Wind storage power generation system black start evaluation method, computer device and related apparatus

Abstract

The present application relates to the technical field of electric power, and discloses a kind of wind storage power generation system black start evaluation method, computer equipment and related devices, the wind storage power generation system black start evaluation method includes: obtaining black start simulation dataset, black start simulation dataset includes the normal simulation data of wind storage power generation system normal time and the fault simulation data of wind storage power generation system fault time;Initial SVDD model is trained using black start simulation dataset, respectively training obtains the first SVDD model of description normal simulation data and the second SVDD model of description fault simulation data, and obtaining the first ball center distance of first SVDD model and the second ball center distance of second SVDD model;The data to be detected of wind storage power generation system is input into first SVDD model and second SVDD model, and target distance is obtained;The state of wind storage power generation system is evaluated based on target distance, first ball center distance and second ball center distance.The present application can make full use of simulation data, improve the accuracy of black start evaluation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric power, in particular to a wind storage power generation system black start evaluation method, a computer device and related devices. BACKGROUND

[0002] In recent years, the large-scale grid connection of new energy has caused the inertia capacity of the power grid to decline, and the power system faces greater power outage risks. Black start is the first step to restore power supply after a large-scale power outage, and more reliable black start power sources are needed to restore power supply as soon as possible. To improve the starting capacity of the system, the possibility of wind storage power generation systems as black start power sources is considered. Black start power sources need to provide stable power for the units to be started and establish stable voltage and frequency during a large-scale power outage. Therefore, during the black start process of the wind storage power generation system, the state of the black start needs to be monitored at all times to support the control of the black start and ensure the safety, stability and reliability of the black start.

[0003] Therefore, the wind storage power generation system needs to be evaluated before use to determine whether it has black start capability. The prior art determines the control limits of different parameters through overvoltage, overcurrent protection, frequency fault protection or statistical methods, and performs state evaluation according to different comparison results, but a single evaluation method leads to inaccurate evaluation results. SUMMARY

[0004] Therefore, the present application provides a wind storage power generation system black start evaluation method, a computer device and related devices, which can make full use of simulation data to improve the accuracy of black start evaluation.

[0005] In a first aspect, the present application provides a wind storage power generation system black start evaluation method, which comprises: obtaining a black start simulation data set, the black start simulation data set comprising normal simulation data of the wind storage power generation system in normal times and fault simulation data of the wind storage power generation system in fault times; training an initial SVDD model using the black start simulation data set, to obtain a first SVDD model describing the normal simulation data and a second SVDD model describing the fault simulation data, and to obtain a first ball center distance of the first SVDD model and a second ball center distance of the second SVDD model; inputting the to-be-detected data of the wind storage power generation system into the first SVDD model and the second SVDD model to obtain a target distance; and evaluating the state of the wind storage power generation system based on the target distance, the first ball center distance and the second ball center distance.

[0006] In the implementation, the application respectively simulates the black start of the wind storage power generation system in normal and fault conditions, and collects corresponding simulation data. The normal simulation data and the fault simulation data are used to train the SVDD model twice, and the first SVDD model describing the normal simulation data and the second SVDD model describing the fault simulation data are obtained. Since the fault simulation data is less, the double-layer SVDD model can make full use of the normal simulation data and the fault simulation data information in the case of unbalanced normal simulation data and fault simulation data, and improve the classification accuracy of the SVDD model. On this basis, the target distance of the to-be-detected data and the SVDD model is compared with the first ball center distance and the second ball center distance to obtain the state information of the wind storage power generation system, which can further improve the accuracy of the black start evaluation.

[0007] In an optional implementation, acquiring the black start simulation data set includes: starting the black start simulation of the wind storage power generation system under normal working conditions, and acquiring normal simulation current data and voltage data at a plurality of preset positions; starting the black start simulation of the wind storage power generation system under a plurality of working condition faults, and acquiring fault simulation current data and voltage data at a plurality of preset positions for each working condition fault.

[0008] In the implementation, a plurality of data acquisition points are set, and a plurality of working condition faults are simulated to respectively acquire fault simulation data, which can fully consider a plurality of scenes in actual application, enrich the black start simulation data set, and further improve the classification and evaluation accuracy of the subsequent SVDD model.

[0009] In an optional implementation, the initial SVDD model is trained by using the black start simulation data set, the first SVDD model describing the normal simulation data and the second SVDD model describing the fault simulation data are respectively trained, and the first ball center distance of the first SVDD model and the second ball center distance of the second SVDD model are acquired, including: setting a first sample label for the normal simulation data and a second sample label for the fault simulation data, the first sample label being 1 and the second sample label being -1; training the initial SVDD model by using the normal simulation data and the fault simulation data to obtain the first SVDD model describing the normal simulation data; acquiring the radius of the first hypersphere of the first SVDD model to obtain the first ball center distance.

[0010] In an optional implementation, the initial SVDD model is trained by using the black start simulation data set, the first SVDD model describing the normal simulation data and the second SVDD model describing the fault simulation data are respectively trained, and the first ball center distance of the first SVDD model and the second ball center distance of the second SVDD model are obtained, including: setting a third sample label for the fault simulation data and a fourth sample label for the normal simulation data, the third sample label being 1 and the fourth sample label being -1; training the initial SVDD model by using the normal simulation data and the fault simulation data to obtain the second SVDD model describing the fault simulation data; obtaining the radius of the second hypersphere of the second SVDD model to obtain the second ball center distance.

