Reactive power balance considering the construction of network hybrid wind farm reactive power voltage optimization control method and system

Abstract

The application discloses a reactive power and voltage optimization control method and system for a follow-and-construct network hybrid wind power plant considering reactive power balance, and the method comprises the following steps: S1, constructing a reactive power characteristic linearization model of a wind turbine generator of the hybrid wind power plant; S2, obtaining typical output scenes of the wind power plant and corresponding probabilities by using a scene analysis method; S3, performing day-ahead random optimization on the follow-and-construct network hybrid wind power plant according to the reactive power characteristic linearization model obtained in step S1, the typical output scenes of the wind power plant and the corresponding probabilities obtained in step S2, and obtaining an action plan of discrete action devices, OLTC and SCB; S4, performing intraday reactive power and voltage optimization control on the follow-and-construct network hybrid wind power plant based on the action plan of the discrete action devices determined in step S3, and performing switching judgment on an optimization control target and a mode according to a wind power output fluctuation level; and S5, controlling reactive power balance inside the follow-and-construct network hybrid wind power plant based on the judgment result of step S4.

Description

Technical Field

[0001] This invention belongs to the field of wind power generation control technology. It proposes a two-stage reactive power and voltage optimization control technology for a hybrid wind farm with grid connection that is based on mixed integer linear programming and considers reactive power balance. More specifically, it is a reactive power and voltage optimization control method and system for a hybrid wind farm with grid connection that considers reactive power balance. Background Technology

[0002] With the deepening of my country's strategic goal of "carbon peaking and carbon neutrality", wind power has seen a significant increase in installed capacity due to its low-carbon, clean, environmentally friendly and widely distributed characteristics. However, the uncertainty and volatility of wind power output have led to increasingly prominent voltage fluctuation problems in wind farms, and even the risk of voltage exceeding the limit, which poses many challenges to the safe and stable operation of the new energy power grid (Liu Chengxi, Zhang Zhaoyi, Lai Qiupin. Optimization of rapid reactive power tracking of wind farms considering short-term fluctuations in wind power [J]. Proceedings of the CSEE, 2023, 43(15): 5850-5863). Currently, the reactive power and voltage optimization control of wind farms generally adopts a centralized decision-making mode, which urgently needs to improve its adaptability to voltage fluctuations caused by wind power output fluctuations (Ouyang Jinxin, Tang Ting, Yao Jun, et al. Active Voltage Control for DFIG-Based Wind Farm Integrated Power System by Coordinating Active and Reactive Powers Under Wind Speed ​​Variations[J]. IEEE Transactions on Energy Conversion, 2019, 34(3): 1504-1511). This mode requires real-time monitoring of the entire field's operating status and the construction of a model characterizing the control rules, operating limits, and response characteristics of wind turbines and other reactive power regulation equipment, thereby achieving reactive power and voltage optimization control that considers the influence of various types of equipment in the wind farm. There are various control methods for wind turbines in a wind farm. The voltage at the turbine port can be controlled by adjusting the reactive power, thereby achieving reasonable regulation of the wind farm voltage and supporting the wind farm voltage. How to effectively utilize its reactive voltage regulation capability to cope with the uncertainty and fluctuation of wind power output leading to voltage fluctuations in wind farms, and thus improve the voltage control level of wind farms, has become a key issue (Wang Yaoxiang, Dai Chaobo, Yang Zhichang, et al. Reactive voltage control strategy for wind farms considering the reactive potential of wind turbine units [J]. Power System Protection and Control, 2022, 50(24): 83-90).

[0003] There are currently two main control methods for doubly-fed induction generator (DFIG) wind turbines: grid-following (GFL) control and grid-forming (GFM) control. Grid-following control cannot quickly and actively adjust voltage, making it difficult to adapt to the randomness and fluctuations in wind power output (Tian Xiaoyu, Zhao Haiyu, Zhai Wenhui, et al. Distributed adaptive decoupling control suitable for frequency and voltage support in hybrid wind farms [J]. Power System Protection and Control, 2025, 53(06): 126-39). To address the many problems faced by grid-following wind turbines, grid-forming control technology can be adopted. Grid-forming turbines can autonomously construct frequency and voltage, and can quickly adjust reactive power when the wind farm voltage fluctuates, effectively regulating the grid connection point voltage while maintaining their own port voltage. Therefore, hybrid wind farms that simultaneously deploy grid-connected and grid-connected turbines within the wind farm significantly improve the dynamic response and voltage support capabilities of the site compared to traditional grid-connected wind farms (Zhang Xing, Zhan Xiangdui, Wu Mengze, et al. A review of grid-connected / grid-connected hybrid mode control of grid-connected converters for high-penetration new energy power generation [J]. Automation of Electric Power Systems, 2024, 48(21): 1-15.). However, due to the complex reactive power characteristics of grid-connected wind turbines, existing research struggles to establish accurate models that characterize their reactive power dynamic characteristics in centralized optimization control. While existing modeling methods can account for the voltage support effect of grid-connected equipment, the reactive power control mode of grid-connected equipment is not clearly characterized, making it difficult to accurately represent the reactive power regulation process of grid-connected equipment.

[0004] Meanwhile, within wind farms, due to the long collection lines and the fact that multiple wind turbines are often connected to a single feeder, there is often an imbalance in reactive power distribution (Qin Xiaohui, Ni Fengyi, Zhang Yuanyuan, et al. A decentralized decision-making refined reactive power and voltage control strategy based on local reactive power balance within wind farms [J]. Power System Technology, 2024, 48(01): 272-282.), resulting in significant voltage deviation and losses. When wind power output fluctuates severely, the difference in reactive power characteristics between grid-connected and grid-connected turbines in hybrid wind farms amplifies the reactive power imbalance. The reactive power imbalance within the farm will expand from the turbines on a certain feeder to the turbines on different feeders. The reactive power regulation capability of grid-connected wind turbines is difficult to fully utilize, while grid-connected wind turbines reduce reactive power margin when frequently adjusting reactive power output, reducing the wind farm's ability to smooth voltage fluctuations and respond to faults caused by sudden events on a short time scale.

[0005] In summary, although existing research has thoroughly studied the reactive power and voltage regulation capabilities of wind turbines, it has not considered the impact of differences in wind turbine control methods. Therefore, it remains difficult to establish a linearized model for hybrid wind farms that can collaboratively optimize grid-connected / network-connected wind turbines. Furthermore, existing research also has certain shortcomings in addressing the reactive power imbalance problem within hybrid wind farms. The unreasonable utilization of reactive power by grid-connected / network-connected wind turbines weakens the wind farm's ability to cope with short-term voltage fluctuations and faults. Summary of the Invention

[0006] To address the aforementioned technical problems, this invention proposes a two-stage reactive power and voltage optimization control method and system for hybrid wind farms with grid connection and integrated operation, considering reactive power balance. The aim is to establish a linearized model for collaboratively optimizing the wind turbine generators with grid connection and integrated operation, effectively addressing the challenges of reactive power and voltage optimization control in wind farms caused by the uncertainty and volatility of wind power output, and to perform reactive power balance control on hybrid wind farms, thus rationally utilizing the reactive power and voltage regulation capabilities of the wind turbine generators with grid connection and integrated operation.

[0007] The present invention adopts the following technical solution:

[0008] A reactive power and voltage optimization control method for a hybrid wind farm with integrated grid and reactive power balance, comprising the following steps:

[0009] S1. Construct a linearized model of the reactive power characteristics of wind turbine units in a hybrid wind farm;

[0010] S2. Use scenario analysis to derive typical power output scenarios of wind farms and their corresponding probabilities;

[0011] S3. Based on the reactive power linearization model obtained in step S1, the typical power output scenarios and corresponding probabilities of the wind farm obtained in step S2, perform day-ahead stochastic optimization of the hybrid wind farm with / with grid connection to obtain the action plan of discrete action equipment OLTC (on-load tap changer transformer) and SCB (shunt capacitor bank).

