Wind power plant current collection line fault positioning method and device, medium and equipment

Abstract

The invention discloses a wind power plant current collection line fault positioning method and device, a medium and equipment. The method comprises the following steps: replacing ridge regression of an output layer of a next-generation reserve pool calculation algorithm with softmax regression to obtain an improved next-generation reserve pool calculation algorithm; constructing a fault interval positioning model based on an improved next-generation reserve pool calculation algorithm; according to a preset starting condition, carrying out single-phase earth fault identification by utilizing the voltage of the voltage of the booster station bus of the wind power plant to be positioned and the effective value of the zero-sequence voltage, and determining a single-phase earth fault; and inputting the electrical quantity parameter of the single-phase earth fault into the fault interval positioning model, and outputting the current collection line fault interval of the wind power plant to be positioned.

Description

Technical Field

[0001] The present invention relates to the technical field of transmission line single-phase grounding fault interval positioning, and more specifically, to a wind farm collector fault positioning method, device, medium and equipment. Background Art

[0002] Wind power generation, as one of the most mature forms of renewable energy generation in my country, is bound to usher in further rapid development. However, most of my country's wind farms are located in areas with complex terrain and harsh environment, and the probability of short-circuit faults in the collector lines is high. Moreover, it is difficult to find the fault point after the line trips, and power transmission cannot be restored in time, which leads to a long period of wind abandonment and power congestion, seriously affecting the benefits of wind farms and the utilization of wind energy resources. At the same time, given that single-phase grounding faults are the most common fault form in the power system, it is very necessary to design a special fault location solution for single-phase grounding faults in wind farm collector lines.

[0003] The wind farm collector is used to collect the electric energy output by the wind turbines. Each wind turbine is connected to the collector through a box transformer. The voltage level of the collector is generally 35kV. Its head end is connected to the busbar of the booster station. The transmitted electric energy is sent to the system after the voltage level is converted to 110kV or 220kV by the main transformer in the station. The length of the collector is short, generally not more than 20km, and a large number of wind turbines will be connected to a collector, so the distance between wind turbines is very short, about a few hundred meters. These structural features make the collector present the characteristics of dense connection of multiple power sources. Because the wind turbine contains power electronic devices inside, its operating characteristics in the event of a fault are not clear, and the collector is densely connected with multiple power sources. These factors make the fault mechanism analysis of the collector complex and it is very difficult to locate the fault. The positioning technology used for transmission and distribution lines in the past is difficult to use. When the wind farm is actually in operation, fault location is mainly based on manual troubleshooting, and because some fault points are difficult to find by naked eye observation, experiments are often required at this time. Therefore, in current projects, positioning work will consume a lot of manpower, material resources and time. Although fault location of power lines is often carried out by fault distance measurement, for collector lines, since the distance between adjacent wind turbines is very short, considering that the error of fault distance measurement is generally around a few hundred meters, it is actually more meaningful to determine the fault interval (the line between the busbar and the wind turbine closest to the busbar or between two adjacent wind turbines is regarded as a line interval). Based on this judgment, some scholars have conducted some theoretical research. Some literatures have proposed to install zero-sequence current measurement points at both ends of each interval, collect zero-sequence current waveforms, and determine the fault interval by calculating the correlation of the waveforms at both ends. However, waveform collection requires a high sampling rate for the measuring device, and the need to upload current waveforms also makes the amount of information transmission very large. Other literatures have proposed to simulate the occurrence of faults in each interval in turn, obtain the calculated values ​​of electrical quantities at the busbar, grounding transformer, and outlets of each wind turbine through short-circuit calculation, and configure measurement points at the above positions of the actual wind farm to collect the actual values ​​of electrical quantities after the fault; determine the fault interval by measuring the difference between the calculated value and the actual value, but this scheme requires complex short-circuit calculation, and the judgment effect of the scheme may be affected due to the accuracy of short-circuit calculation. In addition, the above solutions all require a large number of measurement points to be configured in the wind farm, and the number of measurement points even exceeds the number of wind turbines. Therefore, the economic cost of applying the solution is too high.

[0004] In general, whether it is engineering application or theoretical research, it is still difficult to locate the fault of wind farm collector lines. There are the following technical problems: 1) In actual engineering, the fault location of wind farm collector lines is mainly based on manual investigation. Considering that some fault points are difficult to find with the naked eye, field tests are often required at this time, which makes the location of collector lines time-consuming and laborious. At the same time, the long-term shutdown of many wind turbines due to the failure of collector lines to be eliminated in time will greatly affect the benefits of wind farms. 2) The existing automated scheme for fault location of collector lines often requires a large number of measurement points when implemented, such as the need to install measurement points on each line section and at the outlet of each wind turbine. This will lead to high economic costs for fault location and difficulties in the actual application of the scheme. 3) The neural network algorithm has strong data mining and feature extraction capabilities, and the algorithm calculation results are accurate. It is the most widely used deep learning algorithm at present, but its structure is complex and there are a large number of parameters that need to be determined through training. In order to ensure the calculation effect, a fairly large number of training samples must be invested in model training, which greatly increases the training cost of the model. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides a method, device, medium and equipment for locating a collector fault in a wind farm.