[0011] In this implementation, the influence of the difference in data quantity between the normal simulation data and the fault simulation data on the model can be considered, different sample labels are respectively set for the normal simulation data and the fault simulation data to train the double-layer SVDD model, the first SVDD model describes the normal simulation sample, the second SVDD model describes the fault simulation sample, and the double judgment can further improve the subsequent classification accuracy.

[0012] In an optional implementation, the target distance is obtained by inputting the to-be-detected data into the first SVDD model and the second SVDD model, including: inputting the to-be-detected data into the first SVDD model, calculating the distance between the to-be-detected data and the ball center of the first hypersphere to obtain the first target distance; inputting the to-be-detected data into the second SVDD model, calculating the distance between the to-be-detected data and the ball center of the second hypersphere to obtain the second target distance.

[0013] In an optional implementation, the state of the wind storage and generation system is evaluated based on the target distance, the first ball center distance and the second ball center distance, including: if the first target distance and the second target distance are less than or equal to the first ball center distance and greater than or equal to the second ball center distance, it is judged that the working condition of the wind storage and generation system is normal; if the first target distance and the second target distance are greater than or equal to the first ball center distance and less than or equal to the second ball center distance, it is judged that the working condition of the wind storage and generation system is faulty; if the first target distance and the second target distance are greater than or equal to the first ball center distance and greater than or equal to the second ball center distance, or the first target distance and the second target distance are less than or equal to the first ball center distance and less than or equal to the second ball center distance, it is judged that the working condition of the wind storage and generation system is warning; when a continuous preset number of to-be-detected data are all judged to be the working condition warning of the wind storage and generation system, it is judged that the working condition of the wind storage and generation system is faulty.

[0014] In the implementation mode, the relationship between the first target distance, the second target distance and the first ball center distance and the second ball center distance is combined to further determine whether the working condition of the wind storage power generation system is normal. If the first target distance and the second target distance are greater than or equal to the first ball center distance and greater than or equal to the second ball center distance, or the first target distance and the second target distance are less than or equal to the first ball center distance and less than or equal to the second ball center distance, it is judged that the working condition of the wind storage power generation system is warned; when a continuous preset number of to-be-detected data are all judged to be the working condition warning of the wind storage power generation system, it is judged that the working condition of the wind storage power generation system is faulty, and the working condition detection accuracy can be further improved.

[0015] In a second aspect, the present application provides a wind storage power generation system black start evaluation device, which comprises: an acquisition module for acquiring a black start simulation data set, the black start simulation data set comprising normal simulation data when the wind storage power generation system is normal and fault simulation data when the wind storage power generation system is faulty; a training module for training an initial SVDD model by using the black start simulation data, training a first SVDD model for describing the normal simulation data and a second SVDD model for describing the fault simulation data respectively, and acquiring a first ball center distance of the first SVDD model and a second ball center distance of the second SVDD model; a calculation module for inputting to-be-detected data of the wind storage power generation system into the first SVDD model and the second SVDD model to obtain target distances; and an evaluation module for evaluating the state of the wind storage power generation system based on the target distances, the first ball center distance and the second ball center distance.

[0016] In a third aspect, the present application provides a computer device, which comprises a memory and a processor, the memory and the processor being communicatively connected with each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the wind storage power generation system black start evaluation method of the first aspect or any of the corresponding embodiments thereof.

[0017] In a fourth aspect, the present application provides a computer readable storage medium, which stores computer instructions, and the computer instructions are used to make a computer execute the wind storage power generation system black start evaluation method of the first aspect or any of the corresponding embodiments thereof.

[0018] In a fifth aspect, the present application provides a computer program product, which comprises computer instructions, and the computer instructions are used to make a computer execute the wind storage power generation system black start evaluation method of the first aspect or any of the corresponding embodiments thereof. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the specific embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the specific embodiments or prior art description. Obviously, the drawings described below are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0020] Figure 1 is a schematic diagram of a wind storage power generation system black start evaluation system according to an embodiment of the present application;

[0021] Figure 2 is a schematic diagram of a wind storage power generation system according to an embodiment of the present application;

[0022] Figure 3 is a flow chart of a wind storage power generation system black start evaluation method according to an embodiment of the present application;

[0023] Figure 4 is a flow chart of another wind storage power generation system black start evaluation method according to an embodiment of the present application;

[0024] Figure 5 is a structural block diagram of a wind storage power generation system black start evaluation device according to an embodiment of the present application;

[0025] Figure 6 is a hardware structure schematic diagram of a computer device according to an embodiment of the present application. DETAILED DESCRIPTION

[0026] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the following will combine the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the present application.

[0027] According to an embodiment of the present application, a wind storage power generation system black start evaluation system is provided. Please refer to Figure 1 , Figure 1 is a schematic diagram of a wind storage power generation system black start evaluation system according to an embodiment of the present application. As shown in Figure 1 , the wind storage power generation system black start evaluation system includes a host computer, a simulator and a controller hardware. The simulator is connected with the host computer and the controller hardware respectively. Among them, the controller hardware is connected with the simulator through a cable.