[0012] S4. Based on the day-ahead discrete action plan of the equipment determined in step S3, perform reactive power and voltage optimization control of the intraday hybrid wind farm, and make switching judgments on the optimization control target and mode according to the wind power output fluctuation level.

[0013] Based on the judgment result of step S4, S5 controls the reactive power balance inside the hybrid wind farm.

[0014] Preferably, step S1 is as follows: In order to achieve unified modeling and collaborative optimization of wind turbines in hybrid wind farms, this part takes wind turbines in hybrid doubly fed wind farms as the research object for analysis, and establishes linearized reactive power characteristic models of wind turbines in hybrid and grid-connected wind farms respectively, so as to realize the collaborative optimization control of reactive power and voltage of wind turbines in hybrid and grid-connected wind farms.

[0015] First, the stator-side output power range of the doubly-fed induction generator (DFIG) is affected by the rotor-side converter control (maximum current limit), and the stator-side reactive power limit is:

[0016]

[0017] Among them, Q sGFLmax Q sGFLmin These are the upper and lower limits of the reactive power on the stator side of a grid-fed doubly-fed wind turbine, respectively; P sGFL The active power transmitted from the stator side of a grid-connected doubly-fed induction generator (DFIG) to the power grid; u s I represents the stator voltage amplitude. rmax X represents the maximum value of the rotor-side converter current. m X s These are respectively the magnetizing reactance and the stator reactance;

[0018] Then, ignoring switching losses, line losses, and other factors, the reactive power limit of the grid-side converter can be determined according to... The calculation yielded:

[0019]

[0020] Among them, Q cGFLmax Q cGFLmin These are the upper and lower limits of reactive power for the grid-side converter of a grid-connected doubly-fed wind turbine; P mec s represents the mechanical power input to the grid-connected doubly-fed wind turbine; s represents the slip rate.

[0021] Next, considering both the stator reactive power limit and the grid-side converter reactive power limit, the reactive power adjustment range of the grid-connected doubly-fed wind turbine can be obtained:

[0022]

[0023] Secondly, the grid-type DFIG will actively adjust the reactive power output to regulate the turbine port voltage based on the difference between the voltage and the actual voltage. The reactive power output of the grid-type turbine j in the wind farm at time t can be written as:

[0024]

[0025] in, and These are the voltage values ​​when the grid-type wind turbine reaches the upper and lower limits of reactive power, respectively. For grid-type wind turbine port voltage; K n This is the droop coefficient;

[0026] Then, in order to incorporate the grid-building function into the mixed-integer linear programming, the conditional term for the wind turbine port voltage needs to be avoided. Therefore, a linear model based on binary variables and matrix combinations needs to be established to simulate the grid-building droop function, incorporating the grid-building wind turbines into a centralized decision-making model (including the objective function and constraints):

[0027]

[0028] The above equation simulates the reactive voltage droop function of a grid-type wind turbine by multiplying a binary variable with the reactive power output matrix. The output matrix consists of reactive power under different voltage conditions. As shown in the above equation, the real-time reactive power output of the grid-type DFIG is... It can be obtained through vector α GFM, j, t and Calculated by multiplication. In centralized decision-making processes, α GFM, j,t,1 To α GFM, j,t,4 The sum of these four binary variables is constrained to 1 to ensure that the reactive power output of the grid-connected wind turbine j is unique and correct. Further, by using the Big M method to relax the above equation into inequality constraints related to the binary variables, the port voltage determination and reactive power output calculation of the grid-connected DFIG can be synchronized through the selection of the α value, as shown in the following equation:

[0029]

[0030] Where M is a sufficiently large positive number in the Big M method;

[0031] For the reactive power and voltage control process of wind farms K n The upper and lower limits of reactive power in the grid-type system are both constants. However, considering... When the reactive voltage droop function is in a non-limiting operating state and The product of these terms is a bilinear term, which needs to be linearized using the McCormick method:

[0032]

[0033]

[0034] Where c2 and c3 represent non-decision variables in the linear processing and have no effect on the linear processing, z j,t,2 With z j,t,3 V is an auxiliary variable; GFM,j,max With V GMF,j,minThe upper and lower limits of the voltage at node j of the grid-connected wind turbine.

[0035] Preferably, in step S2, the uncertainty and volatility of wind power output are addressed by day-ahead random optimization, and wind power output is modeled using scenario generation and reduction technology to obtain typical wind farm output scenarios and corresponding probabilities.

[0036] Preferably, in step S2, the Latin hypercube sampling algorithm is used to generate scenes by combining the wind speed probability distribution and historical wind power data. Then, the backward reduction method is used to reduce the scenes to obtain typical power output scenes with different probabilities. The wind speed probability distribution follows a Weibull distribution, and the probability density function of the Weibull distribution is as follows:

[0037]

[0038] Among them, v w (t) represents the wind speed at time t; k s and c s These are the shape and scale parameters of the Weibull distribution, respectively.

[0039] Preferably, in step S3, based on the typical wind power output scenarios and their corresponding probabilities obtained in step S2, and combined with the obtained reactive power characteristic linearization model, day-ahead stochastic optimization of the wind farm hybrid wind farm considering wind farm grid loss and voltage deviation is carried out to obtain the action plan of discrete action equipment OLTC and SCB.

[0040] First, the objective function is as follows:

[0041]

[0042] Where, n s The number of scenarios obtained through scenario analysis; v represents the node voltage of the wind farm, which is not a variable; For the probability corresponding to the scene, and These are the weighting coefficients; P loss,t,s and D v,t,s These represent the network loss and voltage deviation at time t under scenario s, respectively.

[0043] Then, the constraints are as follows:

[0044] Power flow constraints within a wind farm:

[0045]

[0046]

[0047] Where π is the set of all nodes in the wind farm; It is the set of all the terminal nodes connected to all branches that start from node j. P is the set of the first nodes connected to all branches whose final node is node j; GFL,j,t P GFM,j,t Q represents the active power output of node j in the root-type DFIG and the active power output of the network-type DFIG at time t, respectively; GFL,j,t Q GFM,j,t These represent the reactive power outputs of node j at time t for the following DFIG configuration and the network configuration DFIG configuration, respectively; P D,j,t Q D,j,t Let Q be the active load and reactive load at node j at time t, respectively, which are typically 0 for wind farms; SCB,j,t For the SCB reactive power compensation at node j of the wind farm at time t; l ij,t v is the square of the current amplitude of line (i,j) at time t; j,t k is the square of the voltage amplitude at node j of the wind farm at time t; ij,t Let k be the on-load tap-changing transformer (OLTC) ratio of line (i,j) at time t, where k is the ratio when the corresponding line is not equipped with an OLTC. ij,t =1.

[0048] Active power constraints for DFIG network construction:

[0049]

[0050] Among them, P GFL,j,max and P GFL,j,min The active power P of the grid-connected doubly fed wind turbine at node j at time t is respectively. GFL The upper and lower limits of P; GFM,j,max and P GFM,j,min The active power P of the grid-type doubly fed wind turbine at node j at time t is respectively. GFM The upper and lower limits.

[0051] Reactive power constraint with DFIG network

[0052] The reactive power constraint of a grid-type DFIG is mainly affected by the active power output, and is also limited by the stator and rotor current of the unit.

[0053] Among them: Q GFL,j,max and Q GFL,j,min The reactive power Q of the grid-connected doubly fed wind turbine at node j at time t is respectively. GFL The upper and lower limits of reactive power output. The main influencing factor on the reactive power output of a grid-connected DFIG is the active power of the DFIG. When the output of a grid-connected DFIG increases, the upper limit of reactive power will decrease accordingly.

[0054] Networked DFIG reactive power constraint

[0055] The reactive power of grid-connected wind turbines exhibits droop control characteristics. For centralized decision-making using MILP (Mixed-Integer Linear Programming), which is a nonlinear logical relationship based on selection, the reactive power of grid-connected DFIGs is linearized, and the constraints are shown in step S1.