[0006] According to one aspect of the present invention, a method for locating a collector fault in a wind farm is provided, comprising:

[0007] The ridge regression of the output layer of the next generation reserve pool calculation algorithm is replaced with softmax regression to obtain an improved next generation reserve pool calculation algorithm;

[0008] Construct a fault interval location model based on an improved next-generation reserve pool calculation algorithm;

[0009] According to the preset starting conditions, the bus phase voltage and zero-sequence voltage effective value of the booster station of the wind farm to be located are used to identify the single-phase grounding fault and determine the single-phase grounding fault;

[0010] The electrical quantity parameters of the single-phase grounding fault are input into the fault interval location model, and the collector fault interval of the wind farm to be located is output.

[0011] Optionally, a fault interval location model is constructed based on an improved next generation reserve pool calculation algorithm, including:

[0012] The input electrical quantity data and output of the improved next-generation reserve pool calculation algorithm are specified to form a training set, and the algorithm parameters are set for model training to generate a fault interval location model.

[0013] Optionally, the training set is constructed as follows:

[0014] Build a wind farm simulation model based on the structure and parameters of the actual wind farm;

[0015] In the wind farm simulation model, a single-phase grounding fault is set at a preset distance on each section of the collector line, where the fault types set for each single-phase grounding fault include: phase A grounding, phase B grounding and phase C grounding, and the transition resistance includes 0Ω, 50Ω, 100Ω, 150Ω, and 200Ω;

[0016] Perform simulation calculations for each single-phase grounding fault in the wind farm simulation model and collect electrical quantity data under each single-phase grounding fault condition;

[0017] The electrical quantity data obtained under each single-phase grounding fault condition and the corresponding fault interval number constitute a sample, and the set of all samples is used as a training set.

[0018] Optionally, the input electrical quantity data of the improved next-generation reserve pool calculation algorithm include the effective value and phase of the phase voltage and sequence voltage phasor at the bus of the substation, and the effective value and phase of the phase current and sequence current phasor at the beginning, end and midpoint of each collector line, and the output is the fault interval.

[0019] Optionally, the calculation formulas for the sequence voltage phasor and the sequence current phasor are:

[0020]

[0021] In the formula, The three-phase voltage phase quantity collected by the measuring device, The three-phase current phase quantities collected by the measuring device; are the positive, negative and zero sequence voltage phasors at the corresponding positions of the phase voltage, It is the positive sequence, negative sequence and zero sequence current phasor at the corresponding position of the phase current.

[0022] Optionally, the setting of algorithm parameters includes constructing the nonlinear part of the feature vector using a second-order polynomial in the nonlinear vector autoregression part; the number of output categories of the softmax regression is the total number of wind farm intervals, and the optimization algorithm adopted is the stochastic gradient descent algorithm, the number of iterations of the algorithm is 1000, and the learning rate is 0.01.

[0023] Optionally, the start condition includes a first start condition and a second start condition, and

[0024] According to the preset starting conditions, the bus phase voltage and zero-sequence voltage effective value of the booster station of the wind farm to be located are used to identify the single-phase grounding fault and determine the single-phase grounding fault, including:

[0025] Determine whether the zero-sequence voltage effective value satisfies the first starting condition, if so, determine that the wind farm to be located has an asymmetric grounding fault, otherwise determine that the wind farm to be located does not have an asymmetric grounding fault;

[0026] In the case that an asymmetric grounding fault exists in the wind farm to be located, it is determined whether the bus phase voltage of the substation of the wind farm to be located meets the second starting condition. If so, it is determined that a single-phase grounding fault exists in the wind farm to be located; otherwise, it is determined that no single-phase grounding fault exists in the wind farm to be located.

[0027] Optionally, the first starting condition is the zero-sequence voltage of the busbar of the booster station The effective value U m0 Greater than 15% of the effective value of the rated phase voltage of this bus.

[0028] Optionally, the second starting condition is that the effective value of the three-phase voltage exists U C A ≈U B or U B A ≈U C or U A B ≈U C The relationship between U A is the effective value of phase A voltage; U B is the effective value of the B phase voltage; U C is the effective value of the C phase voltage.