[0028] The controller hardware is a controller of the wind storage power generation system, and the simulation machine is an RT-LAB simulation system. The simulation machine is built with the wind storage power generation system with black start capability consistent with the actual project.

[0029] Specifically, the controller hardware provides the simulation machine with actual operation data of the wind storage power generation system, and sends the operation data to the simulation machine as control information. The operation data includes data when the wind storage power generation system is in normal operating condition and data when the wind storage power generation system is in fault operating condition.

[0030] The simulation machine is used to receive the SPWM pulse signal and perform real-time calculation of the model. Specifically, the simulation machine receives the control signal sent by the controller hardware from the data quantity input board card, and performs black start simulation using the control signal. The simulation machine uses real-time calculation of the model to obtain simulation data, and sends the simulation data to the controller hardware through the analog quantity output board card.

[0031] Further, the simulation machine sends the model detection data to the upper computer through TCP / IP using the wind storage power generation system black start evaluation method of the application.

[0032] Further, according to the embodiment of the application, a wind storage power generation system is provided, please refer to Figure 2 , Figure 2 is a schematic diagram of the wind storage power generation system according to the embodiment of the application. As Figure 2 shown, the left device is a wind storage integrated machine, including a fan 1 and a battery 1, which has black start capability. The simulation machine is built with the wind storage power generation system with black start capability consistent with Figure 2 .

[0033] Further, according to the embodiment of the application, a wind storage power generation system black start evaluation method embodiment is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0034] In this embodiment, a wind storage power generation system black start evaluation method is provided, which can be used in the simulation machine described above, Figure 3 is a flowchart of a wind storage power generation system black start evaluation method according to the embodiment of the application. It should be noted that if there are substantially the same results, the embodiment is not limited to the order of the flowchart shown in Figure 3 . As Figure 3 shown, the flowchart includes the following steps:

[0035] Step S301, obtaining black start simulation data set.

[0036] The black start simulation data set includes normal simulation data of the wind storage power generation system in normal state and fault simulation data of the wind storage power generation system in fault state. The simulation data includes current data and voltage data.

[0037] Specifically, the black start simulation is performed on the wind storage power generation system in normal state by using the simulation machine to obtain a plurality of sets of normal simulation current data and normal simulation voltage data in the simulation process; and the black start simulation is performed on the wind storage power generation system in fault state by using the simulation machine to obtain a plurality of sets of fault simulation current data and fault simulation voltage data in the simulation process.

[0038] In step S302, the initial SVDD model is trained by using the black start simulation data set, and a first SVDD model describing the normal simulation data and a second SVDD model describing the fault simulation data are respectively trained, and a first ball center distance of the first SVDD model and a second ball center distance of the second SVDD model are obtained.

[0039] The SVDD (Support Vector Data Descriptor) model is capable of constructing a hyper-sphere describing the distribution of normal data, and marking the points outside the hyper-sphere as abnormal points. The hyper-sphere feature data of the SVDD model includes a hyper-sphere center and a hyper-sphere radius, and the hyper-sphere radius is the hyper-sphere ball center distance.

[0040] Different sample labels are set for the normal simulation data and the abnormal simulation data in the black start simulation data set, and the initial SVDD model is trained by using the normal simulation data and the abnormal simulation data to obtain the first SVDD model describing the normal simulation data; further, the sample labels of the normal simulation data and the abnormal simulation data in the black start simulation data set are exchanged, and the initial SVDD model is trained again by using the normal simulation data and the abnormal simulation data to obtain the second SVDD model describing the normal simulation data.

[0041] It can be understood that the first SVDD model and the second SVDD model are trained according to the same set of black start simulation data set, and therefore the first hyper-sphere center of the first SVDD model and the second hyper-sphere center of the second SVDD model are the same; the first SVDD model and the second SVDD model are trained according to the normal simulation data and the abnormal simulation data with different sample labels, and therefore the hyper-sphere radii of the first SVDD model and the second SVDD model are different.

[0042] Further, the first ball center distance of the first SVDD model and the second ball center distance of the second SVDD model are obtained.

[0043] In step S303, the to-be-detected data of the wind storage power generation system is input into the first SVDD model and the second SVDD model to obtain a target distance.

[0044] The target distance is the distance between the data to be detected and the center of the hypersphere of the SVDD model.

[0045] Specifically, the current data to be tested from the wind-storage power generation system is input into the first SVDD model to calculate the first target distance between the data to be tested and the center of the first hypersphere to the present. The current data to be tested from the wind-storage power generation system is input into the second SVDD model to calculate the second target distance between the data to be tested and the center of the second hypersphere to the present. Since the centers of the first and second hyperspheres are the same, the first target distance and the second target distance are the same.

[0046] Step S304: Evaluate the state of the wind-storage power generation system based on the target distance, the first sphere center distance, and the second sphere center distance.

[0047] By comparing the target distance with the first sphere center distance, it is determined whether the data to be detected is inside the first hypersphere of the first SVDD model; by comparing the target distance with the second sphere center distance, it is determined whether the data to be detected is inside the second hypersphere of the second SVDD model.

[0048] When the data to be tested is inside the supersphere, it indicates that the wind-storage power generation system corresponding to the data is currently operating normally; when the data to be tested is outside the supersphere, it indicates that the wind-storage power generation system corresponding to the data is currently operating abnormally.