[0056] SVG (Static Var Generator) reactive power constraint:

[0057]

[0058] Among them: Q SVG,j,max and Q SVG,j,min These are the upper and lower limits of the SVG reactive power compensation capacity of wind farm node j, respectively.

[0059] OLTC constraints

[0060] The on-load tap-changing transformer is located at branch (i, j) and has a total of taps y. T There are K, respectively. ij,1 , K ij,2 ,…, K ij, yT The following constraints need to be met:

[0061]

[0062] Where: b T,ij,y b is a 0-1 variable. T,ij,y ∈{0,1}.

[0063] in, Since it is a bilinear term, it can be linearized using the Big M method, and the resulting linear form is:

[0064]

[0065] Where: z T,ij,y As an auxiliary variable; due to the limited number of actions N in the OLTC for wind farms. OLTC The number of OLTC actions is limited as follows:

[0066]

[0067] Where: y OLTCij,t Let y be the action marker of the OLTC on line ij at time t. When the OLTC gear changes, y OLTCij,t Actions can be recorded. This represents the maximum number of actions performed by the OLTC on line ij within 24 hours.

[0068] Parallel capacitor bank constraints:

[0069] The number of parallel capacitor banks located at node j that can be switched is n. SCBj,t Each group has a capacity of Q. SCB0,t The following constraints need to be met in the reactive power optimization control of wind farms:

[0070]

[0071] The above equation is a discrete optimization model, which can be transformed into a 0-1 integer programming model using the binary expansion method:

[0072]

[0073] Where: n is determined according to the number of groups set by SCB; δ0, δ1, …, δ n As a binary auxiliary variable, the constraints on the number of SCB actions and the calculation of the number of actions are handled as follows:

[0074]

[0075] Where: n SCBi,t Let n be the action marker of the SCB at time t on node i. When the SCB gear changes, n SCBi,t Actions can be recorded. Let SCB be the maximum number of actions performed on node i within 24 hours.

[0076] Preferably, in step S4, the daytime discrete action plan of the equipment is determined according to step S3, and the reactive power and voltage optimization control of the mixed wind farm is carried out. The switching judgment of the optimization control target and mode is made according to the wind power output fluctuation level.

[0077] Taking into account the economic efficiency of wind farm operation, voltage quality, and reactive power balance of integrated wind farms, this paper introduces a wind power output fluctuation level criterion, dynamically adjusts the objective function structure and scheduling time interval of the optimization model, improves the response capability of integrated wind farms to severe fluctuations in wind power output, and realizes the rational utilization of reactive power resources during operation.

[0078] Specifically, let the fluctuation range of wind power output be... Determine whether it exceeds the set threshold ε (ε=15%).

[0079]

[0080] When wind power output fluctuations are within an acceptable range (i.e.) When the objective function is J1, the optimization objective function focuses on minimizing voltage deviation and network loss. The reactive power optimization control model for a hybrid wind farm with integrated / networked structures can be composed of J1, power flow constraints, SVG, and reactive power constraints with integrated / networked structures. This ensures safe operation without serious reactive power imbalance. The daily dispatch interval is maintained at once every 15 minutes to reduce computational burden while also considering economic efficiency.

[0081] When wind power output fluctuates drastically (i.e.) When optimizing the objective function, a deviation term between the reactive power output of the wind turbine and its reactive power balance reference value is introduced. The objective function adopts J2. Considering the reactive power balance optimization model of the grid-connected / networked system, it can be composed of J2, power flow constraints, SVG, and grid-connected / networked reactive power constraints. The optimization control interval is shortened to once every 5 minutes. Based on the reactive power demand of the reactive power balance area calculated in real time, the reactive power output of the grid-connected wind turbines and SVG is adjusted to ensure that the reactive power output of the grid-connected wind turbines is within a reasonable range. This provides a certain margin for dealing with extreme situations and faults, and improves the ability of the grid-connected / networked hybrid wind farm to cope with sudden accidents.

[0082] The objective functions under different conditions are as follows:

[0083]

[0084] Where: F represents the set of grid-connected wind turbine units, F n To match the number of grid-connected wind turbine units.

[0085] Preferably, step S5 is as follows: by dividing the reactive power balance zone, calculating the command value of reactive power balance for grid-connected wind turbines and SVG, and introducing reactive power balance into the objective function, the reactive power potential of grid-connected wind turbines is fully utilized to avoid the waste of reactive power resources of grid-connected wind turbines, and realizing grid-connected reactive power collaborative balance control under the centralized decision-making of wind farms.

[0086] Reactive power balance control of wind turbine units

[0087] For grid-connected wind turbines, they have the ability to autonomously build voltage and will actively adjust the output reactive power to cope with voltage deviations. In addition, the reactive power response speed of grid-connected wind turbines is fast. Therefore, in the reactive power balance control process of grid-connected wind farms, the reactive power regulation capability of grid-connected units should be fully utilized so that the reactive power generated by grid-connected wind turbines near the collecting bus can meet the zonal balance requirements. There is no need for grid-connected wind turbines to generate additional reactive power support voltage, thus achieving a reactive power balance state and enabling the grid-connected wind farm to achieve reactive power balance.

[0088] This invention uses a two-dimensional numbering method to divide the reactive power balance region of grid-connected wind turbines. The region corresponding to wind turbine j is defined as the m,n balance region, with wind turbine j as the reactive power center, and also includes the wind turbine's box-type transformer and half of the feeder lines between adjacent nodes. When the wind turbine transmits power outward, certain reactive power losses are generated through the feeder lines and transformers. In order to maintain the reactive power balance of the turbine terminal voltage and the wind farm, the reactive power output of the grid-connected wind turbines must meet the reactive power losses during the transmission process.

[0089] The reactive power loss of the box-type transformer in the balance area of ​​the grid-connected wind turbine m,n is:

[0090]

[0091] In the formula: x T, j The reactance of the box-type transformer corresponding to the grid-type wind turbine j.

[0092] The reactive power loss of the feeder lines within the balance area of ​​the grid-connected wind turbine m,n is:

[0093]

[0094] Therefore, the reactive power loss in the balance region of the grid-type wind turbine m,n is:

[0095]

[0096] Based on the compensation of reactive power loss within the reactive power balance zone, the following formula is given as the reference value of reactive power for adjusting the terminal voltage of grid-connected wind turbines. Under severe wind speed fluctuations, the grid-connected / mixed wind farm can achieve reactive power balance within the wind farm while maintaining the terminal voltage of the wind turbines within the allowable range, leaving a certain reactive power margin for grid-connected DFIGs and improving the safety margin of wind farm operation.

[0097]

[0098] Reactive power balance control in wind farms (PCC):

[0099] The reactive power balance area is formed by taking the PCC point of the hybrid wind farm as the core, and also including the main step-up transformer of the wind farm and half of the first section of each feeder on the low-voltage side connected to the collection bus.

[0100] The reactive power loss of the main step-up transformer in the PCC balancing area of ​​the wind farm is:

[0101]

[0102] The reactive power losses of the feeder lines and transmission lines connecting the PCC balance area and the collection bus of the wind farm are as follows:

[0103]

[0104] In the formula: b is the number of grid-connected wind turbine units within the PCC balance area on the side of the connecting busbar.