[0029] According to another aspect of the present invention, a wind farm collector fault location device is provided, comprising:

[0030] A replacement module is used to replace the ridge regression of the output layer of the next generation reserve pool calculation algorithm with softmax regression to obtain an improved next generation reserve pool calculation algorithm;

[0031] A building module for building a fault interval location model based on an improved next generation reserve pool calculation algorithm;

[0032] An identification module is used to identify a single-phase grounding fault by using the bus phase voltage and zero-sequence voltage effective value of the booster station of the wind farm to be located according to preset starting conditions, and determine the single-phase grounding fault;

[0033] The output module is used to input the electrical quantity parameters of the single-phase grounding fault into the fault interval positioning model, and output the collector fault interval of the wind farm to be located.

[0034] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the storage medium stores a computer program, and the computer program is used to execute the method described in any one of the above aspects of the present invention.

[0035] ​​​According to another aspect of the present invention, an electronic device is provided, comprising: a processor; a memory for storing instructions executable by the processor; the processor is configured to read the executable instructions from the memory and execute the instructions to implement the method described in any one of the above aspects of the present invention.

[0036] Therefore, the present invention provides a method for locating the fault of a wind farm collector line, which is based on a wind farm collector line fault locating scheme that improves the calculation of the next generation reserve pool. After the calculation of the next generation reserve pool is improved, a classification model is constructed using it, and the fault interval judgment is realized by classifying the fault interval. The fault interval can be automatically determined and located quickly without the need for manual investigation or on-site testing. It is only necessary to install a voltage measuring device on the busbar of the booster station and a current measuring device at the head end, the end end, and the midpoint of the collector line. Few measuring devices are required, and the measuring device does not require a high sampling rate and a small amount of measurement information transmission, which greatly reduces the economic cost of fault location. The parameters to be trained and the required sample size of the positioning model are significantly reduced, and the results are more interpretable. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] A more complete understanding of exemplary embodiments of the present invention may be obtained by referring to the following drawings:

[0038] Figure 1 It is a flow chart of a method for locating a collector line fault in a wind farm provided by an exemplary embodiment of the present invention;

[0039] Figure 2 is another schematic flow chart of a method for locating a wind farm collector fault provided by an exemplary embodiment of the present invention;

[0040] Figure 3 is a schematic structural diagram of a wind farm collector fault location device provided by an exemplary embodiment of the present invention;

[0041] Figure 4 This is a structure of an electronic device provided by an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0042] Below, the exemplary embodiments according to the present invention will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments of the present invention, and it should be understood that the present invention is not limited to the exemplary embodiments described here.

[0043] It should be noted that the relative arrangement of components and steps, the numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present invention unless specifically stated otherwise.

[0044] Those skilled in the art can understand that the terms "first" and "second" in the embodiments of the present invention are only used to distinguish different steps, devices or modules, etc., and neither represent any specific technical meaning nor indicate the necessary logical order between them.

[0045] It should also be understood that, in the embodiments of the present invention, “plurality” may refer to two or more than two, and “at least one” may refer to one, two or more than two.

[0046] It should also be understood that any component, data or structure mentioned in the embodiments of the present invention can generally be understood as one or more, unless explicitly limited or otherwise indicated in the context.

[0047] In addition, the term "and / or" in the present invention is only a description of the association relationship of the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in the present invention generally indicates that the associated objects before and after are in an "or" relationship.

[0048] It should also be understood that the description of the various embodiments of the present invention focuses on the differences between the various embodiments, and the same or similar aspects thereof can be referenced to each other, and for the sake of brevity, they will not be described one by one.

[0049] At the same time, it should be understood that for the convenience of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship.

[0050] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses.

[0051] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered part of the specification.

[0052] It should be noted that like reference numerals and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0053] Embodiments of the present invention can be applied to electronic devices such as terminal devices, computer systems, servers, etc., which can operate with many other general or special computing system environments or configurations. Examples of well-known terminal devices, computing systems, environments and / or configurations suitable for use with electronic devices such as terminal devices, computer systems, servers, etc. include, but are not limited to: personal computer systems, server computer systems, thin clients, thick clients, handheld or laptop devices, microprocessor-based systems, set-top boxes, programmable consumer electronics, network personal computers, small computer systems, large computer systems, and distributed cloud computing technology environments including any of the above systems, etc.

[0054] Electronic devices such as terminal devices, computer systems, servers, etc. can be described in the general context of computer system executable instructions (such as program modules) executed by computer systems. Generally, program modules can include routines, programs, object programs, components, logic, data structures, etc., which perform specific tasks or implement specific abstract data types. Computer systems / servers can be implemented in a distributed cloud computing environment, where tasks are performed by remote processing devices linked through a communication network. In a distributed cloud computing environment, program modules can be located on local or remote computing system storage media including storage devices.

