A wind power grid-connected regional power grid AVC cooperative control method and system

Abstract

The application relates to the technical field of automatic voltage control of power systems, and discloses a wind power grid-connected regional power grid AVC cooperative control method and system. The method constructs a three-layer dynamic logic architecture: a regional power grid AVC master station layer, a wind power plant cluster logic coordination layer, and a single plant station on-site execution layer. The master station layer constructs a comprehensive coupling degree index based on voltage sensitivity and electrical distance, performs adaptive dynamic partitioning, and generates a corresponding logic coordination unit for each partition; the master station layer periodically solves a multi-objective reactive power optimization model and issues a total reactive power instruction to each partition; the logic coordination unit finely allocates the adjustment amount according to the voltage regulation sensitivity, reactive power margin and response characteristics of each reactive power source; and the on-site execution layer executes the instruction. The application realizes linkage reconstruction of dynamic partitioning and logic coordination units, solves the problems of severe voltage fluctuation and difficult multi-resource coordination under wind power grid connection, and improves voltage stability, economy and equipment operation life.

Description

Technical Field

[0001] This invention relates to the field of automatic voltage control technology for power systems, specifically to an AVC collaborative control method and system for a regional power grid with wind power integration. Background Technology

[0002] Currently, wind power is developing towards large-scale, centralized, and high-proportion grid connection. Regional power grids are gradually forming a wind power cluster grid connection pattern where multiple wind farms, multiple voltage level collection stations, and multiple types of reactive power compensation devices coexist. Wind power output has significant randomness, intermittency, and strong fluctuations. Moreover, the reactive power regulation characteristics and response speed of wind turbines are fundamentally different from those of conventional synchronous motors. Their own reactive power regulation capabilities are limited, which can easily lead to voltage instability problems such as voltage exceeding limits at the wind farm grid connection point, voltage flicker, reactive power backfeed, and local overvoltage or undervoltage, seriously threatening the safe, stable, economical, and efficient operation of the regional power grid.

[0003] Existing traditional AVC control technology is mainly designed for grids dominated by conventional synchronous power sources. When applied to scenarios with a high proportion of wind power grid connection, it has significant technical shortcomings: (1) The control objects are limited to a single wind farm, a single substation or a single node, lacking a regional overall coordination mechanism. Under the high proportion of wind power grid connection, the reactive power regulation of multiple stations is coupled and interferes with each other, which easily leads to voltage regulation command conflicts and grid voltage oscillation problems. (2) The fixed partitioning mode is adopted, and the adaptive partitioning logic is not designed in combination with the characteristics of large fluctuations in wind power output and dynamic changes in grid topology. The partitioning decoupling effect is poor and it cannot adapt to the complex and ever-changing wind power grid connection operation conditions. (3) The control objective is singular, with only the node voltage qualification as the core objective, without comprehensively considering multiple dimensions such as minimizing the active power loss of the whole network, optimizing the number of reactive power equipment operation, and efficiently utilizing the reactive power margin of wind turbines, resulting in insufficient economic efficiency of power grid operation. (4) Using a single time scale control cannot simultaneously address the needs of millisecond-level instantaneous fluctuation smoothing and minute-level global reactive power optimization in wind power, and is prone to problems such as voltage regulation response lag or over-regulation. (5) The coordinated control of various reactive resources such as wind turbines, SVG, and energy storage has not been realized. Various voltage regulating devices operate independently, and the potential for reactive power regulation is not fully explored, which easily leads to resource waste and frequent equipment operation. Summary of the Invention

[0004] The technical problem to be solved by this invention is to provide a regional power grid AVC collaborative control method and system for wind power grid connection, so as to realize global optimization of reactive power resource allocation, precise control of zone decoupling, and efficient execution of hierarchical collaboration.

[0005] The technical solution adopted by this invention to solve its technical problem is a regional power grid AVC collaborative control method for wind power grid connection, which includes the following steps: S1: Construct a three-layer dynamic logic architecture; the three-layer dynamic logic architecture includes a regional power grid AVC master station layer, a wind farm cluster logic coordination layer composed of dynamically generated logic coordination units, and a single-station local execution layer, which are connected in sequence; the single-station local execution layer includes wind farm AVC substations, SVG and energy storage devices; S2: In the AVC master station layer of the regional power grid, a comprehensive coupling index is constructed based on voltage sensitivity and electrical distance, and an improved fuzzy C-means clustering algorithm is used to adaptively and dynamically partition the power grid nodes, generating a corresponding logical coordination unit for each partition; the logical coordination unit is dynamically generated, reconstructed and updated according to the partitioning results, and multiple logical coordination units are arranged in parallel to form the logical coordination layer of the wind farm cluster; S3: In the AVC master station layer of the regional power grid, the global multi-objective reactive power optimization model is solved periodically to generate the total reactive power command for each partition and send it to the corresponding logical coordination unit. S4: In each logic coordination unit, receive the general reactive power command for the zone, and allocate reactive power regulation to each wind farm, SVG and energy storage device according to the voltage regulation sensitivity, reactive power adjustability margin and response characteristics of each reactive power source in the zone. S5: In the local execution layer of a single wind farm, the wind farm AVC substation controller, SVG controller and energy storage device controller send the received single-device adjustment commands from the logic coordination unit to the corresponding execution devices.