[0105] Within the wind farm, the SVG (Static Var Generator) on the low-voltage side of the main step-up transformer at the collection station serves as a controllable reactive power source to provide reactive power. Under severe wind speed fluctuations, reactive power losses in the PCC (Power Process Control) balance area of ​​the wind farm are calculated. Within the adjustable range of the SVG, reactive power losses are compensated to regulate the voltage of the collection bus and PCC points, achieving reactive power balance within the PCC area of ​​the wind farm. This avoids excessive use of reactive power resources by grid-connected wind turbines due to voltage fluctuations at the collection bus, and provides a certain reactive power margin for grid-connected DFIG (Diverterless Power Injection) systems. The reactive power reference value of the SVG is:

[0106]

[0107] This invention also discloses a reactive power and voltage optimization control system for a grid-connected hybrid wind farm that considers reactive power balance, used to execute the above method, which includes the following modules:

[0108] Reactive power characteristic linearization model construction module: Constructs a reactive power characteristic linearization model for wind turbine units in a hybrid wind farm;

[0109] Scenario Analysis Module: Uses scenario analysis to derive typical power output scenarios for wind farms and their corresponding probabilities;

[0110] Action planning module: Based on the reactive power characteristic linearization model, typical power output scenarios of wind farms and corresponding probabilities, day-ahead stochastic optimization of hybrid wind farms with grid connection is performed to obtain the action plan of discrete action equipment;

[0111] Judgment Module: Based on the determined action plan of discrete action equipment, it performs intraday reactive power and voltage optimization control of mixed wind farms, and makes judgments on the switching of optimization control targets and methods according to the wind power output fluctuation level.

[0112] Control module: Based on the judgment results, it controls the reactive power balance inside the hybrid wind farm with grid and ground.

[0113] The present invention has the following technical effects:

[0114] This invention addresses the reactive power and voltage coordination optimization problem of grid-connected and grid-connected wind turbines in hybrid wind farms. It proposes a modeling method based on a linear combination of binary variables and the reactive power output matrix, effectively solving the modeling problem in the coordinated optimization of grid-connected and grid-connected wind turbines. This invention constructs a two-stage reactive power and voltage optimization control model for day-ahead and intraday periods, effectively improving the wind farm's adaptability to random fluctuations and operational safety. Based on the reactive power balance region calculation, this invention fully utilizes the reactive power regulation capabilities of grid-connected wind turbines, achieving a reasonable distribution of reactive power within the wind farm and avoiding the reactive power imbalance problem in hybrid wind farms. Attached Figure Description

[0115] Figure 1 This is a flowchart of a preferred embodiment of the present invention for a reactive power and voltage optimization control method for a hybrid wind farm that considers reactive power balance.

[0116] Figure 2 This is a preferred embodiment of the topology of a hybrid wind farm in a certain region.

[0117] Figure 3 This is a typical power output scenario diagram of a wind turbine unit according to a preferred embodiment of the present invention.

[0118] Figure 4 This is a diagram showing the voltage conditions of a hybrid wind farm after random optimization according to a preferred embodiment of the present invention.

[0119] Figure 5 This is a diagram illustrating the operation of the reactive power compensation devices SVG, OLTC, and SCB according to a preferred embodiment of the present invention.

[0120] Figure 6 This is a power output diagram of a wind farm during the day from 02:00 to 4:00, according to a preferred embodiment of the present invention.

[0121] Figure 7 This is a voltage diagram of a wind farm under hybrid control with grid and ground without reactive power balancing control, according to a preferred embodiment of the present invention.

[0122] Figure 8 This is a reactive power output diagram of a grid-connected wind turbine generator in a wind farm according to a preferred embodiment of the present invention.

[0123] Figure 9 This is a reactive power reference value diagram of a grid-type wind turbine under the reactive power balancing control method of a preferred embodiment of the present invention.

[0124] Figure 10 This is a reactive power output diagram of a grid-type wind turbine generator according to a preferred embodiment of the present invention.

[0125] Figure 11 This is a preferred embodiment of the wind farm voltage diagram considering grid-type reactive power balance control.

[0126] Figure 12 This is a block diagram of a preferred embodiment of the present invention for a reactive power and voltage optimization control system for a hybrid wind farm that considers reactive power balance. Detailed Implementation

[0127] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are merely illustrative of the embodiments of the present invention and are not intended to limit the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of this application. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout.

[0128] It should be noted that the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, such as a process, method, system, product, or server that includes a series of steps or units, not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such process, method, product, or device.

[0129] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0130] like Figure 1 As shown, this embodiment provides a reactive power and voltage optimization control method for a hybrid wind farm that considers reactive power balance, specifically including the following steps:

[0131] S1. Taking the wind turbine units of the hybrid doubly fed wind farm as the research object, we will analyze and establish linearized models of reactive power characteristics of wind turbine units in the hybrid wind farms of the doubly fed and grid-connected wind farms, respectively, to realize the reactive power and voltage coordinated optimization control of the wind turbine units of the doubly fed and grid-connected wind farms.

[0132] First, the stator-side output power range of the doubly-fed induction generator (DFIG) is affected by the rotor-side converter control (maximum current limit), and the stator-side reactive power limit is:

[0133]

[0134] Among them, Q sGFLmax Q sGFLmin These are the upper and lower limits of the reactive power on the stator side of a grid-fed doubly-fed wind turbine, respectively; P sGFL The active power transmitted from the stator side of a grid-connected doubly-fed induction generator (DFIG) to the power grid; u s I represents the stator voltage amplitude. rmaxX represents the maximum value of the rotor-side converter current. m X s These are respectively the magnetizing reactance and the stator reactance;

[0135] Then, ignoring switching losses, line losses, and other factors, the reactive power limit of the grid-side converter can be determined according to... The calculation yielded:

[0136]

[0137] Among them, Q cGFLmax Q cGFLmin These are the upper and lower limits of reactive power for the grid-side converter of a grid-connected doubly-fed wind turbine; P mec s represents the mechanical power input to the grid-connected doubly-fed wind turbine; s represents the slip rate.

[0138] Next, considering both the stator reactive power limit and the grid-side converter reactive power limit, the reactive power adjustment range of the grid-connected doubly-fed wind turbine can be obtained:

[0139]

[0140] Secondly, the grid-type DFIG will actively adjust the reactive power output to regulate the turbine port voltage based on the difference between the voltage and the actual voltage. The reactive power output of the grid-type turbine j in the wind farm at time t can be written as:

[0141]

[0142] in, and These are the voltage values ​​when the grid-type wind turbine reaches the upper and lower limits of reactive power, respectively. For grid-type wind turbine port voltage; K n This is the droop coefficient;

[0143] Then, in order to incorporate the grid-building function into mixed-integer linear programming, the conditional term for wind turbine port voltage needs to be avoided. Therefore, this paper establishes a linear model based on binary variables and matrix combinations to simulate the grid-building droop function, incorporating grid-built wind turbines into a centralized decision-making model:

[0144]

[0145] The above equation simulates the reactive voltage droop function of a grid-type wind turbine by multiplying a binary variable with the reactive power output matrix. The output matrix consists of reactive power under different voltage conditions. As shown in the above equation, the real-time reactive power output of the grid-type DFIG is... It can be obtained through vector α GFM, j, t and Calculated by multiplication. In centralized decision-making processes, α GFM, j,t,1To α GFM, j,t,4 The sum of these four binary variables is constrained to 1 to ensure that the reactive power output of the grid-connected wind turbine j is unique and correct. Further, by using the Big M method to relax the above equation into inequality constraints related to the binary variables, the port voltage determination and reactive power output calculation of the grid-connected DFIG can be synchronized through the selection of the α value, as shown in the following equation:

[0146]

[0147] For the reactive power and voltage control process of wind farms K n The upper and lower limits of reactive power in the grid-type system are both constants. However, considering... When the reactive voltage droop function is in a non-limiting operating state and The product of these terms is a bilinear term, which needs to be linearized using the McCormick method:

[0148]

[0149] Where: z j,t,2 With z j,t,3 V is an auxiliary variable; GFM,j,max With V GFM,j,min The upper and lower limits of the voltage at node j of the grid-connected wind turbine.

[0150] S2. By optimizing the random day-ahead to address the uncertainty and volatility of wind power output, the typical output scenarios of wind farms and their corresponding probabilities are derived using scenario analysis.

