Parameter identification method and device for double-fed wind farm single-machine equivalent grid-connected model

Abstract

The application discloses a parameter identification method and device for a doubly-fed wind farm single-machine equivalent grid-connected model, and the method comprises the following steps: 1, setting an experiment scheme; 2, collecting n groups of detailed model data for identifying to-be-identified parameters; 3, establishing a doubly-fed wind farm single-machine equivalent grid-connected model, and using an adaptive inertia weight particle swarm intelligent algorithm to preliminarily identify the detailed model data to obtain a preliminary parameter identification result; 4, substituting the preliminary parameter identification result corresponding to each group of detailed model data into the doubly-fed wind farm single-machine equivalent grid-connected model, performing experiments in different operating conditions, and selecting optimal parameters; and 5, using the optimal parameters to establish an optimal doubly-fed wind farm single-machine equivalent model, and running the model in different operating conditions to verify the accuracy of the optimal parameters. The application can accurately identify the collection system parameters of the doubly-fed wind farm equivalent model by using multiple groups of detailed model data, so that the doubly-fed wind farm equivalent model suitable for subsynchronous oscillation analysis is established.

Description

Technical Field

[0001] This invention relates to a parameter identification method and apparatus for a single-unit equivalent grid-connected model of a doubly fed wind farm, belonging to the field of power system analysis technology. Background Technology

[0002] Large-scale doubly-fed induction generator (DFIG) wind farms are often located in remote areas, far from load centers, and connected to the main power grid via long transmission lines, often requiring series capacitors in the lines for power transmission. However, using series capacitors to compensate for the lines can lead to subsynchronous oscillations in the DFIG-connected wind farm system, seriously threatening system stability. Considering the large number of turbines in the site, the high model order, and the large computational load, current research on subsynchronous oscillations in DFIG-connected wind farm systems generally adopts equivalent models of DFIG wind farms. However, the current effectiveness analysis of equivalent models of DFIG wind farms focuses on the consistency of the transient characteristics of the wind farm's output power. Whether the equivalent model is suitable for analyzing the subsynchronous oscillation characteristics of wind power grid-connected systems remains inconclusive.

[0003] The collector lines of a wind farm connect all the wind turbines, and their complex grid structure significantly impacts the small-disturbance stability of the wind farm. The accuracy of the equivalent model directly affects the system's stability analysis. When performing equivalent modeling of the wind farm collector lines, the grid structure is typically transformed directly, converting complex trunk or hybrid topologies into simple radial topologies, and then solving for the equivalent line impedance parameters at the outlet of each wind turbine. However, the series-parallel relationships and multi-stage transmission dissipation within the wind farm affect system damping, thus influencing the subsynchronous oscillation characteristics of the wind power grid-connected system before and after the equivalent model. Simply using methods based on equal power loss or equal voltage difference to calculate the equivalent impedance while neglecting the analysis of the system's subsynchronous oscillation characteristics will affect the effectiveness of the equivalent model. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and provide a parameter identification method and device for a single-unit equivalent grid-connected model of a doubly-fed wind farm. This method can accurately identify the impedance parameters of the collector system of a single-unit equivalent grid-connected model of a doubly-fed wind farm using multiple sets of detailed model data, thereby enabling accurate modeling suitable for subsynchronous oscillation analysis.

[0005] To achieve the above objectives, the present invention is implemented using the following technical solution:

[0006] In a first aspect, the present invention provides a parameter identification method for a single-unit equivalent grid-connected model of a doubly-fed induction generator (DFIG) wind farm, comprising:

[0007] The active power data of the transmission line after the doubly fed wind farm entered the subsynchronous oscillation state under multiple experimental conditions were obtained, and the active power data of the transmission line under multiple experimental conditions were used as detailed model data.

[0008] Multiple sets of detailed model data are input into a pre-constructed equivalent grid-connected model of a single doubly fed wind farm unit for preliminary identification, and preliminary results of multiple parameter identification are obtained.

[0009] Substitute the preliminary parameter identification results corresponding to each set of detailed model data into the single-unit equivalent grid-connected model of the doubly fed wind farm to obtain the active power corresponding to each set of detailed model data.

[0010] Calculate the average deviation of active power and the relative error of the average deviation for each set of detailed model data, and select the preliminary parameter identification result corresponding to the set of detailed model data with the smallest relative error as the optimal parameter identification result.