[0049] The black-start assessment method for wind-storage power generation systems provided in this embodiment performs black-start simulations under both normal and fault conditions, and collects corresponding simulation data. The SVDD model is trained twice using both normal and fault simulation data, resulting in a first SVDD model describing the normal simulation data and a second SVDD model describing the fault simulation data. Since there is less fault simulation data, the two-layer SVDD model can fully utilize the information from both normal and fault simulation data when the amount of data is unbalanced, thus improving the classification accuracy of the SVDD model. Based on this, the target distance between the data to be tested and the SVDD model is compared with the first and second sphere center distances to obtain the state information of the wind-storage power generation system, further improving the accuracy of black-start assessment.

[0050] This embodiment provides a black-start evaluation method for wind-storage power generation systems, which can be used in the aforementioned simulator. Figure 4 This is a flowchart of another black-start evaluation method for wind-storage power generation systems according to an embodiment of the present invention. It should be noted that if substantially the same result is obtained, this embodiment does not necessarily reflect that method. Figure 4 The illustrated process sequence is limited. For example... Figure 4 As shown, the process includes the following steps:

[0051] Step S401, obtain black start simulation data set.

[0052] The black start simulation data set includes normal simulation data of the wind storage power generation system in normal state and fault simulation data of the wind storage power generation system in fault state.

[0053] The simulation data includes current data and voltage data of a plurality of preset positions.

[0054] In an implementation manner, please refer to Figure 2 When the black start simulation is performed by using the wind storage power generation system, the simulation data is collected from a plurality of preset positions. Specifically, the current data of the outlet of the energy storage unit, i.e., the current data of the outlet of the battery 1 in the figure, is collected; the current data and the voltage data of the converter machine side, i.e., the current data and the voltage data of the outlet of the wind turbine 1, are collected; the voltage data of the DC bus, i.e., the voltage data on the DC bus DC bus 1, is collected; and the current data and the voltage data of the converter grid side, i.e., the current data and the voltage data of the outlet of the voltage source converter VSC 1, are collected.

[0055] Specifically, the above step S401 includes:

[0056] Step S4011, start the black start simulation under the normal working condition of the wind storage power generation system, and obtain normal simulation data of a plurality of preset positions.

[0057] Specifically, when the working condition of the wind storage power generation system is normal, the black start simulation is performed, and the normal simulation current data and the normal simulation voltage data of a plurality of preset positions in the black start simulation process are collected.

[0058] In an implementation manner, when the working condition of the wind storage power generation system is normal, the wind speed is randomly set between the cut-in wind speed and the cut-out wind speed, 100 groups are set, the black start simulation is performed, each group of simulation is sampled at an interval of 5 seconds, and the current data of the outlet of the energy storage unit, the current data and the voltage data of the converter machine side, the voltage data of the DC bus, and the current data and the voltage data of the converter grid side in the black start process are collected. 5 minutes is taken as one batch, 6 60 matrices are constructed, 6 batches are collected, and a total of 600 groups of normal simulation data are collected.

[0059] Step S4012, start the black start simulation under a plurality of fault working conditions of the wind storage power generation system, and obtain fault simulation data of a plurality of preset positions for each fault working condition.

[0060] Specifically, when the working condition of the wind storage power generation system is fault, the black start simulation is performed, and the fault simulation current data and the fault simulation voltage data of a plurality of preset positions in the black start simulation process are collected.

[0061] The wind storage power generation system working condition fault type has multiple types, and for each working condition fault, multiple preset position fault simulation current data and voltage data are obtained.

[0062] Specifically, the wind storage power generation system working condition fault type includes simulating insufficient energy storage capacity to support black start, simulating network side short circuit, simulating DC bus grounding short circuit, and the like.

[0063] In an implementation mode, when simulating the scenario of insufficient energy storage capacity to support black start, between the cut-in wind speed and the cut-out wind speed of the wind turbine, the wind speed is randomly set, 10 groups are set, black start simulation is carried out, each group of simulation is sampled at an interval of 5 seconds, and the current data of the energy storage unit outlet, the current data and voltage data of the machine side of the converter, the voltage data of the DC bus, and the current data and voltage data of the network side of the converter are collected during the black start process. 5 minutes as one batch, 6 60 matrix, 6 batches are collected, and a total of 60 groups of fault simulation data are collected.

[0064] When simulating the scenario of network side short circuit, between the cut-in wind speed and the cut-out wind speed of the wind turbine, the wind speed is randomly set, 10 groups are set, black start simulation is carried out, each group of simulation is sampled at an interval of 5 seconds, and the current data of the energy storage unit outlet, the current data and voltage data of the machine side of the converter, the voltage data of the DC bus, and the current data and voltage data of the network side of the converter are collected during the black start process. 5 minutes as one batch, 6 60 matrix, 6 batches are collected, and a total of 60 groups of fault simulation data are collected.

[0065] When simulating the scenario of DC bus grounding short circuit, between the cut-in wind speed and the cut-out wind speed of the wind turbine, the wind speed is randomly set, 10 groups are set, black start simulation is carried out, each group of simulation is sampled at an interval of 5 seconds, and the current data of the energy storage unit outlet, the current data and voltage data of the machine side of the converter, the voltage data of the DC bus, and the current data and voltage data of the network side of the converter are collected during the black start process. 5 minutes as one batch, 6 60 matrix, 6 batches are collected, and a total of 60 groups of fault simulation data are collected.