[0055] Exemplary Methods

[0056] Figure 1 FIG. 1 is a flow chart of a method for locating a wind farm collector fault provided by an exemplary embodiment of the present invention. This embodiment can be applied to electronic devices, such as Figure 1 As shown, the wind farm collector fault location method 100 includes the following steps:

[0057] Step 101, replacing the ridge regression of the output layer of the next generation reserve pool calculation algorithm with softmax regression to obtain an improved next generation reserve pool calculation algorithm;

[0058] Step 102, constructing a fault interval location model based on an improved next generation reserve pool calculation algorithm;

[0059] Step 103, according to the preset starting conditions, the bus phase voltage and the effective value of the zero-sequence voltage of the booster station of the wind farm to be located are used to identify the single-phase grounding fault and determine the single-phase grounding fault;

[0060] Step 104: input the electrical quantity parameters of the single-phase grounding fault into the fault section location model, and output the collector fault section of the wind farm to be located.

[0061] Specifically, the present invention provides a wind farm collector line fault location solution based on improved next generation reserve pool calculation. After improving the next generation reserve pool calculation, a classification model is constructed therewith to realize fault interval judgment by classifying the fault interval.

[0062] The present invention proposes a wind farm collector fault location solution based on improved next generation reserve pool calculation, referring to Figure 2 As shown, the specific steps include:

[0063] Step 1: Improve the next generation reserve pool calculation algorithm. As a new deep learning algorithm, the next generation reserve pool calculation reduces training parameters and training sample requirements while retaining excellent feature extraction capabilities compared to traditional neural network algorithms. However, when this algorithm was proposed, it was aimed at regression prediction problems, so ridge regression was used in the algorithm output layer. The present invention models the fault interval positioning problem (the line between the busbar and the wind turbine closest to the busbar or the line between two adjacent wind turbines is regarded as a line interval) as a classification problem, and it is necessary to use the next generation reserve pool calculation to build a classification model. Therefore, the algorithm must be improved so that it can complete the classification function. In order to realize the classification function, the present invention modifies the output layer of the next generation reserve pool calculation and replaces the ridge regression with softmax regression.

[0064] Step 2: Build a fault interval location model based on the improved next-generation reserve pool calculation. The improved next-generation reserve pool calculation is a fault interval location model after input and output specification, parameter setting, and model training, which specifically includes:

[0065] Step 2-1: Specify the input and output of the improved next-generation reserve pool calculation. The input of the improved next-generation reserve pool calculation is the effective value and phase of the phase voltage (i.e., A phase voltage, B phase voltage, C phase voltage) and sequence voltage (i.e., positive sequence voltage, negative sequence voltage, zero sequence voltage) phasors at the busbar of the booster station, and the effective value and phase of the phase current (i.e., A phase current, B phase current, C phase current) and sequence current (i.e., positive sequence current, negative sequence current, zero sequence current) phasors at the beginning, end and midpoint of each collector line, and its output is the fault interval. All phase voltages and phase currents described in the present invention are collected by measuring devices, and all sequence voltages and sequence currents are calculated from phase voltages and phase currents. The calculation formula is as follows:

[0066]

[0067] In the formula, The three-phase voltage phase quantity collected by the measuring device, The three-phase current phase quantities collected by the measuring device; are the positive, negative and zero sequence voltage phasors at the corresponding positions of the phase voltage, are the positive-sequence, negative-sequence, and zero-sequence current phasors at the corresponding positions of the phase current;

[0068] Step 2-2: Set the parameters for improving the next generation reserve pool calculation. The parameter values ​​are as follows: In the nonlinear vector autoregression part, a second-order polynomial is used to construct the nonlinear part of the feature vector; the number of output categories of the softmax regression is the total number of wind farm intervals, and the optimization algorithm used is the stochastic gradient descent algorithm, the number of iterations of the algorithm is 1000, and the learning rate is 0.01.

[0069] Step 2-3: Train the improved next generation reserve pool calculation. First, form a training set, and then use the training set to determine the internal parameters of the improved next generation reserve pool calculation and fix the algorithm structure. The training set is formed by simulation, specifically:

[0070] The simulation model is built based on the structure and parameters of the actual wind farm, and then a single-phase grounding fault can be set every 100m in each interval (users can set it according to their needs, and it is not limited here). The fault types set for each fault include A-phase grounding, B-phase grounding, and C-phase grounding, and the transition resistance includes 0Ω, 50Ω, 100Ω, 150Ω, and 200Ω. Simulate all the above fault conditions and collect the electrical quantities described in step 21 under each fault condition. The electrical quantities obtained under each fault condition and the corresponding fault interval number together constitute a sample, and the set of all samples is the training set.

[0071] Step 3: Identify single-phase grounding fault. When the identification is successful, execute step 4. Obtain the phase voltage and zero-sequence voltage effective value of the booster station bus voltage (the acquisition method is the same as step 21) to realize the single-phase grounding fault identification, which specifically includes:

[0072] Step 3-1: Use the zero-sequence voltage of the booster station bus to determine whether the start condition 1 is met. When the start condition 1 is met, it indicates that an asymmetric grounding fault occurs in the wind farm, and then proceed to step 32 to further determine whether the asymmetric grounding fault is a single-phase grounding fault. Start condition 1 refers to the zero-sequence voltage of the booster station bus. The effective value Um0 is greater than 15% of the effective value of the rated phase voltage of this bus.

