Wind power plant equivalence method and system based on fast DB-SCAN algorithm and considering low-crossing characteristic

Abstract

The invention relates to the field of new energy modeling equivalence, in particular to a wind power plant equivalence method and system based on a fast DB-SCAN algorithm and considering a low-crossing characteristic. According to the method, the time constants of the generator terminal voltage, the active power, the reactive power and the generator stator winding are selected as grouping indexes, the steady-state operation state and the transient change characteristic can be reflected at the same time, and theoretical verification is passed. Then, a traditional DB-SCAN algorithm is improved, a particle swarm algorithm is introduced for automatic optimization, the practicability of the algorithm is enhanced, a K-D decision tree is introduced for preliminary data division, the time complexity of the algorithm is reduced, and the final fast DB-SCAN algorithm has high practicability.

Description

Technical Field

[0001] The present invention belongs to the technical field of multi-machine equivalent value of wind farms, and in particular relates to a wind farm equivalent value method based on a fast DB-SCAN algorithm and taking low penetration characteristics into consideration. Background Art

[0002] In recent years, my country's renewable energy has developed rapidly, the penetration rate of renewable energy in regional power grids has continued to rise, the scale of wind farms has expanded rapidly, and the capacity of wind turbines has increased significantly. At the same time, the operating characteristics of wind power show strong non-synchronous characteristics, which are significantly different from the operating characteristics of traditional synchronous units. In order to study the impact of wind power grid connection under high penetration rate on the power grid, it is necessary to build an electromagnetic model for expansion mechanism and impact analysis. However, if each wind turbine is modeled in detail when building the model, the model will be very complex and the system order will be too high, which is not only not conducive to the analysis of the research mechanism, but also the simulation speed is difficult to meet the actual requirements.

[0003] The equivalent value of a wind farm is divided into single-machine multiplication equivalent value and multi-machine aggregation equivalent value. After the single-machine multiplication equivalent value is established, the multiplication module equivalent value is detailed in the wind farm, and the equivalent accuracy is low, especially it cannot accurately reflect the transient characteristics of the wind farm, and it is difficult to meet actual needs. The multi-machine aggregation equivalent value is to select appropriate classification indicators, group all wind turbines according to the similarity of operating characteristics through a clustering algorithm, perform single-machine multiplication on each group, and finally obtain a multi-machine equivalent model.

[0004] For the selection of clustering indicators, the multi-machine equivalent method suitable for fault analysis needs to take into account both the steady-state operating state and the transient characteristics during the transient process. The current technical solution mainly uses state quantities and change trajectories for clustering. The current indicator selection mainly uses only steady-state state quantities, or multiple state quantities during the steady-state period and the fault period for clustering. At the same time, the impact of the fan control strategy during the fault period has not been considered in detail.

[0005] As for clustering algorithms, the current technologies used are K-means clustering algorithm, fuzzy C-means clustering, and Gaussian mixture model. Among them, K-means algorithm has simple principle and fast calculation speed and is widely used.

[0006] The selection of clustering indicators in the existing technology cannot simultaneously and well reflect the steady-state operation status and fault transient characteristics of the unit. In addition, the current wind turbines all have low voltage ride-through capabilities, and the low voltage ride-through control strategies of doubly fed units and permanent magnet synchronous units are significantly different. The existing technology does not consider the impact of the control strategy of the wind turbine during the fault on its fault transient characteristics, and cannot reflect the transient characteristics during the low voltage ride-through period.

[0007] The existing clustering algorithms are mainly traditional K-means clustering algorithms. The traditional K-means algorithm not only has high requirements for the selection of initial cluster centers, but also has poor clustering results if the selection is inappropriate. The cluster centers need to be manually determined, which lacks practical applicability. In addition, the K-means algorithm itself has strict requirements on the distribution of data points and cannot process data with arbitrary distribution. If the data is distributed in a ring or circle in a certain dimension, its clustering effect is difficult to meet the requirements. Although the traditional DB-SCAN algorithm can process data with arbitrary distribution and has good clustering effect, the algorithm is complex and needs to traverse to find expandable neighborhood points, which is time-consuming and not suitable for large-scale wind farm clustering. Summary of the invention

[0008] In view of the above problems existing in the prior art, the present invention is proposed.