[0066] In this implementation mode, when the wind storage power generation system working condition fails, a total of 180 groups of fault simulation data are collected.

[0067] In this implementation mode, multiple data collection points are set, and multiple working condition faults are simulated to obtain fault simulation data, which can fully consider multiple scenarios in actual application, enrich the black start simulation data set, and further improve the classification and evaluation accuracy of the subsequent SVDD model.

[0068] Step S402, training the initial SVDD model by using the black start simulation data set, training the first SVDD model describing the normal simulation data and the second SVDD model describing the fault simulation data respectively, and obtaining the first ball center distance of the first SVDD model and the second ball center distance of the second SVDD model.

[0069] Before training the initial SVDD model by using the black start simulation data set, the normal simulation data and the fault simulation data in the black start simulation data set are preprocessed, normalized to make the transformed simulation data range [0, 1], and the data in the black start simulation data set is rearranged into a single column matrix , for the dimension.

[0070] The initial SVDD model is trained by using the black start simulation data set after preprocessing.

[0071] Specifically, the above step S402 includes:

[0072] Step S4021, setting the first sample label for the normal simulation data and the second sample label for the fault simulation data, and training the initial SVDD model by using the normal simulation data and the fault simulation data to obtain the first SVDD model describing the normal simulation data.

[0073] Wherein, the first label is 1, and the second label is -1. The sample label is set for the black start simulation data set, the first sample label of the normal simulation data is set to 1, and the sample label of the fault simulation data is set to -1. The initial SVDD model is trained for the first time by using the black start simulation data set and the corresponding sample label, a hypersphere containing the normal simulation data in the black start simulation data set is constructed, and the first SVDD model describing the normal simulation data is obtained.

[0074] Specifically, the optimization problem is constructed as follows:

[0075]

[0076] s.t

[0077] Wherein, i is the index of the black start simulation data, n is the total number of the black start simulation data, and are the penalty factor and the relaxation factor describing the normal simulation data respectively, is the hypersphere center describing the normal simulation data, is the hypersphere radius describing the normal simulation data.

[0078] Wherein, the penalty factor is:

[0079] .

[0080] wherein, is the percentage of black-start simulation data outside the hypersphere, using The hypersphere volume can be coordinated with the misclassification rate of black-start simulation data.

[0081] Construct the dual form of the above problem:

[0082]

[0083]

[0084] Further, introduce the Gaussian kernel function to simplify the inner product operation and complete the feature space mapping:

[0085] .

[0086] wherein, is the Lagrange multiplier, is the Gaussian kernel function, is the width parameter of the Gaussian kernel function. The corresponding Lagrange multiplier satisfies Black-start simulation data is the support vector on the surface of the hypersphere. The support vector index i that satisfies this condition is screened to form an integer set s.

[0087] Further, calculate the hypersphere radius of the support vector, that is, the distance from the hypersphere center:

[0088] .

[0089] wherein, is any support vector, .

[0090] Through the above method, the initial SVDD model is trained using the black-start simulation data set and the corresponding sample labels, a first SVDD model describing normal simulation data is obtained, and a first hypersphere center distance of the first SVDD model is determined.

[0091] Step S4022, set the fault simulation data as a third label and the normal simulation data as a fourth label, train the initial SVDD model using the normal simulation data and the fault simulation data, and obtain a second SVDD model describing the fault simulation data.

[0092] The third label is 1, and the fourth label is -1. The sample labels are set for the black start simulation data set, the third sample label of the fault simulation data is set to 1, and the sample label of the normal simulation data is set to -1. The initial SVDD model is trained for the second time by using the black start simulation data set and the corresponding sample labels, a hypersphere containing the fault simulation data in the black start simulation data set is constructed, and a second SVDD model describing the fault simulation data is obtained.

[0093] Specifically, an optimization problem is constructed:

[0094]

[0095] s.t

[0096] wherein i is the index of the black start simulation data, n is the total number of the black start simulation data, and are respectively a penalty factor and a relaxation factor describing the fault simulation data, is a hypersphere center describing the fault simulation data, is a hypersphere radius describing the fault simulation data.

[0097] wherein the penalty factor is:

[0098] .

[0099] wherein is the percentage of the black start simulation data outside the hypersphere, and can coordinate the sphere volume and the misclassification rate of the black start simulation data.

[0100] The dual form of the above problem is constructed:

[0101]

[0102] s.t

[0103] Further, a Gaussian kernel function is introduced to simplify the inner product operation, and the feature space mapping is completed:

[0104] .

[0105] wherein is a Lagrange multiplier, is a Gaussian kernel function, is a width parameter of the Gaussian kernel function. The corresponding Lagrange multiplier satisfies the black start simulation data For the support vector on the surface of the sphere, the support vector index i satisfying the condition is screened to form an integer set s.

[0106] Further, the support vector hypersphere radius, i.e. the hypersphere center distance, is calculated.

[0107] .

[0108] wherein, is any support vector, .

[0109] By the above method, the initial SVDD model is trained by using the black start simulation data set and the corresponding sample label, the second SVDD model describing the fault simulation data is obtained, and the second hypersphere center distance of the second SVDD model is determined.