[0151] In this step, combining the wind speed probability distribution and historical wind power data, the Latin hypercube sampling algorithm is used to generate scenarios. Then, the backward reduction method is used to reduce the scenarios to obtain typical power output scenarios with different probabilities. The wind speed probability distribution follows a Weibull distribution, and the probability density function of the Weibull distribution is as follows:

[0152]

[0153] Among them, v w (t) represents the wind speed at time t; k s and c s These are the shape and scale parameters of the Weibull distribution, respectively.

[0154] S3. Combining the reactive power linearization model obtained in step S1, the typical power output scenarios and corresponding probabilities of the wind farm obtained in step S2, and considering the wind farm grid loss and voltage deviation, the day-ahead stochastic optimization of the wind farm with the grid connection / connection hybrid type is carried out to obtain the action plan of the discrete action equipment OLTC and SCB.

[0155] First, the objective function is as follows:

[0156]

[0157] in, and These are the weighting coefficients; P loss,t,s and D v,t,s These represent the network loss and voltage deviation at time t under scenario s, respectively.

[0158] Then, the constraints are as follows:

[0159] Power flow constraints within a wind farm:

[0160]

[0161] Where π is the set of all nodes in the wind farm; It is the set of all the terminal nodes connected to all branches that start from node j. P is the set of the first nodes connected to all branches whose final node is node j; GFL,j,t P GFM,j,t Q represents the active power output of node j in the root-type DFIG and the active power output of the network-type DFIG at time t, respectively; GFL,j,t Q GFM,j,t These represent the reactive power outputs of node j at time t for the following DFIG configuration and the network configuration DFIG configuration, respectively; P D,j,t Q D,j,t Let Q be the active load and reactive load at node j at time t, respectively, which are typically 0 for wind farms; SCB,j,t For the SCB reactive power compensation at node j of the wind farm at time t; l ij,t v is the square of the current amplitude of line (i,j) at time t; j,t k is the square of the voltage amplitude at node j of the wind farm at time t; ij,t Let k be the on-load tap-changing transformer (OLTC) ratio of line (i,j) at time t, where k is the ratio when the corresponding line is not equipped with an OLTC. ij,t =1.

[0162] Active power constraints for DFIG network construction:

[0163]

[0164] Among them, P GFL,j,max and P GFL,j,min The active power P of the grid-connected doubly fed wind turbine at node j at time t is respectively. GFL The upper and lower limits of P; GFM,j,max and P GFM,j,min The active power P of the grid-type doubly fed wind turbine at node j at time t is respectively. GFMThe upper and lower limits.

[0165] Reactive power constraints with DFIG network type:

[0166] The reactive power constraint of a grid-type DFIG is mainly affected by the active power output, and is also limited by the stator and rotor current of the unit.

[0167]

[0168] Among them: Q GFL,j,max and Q GFL,j,min The reactive power Q of the grid-connected doubly fed wind turbine at node j at time t is respectively. GFL The upper and lower limits of reactive power output. The main influencing factor on the reactive power output of a grid-connected DFIG is the active power of the DFIG. When the output of a grid-connected DFIG increases, the upper limit of reactive power will decrease accordingly.

[0169] Networked DFIG reactive power constraints:

[0170] The reactive power of grid-connected wind turbines exhibits droop control characteristics. For centralized decision-making using MILP, which is a nonlinear logic relationship based on selection, the reactive power of grid-connected DFIGs is linearized, and the constraints are shown in step S1.

[0171] SVG reactive power constraints:

[0172]

[0173] in: and These are the upper and lower limits of the SVG reactive power compensation capacity of wind farm node j, respectively.

[0174] OLTC constraints:

[0175] The on-load tap-changing transformer is located at branch (i, j) and has a total of taps y. T One, respectively The following constraints need to be met:

[0176]

[0177] Where: b T,ij,y b is a 0-1 variable. T,ij,y ∈{0,1}.

[0178] in, Since it is a bilinear term, it can be linearized using the Big M method, and the resulting linear form is:

[0179]

[0180] Where: z T,ij,yAs an auxiliary variable; due to the limited number of actions N in the OLTC for wind farms. OLTC The number of OLTC actions is limited as follows:

[0181]

[0182] Where: y OLCTij,t Let y be the action marker of the OLTC on line ij at time t. When the OLTC gear changes, y OLCTij,t Actions can be recorded. This represents the maximum number of actions performed by the OLTC on line ij within 24 hours.

[0183] Parallel capacitor bank constraints:

[0184] The number of parallel capacitor banks located at node j that can be switched is n. SCBj,t Each group has a capacity of Q. SCB0,t The following constraints need to be met in the reactive power optimization control of wind farms:

[0185]

[0186] The above equation is a discrete optimization model, which can be transformed into a 0-1 integer programming model using the binary expansion method:

[0187]

[0188] Where: n is determined according to the number of groups set by SCB; δ0, δ1, …, δ n As a binary auxiliary variable, the constraints on the number of SCB actions and the calculation of the number of actions are handled as follows:

[0189]

[0190] Where: n SCBi,t Let n be the action marker of the SCB at time t on node i. When the SCB gear changes, n SCBi,t Actions can be recorded. Let SCB be the maximum number of actions performed on node i within 24 hours.

[0191] S4. Based on the determined day-ahead discrete action equipment operation plan, perform intraday reactive power and voltage optimization control for hybrid wind farms, and determine the switching of optimization control targets and methods based on the wind power output fluctuation level.

[0192] Taking into account the economic efficiency of wind farm operation, voltage quality, and reactive power balance of integrated wind farms, this paper introduces a wind power output fluctuation level criterion, dynamically adjusts the objective function structure and scheduling time interval of the optimization model, improves the response capability of integrated wind farms to severe fluctuations in wind power output, and realizes the rational utilization of reactive power resources during operation.

[0193] Specifically, let the fluctuation range of wind power output be... Determine whether it exceeds the set threshold ε.

[0194]

[0195] When wind power output fluctuations are within an acceptable range (i.e.) When the objective function is J1, the optimization objective function focuses on minimizing voltage deviation and network loss. The reactive power optimization control model for a hybrid wind farm with integrated / networked structures can be composed of J1, power flow constraints, SVG, and reactive power constraints with integrated / networked structures. This ensures safe operation without serious reactive power imbalance. The daily dispatch interval is maintained at once every 15 minutes to reduce computational burden while also considering economic efficiency.

[0196] When wind power output fluctuates drastically (i.e.) When optimizing the objective function, a deviation term between the reactive power output of the wind turbine and its reactive power balance reference value is introduced. The objective function adopts J2. Considering the reactive power balance optimization model of the grid-connected / networked system, it can be composed of J2, power flow constraints, SVG, and grid-connected / networked reactive power constraints. The optimization control interval is shortened to once every 5 minutes. Based on the reactive power demand of the reactive power balance area calculated in real time, the reactive power output of the grid-connected wind turbines and SVG is adjusted to ensure that the reactive power output of the grid-connected wind turbines is within a reasonable range. This provides a certain margin for dealing with extreme situations and faults, and improves the ability of the grid-connected / networked hybrid wind farm to cope with sudden accidents.

[0197] The objective functions under different conditions are as follows:

[0198]

[0199] Where: F represents the set of grid-connected wind turbine units, F n To match the number of grid-connected wind turbine units.

[0200] S5. Based on the judgment result of step S4, when the wind power output fluctuates drastically, strengthen the reactive power balance control inside the hybrid wind farm, shorten the optimized control interval to once every 5 minutes, and carry out reactive power balance control on the hybrid wind farm by dividing the reactive power balance zone and setting the reactive power balance reference value of the grid-connected wind turbine.

[0201] In this step, by dividing the reactive power balance into zones, the command values ​​for reactive power balance of grid-connected wind turbines and SVG are calculated, and reactive power balance is introduced into the objective function. By making full use of the reactive power potential of grid-connected wind turbines, the waste of reactive power resources of grid-connected wind turbines is avoided, and grid-connected reactive power collaborative balance control under the centralized decision-making of wind farms is realized.