[0073] The principle of the above starting condition 1 is:

[0074] Wind farms generally use small resistance grounding, or the collector side of the main transformer uses star connection, directly leads the neutral point to ground through a small resistor, or uses a grounding transformer to create a neutral point grounded through a small resistor. When an asymmetric grounding short circuit occurs in the collector, a complete zero-sequence loop (fault point-collector-main transformer or grounding transformer) will be formed inside the wind farm, causing zero-sequence voltage to appear on the bus of the booster station, which does not exist in normal conditions or when other short-circuit faults occur. Therefore, when a large zero-sequence voltage appears on the bus, it can be considered that an asymmetric grounding short circuit has occurred in the collector.

[0075] Step 3-2: From the effective value of the bus voltage U A , U B , U C Determine whether the start condition 2 is met. When the start condition 2 is met, it indicates that the current asymmetric ground fault is a single-phase ground fault, and thus the identification is successful. Start condition 2 means that the effective values ​​of two phase voltages are close and greater than the effective value of the third phase voltage, that is, the effective value of the three-phase voltage exists U C A ≈U B or U B A ≈U C or U A B ≈U C The specific judgment method is: for the effective value U A , U B , U C Sort by size, and the minimum value is recorded as U in The maximum value is denoted as U max , and the other effective value is recorded as U mid , and then make the ratio U max / U min , U mid / U min , U max / U mid , if U max / U min , U mid / U min are greater than 1.1, and U max / U mid Less than 1.1, indicating that U A , U B , U C There exists U C A ≈U B or U B A ≈U C or U A B ​​​​​​≈U C relationship, otherwise it does not exist.

[0076] The principle of the above starting condition 2 is:

[0077] The collector grounding fault will cause the fault phase voltage to drop at the fault point, and because the collector is short (generally within 20km), the voltage of this phase of the booster station bus will also drop. Therefore, the effective value of the bus voltage can be collected to find the phase with the largest voltage drop, which is the fault phase. When there is only one fault phase, it indicates that the current fault is a single-phase grounding fault.

[0078] Step 4: Obtain the effective value and phase of the phase voltage and sequence voltage phasor at the busbar of the booster station in real time, as well as the effective value and phase of the phase current and sequence current phasor at the beginning, end and midpoint of each collector line, and input them into the fault interval location model, and obtain the fault interval from the fault interval location model.

[0079] The principle of using the above fault interval location model to realize fault interval judgment is:

[0080] When a single-phase grounding fault occurs in a certain section of the collector line, the phase voltage of the busbar of the booster station and the phase current of each point on the collector line will change, and the change amount of the phase voltage and phase current at each point is different for different fault sections. At the same time, since the wind farm is no longer in symmetrical operation after the fault, the negative sequence voltage, zero sequence voltage, and negative sequence current at various locations in the field are no longer 0, and the zero sequence current at some locations is no longer 0; the negative sequence and zero sequence components of voltage and current are fault components, and their distribution in the wind farm is closely related to the fault location. However, since the fault characteristics of wind turbines are very complex and there are many particularities in the structure of wind farms, it is difficult to establish the connection between these electrical quantities and the fault section by using the fault analysis method. Therefore, the phase voltage and sequence voltage of the busbar of the booster station under various fault conditions, as well as the phase current and sequence current at the head end, end end, and midpoint of the collector line can be obtained to form a training set, and a deep learning algorithm is used to mine the mapping relationship between the above electrical quantities and the fault section from the training set, thereby obtaining a fault section positioning model. The present invention uses the next generation reserve pool calculation to achieve the above purpose. Compared with general deep learning algorithms, reserve pool calculations require fewer training samples and use linear optimization, which requires less computing resources. At the same time, compared with previous reserve pool calculations, the next-generation reserve pool calculations used in the present invention do not require random matrices, require fewer parameters to be adjusted, greatly improve the calculation speed, and provide better interpretability of the results.

[0081] Step 5: Reset the plan and return to step 3.

[0082] Therefore, the present invention has the following beneficial effects:

[0083] (1) The present invention can automatically determine the fault range and locate it quickly without the need for manual investigation or on-site testing.

[0084] The present invention can establish an "end-to-end" collector fault interval location model by improving the next generation reserve pool calculation algorithm. After collecting the electrical quantity information of a small number of measuring points on the collector line, the model can be directly input to output the fault interval and automatically determine the fault interval.

[0085] (2) The present invention only requires the installation of a voltage measuring device on the busbar of the substation and the installation of a current measuring device at the head end, the end end, and the midpoint of the collector line. A small number of measuring devices are required, and the measuring devices do not require a high sampling rate and the amount of measurement information transmitted is small, which greatly reduces the economic cost of fault location.