[0011] Furthermore, the active power data of the transmission line after the doubly-fed wind farm has entered a subsynchronous oscillation state, simulated under multiple experimental conditions, is used as detailed model data, thereby obtaining multiple sets of detailed model data, including:

[0012] Calculate the relationship between the subsynchronous oscillation frequency of a doubly fed wind farm connected to the grid via series compensation and the reactance of the collector system; calculate the relationship between the subsynchronous oscillation amplitude of a doubly fed wind farm connected to the grid via series compensation and the resistance of the collector system.

[0013] Based on the calculation results, multiple sets of experimental conditions were set up, and the experimental point was determined to be the AC side grid connection point; among them, any set of experimental conditions includes: active power command, reactive power command, and the time of connecting the line series compensation capacitor.

[0014] Set the initial active and reactive power of the doubly fed wind farm, and set the disturbance parameters of the doubly fed wind farm through the series compensation capacitor time in any set of experimental conditions to perform transient simulation.

[0015] Once the simulation enters the subsynchronous oscillation state, the active power data of the doubly fed wind farm transmission line is recorded and used as a set of detailed model data to obtain multiple sets of detailed model data.

[0016] Furthermore, the relationship between the subsynchronous oscillation frequency of the doubly-fed wind farm connected to the grid via series compensation and the reactance of the collector system is calculated, as shown in Equation 1:

[0017]

[0018] In the formula, f s For synchronization frequency, X C For the capacitive reactance of the series capacitor, K C For the series compensation degree of transmission lines, X L Let X be the reactance of the transmission line, and ∑X be the equivalent total reactance of the entire system, including the generator, the collector system, and the transmission line.

[0019] Furthermore, the relationship between the subsynchronous oscillation amplitude of the doubly-fed wind farm connected to the grid via series compensation and the resistance of the collector system is calculated, as shown in Equation 2:

[0020]

[0021] In the formula, R req R is the equivalent rotor resistance. r ' is the rotor resistance, s sso R is the slip of the generator at the subsynchronous frequency. B R is the total resistance of the doubly-fed wind farm grid-connected system. s R is the stator resistance of the generator. J R is the resistance of the collector system in a doubly-fed wind farm. L This represents the resistance of the power transmission line.

[0022] Furthermore, the multiple sets of detailed model data are input into a pre-constructed equivalent grid-connected model of a single doubly-fed wind farm unit for preliminary identification, resulting in preliminary identification results for multiple parameters, including:

[0023] Based on the detailed model of the doubly fed wind farm obtained in advance, an equivalent grid-connected model of a single doubly fed wind farm unit is established. The parameters of the equivalent doubly fed wind turbine and transformer are obtained by the capacity weighting method. The parameters corresponding to the equivalent grid-connected model of a single doubly fed wind farm unit are set with the criterion that the active power and reactive power output of the wind farm before and after equivalence are equal.

[0024] The complex trunk-type grid structure is transformed into a radial pure parallel structure. The impedance value of each doubly fed wind turbine collector line in the radial pure parallel structure is obtained. The impedance value of the collector system in the single-unit equivalent grid-connected model of the doubly fed wind farm is calculated by combining the principle that the total loss after equivalence is equal to the sum of the losses of each branch before equivalence.

[0025] The impedance value of the collection system in the single-unit equivalent grid-connected model of a doubly fed wind farm is used as the initial value for the iteration of the adaptive inertial weighted particle swarm intelligent algorithm.

[0026] The adaptive inertial weighted particle swarm intelligence algorithm and the equivalent grid-connected model of a single doubly fed wind farm were used to perform preliminary identification of detailed model data, and preliminary results of parameter identification were obtained.

[0027] Furthermore, the impedance value of each doubly-fed wind turbine collector line in the radial pure parallel structure is calculated using the formula shown in Equation 3:

[0028]

[0029] In the formula, w represents any doubly fed fan in a radial pure parallel structure, and Z... j Let P be the actual impedance of the j-th doubly-fed wind turbine collector line in this trunk topology. j For the current flowing through the collector line impedance Zj The power of Z eqw Let be the equivalent impedance of the current collector line of the w-th doubly fed wind turbine in a radially parallel structure.

[0030] Furthermore, the impedance value of the collector system in the equivalent grid-connected model of a single doubly-fed wind farm is calculated based on the principle that the total loss after equivalence equals the sum of the losses of each branch before equivalence. The calculation formula is shown in Equation 4:

[0031]

[0032] In the formula, Z eqh P is the equivalent impedance of the h-th doubly-fed wind turbine collector line in the radial, purely parallel structure of the entire doubly-fed wind farm; h For the current flowing through the collector line impedance Z eqh The power of Z; eq This represents the equivalent impedance of the collector line of the doubly fed wind turbine after aggregation.