[0009] In order to solve the above technical problems, the present invention provides the following technical solutions, which are wind farm equivalent methods based on fast DB-SCAN algorithm and considering low voltage ride-through characteristics, including: by respectively establishing models for the doubly fed units and permanent magnet direct drive units during steady state and low voltage ride-through, obtaining indicators of transient characteristics of wind turbines during low voltage ride-through, and combining the state quantity before steady state to obtain a grouping index that takes into account steady state operation and transient change characteristics;

[0010] Import the clustering index data into the fast DB-SCAN clustering algorithm for automatic optimization to obtain the optimal clustering result;

[0011] According to the optimal clustering results, each type of result is multiplied by a single machine to obtain a multi-machine equivalent model.

[0012] As a preferred scheme of the wind farm equivalence method based on the fast DB-SCAN algorithm and considering the low-throughput characteristics described in the present invention, the grouping indicators include selecting the type of wind turbines in the wind farm, the terminal voltage, the active power, the reactive power, and the time constant of the stator winding of the generator as grouping indicators.

[0013] As a preferred solution of the wind farm equivalent method based on the fast DB-SCAN algorithm and considering the low penetration characteristics described in the present invention, the grouping index can reflect the operating status and transient characteristics of the unit.

[0014] As a preferred solution of the wind farm equivalent method based on the fast DB-SCAN algorithm and considering the low-through characteristics described in the present invention, wherein: during the transient process of the doubly fed unit, the short-circuit current reaches the fault steady-state value from the initial value through the RLC discharge process, and the transient characteristics are only related to the terminal voltage, the active and reactive power before the fault, and the time constant of the unit.

[0015] As a preferred solution of the wind farm equivalence method based on the fast DB-SCAN algorithm and considering the low-through characteristics described in the present invention, wherein: the short-circuit current of the permanent magnet unit jumps from an initial value to a fault steady-state value during the transient process, and the transient characteristics are only related to the machine-end voltage and the active and reactive power before the fault; the machine-end voltage of the unit, the active and reactive power before the fault, the time constant of the unit and the machine-end voltage, and the active and reactive power before the fault are selected as clustering indicators to accurately represent the steady-state operation state and transient change characteristics of the unit.

[0016] As a preferred solution of the wind farm equivalence method based on the fast DB-SCAN algorithm and considering the low-penetration characteristics described in the present invention, the introduction of the fast DB-SCAN clustering algorithm for automatic optimization includes processing indicator data with arbitrary distribution in each dimension through density-based scalable neighborhood point search, introducing the KD decision tree for preliminary processing of the data, so that data with high similarity are densely distributed on the KD decision tree, and reducing the time required for DB-SCAN algorithm clustering.

[0017] As a preferred solution of the wind farm equivalent method based on the fast DB-SCAN algorithm and considering the low penetration characteristics described in the present invention, wherein: the optimal clustering result includes giving a density threshold and a neighborhood radius key parameter, and automatically searching for the optimal solution by introducing a particle swarm algorithm, and optimizing the density threshold and the neighborhood radius key parameters through continuous iteration, so that the clustering effect is optimal.

[0018] As a preferred solution of the wind farm equivalent system based on the fast DB-SCAN algorithm and considering the low penetration characteristics described in the present invention, it includes: a steady-state transient modeling module, a fast clustering analysis module, and a multi-level equivalent module.

[0019] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and is characterized in that when the processor executes the computer program, the steps of any one of the wind farm equivalent methods based on a fast DB-SCAN algorithm and considering low penetration characteristics are implemented.

[0020] A computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of any one of the methods for wind farm equivalence based on a fast DB-SCAN algorithm and considering low penetration characteristics are implemented.