[0110] It can be understood that the first SVDD model and the second SVDD model are trained according to the same set of black simulation data sets, so the first hypersphere center of the first SVDD model and the second hypersphere center of the second SVDD model are the same; The first SVDD model and the second SVDD model are trained according to the normal simulation data and the abnormal simulation data with different sample labels, so the hypersphere radii of the first SVDD model and the second SVDD model are different.

[0111] In this implementation, the influence of the data quantity difference between the normal simulation data and the fault simulation data on the model can be considered, different sample labels are set for the normal simulation data and the fault simulation data respectively, and a double-layer SVDD model is trained, the first SVDD model describes the normal simulation sample, and the second SVDD model describes the fault simulation sample, and the double judgment can further improve the subsequent classification accuracy.

[0112] In step S403, the to-be-detected data of the wind storage power generation system is input into the first SVDD model and the second SVDD model to obtain a target distance.

[0113] The to-be-detected data of the wind storage power generation system is normalized, and the normalized to-be-detected data is input into the first SVDD model, the distance between the to-be-detected data and the center of the first hypersphere is calculated, and a first target distance is obtained.

[0114] .

[0115] The normalized to-be-detected data is input into the second SVDD model, the distance between the to-be-detected data and the center of the second hypersphere is calculated, and a second target distance is obtained.

[0116] .

[0117] Wherein, since the first SVDD model and the second SVDD model have the same training data, only the training sample labels are different, the first target distance and the second target distance are the same, that is .

[0118] Step S404, based on the target distance, the first sphere center distance and the second sphere center distance, the state of the wind storage power generation system is evaluated.

[0119] By comparing the target distance and the first sphere center distance, it is judged whether the to-be-detected data is inside the first hypersphere of the first SVDD model; by comparing the target distance and the second sphere center distance, it is judged whether the to-be-detected data is inside the second hypersphere of the second SVDD model.

[0120] Specifically, the sizes of the first target distance, the second target distance, the first sphere center distance and the second sphere center distance are compared. If the first target distance and the second target distance are less than or equal to the first sphere center distance and greater than or equal to the second sphere center distance, it is judged that the working condition of the wind storage power generation system is normal. If the first target distance and the second target distance are greater than or equal to the first sphere center distance and less than or equal to the second sphere center distance, it is judged that the working condition of the wind storage power generation system is faulty.

[0121] If the first target distance and the second target distance are greater than or equal to the first sphere center distance and greater than or equal to the second sphere center distance, or the first target distance and the second target distance are less than or equal to the first sphere center distance and less than or equal to the second sphere center distance, it is judged that the working condition of the wind storage power generation system is warning; when a continuous preset number of to-be-detected data are all judged to be the working condition warning of the wind storage power generation system, it is judged that the working condition of the wind storage power generation system is faulty. In a specific implementation mode, the preset number is 10.

[0122] In the implementation mode, the application respectively simulates the black start of the wind storage power generation system in normal and fault conditions, and collects corresponding simulation data. The normal simulation data and the fault simulation data are used for training the SVDD model twice, and the first SVDD model describing the normal simulation data and the second SVDD model describing the fault simulation data are obtained respectively. Since the fault simulation data is less, the double-layer SVDD model can make full use of the normal simulation data and the fault simulation data information in the case of unbalanced amount of normal simulation data and fault simulation data, and improve the classification accuracy of the SVDD model. On this basis, the relationship between the first target distance, the second target distance, the first ball center distance and the second ball center distance is combined to further determine whether the working condition of the wind storage power generation system is normal. If the first target distance and the second target distance are greater than or equal to the first ball center distance and greater than or equal to the second ball center distance, or the first target distance and the second target distance are less than or equal to the first ball center distance and less than or equal to the second ball center distance, it is judged that the working condition of the wind storage power generation system is warning; when a continuous preset number of to-be-detected data are judged to be the working condition warning of the wind storage power generation system, it is judged that the working condition of the wind storage power generation system is fault, which can further improve the working condition detection accuracy.

[0123] In the embodiment, a wind storage power generation system black start evaluation device is also provided, which is used to implement the above-mentioned embodiments and preferred embodiments, and will not be described again. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware, or a combination of software and hardware is also possible and contemplated.

[0124] The embodiment provides a wind storage power generation system black start evaluation device, which is applied to Figure 1 The simulation machine is shown in FIG. 1, please refer to Figure 5 , Figure 5 The structure block diagram of the wind storage power generation system black start evaluation device according to the embodiment of the application is shown in FIG. 2, and the wind storage power generation system black start evaluation device includes: Figure 5

[0125] The acquisition module 501 is used to acquire the black start simulation data set, and the black start simulation data set includes the normal simulation data when the wind storage power generation system is normal and the fault simulation data when the wind storage power generation system is fault.

[0126] The training module 502 is used to train the initial SVDD model by using the black start simulation data, and the first SVDD model describing the normal simulation data and the second SVDD model describing the fault simulation data are respectively trained, and the first ball center distance of the first SVDD model and the second ball center distance of the second SVDD model are acquired.

[0127] ​The calculation module 503 is used to input the data to be detected from the wind-storage power generation system into the first SVDD model and the second SVDD model to obtain the target distance.