[0033] Furthermore, the formula for calculating the average deviation is as shown in Equation 5:

[0034]

[0035] In the formula, F is the average deviation; X S X represents the per-unit value of the single-machine equivalent model data for the electrical quantities to be assessed; M The per-unit values ​​of the detailed model data for the electrical quantities to be assessed; K S_START K S_End These are the first and last indexes of the single-machine equivalent model data within the calculation error interval, respectively; K M_START K M_End These are the first and last serial numbers of the detailed model data within the calculation error interval, respectively;

[0036] The formula for calculating the relative error of the average deviation is shown in Equation 6:

[0037]

[0038] In the formula, error is the relative error of the average deviation; x1 is the average deviation of active power; x min Substitute the preliminary results of the identification of n sets of parameters into the minimum value of the average deviation of active power under the same set of operating conditions.

[0039] Furthermore, it also includes:

[0040] The impedance value of the collection system in the single-unit equivalent grid-connected model of a doubly fed wind farm is set as the optimal parameter identification result to establish the optimal single-unit equivalent model of the doubly fed wind farm.

[0041] The optimal single-machine equivalent model is run under multiple experimental conditions. The active power command, reactive power command, and connection time of the series compensation capacitor are input for multiple experimental conditions. The active power during the operation is recorded as the optimal equivalent model data. The average deviation between the optimal equivalent model data and the detailed model data is calculated to analyze the effectiveness of the optimal single-machine equivalent model.

[0042] Secondly, the present invention provides a collector system parameter identification device suitable for equivalent modeling of doubly-fed wind farms, comprising:

[0043] The detailed model data acquisition module is used to acquire the active power data of the transmission line after the doubly fed wind farm enters the subsynchronous oscillation state under multiple experimental conditions, and uses the active power data of the transmission line under multiple experimental conditions as detailed model data.

[0044] The preliminary identification module is used to input multiple sets of detailed model data into a pre-constructed doubly fed wind farm single-unit equivalent grid-connected model for preliminary identification, and obtain preliminary identification results for multiple parameters;

[0045] The active power acquisition module is used to substitute the preliminary parameter identification results corresponding to each set of detailed model data into the single-unit equivalent grid-connected model of the doubly fed wind farm to obtain the active power corresponding to each set of detailed model data.

[0046] The optimal parameter identification result acquisition module is used to calculate the average deviation of active power and the relative error of the average deviation corresponding to each set of detailed model data, and select the preliminary parameter identification result corresponding to the set of detailed model data with the smallest relative error as the optimal parameter identification result.

[0047] Compared with the prior art, the beneficial effects achieved by the present invention are as follows:

[0048] 1. This invention considers the impact of different active power of doubly fed wind farms on the identification results, identifies multiple sets of detailed model data, extracts the optimal parameters from the preliminary results of multiple parameter identification, and improves the accuracy of parameter identification results.

[0049] 2. Based on the calculation results of the subsynchronous oscillation characteristic quantities, the present invention sets up an experimental scheme, providing detailed model data under typical working conditions for parameter identification.

[0050] 3. This invention uses the equal power loss method as the initial value and uses the adaptive inertial weighted particle swarm intelligent algorithm to identify the impedance of the power collection system, making the results more accurate and able to reflect the actual operation of wind turbine units. Attached Figure Description

[0051] Figure 1 This is a structural diagram of the experimental platform for obtaining detailed model data provided in an embodiment of the present invention;

[0052] Figure 2 This is a detailed grid-connected topology diagram of a doubly-fed wind farm provided in an embodiment of the present invention;

[0053] Figure 3 The grid topology diagram of the single-unit equivalent grid-connected model of a doubly fed wind farm provided in the embodiments of the present invention.

[0054] Figure 4 The flowchart illustrates the parameter identification method for a single-unit equivalent grid-connected model of a doubly fed wind farm provided in this embodiment of the invention. Detailed Implementation

[0055] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.

[0056] Example 1

[0057] This embodiment introduces a parameter identification method for a single-unit equivalent grid-connected model of a doubly-fed wind farm, including:

[0058] The active power data of the transmission line after entering the subsynchronous oscillation state in the simulation of the doubly fed wind farm under multiple experimental conditions were obtained as detailed model data, and then multiple sets of detailed model data were obtained.