[0021] The beneficial effects of the present invention are as follows: the beneficial effects of the present method are that a clustering index that can better reflect the steady-state and transient characteristics of a wind turbine is obtained through detailed theoretical derivation, and the multi-machine equivalent model obtained by clustering using the index has steady-state and transient characteristics that are closer to the detailed model than the model of the prior art, and can better reflect the external operating characteristics of the wind farm; at the same time, the invented fast DB-SCAN algorithm can not only process data with arbitrary distribution in each dimension with stronger applicability, but also the algorithm does not need to set parameters, can achieve automatic optimization, and has practical engineering value; finally, the time complexity of the algorithm is much lower than that of the prior art, and for large wind turbines, its operating speed can still meet engineering requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0023] Figure 1 An operation flow chart of a wind farm equivalent method based on a fast DB-SCAN algorithm and considering low penetration characteristics provided by one embodiment of the present invention.

[0024] Figure 2 A schematic diagram of an equivalent circuit of a doubly-fed wind turbine generator set after a fault according to a wind farm equivalent method based on a fast DB-SCAN algorithm and taking into account low-thru characteristics provided by an embodiment of the present invention.

[0025] Figure 3 A schematic diagram of a control principle diagram of a doubly-fed wind turbine according to an embodiment of the present invention, based on a fast DB-SCAN algorithm and a wind farm equivalent method taking into account low penetration characteristics.

[0026] Figure 4 A control principle diagram of a permanent magnet wind turbine according to an embodiment of the present invention, based on a fast DB-SCAN algorithm and a wind farm equivalent method taking low penetration characteristics into consideration. DETAILED DESCRIPTION

[0027] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0028] Example 1

[0029] Reference Figure 1-Figure 4 , which is the first embodiment of the present invention, provides a wind farm equivalent method based on a fast DB-SCAN algorithm and taking into account low penetration characteristics, such as Figure 1 As shown, including:

[0030] In view of the above-mentioned defects in the prior art, the present invention provides a multi-machine equivalent method based on a fast DB-SCAN clustering algorithm and considering low voltage ride-through characteristics, taking into account both the simulation steady-state operation state and transient change characteristics, and taking into account both simulation accuracy and engineering practicality. The wind farm equivalent method based on the fast DB-SCAN algorithm and considering low voltage ride-through characteristics, first of all, should select physical quantities that can accurately describe the steady-state operation state and transient change characteristics of the wind farm, and then use them as indicators for clustering. At the same time, the selection of indicator physical quantities should be independent of each other, and do not repeat the description of the same characteristics of the wind farm. The overlap of the characteristics reflected by the indicator physical quantities will also directly affect the clustering effect. In addition, the selected indicators should be able to reflect the transient change characteristics of the wind turbine during the low voltage ride-through period, so as to consider the impact of the low voltage ride-through strategy of the wind turbine on the transient characteristics of the wind turbine. This method selects the terminal voltage, active power, reactive power, and time constant of the stator winding of all wind turbines in the wind farm as wind farm clustering indicators; secondly, considering the distribution state of the selected indicators in each dimension, there may be a circular or annular distribution in a certain dimension. This method adopts the fast DB-SCAN algorithm as the clustering algorithm, and processes the indicator data with arbitrary distribution in each dimension through density-based scalable neighborhood point search. At the same time, in order to improve the complex search traversal and high time complexity of the traditional DB-SCAN clustering algorithm, the KD decision tree is introduced to preliminarily process the data, so that the data with high similarity is more closely distributed on the KD decision tree, which greatly reduces the time required for DB-SCAN algorithm clustering. In addition, the DB-SCAN algorithm needs to give the key parameters of density threshold and neighborhood radius. The setting of the parameters will directly determine the quality of the clustering effect. This method automatically searches for the best by introducing the particle swarm algorithm, and optimizes the key parameters of density threshold and neighborhood radius through continuous iteration, so that the clustering effect is optimal. No manual setting is required, which enhances the engineering practicality of this method. Finally, according to the optimal clustering result obtained after importing the clustering index data into the algorithm, the single machine multiplication is performed to obtain the multi-machine equivalent model, taking into account the equivalent accuracy and simulation speed. The overall wind farm equivalent method includes the following steps:

[0031] Step (1) selecting the type of each wind turbine in the wind farm, terminal voltage, active power, reactive power, and time constant of the stator winding of the generator as grouping indicators;

[0032] Step (2) importing the clustering index data into the fast DB-SCAN clustering algorithm for automatic optimization to obtain the optimal clustering result;

[0033] Step (3) Based on the optimal clustering result, each type of result is multiplied by a single machine to obtain a multi-machine equivalent model.