[0202] Reactive power balance control of wind turbine units:

[0203] For grid-connected wind turbines, they have the ability to autonomously build voltage and will actively adjust the output reactive power to cope with voltage deviations. In addition, the reactive power response speed of grid-connected wind turbines is fast. Therefore, in the reactive power balance control process of grid-connected wind farms, the reactive power regulation capability of grid-connected units should be fully utilized so that the reactive power generated by grid-connected wind turbines near the collecting bus can meet the zonal balance requirements. There is no need for grid-connected wind turbines to generate additional reactive power support voltage, thus achieving a reactive power balance state and enabling the grid-connected wind farm to achieve reactive power balance.

[0204] This embodiment uses a two-dimensional numbering method to divide the reactive power balance area of ​​the grid-connected wind turbine. The area corresponding to wind turbine j is defined as the m,n balance area. This area takes wind turbine j as the reactive power center and includes the wind turbine's box-type transformer and half of the feeder lines between adjacent nodes. When the wind turbine transmits power outward, certain reactive power losses will occur through the feeder lines and transformers. In order to maintain the reactive power balance of the turbine terminal voltage and the wind farm, the reactive power output of the grid-connected wind turbine must meet the reactive power losses during the transmission process.

[0205] The reactive power loss of the box-type transformer in the balance area of ​​the grid-connected wind turbine m,n is:

[0206]

[0207] In the formula: x T, j The reactance of the box-type transformer corresponding to the grid-type wind turbine j.

[0208] The reactive power loss of the feeder lines within the balance area of ​​the grid-connected wind turbine m,n is:

[0209]

[0210] Therefore, the reactive power loss in the balance region of the grid-type wind turbine m,n is:

[0211]

[0212] Based on the reactive power loss within the reactive power balance zone, the reactive power reference value for adjusting the terminal voltage of the grid-connected wind turbine is given as shown in Equation (36). In the case of severe wind speed fluctuations, the grid-connected / mixed wind farm can achieve reactive power balance within the wind farm while maintaining the terminal voltage of the wind turbine within the allowable range, leaving a certain reactive power margin for the grid-connected DFIG and improving the safety margin of wind farm operation.

[0213]

[0214] Reactive power balance control in wind farms (PCC):

[0215] The reactive power balance area is formed by taking the PCC point of the hybrid wind farm as the core, and also including the main step-up transformer of the wind farm and half of the first section of each feeder on the low-voltage side connected to the collection bus.

[0216] The reactive power loss of the main step-up transformer in the PCC balancing area of ​​the wind farm is:

[0217]

[0218] The reactive power losses of the feeder lines and transmission lines connecting the PCC balance area and the collection bus of the wind farm are as follows:

[0219]

[0220] In the formula: b is the number of grid-connected wind turbine units within the PCC balance area on the side of the connecting busbar.

[0221] Within the wind farm, the SVG (Static Var Generator) on the low-voltage side of the main step-up transformer at the collection station serves as a controllable reactive power source to provide reactive power. Under severe wind speed fluctuations, reactive power losses in the PCC (Power Process Control) balance area of ​​the wind farm are calculated. Within the adjustable range of the SVG, reactive power losses are compensated to regulate the voltage of the collection bus and PCC points, achieving reactive power balance within the PCC area of ​​the wind farm. This avoids excessive use of reactive power resources by grid-connected wind turbines due to voltage fluctuations at the collection bus, and provides a certain reactive power margin for grid-connected DFIG (Diverterless Power Injection) systems. The reactive power reference value of the SVG is:

[0222]

[0223] The technical advantages of this invention are illustrated below with simulation examples.

[0224] The analysis and verification were carried out using a real hybrid wind farm in a certain region as an example. The topological results are as follows: Figure 2 As shown. The wind farm comprises 34 doubly-fed induction generator (DFIG) wind turbines with a rated capacity of 2.5MW, including 26 grid-connected turbines and 8 grid-connected turbines. The turbine port voltage is 690V, connected to a 35kV collector line via a step-up transformer. The collector line is then connected to a 220kV main grid via another step-up transformer. A 50Mvar static var compensator is installed on the 35kV busbar. The base voltage is 35kV, and the base power is 100MVA. Simultaneously, by employing scenario generation and reduction techniques to acquire wind power output scenarios and performing day-ahead random optimization, typical wind turbine output scenarios are obtained, including four typical wind power output scenarios, such as... Figure 3 As shown.

[0225] During the day-ahead optimization phase, control variables include SCB, SVG, OLTC action plans, and reactive power of each wind turbine within the wind farm. The voltage situation after applying day-ahead stochastic optimization control to a hybrid wind farm (with / with grid) is as follows: Figure 4 As shown, the network loss and voltage offset are 74.1783 MW·h and 189.350 kV, respectively.

[0226] The operation status of the recently randomly optimized reactive power compensation device SVG, OLTC, and SCB are as follows: Figure 5 As shown. During the early morning and nighttime periods of higher power output, the SVG and SCB at the busbar maintain high reactive power output, and the OLTC conservatively regulates the voltage. During the midday period of lower power output, the site does not require excessive reactive power to regulate the voltage, and the OLTC adjusts the transformer ratio to adapt to voltage changes. Due to the constraint of the number of operations, the daily operations of the OLTC and SCB are controlled to within 10 times.

[0227] To verify the day-ahead and intraday optimization coordination effect proposed in this invention, considering the intraday optimization control effect of wind farms with reactive power balance in a grid-connected DFIG system, the analysis focused on the period from 02:00 to 04:00, during which power output fluctuations are relatively strong. The specific power output during this period is as follows: Figure 6 As shown, the wind turbine output of the wind farm began to fluctuate drastically at 02:55, increasing sharply from 1.525MW to 2.075MW every 5 minutes before decreasing to 1.302MW. All output fluctuations were above the critical fluctuation value. Using a 15-minute optimized control interval, the wind farm voltage curve from 02:00 to 04:00 without considering reactive power balance in a grid-connected DFIG system is shown below. Figure 7 As shown, the reactive power output of the grid-connected wind turbines and the output of the grid-connected wind turbines in the entire wind farm at this stage are as follows: Figure 8 As shown. Combining the wind farm voltage situation and the output of the grid-connected wind turbines, it can be seen that, although the intraday optimization control without considering reactive power balance control can keep the wind farm voltage within a safe range with the cooperation of other reactive power and voltage regulation equipment within the wind farm, it can also... Figure 8 The data shows that the reactive power generated by the grid-connected wind turbines 30, 31, and 32, located closer to the wind farm's collector bus, is significantly higher than that of other wind turbines located closer to the collector bus and grid-connected wind turbines. During the period from 02:00 to 04:00, they generated an average of 0.085 Mvar more than other wind turbines located closer to the collector bus and grid-connected wind turbines. This reactive power imbalance is more severe when the wind farm's output fluctuates above a certain level, necessitating reactive power balancing control measures for both grid-connected and grid-connected wind turbines to suppress this imbalance.

[0228] Based on the reactive power balance control method for grid-connected / integrated wind farms, and combining wind farm parameters and wind farm output data during periods of severe power fluctuations, the reactive power reference value of grid-connected wind turbines under the grid-connected / integrated reactive power balance control method is obtained as follows: Figure 9 As shown. This method of calculating reactive power reference based on the active power distribution of a wind farm can well adapt to the changes in reactive power demand caused by short-term power output fluctuations in wind farms. Grid-connected wind turbines can flexibly adjust reactive power output within a short period to match the reactive power demand of their respective reactive power balance zones.