[0059] The multiple sets of detailed model data are input into the pre-constructed single-unit equivalent grid-connected model of the doubly fed wind farm for preliminary identification, and preliminary results of multiple parameter identification are obtained.

[0060] Substitute the preliminary parameter identification results corresponding to each set of detailed model data into the single-unit equivalent grid-connected model of the doubly fed wind farm to obtain the active power corresponding to each set of detailed model data.

[0061] Calculate the average deviation of active power and the relative error of the average deviation for each set of detailed model data, and select the preliminary parameter identification result corresponding to the set of detailed model data with the smallest relative error as the optimal parameter identification result.

[0062] like Figure 4 As shown in this embodiment, the parameter identification method for the equivalent grid-connected model of a single doubly fed wind farm unit involves the following steps in its application process:

[0063] Step 1: Set up the experimental plan:

[0064] Step 1.1: Calculate the relationship between the subsynchronous oscillation frequency of the doubly-fed wind farm connected to the grid via series compensation and the reactance of the collector system, as shown in Equation 1:

[0065]

[0066] In the formula, fs For synchronization frequency, X C For the capacitive reactance of the series capacitor, K C For the series compensation degree of transmission lines, X L Let X be the reactance of the transmission line, and ∑X be the equivalent total reactance of the entire system, including the generator, the collector system, and the transmission line.

[0067] The relationship between the subsynchronous oscillation amplitude of a doubly-fed wind farm connected to the grid via series compensation and the resistance of the collector system is calculated as shown in Equation 2:

[0068]

[0069] In the formula, R req R is the equivalent rotor resistance. r ' is the rotor resistance, s sso R is the slip of the generator at the subsynchronous frequency. B R is the total resistance of the doubly-fed wind farm grid-connected system. s R is the stator resistance of the generator. J R is the resistance of the collector system in a doubly-fed wind farm. L The resistance of the transmission line is denoted as . When the amplitude of the equivalent rotor resistance exceeds the sum of the equivalent resistances of the stator, collector system, and transmission line at that resonant frequency, the resistance of the entire system will be negative. This will cause the resonant current to continuously diverge and oscillate. The larger the absolute value of the negative resistance, the more severe the subsynchronous oscillation phenomenon will occur in the system.

[0070] Step 1.2: Based on the calculation results, set up n sets of experimental operating conditions and determine the experimental point as the AC side grid connection point; n sets of experimental schemes are composed of n sets of experimental operating conditions and experimental points; any set of experimental operating conditions includes: active power command, reactive power command, and series compensation activation time;

[0071] Step 2: Collect n sets of detailed model data for identifying the parameters to be identified:

[0072] Step 2.1: Set the initial active and reactive power of the doubly-fed wind farm in the experimental platform. Based on the connection time of the series compensation capacitor in any set of experimental conditions, set the disturbance parameters of the doubly-fed wind farm collection system in the experimental platform so that the experimental platform can perform transient simulation.

[0073] Step 2.2: After the experimental platform simulation begins to enter the subsynchronous oscillation state, record the active power data of the doubly fed wind farm transmission line and use it as a set of detailed model data to obtain n sets of detailed model data.

[0074] Step 3: Obtain preliminary results of parameter identification:

[0075] Step 3.1: Establish an equivalent grid-connected model of a single doubly-fed wind farm based on the pre-obtained detailed model of the doubly-fed wind farm. Obtain the parameters of the equivalent doubly-fed wind turbine and transformer through the capacity weighting method. Set the parameters corresponding to the equivalent grid-connected model of a single doubly-fed wind farm based on the criterion that the active power and reactive power output of the wind farm before and after equivalence are equal.

[0076] Step 3.2: Transform the complex trunk-type grid structure into a radial pure parallel structure. The impedance value of each doubly-fed wind turbine collector line in the radial topology is calculated using Equation 3:

[0077]

[0078] In the formula, w represents any doubly-fed wind turbine in the radial topology, and Z... j Let P be the actual impedance of the j-th doubly-fed wind turbine collector line in this trunk topology. j For the current flowing through the collector line impedance Z j The power of Z eqw Let be the equivalent impedance of the collector line of the w-th doubly-fed wind turbine in the radial topology. Then, combining this with the principle that the total loss after equivalence equals the sum of the losses of each branch before equivalence, the collector system impedance in the equivalent grid-connected model of a single doubly-fed wind farm is calculated using the formula shown in Equation 4:

[0079]

[0080] In the formula, Z eqh P is the equivalent impedance of the collector line of the h-th doubly-fed wind turbine in the radial topology of the entire doubly-fed wind farm; h For the current flowing through the collector line impedance Z eqh The power of Z; eq This represents the equivalent impedance of the collector line of the doubly fed wind turbine after aggregation.