[0034] The technical roadmap of the equivalent method is as attached Figure 1 :

[0035] The principle that the grouping index selected in step (1) can accurately reflect the unit operating status and transient characteristics is as follows:

[0036] For doubly fed units:

[0037] First of all, the basic principle of the doubly fed unit is that the DC voltage udc is inverted to the excitation voltage Ur through the rotor side converter. Ur is controlled by the modulation ratio, so the external characteristics of the rotor excitation voltage can be expressed as:

[0038]

[0039] Where: A, n, K dc are the three-phase bridge current inversion coefficient, stator-rotor turns ratio and rotor-side converter modulation ratio of the rotor-side converter respectively; i dc , X rsc , Δu rsc They are DC bus current, converter equivalent arc-commutating reactance and IGBT voltage drop respectively.

[0040] When a transient fault occurs, Udc and speed remain basically unchanged under the corresponding control strategy. Using the motor convention, the space vector model of the fan in the dq coordinate system can be expressed as:

[0041]

[0042] Where: u s 、u r 、i s 、i r , s , r are the stator and rotor voltage, current and flux converted to the stator side respectively; L s , L r , L m are stator inductance, rotor inductance and excitation inductance respectively; ω s ,ω s-r They are the synchronous frequency and the slip angular frequency respectively.

[0043]

[0044] When a three-phase short circuit occurs in the power grid, the grid-side system is equivalent to the Thevenin equivalent circuit, where the equivalent potential on the system side is EL, the equivalent impedance from the system to the fault point is Z1L, and the equivalent impedance from the doubly fed wind turbine generator set to the fault point is Z 2L, the excess impedance is Zf, and the terminal voltage of the doubly-fed wind turbine generator is u s， L s , L r are stator and rotor leakage inductance, L g , R g are the filter inductance and resistance respectively. According to equations (1) to (3), we can get Figure 1 The equivalent circuit of the doubly-fed wind turbine generator after the fault is shown in the attached figure. Figure 2 The overall control strategy of the doubly fed unit is shown in the attached Figure 3 , 4 shown.

[0045] When a fault occurs, the unit control strategy will switch to low-throw control, change the reactive current to support the voltage, change the modulation ratio Kdc of the rotor-side converter to change the rotor excitation voltage u r , and then adjust the rotor excitation current. Finally, when the fault reaches a steady state, the rotor excitation current reaches a steady-state current reference value through a dynamic process. At this time, i s =i s∞ ,i r =i r∞ From the stator voltage equation (2), we can get:

[0046]

[0047] where i r∞ =i rd_ref +ji rq_ref , the steady-state rotor excitation current consists of the rotor active and reactive current reference values.

[0048]

[0049] When the fault reaches steady state, the short-circuit current value will be rq_ref 、i rd_ref , R s ,X,EL,Z 1L , Z 2L , Z f , Lm, ω s Determine, where each impedance constitutes the time constant of the unit transient, i rq_ref 、i rd_ref Is an unknown quantity.

[0050] The following analysis is about the steady state and low penetration irq_ref ,i rd_ref

[0051] According to the instantaneous power calculation formula:

[0052]

[0053] The rotor-side converter of the doubly-fed wind turbine generator adopts stator flux orientation, ψ sq =0, Then formula (3) can be written as

[0054]

[0055] According to equations (6) and (7), eliminating the stator current, the relationship between the rotor current at the initial moment and the active and reactive power output of the doubly-fed wind turbine generator set before the fault and the voltage before the fault can be expressed as:

[0056]

[0057] During the low-voltage breakout period, according to GB / T 19963-2011 Technical Regulations for Wind Farm Access to Power System,

[0058]

[0059] Combining equations (8) and (9), we get:

[0060]

[0061] It can be seen from formula (10) that the rotor active and reactive current reference value during the low-power run-through period is only related to the active and reactive power and the terminal voltage before the fault. Combined with formula (5), it can be seen that the steady-state short-circuit current after the fault is only related to the rotor active and reactive current reference value and the time constant.

