needs to be concerned. Under the radial topological structure, first determine the feeder connected to the fault area, collect the PMU measurement data of this feeder, calculate the measurement residuals of this feeder, and finally, obtain the local measurement residual index by summing the squared measurement residuals of the feeder. The local measurement residual index is obtained by summing the squared values of all measurement residuals within the selected calculation range. This calculation method ensures that the calculation result of the residual can reflect the electrical characteristics of the entire area, thereby helping to accurately locate the fault. Whether it is a meshed topology or a radial topology, the calculation formula of the local measurement residual index is the same, that is, square and sum all measurement residual values within the calculation range. In this way, a comprehensive residual index can be obtained, indicating the degree of difference between the measurement values and the estimated values within this area. The calculated local measurement residual index will help the system judge the specific location of the fault. In the case of a more complex network topology, the residuals of the meshed topological structure will be more complex, while the residuals of the radial topological structure will be more concentrated, and the fault point can be located more quickly. The local measurement residual index reflects the deviation between the electrical state and the estimated state of the fault area. The fault location can be determined by comparing the local measurement residual indexes within different calculation ranges, and finally accurate short-circuit fault location can be achieved.

[0119] Furthermore, the determination conditions for the meshed topology and the radial topology include:

[0120] If there are two or more independent power supply paths in the fault area connecting to the same power supply node, it is determined that the fault area belongs to the meshed topology;

[0121] If all the buses in the fault area are connected to the power source node through a single feeder, it is determined that the fault area belongs to the radial topology.

[0122] As a preference of the above embodiment, the meshed topology structure means that there are multiple independent power supply paths in the distribution network, and these paths can be interconnected to the same power source node to form an electrical loop. In the meshed topology structure, the distribution of current and voltage is affected by multiple independent power supply paths and usually has a high complexity. When there are two or more independent power supply paths connected to the same power source node in the fault area, it can be determined that the fault area belongs to the meshed topology. At this time, the topology of the distribution network has multiple intertwined electrical paths, and the current can flow through different paths, resulting in the complexity of the electrical state in the network. By analyzing the topology structure of the distribution network, all the power source nodes and connection paths are identified. If in the fault area, at least two different lines (feeders or power supply paths) are connected to the same power source node, then this area is considered to have a meshed topology structure. This determination can be achieved through a topology analysis algorithm or line measurement data; the radial topology structure means that all power sources are connected to each node in the distribution network through a single path. In this topology, the path of current flow is unidirectional, and all electrical loads are supplied by a single feeder, which makes this structure relatively simple and the fault analysis relatively easy. If all the buses in the fault area are connected to the same power source node through a single feeder, it can be determined that the fault area belongs to the radial topology. In the radial topology, each load point has only one power source path, which makes the current distribution and fault location more direct and clear. By analyzing the topology structure of the distribution network, it is judged whether all the buses in the fault area are connected to the power source node through a single feeder. If so, this area is considered to have a radial topology structure, and this determination can be achieved through the automatic detection of topology information and feeder analysis. An automatic topology determination mechanism is used to determine the topology type of the distribution network and, based on this, select an appropriate fault location method. If it is determined to be a meshed topology, the fault location will consider the measured values of the entire distribution network, while if it is determined to be a radial topology, the fault location will focus on the specific feeder in the fault area; first, the topology structure of the distribution network is analyzed in real time, the power source nodes and all connection paths are identified, and it is judged whether there are multiple independent power supply paths in the fault area or it is only connected to the power source node through a single feeder. According to the analysis results, the appropriate topology type is automatically selected, so as to optimize the accuracy and efficiency of short-circuit fault location.

[0123] Embodiment 2;

[0124] Based on the same inventive concept as the active distribution network short-circuit fault location method based on PMU state estimation in the foregoing embodiment, the present invention also provides an active distribution network short-circuit fault location system based on PMU state estimation. The system includes:

[0125] A data acquisition unit, which acquires the real-time measurement data of PMUs in the distribution network and the pseudo-measurement data before a fault.

[0126] A state estimation unit, which dynamically updates the pseudo-measurement values based on the real-time measurement data and the pseudo-measurement data according to the state estimation algorithm, and generates the real-time state estimation result of the distribution network.

[0127] A residual calculation unit, which divides the distribution network into multiple candidate regions according to the topological structure characteristics, performs state estimation on each candidate region, and calculates the measurement residual index of each candidate region.

[0128] A fault location unit, which determines the fault region according to the measurement residual index, performs state estimation on all buses in the fault region, calculates the local measurement residual index, and locates the specific fault bus or line segment.

[0129] The above adjustment system in the present invention can be effectively implemented, and the technical effects it can achieve are as described in the above embodiments, which will not be elaborated here.

[0130] Embodiment III;

[0131] Based on the same inventive concept as the active distribution network short-circuit fault location method based on PMU state estimation in the foregoing embodiments, the present invention also provides an active distribution network short-circuit fault location device based on PMU state estimation, which is used to implement the active distribution network short-circuit fault location method based on PMU state estimation.

[0132] The above device in the present invention can effectively implement the active distribution network short-circuit fault location method based on PMU state estimation, and the technical effects it can achieve are as described in the above embodiments, which will not be elaborated here.

[0133] Although the present application has been described in combination with specific features and their embodiments, it is obvious that various modifications and combinations can be made without departing from the spirit and scope of the present application. Accordingly, this specification and the drawings are merely exemplary descriptions of the present application defined by the appended claims, and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of the present application. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is intended to include these changes and modifications.