[0128] Evaluation module 504 is used to evaluate the status of the wind-storage power generation system based on the target distance, the first center-to-center distance, and the second center-to-center distance.

[0129] Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0130] In this embodiment, the black start evaluation device for the wind-storage power generation system is presented in the form of a functional unit. Here, a unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0131] This invention also provides a computer device having the above-described features. Figure 5 The black start evaluation device for the wind-storage power generation system is shown.

[0132] Please see Figure 6 , Figure 6 This is a schematic diagram of the structure of a computer device provided in an optional embodiment of the present invention, such as... Figure 6 As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 6 Take a processor 10 as an example.

[0133] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0134] The memory 20 stores instructions executable by the at least one processor 10 to cause the at least one processor 10 to perform the methods implemented by the above-described embodiments.

[0135] The memory 20 can include a program region and a data region. The program region can store an operating system and application programs required by at least one function. The data region can store data created according to use of the computer device, and the like. In addition, the memory 20 can include a high-speed random access memory, and can further include a non-transitory memory such as at least one of a magnetic disk storage device, a flash memory device, or other non-transitory solid-state memory device. In some alternative embodiments, the memory 20 can optionally include a memory disposed remotely with respect to the processor 10, and these remote memories can be connected to the computer device through a network. Examples of the network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0136] The memory 20 can include a volatile memory such as a random access memory, and can further include a non-volatile memory such as a flash memory, a hard disk, or a solid state disk. The memory 20 can also include a combination of the above-mentioned types of memories.

[0137] The computer device further includes an input device 30 and an output device 40. The processor 10, the memory 20, the input device 30, and the output device 40 can be connected by a bus or other means, Figure 6 The connection by the bus is taken as an example.

[0138] The input device 30 can receive input digital or character information, and generate key signal input related to user settings and function control of the computer device, such as a touch screen, a keypad, a mouse, a trackpad, a touchpad, a pointing stick, one or more mouse buttons, a trackball, a joystick, and the like. The output device 40 can include a display device, an auxiliary lighting device (e.g., an LED), a tactile feedback device (e.g., a vibration motor), and the like. The display device includes, but is not limited to, a liquid crystal display, a light-emitting diode, a display, and a plasma display. In some alternative embodiments, the display device can be a touch screen.

[0139] The embodiments of the present application further provide a computer readable storage medium, and the method according to the embodiments of the present application can be implemented in hardware, firmware, or recorded in a storage medium, or stored in a remote storage medium or a non-transitory machine readable storage medium and downloaded to a local storage medium through network, so that the method described herein can be processed by such software on a storage medium using a general purpose computer, a special purpose processor, or programmable or special hardware. The storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, or a solid state disk, etc. Further, the storage medium can also include a combination of the above-mentioned memories. It can be understood that the computer, the processor, the microprocessor controller, or the programmable hardware includes a storage component that can store or receive software or computer code, when the software or computer code is accessed and executed by the computer, the processor, or the hardware, the method shown in the above embodiments is implemented.

[0140] Part of the present application can be applied as a computer program product, for example, computer program instructions, when executed by a computer, through the operation of the computer, the method and / or technical solutions according to the present application can be called or provided. Those skilled in the art should understand that the form of computer program instructions in computer readable medium includes but is not limited to source file, executable file, installation package file, etc. Correspondingly, the way of computer program instructions executed by computer includes but is not limited to: the computer directly executes the instructions, or the computer compiles the instructions and then executes the corresponding compiled program, or the computer reads and executes the instructions, or the computer reads and installs the instructions and then executes the corresponding installed program. Here, the computer readable medium can be any available computer readable storage medium or communication medium accessible to the computer.

[0141] Although the embodiments of the present application are described in conjunction with the accompanying drawings, various modifications and changes can be made by those skilled in the art without departing from the spirit and scope of the present application, and such modifications and changes fall within the scope defined by the appended claims.