[0062] The following analyzes the changes in short-circuit current during transient process:

[0063] During low voltage ride-through, the voltage needs to be changed by changing the excitation voltage Ur, so to analyze the change of short-circuit current, it is only necessary to analyze the change process of Ur.

[0064] Eliminating the rotor current from equation (3) yields the stator flux ψs, which is then substituted into the rotor voltage equation of equation (2) to yield:

[0065]

[0066] Convert equation (11) to the rotor rotating coordinate system:

[0067]

[0068] For the transient process of the doubly fed unit, the influence of the stator flux is much smaller than that of the rotor flux, and the stator flux time constant is small, so the influence of the stator flux in the transient process is ignored. At the same time, the time constant of the IGBT is much smaller than the generator winding flux time constant, so it can be considered that Kdc changes in steps during the transient process. In summary, from equations (1) and (12), it can be obtained that the rotor current and the rotor excitation voltage constitute an RLC discharge process during the transient process, and the rotor excitation voltage jumps from the initial value before the fault to the low-through control reference value. It can be expressed as:

[0069]

[0070] Considering the actual double-fed wind turbine parameters, equation (3) can be expressed as:

[0071]

[0072] From formula (14), we can see that the stator current and rotor current change in the same way during the transient process. Therefore, the stator short-circuit current can be expressed as:

[0073]

[0074] Where: i′ S0 、i S∞ It is the initial short-circuit current of the fault and the steady-state short-circuit current of the fault.

[0075] For permanent magnet synchronous direct drive fans:

[0076] The permanent magnet generator set has a full power converter, and the electrical disturbance of the power grid will not directly affect the permanent magnet generator set. Therefore, its short-circuit current transient characteristics are mainly determined by the grid-side converter control strategy during the fault period. According to the standard, the current reference value during the low-pass period is:

[0077]

[0078] i gd0 and i gq0 It can be solved by the power relationship before the fault. The grid-side converter of the permanent magnet wind turbine generator adopts grid voltage orientation, that is, u gd =u g ,u gd =0. Then the instantaneous power formula can be written as:

[0079]

[0080] Combining equation (17) and equation (18), we get:

[0081]

[0082] Order I g =i gd-ref +jigq-ref is the steady-state short-circuit current output by the permanent magnet wind turbine generator set after a fault, and is obtained from formula (13):

[0083]

[0084] It can be seen from formula (19) that the steady-state short-circuit current Ig output by the permanent magnet wind turbine generator set is determined by the terminal voltage ug and the fault active and reactive power.

[0085] For the transient change characteristics of permanent magnet units:

[0086] Similarly, the permanent magnet generator set is isolated from the grid by the full-power converter, and the impact of the grid does not directly affect the generator set, so the transient characteristics of the permanent magnet generator set mainly depend on the characteristics of the grid-side converter during the low-pass period. In the grid-side converter, due to the PI link of the voltage and current double loop, the dq axis current can quickly keep up with the current reference value, and the time constant of the IGBT is very small, so the transient process of the permanent magnet generator set can be considered as a step change.

[0087] From the above analysis, it can be seen that during the transient process of the doubly fed unit, the short-circuit current jumps from the initial value to the fault steady-state value through the RLC discharge process, and its transient characteristics are only related to the terminal voltage, the active and reactive power before the fault, and the time constant of the unit. During the transient process of the permanent magnet unit, the short-circuit current jumps from the initial value to the fault steady-state value, and its transient characteristics are only related to the terminal voltage and the active and reactive power before the fault. Therefore, selecting the terminal voltage of the unit, the active and reactive power before the fault, the time constant of the unit and the terminal voltage, and the active and reactive power before the fault as clustering indicators can more accurately represent the steady-state operation status and transient change characteristics of the unit.