[0081] Step 3.3: Use the collector system impedance value in the single-unit equivalent grid-connected model of a doubly fed wind farm as the initial value for the iteration of the adaptive inertial weighted particle swarm intelligent algorithm;

[0082] Step 3.4: Use the adaptive inertial weighted particle swarm intelligence algorithm and the equivalent grid-connected model of a single doubly fed wind farm unit to perform preliminary identification of detailed model data and obtain preliminary results of parameter identification.

[0083] Step 4: Select the optimal parameters:

[0084] Step 4.1: Substitute the preliminary parameter identification results corresponding to each set of detailed model data into the single-unit equivalent grid-connected model of the doubly fed wind farm, and then conduct experiments under different operating conditions corresponding to the detailed model data to obtain the active power corresponding to each set of detailed model data.

[0085] Step 4.2 Calculate the average deviation of active power and the relative error of the average deviation for each set of detailed model data, and select the preliminary parameter identification result corresponding to the set of detailed model data with the smallest relative error as the optimal parameter identification result.

[0086] The formula for calculating the average deviation is shown in Equation 5:

[0087]

[0088] In the formula, F is the average deviation; X S X represents the per-unit value of the single-machine equivalent model data for the electrical quantities to be assessed; M The per-unit values ​​of the detailed model data for the electrical quantities to be assessed; K S_START K S_End These are the first and last indices of the single-machine equivalent model data within the calculation error interval, respectively; K M_START K M_End These are the first and last serial numbers of the detailed model data within the calculation error interval, respectively;

[0089] The formula for calculating the relative error of the average deviation is shown in Equation 6:

[0090]

[0091] In the formula, error is the relative error of the average deviation; x1 is the average deviation of active power; x min Substitute the preliminary results of the identification of n sets of parameters into the minimum value of the average deviation of active power under the same set of operating conditions.

[0092] Step 5: Establish the optimal single-machine equivalent model and perform an effectiveness analysis to verify the accuracy of the parameter identification results.

[0093] Step 5.1: Set the impedance value of the collection system in the single-unit equivalent grid-connected model of the doubly-fed wind farm as the optimal parameter identification result, and establish the optimal single-unit equivalent model of the doubly-fed wind farm;

[0094] Step 5.2: Run the optimal single-machine equivalent model under the operating conditions specified in Step 1. Input the active power command, reactive power command and series compensation input time of n sets of experimental operating conditions. Record the active power during the operation as the optimal equivalent model data. Calculate the average deviation between the optimal equivalent model data and the detailed model data according to Equation 5. Analyze the effectiveness of the optimal single-machine equivalent model to verify the accuracy of the parameter identification results.

[0095] The following description, in conjunction with a preferred embodiment, illustrates the content designed in the above embodiments.

[0096] 1. The experimental plan was developed according to step 1. The detailed operating conditions of the five groups of doubly fed wind farm models are shown in Table 1.

[0097] Table 1 Detailed Model Operating Conditions

[0098] P / pu Q / pu Series compensation input time / t Data collection Operating Condition 1 0.4 0 0.5 P Operating Condition 2 0.5 0 0.5 P Operating Condition 3 0.6 0 0.5 P Operating Condition 4 0.7 0 0.5 P Operating Condition 5 0.8 0 0.5 P

[0099] 2. Follow step 2 in Figure 1 The experimental platform shown in Table 1 completed the experimental scheme and collected active power data of the transmission lines of five sets of detailed models of doubly-fed induction generator (DFIG) wind farms. The detailed model of the DFIG grid-connected system can be found in [reference needed]. Figure 2 The model parameters are shown in Table 2.

[0100] Figure 1 The experimental platform shown consists of a doubly-fed wind farm, a transmission line series compensation device, a power grid simulator, and a data acquisition device. The transmission line series compensation device is used to set up series capacitors under five operating conditions, and the data acquisition device is used to collect detailed model data.

[0101] Table 2 Model Parameters

[0102]

[0103]

[0104] 3. Construct an equivalent grid-connected model of a single doubly-fed induction generator (DFIG) wind farm without predefined parameters. Assign values ​​to its internal parameters according to step 3, and use the algorithm to identify five sets of detailed model data, obtaining preliminary results for parameter identification. The topology of the DFIG wind farm grid-connected system single-unit equivalent model is shown below. Figure 3 The identification results are shown in Table 3, and the preliminary results of parameter identification are shown in Table 4.