[0229] After optimization control based on grid-connected reactive power balancing, the reactive power generated by the grid-connected wind turbines is close to the reactive power command value calculated to achieve reactive power balancing. The average error is controlled within 0.0086 Mvar during periods of drastic fluctuations in wind power output, thus achieving reactive power balance for the grid-connected wind turbines within their equilibrium range. By fully utilizing the reactive power regulation capability of the grid-connected wind turbines and rationally controlling the reactive power output of the grid-connected wind turbines, a better grid-connected reactive power balancing effect is achieved, while also keeping the wind farm voltage within a safe range. The reactive power output of the grid-connected wind turbines is as follows: Figure 10 As shown, the wind farm voltage is as follows Figure 11 As shown. Combined with Figure 10 , 11 It can be seen that when the grid-connected wind turbines generate reactive power to meet their own reactive power balance requirements, the four grid-connected wind turbines near the busbar in the wind farm reduce their output by an average of 0.0448 Mvar. During the process of severe fluctuations in wind power output, reactive power balance between the grid-connected and grid-connected wind turbines can be achieved, effectively avoiding the waste of reactive power resources of some grid-connected wind turbines in the wind farm, and further improving the ability of the grid-connected / grid-connected hybrid wind farm to cope with sudden accidents.

[0230] like Figure 12 As shown, this embodiment discloses a reactive power and voltage optimization control system for a grid-connected hybrid wind farm that considers reactive power balance, used to execute the above method, and includes the following modules:

[0231] Reactive power characteristic linearization model construction module: Constructs a reactive power characteristic linearization model for wind turbine units in a hybrid wind farm;

[0232] Scenario Analysis Module: Uses scenario analysis to derive typical power output scenarios for wind farms and their corresponding probabilities;

[0233] Action planning module: Based on the reactive power characteristic linearization model, typical power output scenarios of wind farms and corresponding probabilities, day-ahead stochastic optimization of hybrid wind farms with grid connection is performed to obtain the action plan of discrete action equipment;

[0234] Judgment Module: Based on the determined action plan of discrete action equipment, it performs intraday reactive power and voltage optimization control of mixed wind farms, and makes judgments on the switching of optimization control targets and methods according to the wind power output fluctuation level.

[0235] Control module: Based on the judgment results, it controls the reactive power balance inside the hybrid wind farm with grid and ground.

[0236] Other aspects of this embodiment can be referred to in the above method embodiments.

[0237] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A reactive power and voltage optimization control method for wind farms with integrated grid and reactive power balance, characterized by: The specific steps are as follows: S1. Construct a linearized model of the reactive power characteristics of wind turbine units in a hybrid wind farm; S2. Use scenario analysis to derive typical power output scenarios of wind farms and their corresponding probabilities; S3. Based on the reactive power linearization model obtained in step S1, the typical power output scenarios and corresponding probabilities of the wind farm obtained in step S2, perform day-ahead stochastic optimization of the hybrid wind farm with grid connection and take-off, and obtain the action plans of the discrete action devices OLTC and SCB. S4. Based on the day-ahead discrete action plan of the equipment determined in step S3, perform reactive power and voltage optimization control of the mixed wind farm during the day, and make judgment on the switching of optimization control target and mode according to the wind power output fluctuation level. S5. Based on the judgment result of step S4, control the reactive power balance inside the hybrid wind farm with grid and ground.

2. The reactive power and voltage optimization control method for a hybrid wind farm considering reactive power balance as described in claim 1, characterized in that the steps are as follows: In S1, the wind turbine generators in the hybrid wind farm are hybrid doubly fed wind farm wind turbine generators that are integrated with the grid. S1.1, the reactive power limit on the stator side of the wind turbine in a doubly-fed wind farm is: Among them, Q sGFLmax Q sGFLmin These are the upper and lower limits of the reactive power on the stator side of a grid-fed doubly-fed wind turbine, respectively; P sGFL The active power transmitted from the stator side of a grid-connected doubly-fed induction generator (DFIG) to the power grid; u s I represents the stator voltage amplitude. rmax X represents the maximum value of the rotor-side converter current. m X s These are the magnetizing reactance and the stator reactance, respectively. S1.2, the reactive power limit of the grid-side converter is based on The calculation yielded: Among them, Q cGFLmax Q cGFLmin These are the upper and lower limits of reactive power for the grid-side converter of a grid-connected doubly-fed wind turbine; P mec s represents the mechanical power input to the grid-connected doubly-fed wind turbine; s represents the slip rate. S1.3, based on considering the stator reactive power limit and the grid-side converter reactive power limit, the reactive power adjustment range of the grid-connected doubly-fed induction generator is obtained: S1.4, the grid-connected doubly-fed induction generator (DFIG) actively adjusts the reactive power output of the turbine port voltage based on the difference between the voltage and the actual voltage. The reactive power output of the grid-connected turbine j in the wind farm at time t is: in, It is the maximum reactive power output of the grid-connected wind turbine located at node j. This is the reference value for the reactive power output of the grid-connected wind turbine located at node j. This is the voltage reference value for the grid-connected wind turbine located at node j, where j is the node and t is time. and These are the voltage values ​​when the grid-type wind turbine reaches the upper and lower limits of reactive power, respectively. For grid-type wind turbine port voltage; K n This is the droop coefficient; S1.5, establish a linear model based on binary variables and matrix combinations to simulate the grid droop function, and incorporate grid-type wind turbines into the centralized decision-making model: in, The meaning is the binary variable corresponding to node j at time t; The inequality constraints in the above equation are relaxed to binary variables using the Big M method. By selecting the value of α, the port voltage determination and reactive power output calculation of the DFIG network are synchronized, as shown in the following equation: Where M is a positive number, K n Both the upper and lower limits of reactive power in the grid-type system are constants; when the reactive voltage droop function is in a non-limit operating state... and The product of these terms is a bilinear term, which is linearized using the McCormick relaxation method: Where c2 and c3 represent non-decision variables in the linear processing, z j,t,2 With z j,t,3 V is an auxiliary variable; GFM,j,max With V GMF,j,min The upper and lower limits of the voltage at node j of the grid-connected wind turbine.

3. The reactive power and voltage optimization control method for a grid-connected hybrid wind farm considering reactive power balance as described in claim 2, characterized in that, In step S2, the wind power output model is performed using the scene generation and reduction method to obtain typical wind farm output scenarios and their corresponding probabilities.

4. The reactive power and voltage optimization control method for a grid-connected hybrid wind farm considering reactive power balance as described in claim 3, characterized in that, In step S2, combining the wind speed probability distribution and historical wind power data, the Latin hypercube sampling algorithm is used to generate scenarios. Then, the backward reduction method is used to reduce the scenarios to obtain typical power output scenarios of wind farms with different probabilities. The wind speed probability distribution follows a Weibull distribution, and the probability density function of the Weibull distribution is as follows: Among them, v w (t) represents the wind speed at time t; k s and c s These are the shape and scale parameters of the Weibull distribution, respectively.