[0088] Example 2

[0089] The second embodiment of the present invention is different from the previous embodiment in that:

[0090] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0091] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0092] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0093] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0094] Example 3

[0095] The third embodiment of the present invention provides a wind farm equivalent system based on a fast DB-SCAN algorithm and taking into account low penetration characteristics, characterized in that it includes a steady-state transient modeling module, a fast clustering analysis module, and a multi-level equivalent module.

[0096] The steady-state transient modeling module builds models for the doubly-fed units and permanent magnet direct-drive units during steady-state and low-voltage ride-through, respectively, to obtain indicators of transient characteristics of wind turbines during low-voltage ride-through, and combines the state quantity before steady-state to obtain clustering indicators that take into account both steady-state operation and transient change characteristics;

[0097] The fast clustering analysis module imports the clustering index data into the fast DB-SCAN clustering algorithm for automatic optimization to obtain the optimal clustering result;

[0098] The multi-level equivalent module multiplies each type of result by a single machine according to the optimal clustering result to obtain a multi-machine equivalent model.

[0099] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A wind farm equivalent method based on a fast DB-SCAN algorithm and considering low penetration characteristics, characterized by: include, By building models for the doubly-fed units and permanent magnet direct-drive units during steady state and low voltage ride-through, we obtain indicators of transient characteristics of wind turbines during low voltage ride-through, and combine the state quantities before steady state to obtain grouping indicators that take into account both steady state operation and transient change characteristics. Import the clustering index data into the fast DB-SCAN clustering algorithm for automatic optimization to obtain the optimal clustering result; According to the optimal clustering results, each type of result is multiplied by a single machine to obtain a multi-machine equivalent model.

2. The wind farm equivalent method based on the fast DB-SCAN algorithm and considering the low penetration characteristics according to claim 1, characterized in that: The grouping index includes selecting the type of each wind turbine in the wind farm, the terminal voltage, the active power, the reactive power, and the time constant of the stator winding of the generator as the grouping index.

3. The wind farm equivalent method based on the fast DB-SCAN algorithm and considering the low penetration characteristics according to claim 2, characterized in that: The grouping index reflects the operating status and transient characteristics of the unit.

4. The wind farm equivalent method based on the fast DB-SCAN algorithm and considering the low penetration characteristics according to claim 3, characterized in that: During the transient process of the doubly fed unit, the short-circuit current reaches the fault steady-state value from the initial value through the RLC discharge process. The transient characteristics are only related to the terminal voltage, the active and reactive power before the fault, and the time constant of the unit.

5. The wind farm equivalent method based on the fast DB-SCAN algorithm and considering the low penetration characteristics according to claim 4, characterized in that: During the transient process of the permanent magnet unit, the short-circuit current jumps from the initial value to the fault steady-state value. The transient characteristics are only related to the terminal voltage and the active and reactive power before the fault. The terminal voltage of the unit, the active and reactive power before the fault, the time constant of the unit and the terminal voltage, and the active and reactive power before the fault are selected as clustering indicators to accurately represent the steady-state operation status and transient change characteristics of the unit.

6. The wind farm equivalent method based on the fast DB-SCAN algorithm and considering the low penetration characteristics according to claim 5, characterized in that: The method of importing the fast DB-SCAN clustering algorithm for automatic optimization includes processing the indicator data with arbitrary distribution in each dimension through density-based scalable neighborhood point search, introducing the KD decision tree for preliminary data processing, so that the data with high similarity is densely distributed on the KD decision tree, and reducing the time required for DB-SCAN algorithm clustering.

7. The wind farm equivalent method based on the fast DB-SCAN algorithm and considering the low penetration characteristics according to claim 6, characterized in that: The method of obtaining the optimal clustering result includes giving a density threshold and a neighborhood radius key parameter, automatically searching for the optimal solution by introducing a particle swarm algorithm, and optimizing the density threshold and the neighborhood radius key parameter through iteration, so that the clustering effect is optimal.

8. A system based on the wind farm equivalent method based on the fast DB-SCAN algorithm and considering low penetration characteristics according to any one of claims 1 to 7, characterized in that: Including steady-state transient modeling module, fast clustering analysis module, and multi-level equivalent module.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images