[0105] Table 3 Comparison of Identification Results

[0106] result Error after identification Unidentified error Result 1 0.1791 0.4768 Result 2 0.1867 0.8519 Result 3 0.1354 0.8509 Result 4 0.0947 0.7395 Result 5 0.038 0.4675

[0107] Table 4 Preliminary Identification Results

[0108] result <![CDATA[Equivalent resistance of the collector system (R req / Ω)]]> <![CDATA[Equivalent reactance (X req / Ω) of the collector system]]> Result 1 0.3354 0.8369 Result 2 0.3467 1.1490 Result 3 0.3282 1.1947 Result 4 0.2277 1.2555 Result 5 0.2257 1.0357

[0109] 4. Following step 4, substitute the preliminary parameter identification results corresponding to each set of detailed model data into the single-unit equivalent model of the doubly-fed induction generator (DFIG) wind farm grid-connected system, and conduct experiments under different operating conditions corresponding to the detailed model data to obtain the active power corresponding to each set of detailed model data; calculate the average deviation and relative error of the active power corresponding to each set of detailed model data, calculate the sum of the relative errors of the average deviation of each set of preliminary parameter identification results under each set of operating conditions, and select the preliminary parameter identification results corresponding to the detailed model data with the smallest total relative error as the optimal parameter identification results; the optimal parameter identification results are: equivalent resistance of the collector system (R... req) = 0.3467Ω, the equivalent reactance of the collector system (X) req = 1.1490Ω. The average deviation of active power is shown in Table 5, and the relative error of the average deviation of active power is shown in Table 6.

[0110] Table 5 shows the average deviation of active power for each set of parameter identification results under each set of operating conditions.

[0111] Data / Results Result 1 Result 2 Result 3 Result 4 Result 5 Operating Condition 1 0.1791 1.0369 1.0536 0.8917 0.7261 Operating Condition 2 0.8422 0.1867 0.2406 0.3269 0.7439 Operating Condition 3 0.5391 0.1348 0.1354 0.1168 0.4317 Operating Condition 4 0.6127 0.1611 0.1155 0.0947 0.4653 Operating Condition 5 0.2649 0.2855 0.3116 0.3692 0.038

[0112] Table 6. Relative Error of Average Deviation of Active Power

[0113] Data / Results Result 1 Result 2 Result 3 Result 4 Result 5 Operating Condition 1 0 4.7895 4.8827 3.9787 3.0541 Operating Condition 2 3.5109 0 0.2886 0.7509 2.9844 Operating Condition 3 3.6155 0.1541 0 0.1592 2.6960 Operating Condition 4 5.4699 0.7011 0.2196 0 3.9134 Operating Condition 5 5.9710 6.5131 7.2 8.7157 0

[0114] 5. Substitute the optimal parameter identification results into the single-unit equivalent model of the doubly-fed induction generator (DFIG) wind farm grid-connected system according to step 5 to obtain the optimal single-unit equivalent model of the DFIG wind farm. Then, run the optimal single-unit equivalent model under the operating conditions corresponding to the five detailed model data, record the active power during the operation as the optimal equivalent model data, and calculate the average deviation between the optimal equivalent model data and the detailed model data. The results are shown in Table 7 below:

[0115] Table 7. Operation results of the optimal single-machine equivalent model under different operating conditions.

[0116] data Average deviation of active power Operating Condition 1 1.0369 Operating Condition 2 0.1867 Operating Condition 3 0.1348 Operating Condition 4 0.1611 Operating Condition 5 0.2855

[0117] According to the "NBT 31066-2015 Wind Turbine Electrical Simulation Model Modeling Guidelines", the deviation is within the allowable range, thus verifying the accuracy of the identification results of this invention.

[0118] Example 2

[0119] This embodiment provides a collector system parameter identification device suitable for equivalent modeling of doubly-fed wind farms, including:

[0120] The detailed model data acquisition module is used to acquire the active power data of the transmission line after the doubly fed wind farm enters the subsynchronous oscillation state under multiple experimental conditions, and uses the active power data of the transmission line under multiple experimental conditions as detailed model data.