5. The reactive power and voltage optimization control method for a grid-connected hybrid wind farm considering reactive power balance as described in claim 4, characterized in that, In step S3, the objective function is specifically: Where, n s The number of scenarios obtained through scenario analysis; v is the node voltage of the wind farm; For the probability corresponding to the scene, and These are the weighting coefficients; P loss,t,s and D v,t,s These represent the network loss and voltage deviation at time t under scenario s, respectively. The constraints are as follows: Power flow constraints within a wind farm: in, It is the active power flowing from node j to node k at time t. It is the active power flowing from node i to node j at time t. It is the reactive power flowing from node j to node k at time t. It is the reactive power flowing from node i to node j at time t. It is the resistance value from node i to node j. It is the reactance value from node i to node j. It is the reference voltage of the wind farm at time t, and π is the set of all nodes in the wind farm; It is the set of all the terminal nodes connected to all branches that start from node j; P is the set of the first nodes connected to all branches whose final node is node j; GFL,j,t P GFM,j,t These represent the active power output of node j in the root-type DFIG and the active power output of the network-type DFIG at time t, respectively; Q GFL,j,t Q GFM,j,t These represent the reactive power outputs of node j at time t for the following DFIG configuration and the network configuration DFIG configuration, respectively; P D,j,t Q D,j,t These represent the active and reactive loads at node j at time t, respectively; Q SCB,j,t For the SCB reactive power compensation at node j of the wind farm at time t; l ij,t v is the square of the current amplitude of line (i,j) at time t; j,t k is the square of the voltage amplitude at node j of the wind farm at time t; ij,t Let be the on-load tap-changing transformer OLTC ratio of line (i,j) at time t; Active power constraints for DFIG network construction: Among them, P GFL,j,max and P GFL,j,min The active power P of the grid-connected doubly fed wind turbine at node j at time t is respectively. GFL The upper and lower limits of P; GFM,j,max and P GFM,j,min The active power P of the grid-type doubly fed wind turbine at node j at time t is respectively. GFM The upper and lower limits; Reactive power constraints with DFIG network type: Among them: Q GFL,j,max and Q GFL,j,min The reactive power Q of the grid-connected doubly fed wind turbine at node j at time t is respectively. GFL The upper and lower limits; The specific reactive power constraints of the grid-type DFIG are as follows: Static Var Generator (SVG) reactive power constraints: in, and These are the upper and lower limits of the SVG reactive power compensation capacity of wind farm node j, respectively. OLTC constraints: The on-load tap-changing transformer is located at branch (i, j) and has a total of taps. One, respectively , , ... , Satisfying the constraints: in, It is the voltage at node i at time t. In the ij branch, y is the auxiliary variable for OLTC action, where y is the number of gears in the OLTC. It is a 0-1 variable, and at the same time ; in, Since it is a bilinear term, we use the Big M method to linearize it, and the resulting linear form is: in, It is the auxiliary variable for OLTC action in the ij branch, where K is the gear of OLTC. As an auxiliary variable; the limit N for the number of OLTC actions in a wind farm. OLTC as follows: in, It is the OLTC action auxiliary variable in the ij branch at time t. It is the gear selection at the previous moment. The action marker of the OLTC on line ij at time t; The maximum number of actions performed by the OLTC on line ij within 24 hours; Parallel capacitor bank constraints: The number of parallel capacitor banks located at node j that can be switched is n. SCBj,t Each group has a capacity of Q. SCB0,t The following constraints must be satisfied in the reactive power optimization control process of the wind farm: in, It is the reactive power compensation amount of SCB at node j at time t. It is the reactive power compensation amount of one gear of SCB; The above equation is a discrete optimization model, which can be transformed into a 0-1 integer programming model using the binary expansion method: in, δi represents the maximum number of SCB actions at node i at time t, where n is determined by the set number of SCB groups; δ0, δ1, ..., δi n As a binary auxiliary variable, the constraint on the number of SCB actions and the calculation of the number of actions are as follows: in, The action label of SCB at node i at time t; Let SCB be the maximum number of actions performed on node i within 24 hours.

6. The reactive power and voltage optimization control method for a hybrid wind farm considering reactive power balance as described in claim 5, characterized in that, In step S4, the fluctuation range of wind power output is assumed to be... Determine whether it exceeds a set threshold ε; where: in, The active power output of the GFL type wind turbine at time t is... It is the rated active power output of the GFL type fan; when When the objective function is J1, the optimization objective function is to minimize voltage deviation and network loss. The reactive power optimization control model of the hybrid wind farm with / with network consists of J1, power flow constraints, SVG, and reactive power constraints with / with network. when The objective function is optimized by introducing a deviation term between the reactive power output of the wind turbine and its reactive power balance reference value. The objective function adopts J2. The reactive power balance optimization model of the grid consists of J2, power flow constraints, SVG, and grid reactive power constraints. The objective function is as follows: Where w1 is the weighting coefficient corresponding to voltage deviation, w2 is the weighting coefficient corresponding to network loss, w3 is the weighting coefficient corresponding to reactive power balance, and n is the number of wind farm nodes. It is the voltage at node j of the wind farm at time t. It is the reference voltage for the wind farm. This represents the wind farm network loss at time t. It is the reactive power of node j at time t. F is the reference value of reactive power at time t, and F is the set of grid-connected wind turbines. n To match the number of grid-connected wind turbine units.

7. The reactive power and voltage optimization control method for a grid-connected hybrid wind farm considering reactive power balance as described in claim 6, characterized in that, Step S5 is as follows: Based on the judgment result of step S4, when the fluctuation range of wind power output is greater than the set value, the reactive power balance control inside the hybrid wind farm is strengthened, the optimized control interval is shortened to once every 5 minutes, and reactive power balance control is carried out on the hybrid wind farm by dividing the reactive power balance zone and setting the reactive power balance reference value of the grid-connected wind turbine.

8. The reactive power and voltage optimization control method for a grid-connected hybrid wind farm considering reactive power balance as described in claim 7, characterized in that, In step S5, by dividing the reactive power balance into zones, the command values ​​for reactive power balance of grid-connected units and SVG are calculated, and reactive power balance is introduced into the objective function to achieve grid-connected reactive power collaborative balance control under the centralized decision-making mode of wind farm.

9. The reactive power and voltage optimization control method for a grid-connected hybrid wind farm considering reactive power balance as described in claim 8, characterized in that, In step S5, the reactive power balancing control of the wind turbine is as follows: The two-dimensional numbering method is used to divide the reactive power balance area of ​​the grid-type wind turbine. The area corresponding to wind turbine j is defined as the m,n balance area. This area takes wind turbine j as the reactive power center and includes the box-type transformer of the wind turbine and half of the feeder line between adjacent nodes. The reactive power loss of the box-type transformer in the balance area of ​​the grid-connected wind turbine m,n is: In the formula, It is the square of the active power of the GFL corresponding to node j. It is the gear position of the OLTC located on branch ij at time t, x T, j The reactance of the box-type transformer corresponding to the grid-type wind turbine j; The reactive power loss of the feeder lines within the balance area of ​​the grid-connected wind turbine m,n is: Where w represents the number of wind turbines in the area. It is the reactance of the branch on the nth branch in the equilibrium region. The equilibrium region is located in the reactance of the next branch n. Therefore, the reactive power loss in the balance region of the grid-type wind turbine m,n is: in, This refers to the reactive power loss of the transformer corresponding to wind turbine j. It represents the line loss within the mn equilibrium region; A set reactive power margin is reserved for the grid-type DFIG; the calculated reactive power loss within the area is equal to the reactive power balance reference value: in, This is the reactive power balance reference value for GFL type wind turbine j; The specific details of reactive power balance control (PCC) in wind farms are as follows: The reactive power loss of the main step-up transformer in the PCC balancing area of ​​the wind farm is: in, It is the transformer reactance value; The reactive power losses of the feeder lines and transmission lines connecting the PCC balance area and the collection bus of the wind farm are as follows: Where: b is the number of grid-connected wind turbine units within the PCC balance area on the side of the connecting busbar; The reactive power reference value for SVG is: in, It refers to the reactive power loss of the lines within the balance area.

10. A reactive power and voltage optimization control system for a grid-connected hybrid wind farm considering reactive power balance, used to execute the method as described in any one of claims 1-9, characterized in that, Includes the following modules: Reactive power characteristic linearization model construction module: Constructs a reactive power characteristic linearization model for wind turbine units in a hybrid wind farm; Scenario Analysis Module: Uses scenario analysis to derive typical power output scenarios for wind farms and their corresponding probabilities; Action planning module: Based on the reactive power characteristic linearization model, typical power output scenarios of wind farms and corresponding probabilities, day-ahead stochastic optimization of hybrid wind farms with grid connection is performed to obtain the action plan of discrete action equipment; Judgment Module: Based on the determined action plan of discrete action equipment, it performs intraday reactive power and voltage optimization control of mixed wind farms, and makes judgments on the switching of optimization control targets and methods according to the wind power output fluctuation level. Control module: Based on the judgment results, it controls the reactive power balance inside the hybrid wind farm with grid and ground.

Images