[0086] Deep learning algorithms have powerful feature extraction and data mining capabilities, and can make full use of limited information to obtain fault features. Therefore, the present invention designs a positioning solution based on a deep learning algorithm. It only needs to set up a small number of measuring points at key locations to collect information, so as to establish a connection between the input information and the fault interval, and realize fault location with a small number of measuring points. At the same time, the measuring points only need to obtain the effective value and phase of the voltage and current, so there is no need to perform high-frequency sampling on the signal, and the amount of information transmission is also small.

[0087] (3) The present invention can significantly reduce the parameters to be trained and the required sample size of the positioning model, and the results are more interpretable.

[0088] Compared with traditional neural networks, the next generation of reserve pool calculation algorithm has a simpler structure and fewer parameters to be trained, so the number of samples required for model training is greatly reduced, which can greatly reduce the training cost of the positioning model. The present invention introduces this algorithm to realize the construction of the positioning model, and improves it so that it can adapt to the solution of classification problems, which can effectively avoid the problems of traditional neural networks with many parameters to be trained and large sample requirements. At the same time, the next generation of reserve pool calculation is based on nonlinear vector autoregression, and its interpretability is better than that of traditional neural networks.

[0089] Exemplary Devices

[0090] Figure 3 FIG. 1 is a schematic diagram of a wind farm collector fault location device provided by an exemplary embodiment of the present invention. Figure 3 As shown, the device 300 includes:

[0091] A replacement module 310 is used to replace the ridge regression of the output layer of the next generation reserve pool calculation algorithm with a softmax regression to obtain an improved next generation reserve pool calculation algorithm;

[0092] A construction module 320 is used to construct a fault interval location model based on the improved next generation reserve pool calculation algorithm;

[0093] The identification module 330 is used to identify the single-phase grounding fault by using the bus phase voltage and zero-sequence voltage effective value of the booster station of the wind farm to be located according to the preset starting conditions, and determine the single-phase grounding fault;

[0094] The output module 340 is used to input the electrical quantity parameters of the single-phase grounding fault into the fault interval positioning model, and output the collector line fault interval of the wind farm to be located.

[0095] Optionally, the construction module 320 includes:

[0096] The generation submodule is used to specify the input electrical quantity data and output of the improved next-generation reserve pool calculation algorithm to form a training set and set algorithm parameters for model training to generate the fault interval location model.

[0097] Optionally, the training set is formed as follows:

[0098] Build a wind farm simulation model based on the structure and parameters of the actual wind farm;

[0099] In the wind farm simulation model, a single-phase grounding fault is set at a preset distance on each section of the collector line, wherein the fault type set for each single-phase grounding fault includes: phase A grounding, phase B grounding and phase C grounding, and the transition resistance includes 0Ω, 50Ω, 100Ω, 150Ω, and 200Ω;

[0100] Performing simulation calculation for each single-phase grounding fault in the wind farm simulation model, and collecting electrical quantity data under each single-phase grounding fault condition;

[0101] The electrical quantity data obtained under each single-phase grounding fault condition and the corresponding fault interval number constitute a sample, and the set of all samples is used as the training set.

[0102] Optionally, the input electrical quantity data of the improved next-generation reserve pool calculation algorithm include the effective value and phase of the phase voltage and sequence voltage phasor at the bus of the substation, and the effective value and phase of the phase current and sequence current phasor at the beginning, end and midpoint of each collector line, and the output is the fault interval.

[0103] Optionally, the calculation formulas of the sequence voltage phasor and the sequence current phasor are:

[0104]

[0105] In the formula, The three-phase voltage phase quantity collected by the measuring device, The three-phase current phase quantities collected by the measuring device; are the positive, negative and zero sequence voltage phasors at the corresponding positions of the phase voltage, It is the positive sequence, negative sequence and zero sequence current phasor at the corresponding position of the phase current.

[0106] Optionally, the setting of algorithm parameters includes constructing the nonlinear part of the feature vector using a second-order polynomial in the nonlinear vector autoregression part; the number of output categories of the softmax regression is the total number of wind farm intervals, and the optimization algorithm adopted is the stochastic gradient descent algorithm, the number of iterations of the algorithm is 1000, and the learning rate is 0.01.

[0107] Optionally, the start condition includes a first start condition and a second start condition, and

[0108] The identification module 330 includes:

[0109] A first judgment submodule is used to judge whether the zero-sequence voltage effective value satisfies the first starting condition, and if so, it is judged that an asymmetric grounding fault exists in the wind farm to be located, otherwise, it is judged that an asymmetric grounding fault does not exist in the wind farm to be located;

[0110] The second judgment submodule is used to judge whether the bus phase voltage of the substation of the wind farm to be located meets the second starting condition when an asymmetric grounding fault exists in the wind farm to be located; if so, it is determined that a single-phase grounding fault exists in the wind farm to be located; otherwise, it is determined that no single-phase grounding fault exists in the wind farm to be located.