[0121] The preliminary identification module is used to input multiple sets of detailed model data into a pre-constructed doubly fed wind farm single-unit equivalent grid-connected model for preliminary identification, and obtain preliminary identification results for multiple parameters;

[0122] The active power acquisition module is used to substitute the preliminary parameter identification results corresponding to each set of detailed model data into the single-unit equivalent grid-connected model of the doubly fed wind farm to obtain the active power corresponding to each set of detailed model data.

[0123] The optimal parameter identification result acquisition module is used to calculate the average deviation of active power and the relative error of the average deviation corresponding to each set of detailed model data, and select the preliminary parameter identification result corresponding to the set of detailed model data with the smallest relative error as the optimal parameter identification result.

[0124] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for parameter identification in a single-unit equivalent grid-connected model of a doubly-fed induction generator (DFIG) wind farm, characterized in that, include: The active power data of the transmission line after the doubly fed wind farm entered the subsynchronous oscillation state under multiple experimental conditions were obtained, and the active power data of the transmission line under multiple experimental conditions were used as detailed model data. Multiple sets of detailed model data are input into a pre-constructed equivalent grid-connected model of a single doubly fed wind farm unit for preliminary identification, and preliminary results of multiple parameter identification are obtained. Substitute the preliminary parameter identification results corresponding to each set of detailed model data into the single-unit equivalent grid-connected model of the doubly fed wind farm to obtain the active power corresponding to each set of detailed model data. Calculate the average deviation of active power and the relative error of the average deviation for each set of detailed model data, and select the preliminary parameter identification result corresponding to the set of detailed model data with the smallest relative error as the optimal parameter identification result. The active power data of the transmission line after the doubly-fed wind farm has entered a subsynchronous oscillation state, simulated under multiple experimental conditions, is used as detailed model data, thereby obtaining multiple sets of detailed model data, including: Calculate the relationship between the subsynchronous oscillation frequency of a doubly fed wind farm connected to the grid via series compensation and the reactance of the collector system; calculate the relationship between the subsynchronous oscillation amplitude of a doubly fed wind farm connected to the grid via series compensation and the resistance of the collector system. Based on the calculation results, multiple sets of experimental conditions were set up, and the experimental point was determined to be the AC side grid connection point; among them, any set of experimental conditions includes: active power command, reactive power command, and the time of connecting the line series compensation capacitor. Set the initial active and reactive power of the doubly fed wind farm, and set the disturbance parameters of the doubly fed wind farm through the series compensation capacitor time in any set of experimental conditions to perform transient simulation. Once the simulation enters the subsynchronous oscillation state, the active power data of the doubly fed wind farm transmission line is recorded and used as a set of detailed model data to obtain multiple sets of detailed model data. The multiple sets of detailed model data are input into a pre-constructed equivalent grid-connected model of a single doubly-fed wind farm unit for preliminary identification, resulting in preliminary identification results for multiple parameters, including: Based on the detailed model of the doubly fed wind farm obtained in advance, an equivalent grid-connected model of a single doubly fed wind farm unit is established. The parameters of the equivalent doubly fed wind turbine and transformer are obtained by the capacity weighting method. The parameters corresponding to the equivalent grid-connected model of a single doubly fed wind farm unit are set with the criterion that the active power and reactive power output of the wind farm before and after equivalence are equal. The complex trunk-type grid structure is transformed into a radial pure parallel structure. The impedance value of each doubly fed wind turbine collector line in the radial pure parallel structure is obtained. The impedance value of the collector system in the single-unit equivalent grid-connected model of the doubly fed wind farm is calculated by combining the principle that the total loss after equivalence is equal to the sum of the losses of each branch before equivalence. The impedance value of the collection system in the single-unit equivalent grid-connected model of a doubly fed wind farm is used as the initial value for the iteration of the adaptive inertial weighted particle swarm intelligent algorithm. The adaptive inertial weighted particle swarm intelligence algorithm and the equivalent grid-connected model of a single doubly fed wind farm were used to perform preliminary identification of detailed model data, and preliminary results of parameter identification were obtained.

2. The parameter identification method for a single-unit equivalent grid-connected model of a doubly-fed wind farm according to claim 1, characterized in that, The relationship between the subsynchronous oscillation frequency of a doubly-fed wind farm connected to the grid via series compensation and the reactance of the collector system is calculated as shown in Equation 1: （1） In the formula, For synchronization frequency, For the capacitive reactance of the series capacitor, For the series compensation degree of transmission lines, For the reactance of the transmission line, It is the equivalent total reactance of the entire system, including the generator, the collector system, and the transmission lines.