Claims

1. A short-circuit fault location method for active distribution networks based on PMU state estimation, characterized in that The method includes: Collecting real-time measurement data of PMUs in the distribution network and pseudo-measurement data before a fault; Based on the real-time measurement data and pseudo-measurement data, dynamically updating the pseudo-measurement values according to a state estimation algorithm to generate a real-time state estimation result of the distribution network; Dividing the distribution network into multiple candidate regions according to topological structure characteristics, performing state estimation on each candidate region, and calculating the measurement residual index of each candidate region; Determining the fault region according to the measurement residual index, performing the state estimation on all buses in the fault region, calculating the local measurement residual index, and locating the specific fault bus or line segment.

2. The active distribution network short-circuit fault location method based on PMU state estimation according to claim 1, wherein, Dynamically updating the pseudo-measurement values according to a state estimation algorithm, including: In each iteration of the state estimation, updating the pseudo-measurement values according to the voltage magnitude estimation result of the current iteration and the final voltage magnitude of the state estimation before the fault; If the voltage magnitude estimation result of a load bus is greater than or equal to 0.85 per unit value, updating the pseudo-measurement value of the bus based on the ZIP model; If the voltage magnitude estimation result of the load bus is less than 0.85 per unit value, updating the pseudo-measurement value of the bus based on the constant impedance model in the ZIP model; Calculating the injection current at the fault location according to the updated pseudo-measurement values, source injection current, and the real-time measurement data, and correcting the iteration result of the state estimation.

3. The method for locating short-circuit faults in an active distribution network based on PMU state estimation according to claim 2, wherein, Updating the pseudo-measurement value of the bus based on the ZIP model, including: Obtaining the historical state estimation result of the bus before the fault, and extracting the correlation data between the power consumption of the bus and the voltage magnitude estimation result; Based on the correlation data, constructing an optimization objective function of the ZIP model by weighted least squares method; Solving the optimization objective function to determine the weight coefficients of the constant impedance, constant current, and constant power components in the ZIP model; Dynamically calculating the pseudo-measurement value of the bus according to the weight coefficients and the voltage magnitude estimation result of the current iteration.

4. The method for locating short - circuit faults in an active distribution network based on PMU state estimation according to claim 3, wherein, Constructing an optimization objective function of the ZIP model by weighted least squares method, including: Obtaining the historical state estimation results of the bus under at least four voltage magnitude estimation results before the fault, and each historical state estimation result includes the voltage magnitude estimation result of the bus and the corresponding active power estimation value and reactive power estimation value; According to the power expression of the ZIP model, generating a theoretical power calculation value of the bus under the voltage magnitude estimation result, and the theoretical power calculation value is the weighted sum of the constant impedance component, constant current component, and constant power component; Taking the minimization of the sum of squared errors between the theoretical power calculation value and the power estimation value in the historical state estimation result as the optimization objective, and constructing the optimization objective function.

5. The active distribution network short-circuit fault location method based on PMU state estimation according to claim 1, characterized in that Locating the specific fault bus or line segment, including: Inserting virtual buses on each line segment connecting to the buses in the fault region to simulate the fault location; Performing the state estimation on each line segment containing the virtual bus, and calculating the corresponding local measurement residual index; Compare the local measurement residual index of the virtual bus with the local measurement residual index of the bus in the fault area; If the local measurement residual index of the virtual bus is low, it is determined that the fault is located in the corresponding line segment, otherwise it is determined that the fault is located in the bus of the fault area.

6. The method for locating short - circuit faults in an active distribution network based on PMU state estimation according to claim 5, wherein, Perform the state estimation for each line segment containing the virtual bus, including: Modify the state vector of the distribution network, and use the voltage amplitude and injected current of the virtual bus as new state variables; Based on the injected current of the virtual bus, the source injected current in the fault area, and the real-time measurement data, calculate the current distribution of the line segment; According to the current distribution and the voltage amplitude of the virtual bus, iteratively update the state estimation result until the convergence condition is met; Among them, the injected current of the virtual bus is calculated through the conversion result of the phasor information and the pseudo-measurement value in the real-time measurement data.

7. The method for locating short - circuit faults in an active distribution network based on PMU state estimation according to claim 1, wherein, Calculate the local measurement residual index, including: Select the calculation range according to the topological structure type of the fault area; If the fault area belongs to the mesh topology, calculate the local measurement residual index based on the residuals of all measurement values of the distribution network; If the fault area belongs to the radial topology, calculate the local measurement residual index based on the measurement residuals of the feeder where the bus in the fault area is located; The local measurement residual index is the sum of the squares of all measurement residuals within the selected calculation range.

8. The method for locating short-circuit faults in an active distribution network based on PMU state estimation according to claim 7, characterized in that, The determination conditions for the mesh topology and the radial topology include: If there are two or more independent power supply paths in the fault area connecting to the same power supply node, it is determined that the fault area belongs to the mesh topology; If all buses in the fault area are connected to the power supply node through a single feeder, it is determined that the fault area belongs to the radial topology.

9. An active distribution network short-circuit fault location system based on PMU state estimation, characterized in that, The system includes: A data acquisition unit that acquires the real-time measurement data of PMUs in the distribution network and the pseudo-measurement data before the fault; A state estimation unit that, based on the real-time measurement data and the pseudo-measurement data, dynamically updates the pseudo-measurement values according to the state estimation algorithm to generate the real-time state estimation result of the distribution network; A residual calculation unit that divides the distribution network into multiple candidate areas according to the topological structure characteristics, performs state estimation on each candidate area, and calculates the measurement residual index of each candidate area; A fault location unit that determines the fault area according to the measurement residual index, performs state estimation on all buses in the fault area, calculates the local measurement residual index, and locates the specific fault bus or line segment.

10. An active distribution network short-circuit fault location device based on PMU state estimation, which is used to implement any one of the methods in claims 1-8.

Images