[0111] Optionally, the first starting condition is the zero-sequence voltage of the busbar of the booster station The effective value U m0 Greater than 15% of the effective value of the rated phase voltage of this bus.

[0112] Optionally, the second starting condition is that the effective value of the three-phase voltage exists C A ≈U B or U B A ≈U C or U A B ≈U C The relationship between U A is the effective value of phase A voltage; U B is the effective value of the B phase voltage; U C is the effective value of the C phase voltage.

[0113] Exemplary Electronic Devices

[0114] Figure 4 ​​​This is a structure of an electronic device provided by an exemplary embodiment of the present invention. Figure 4 As shown, the electronic device 40 includes one or more processors 41 and a memory 42 .

[0115] The processor 41 may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device to perform desired functions.

[0116] The memory 42 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory (cache), etc. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 41 may run the program instructions to implement the methods of the software programs of the various embodiments of the present invention described above and / or other desired functions. In one example, the electronic device may also include: an input device 43 and an output device 44, which are interconnected via a bus system and / or other forms of connection mechanisms (not shown).

[0117] In addition, the input device 43 may also include, for example, a keyboard, a mouse, etc.

[0118] The output device 44 can output various information to the outside, and can include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto.

[0119] Of course, to simplify, Figure 4 Only some of the components related to the present invention in the electronic device are shown, and components such as a bus, an input / output interface, etc. are omitted. In addition, the electronic device may further include any other appropriate components according to specific application conditions.

[0120] Exemplary computer program products and computer-readable storage media

[0121] In addition to the above-mentioned methods and devices, an embodiment of the present invention may also be a computer program product, which includes computer program instructions, which, when executed by a processor, enable the processor to execute the steps of the method according to various embodiments of the present invention described in the above-mentioned "Exemplary Method" section of this specification.

[0122] The computer program product may be written in any combination of one or more programming languages ​​to write program code for performing the operations of the embodiments of the present invention, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0123] In addition, an embodiment of the present invention may also be a computer-readable storage medium having computer program instructions stored thereon, which, when executed by a processor, enable the processor to execute the steps of the method according to various embodiments of the present invention described in the above “Exemplary Method” section of this specification.

[0124] The computer readable storage medium can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can include, for example, but is not limited to, a system, system or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0125] The basic principle of the present invention is described above in conjunction with specific embodiments. However, it should be pointed out that the advantages, strengths, effects, etc. mentioned in the present invention are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. must be possessed by each embodiment of the present invention. In addition, the specific details disclosed above are only for the purpose of illustration and facilitation of understanding, rather than limitation, and the above details do not limit the present invention to being implemented by adopting the above specific details.

[0126] Each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system embodiment, since it basically corresponds to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0127] The block diagrams of the devices, systems, equipment, and systems involved in the present invention are only illustrative examples and are not intended to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagram. As will be appreciated by those skilled in the art, these devices, systems, equipment, and systems can be connected, arranged, and configured in any manner. Words such as "including", "comprising", "having", etc. are open words, referring to "including but not limited to", and can be used interchangeably with them. The words "or" and "and" used here refer to the words "and / or" and can be used interchangeably with them, unless the context clearly indicates otherwise. The word "such as" used here refers to the phrase "such as but not limited to", and can be used interchangeably with it.

[0128] The method and system of the present invention may be implemented in many ways. For example, the method and system of the present invention may be implemented by software, hardware, firmware or any combination of software, hardware, firmware. The above order of steps for the method is only for illustration, and the steps of the method of the present invention are not limited to the order specifically described above, unless otherwise specifically stated. In addition, in some embodiments, the present invention may also be implemented as a program recorded in a recording medium, which includes machine-readable instructions for implementing the method according to the present invention. Thus, the present invention also covers a recording medium storing a program for executing the method according to the present invention.

[0129] It should also be noted that in the system, device and method of the present invention, each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. The above description of the disclosed aspects is provided to enable any technician in the field to make or use the present invention. Various modifications to these aspects are very obvious to those skilled in the art, and the general principles defined here can be applied to other aspects without departing from the scope of the present invention. Therefore, the present invention is not intended to be limited to the aspects shown here, but in accordance with the widest range consistent with the principles and novel features disclosed here.

[0130] The above description has been given for the purpose of illustration and description. In addition, this description is not intended to limit the embodiments of the present invention to the forms disclosed herein. Although a number of example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, additions and sub-combinations thereof.