3. The parameter identification method for a single-unit equivalent grid-connected model of a doubly-fed wind farm according to claim 1, characterized in that, The relationship between the subsynchronous oscillation amplitude of a doubly-fed wind farm connected to the grid via series compensation and the resistance of the collector system is calculated as shown in Equation 2: （2） In the formula, This is the equivalent rotor resistance. For rotor resistance, This refers to the slip of the generator at the subsynchronous frequency. The total resistance of the doubly-fed wind farm grid-connected system. For generator stator resistance, For the collector system resistance of a doubly-fed wind farm, This represents the resistance of the power transmission line.

4. The parameter identification method for a single-unit equivalent grid-connected model of a doubly-fed wind farm according to claim 1, characterized in that, The impedance value of each doubly fed wind turbine collector line in the radial pure parallel structure is calculated using the formula shown in Equation 3: （3） In the formula, w represents any doubly fed wind turbine in a radial, purely parallel structure. Let be the actual impedance of the collector line of the j-th doubly-fed wind turbine in this trunk topology. For the impedance of the collector line power, Let be the equivalent impedance of the collector line of the w-th doubly fed wind turbine in a radial pure parallel structure.

5. The parameter identification method for a single-unit equivalent grid-connected model of a doubly-fed wind farm according to claim 1, characterized in that, The impedance value of the collection system in the equivalent grid-connected model of a single unit of a doubly-fed wind farm is calculated based on the principle that the total loss after equivalence equals the sum of the losses of each branch before equivalence. The calculation formula is shown in Equation 4: （4） In the formula, Let be the equivalent impedance of the collector line of the h-th doubly-fed wind turbine in the radial pure parallel structure of the entire doubly-fed wind farm; For the impedance of the collector line The power; This represents the equivalent impedance of the collector line of the doubly fed wind turbine after aggregation.

6. The parameter identification method for a single-unit equivalent grid-connected model of a doubly-fed wind farm according to claim 1, characterized in that, The formula for calculating the average deviation is as shown in Equation 5: （5） In the formula, The average deviation; The per-unit values ​​of the single-machine equivalent model data for the electrical quantities to be assessed; The per-unit values ​​are the detailed model data for the electrical quantities to be assessed; , These are the first and last sequence numbers of the single-machine equivalent model data within the calculation error interval, respectively. , These are the first and last serial numbers of the detailed model data within the calculation error interval, respectively; The formula for calculating the relative error of the average deviation is shown in Equation 6: （6） In the formula, The relative error is the average deviation. This represents the average deviation of active power. Substitute the preliminary results of the identification of n sets of parameters into the minimum value of the average deviation of active power under the same set of operating conditions.

7. The parameter identification method for a single-unit equivalent grid-connected model of a doubly-fed wind farm according to claim 6, characterized in that, Also includes: The impedance value of the collection system in the single-unit equivalent grid-connected model of a doubly fed wind farm is set as the optimal parameter identification result to establish the optimal single-unit equivalent model of the doubly fed wind farm. The optimal single-machine equivalent model is run under multiple experimental conditions. The active power command, reactive power command, and connection time of the series compensation capacitor are input for multiple experimental conditions. The active power during the operation is recorded as the optimal equivalent model data. The average deviation between the optimal equivalent model data and the detailed model data is calculated to analyze the effectiveness of the optimal single-machine equivalent model.

8. A collector system parameter identification device suitable for equivalent modeling of doubly-fed induction generator (DFIG) wind farms, used to implement the parameter identification method for the single-unit equivalent grid-connected model of a DFIG wind farm as described in any one of claims 1-7, characterized in that, include: The detailed model data acquisition module is used to acquire the active power data of the transmission line after the doubly fed wind farm enters the subsynchronous oscillation state under multiple experimental conditions, and uses the active power data of the transmission line under multiple experimental conditions as detailed model data. The preliminary identification module is used to input multiple sets of detailed model data into a pre-constructed doubly fed wind farm single-unit equivalent grid-connected model for preliminary identification, and obtain preliminary identification results for multiple parameters; The active power acquisition module is used to substitute the preliminary parameter identification results corresponding to each set of detailed model data into the single-unit equivalent grid-connected model of the doubly fed wind farm to obtain the active power corresponding to each set of detailed model data. The optimal parameter identification result acquisition module is used to calculate the average deviation of active power and the relative error of the average deviation corresponding to each set of detailed model data, and select the preliminary parameter identification result corresponding to the set of detailed model data with the smallest relative error as the optimal parameter identification result.

Images