Claims

1. A method for black start assessment of a wind energy storage and generation system, characterized in that, The method comprises: obtaining a black start simulation data set, the black start simulation data set comprising normal simulation data of a wind storage power generation system in normal state and fault simulation data of the wind storage power generation system in fault state; training an initial SVDD model using the black start simulation data set, training a first SVDD model describing the normal simulation data and a second SVDD model describing the fault simulation data respectively, and obtaining a first ball center distance of the first SVDD model and a second ball center distance of the second SVDD model; inputting to-be-detected data of the wind storage power generation system into the first SVDD model and the second SVDD model to obtain a target distance; evaluating a state of the wind storage power generation system based on the target distance, the first ball center distance and the second ball center distance; the training of the initial SVDD model using the black start simulation data set, the training of the first SVDD model describing the normal simulation data and the second SVDD model describing the fault simulation data respectively, and the obtaining of the first ball center distance of the first SVDD model and the second ball center distance of the second SVDD model comprise: setting a first sample label for the normal simulation data and a second sample label for the fault simulation data, the first sample label being 1 and the second sample label being -1; training the initial SVDD model using the normal simulation data and the fault simulation data to obtain the first SVDD model describing the normal simulation data; obtaining a radius of a first hypersphere of the first SVDD model to obtain the first ball center distance; setting a third sample label for the fault simulation data and a fourth sample label for the normal simulation data, the third sample label being 1 and the fourth sample label being -1; training the initial SVDD model using the normal simulation data and the fault simulation data to obtain the second SVDD model describing the fault simulation data; obtaining a radius of a second hypersphere of the second SVDD model to obtain the second ball center distance; the inputting of the to-be-detected data into the first SVDD model and the second SVDD model to obtain the target distance comprises: inputting the to-be-detected data into the first SVDD model to calculate a distance between the to-be-detected data and a ball center of the first hypersphere to obtain a first target distance; inputting the to-be-detected data into the second SVDD model to calculate a distance between the to-be-detected data and a ball center of the second hypersphere to obtain a second target distance; the evaluation of the state of the wind storage power generation system based on the target distance, the first ball center distance and the second ball center distance comprises: if the first target distance and the second target distance are less than or equal to the first ball center distance and greater than or equal to the second ball center distance, it is determined that the wind storage power generation system is in normal working condition; if the first target distance and the second target distance are greater than or equal to the first ball center distance and less than or equal to the second ball center distance, it is determined that the wind storage power generation system is in fault working condition. if the first target distance and the second target distance are greater than or equal to the first ball center distance and greater than or equal to the second ball center distance, or the first target distance and the second target distance are less than or equal to the first ball center distance and less than or equal to the second ball center distance, it is determined that the wind storage and generation system operating condition warning; when a continuous preset number of to-be-detected data are all determined to be the wind storage and generation system operating condition warning, it is determined that the wind storage and generation system operating condition fault.

2. The wind energy storage system black start assessment method of claim 1, wherein, The black start simulation data set includes: starting the black start simulation under the normal condition of the wind storage and generation system, obtaining normal simulation current data and voltage data of a plurality of preset positions; starting the black start simulation under a plurality of operating condition faults of the wind storage and generation system, for each operating condition fault, obtaining fault simulation current data and voltage data of a plurality of preset positions.

3. A wind energy storage and generation system black start assessment device, characterized in that, The device includes: an acquisition module configured to acquire a black start simulation data set, the black start simulation data set including normal simulation data when the wind storage and generation system is normal and fault simulation data when the wind storage and generation system fails; a training module configured to train an initial SVDD model using the black start simulation data, respectively training a first SVDD model describing the normal simulation data and a second SVDD model describing the fault simulation data, and obtaining a first ball center distance of the first SVDD model and a second ball center distance of the second SVDD model; a calculation module configured to input to-be-detected data of the wind storage and generation system into the first SVDD model and the second SVDD model to obtain target distances; an evaluation module configured to evaluate the state of the wind storage and generation system based on the target distances, the first ball center distance, and the second ball center distance; The training of the initial SVDD model using the black start simulation data set, the training of the first SVDD model describing the normal simulation data, and the training of the second SVDD model describing the fault simulation data, and the obtaining of the first ball center distance of the first SVDD model and the second ball center distance of the second SVDD model include: setting a first sample label for the normal simulation data and a second sample label for the fault simulation data, the first sample label being 1 and the second sample label being -1; training the initial SVDD model using the normal simulation data and the fault simulation data to obtain the first SVDD model describing the normal simulation data; obtaining the radius of the first hypersphere of the first SVDD model to obtain the first ball center distance; setting a third sample label for the fault simulation data and a fourth sample label for the normal simulation data, the third sample label being 1 and the fourth sample label being -1; training the initial SVDD model using the normal simulation data and the fault simulation data to obtain the second SVDD model describing the fault simulation data; obtaining the radius of the second hypersphere of the second SVDD model to obtain the second ball center distance; The inputting of the to-be-detected data into the first SVDD model and the second SVDD model to obtain target distances includes: inputting the to-be-detected data into the first SVDD model, calculating a distance between the to-be-detected data and a center of the first hypersphere, and obtaining a first target distance; inputting the to-be-detected data into the second SVDD model, calculating a distance between the to-be-detected data and a center of the second hypersphere, and obtaining a second target distance; the evaluating the state of the wind storage power generation system based on the target distance, the first center distance and the second center distance comprises: if the first target distance and the second target distance are less than or equal to the first center distance and greater than or equal to the second center distance, it is determined that the wind storage power generation system is in a normal working condition; if the first target distance and the second target distance are greater than or equal to the first center distance and less than or equal to the second center distance, it is determined that the wind storage power generation system is in a fault working condition; if the first target distance and the second target distance are greater than or equal to the first center distance and greater than or equal to the second center distance, or the first target distance and the second target distance are less than or equal to the first center distance and less than or equal to the second center distance, it is determined that the wind storage power generation system is in a warning working condition; if a continuous preset number of to-be-detected data are all determined to be in the warning working condition, it is determined that the wind storage power generation system is in the fault working condition.

4. A computer device, comprising: comprise: a memory and a processor, which are communicatively connected, the memory stores computer instructions, and the processor executes the computer instructions to perform the wind storage power generation system black start evaluation method of claim 1 or 2.

5. A computer readable storage medium, characterized in that, the computer readable storage medium stores computer instructions, and the computer instructions are used to make a computer execute the wind storage power generation system black start evaluation method of claim 1 or 2.

6. A computer program product, characterised in that, comprise computer instructions, and the computer instructions are used to make a computer execute the wind storage power generation system black start evaluation method of claim 1 or 2.

Images