Claims

1. A method for locating a collector fault in a wind farm, characterized in that: include: The ridge regression of the output layer of the next generation reserve pool calculation algorithm is replaced with softmax regression to obtain an improved next generation reserve pool calculation algorithm; Constructing a fault interval location model based on the improved next generation reserve pool calculation algorithm; According to the preset starting conditions, the bus phase voltage and zero-sequence voltage effective value of the booster station of the wind farm to be located are used to identify the single-phase grounding fault and determine the single-phase grounding fault; The electrical quantity parameters of the single-phase grounding fault are input into the fault section location model, and the collector fault section of the wind farm to be located is output.

2. The method according to claim 1, characterized in that A fault interval location model is constructed based on the improved next generation reserve pool calculation algorithm, including: The input electrical quantity data and output of the improved next-generation reserve pool calculation algorithm are designated to form a training set and algorithm parameters are set to perform model training to generate the fault interval location model.

3. The method according to claim 2, characterized in that The training set is constructed as follows: Build a wind farm simulation model based on the structure and parameters of the actual wind farm; In the wind farm simulation model, a single-phase grounding fault is set at a preset distance on each section of the collector line, wherein the fault type set for each single-phase grounding fault includes: phase A grounding, phase B grounding and phase C grounding, and the transition resistance includes 0Ω, 50Ω, 100Ω, 150Ω, and 200Ω; Performing simulation calculation for each single-phase grounding fault in the wind farm simulation model, and collecting electrical quantity data under each single-phase grounding fault condition; The electrical quantity data obtained under each single-phase grounding fault condition and the corresponding fault interval number constitute a sample, and the set of all samples is used as the training set.

4. The method according to claim 2, characterized in that: The input electrical quantity data of the improved next-generation reserve pool calculation algorithm include the effective value and phase of the phase voltage and sequence voltage phasor at the bus of the substation, the effective value and phase of the phase current and sequence current phasor at the beginning, end and midpoint of each collector line, and the output is the fault interval.

5. The method according to claim 4, characterized in that The calculation formulas for the sequence voltage phasor and the sequence current phasor are: In the formula, The three-phase voltage phase quantity collected by the measuring device is: The three-phase current phase quantities collected by the measuring device; are the positive, negative and zero sequence voltage phasors at the corresponding positions of the phase voltage, It is the positive sequence, negative sequence and zero sequence current phasor at the corresponding position of the phase current.

6. The method according to claim 2, characterized in that The setting of algorithm parameters includes the use of second-order polynomial to construct the nonlinear part of the feature vector in the nonlinear vector autoregression part; the number of output categories of softmax regression is the total number of wind farm intervals, and the optimization algorithm used is the stochastic gradient descent algorithm with 1000 iterations and a learning rate of 0.

01.

7. The method according to claim 1, characterized in that The start condition includes a first start condition and a second start condition, and According to the preset starting conditions, the bus phase voltage and zero-sequence voltage effective value of the booster station of the wind farm to be located are used to identify the single-phase grounding fault and determine the single-phase grounding fault, including: Determining whether the zero-sequence voltage effective value satisfies the first starting condition, and if so, determining that an asymmetric grounding fault exists in the wind farm to be located, and otherwise determining that an asymmetric grounding fault does not exist in the wind farm to be located; In the case that an asymmetric grounding fault exists in the wind farm to be located, it is determined whether the bus phase voltage of the substation of the wind farm to be located meets the second starting condition. If so, it is determined that a single-phase grounding fault exists in the wind farm to be located; otherwise, it is determined that no single-phase grounding fault exists in the wind farm to be located.

8. The method according to claim 7, characterized in that The first starting condition is the zero-sequence voltage U m0 The effective value U m0 Greater than 15% of the effective value of the rated phase voltage of this bus.

9. The method according to claim 7, characterized in that: The second starting condition is that the effective value of the three-phase voltage exists U C A ≈U B or U B A ≈U C or U A B ≈U C The relationship between U A is the effective value of phase A voltage; U B is the effective value of the B phase voltage; U C is the effective value of the C phase voltage.​​​ 10. A wind farm collector fault location device, characterized in that: include: A replacement module is used to replace the ridge regression of the output layer of the next generation reserve pool calculation algorithm with softmax regression to obtain an improved next generation reserve pool calculation algorithm; A construction module, used to construct a fault interval location model based on the improved next generation reserve pool calculation algorithm; An identification module is used to identify a single-phase grounding fault by using the bus phase voltage and zero-sequence voltage effective value of the booster station of the wind farm to be located according to preset starting conditions, and determine the single-phase grounding fault; The output module is used to input the electrical quantity parameters of the single-phase grounding fault into the fault interval positioning model, and output the collector line fault interval of the wind farm to be located.

11. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and the computer program is used to execute the method according to any one of claims 1 to 9.

12. An electronic device, characterized in that: The electronic device comprises: processor; a memory for storing instructions executable by the processor; The processor is used to read the executable instructions from the memory and execute the instructions to implement the method described in any one of claims 1 to 9.

Images