[0006] Furthermore, the method for constructing the comprehensive coupling degree described in S2 is as follows: first calculate the voltage sensitivity between nodes. and electrical distance Then, through the weighted fusion formula Calculate and form a comprehensive coupling degree matrix; where, For nodes For nodes Voltage sensitivity, representing the node Changes in injected reactive power cause node Rate of change of voltage; For nodes The voltage amplitude; For nodes Injected reactive power, The symbol is for partial differentials. This indicates that, with other variables remaining constant, the node voltage amplitude For nodes Injected reactive power The partial derivatives; node With nodes The electrical distance between them; the smaller the value, the closer the electrical connection. For nodes Self-impedance; For nodes Self-impedance; For nodes With nodes Mutual impedance between them; For nodes With nodes The overall coupling degree between them; the larger the value, the stronger the coupling and the more suitable they are to be assigned to the same partition. Voltage sensitivity The absolute value; For nodes With nodes Electrical distance between them; , Let be the weighting coefficient, satisfying This is used to balance the contributions of voltage sensitivity and electrical distance to the overall coupling.

[0007] Furthermore, the improved fuzzy C-means clustering algorithm described in S2 includes the following sub-steps: Based on the preset number of partitions K in the wind power cluster layout, the hub node with the central electrical location is selected as the initial cluster center; Compute the electrical characteristic distance between nodes and cluster centers ;in, For nodes With the Cluster centers The smaller the electrical characteristic distance between nodes, the higher the fit between the node and the partition; Total number of critical nodes; For nodes With nodes The overall coupling degree; For the first Cluster centers With nodes The equivalent coupling degree; Calculate the node affiliation probability ,satisfy Where m is the fuzzy weighted index; For nodes Belonging to the The probability of each partition; The preset number of partitions; For nodes With the Electrical feature distance between cluster centers; For nodes With the Electrical feature distance between cluster centers; For fuzzy weighted index, it is usually taken as This controls the degree of ambiguity in the partitioning; For the summation index, from 1 to Satisfy normalization constraints ; Update cluster center ;in, For the updated number The first cluster center One component; : No. The set of nodes contained in each partition; For nodes Belongs to the partition The probability of; It is a fuzzy weighted index; For nodes With nodes The overall coupling degree; Iterate until the node affiliation is stable, and generate the corresponding logical coordination unit.

[0008] Furthermore, the objective function of the global multi-objective reactive power optimization model described in S3 includes: minimizing node voltage deviation, minimizing the total active power loss, and minimizing the number of reactive power equipment operations; the constraints include: node voltage constraints, reactive power capacity constraints, and line power flow constraints.

[0009] Furthermore, the allocation of reactive power regulation in S4 includes: Calculate the voltage regulation sensitivity weight ;in, For the first in the partition The voltage regulation sensitivity weight of the reactive power source to the central point voltage; For the first The absolute value of the voltage sensitivity of the reactive power source; It is the sum of the absolute values ​​of the voltage sensitivity of all reactive power sources within the partition; Calculate the reactive power adjustability margin factor ;in, For the first The reactive power adjustable margin coefficient of a reactive power source reflects its residual regulation capacity. For the first The upper limit of reactive power output of the Taiwan reactive power source; For the first The lower limit of reactive power output of the Taiwan reactive power source; For the first Rated capacity of the reactive power source; Constructing a comprehensive allocation coefficient ,in, For the first The overall allocation coefficient of the reactive power source; This is a weighting coefficient, ranging from 0.5 to 0.7, used to balance the contributions of sensitivity weight and margin coefficient. Weights for voltage regulation sensitivity; This is the reactive power adjustability margin coefficient; For the comprehensive allocation coefficient Normalization is performed:

[0010] in, The normalized comprehensive allocation coefficient satisfies ; This is the sum of the comprehensive allocation coefficients for all reactive power sources within the partition; Calculate the reactive power regulation of a single device ;in, No. The reactive power regulation of the reactive power source allocation; The normalized comprehensive allocation coefficient; For total reactive power command of the partition; Limit the adjustment amount that exceeds the limit, and redistribute the difference to the remaining equipment that does not exceed the limit; Assign steady-state base value components to the AVC substations of the wind farm, and assign rapid fluctuation and correction components to the SVG and energy storage devices.

[0011] A regional power grid AVC collaborative control system for wind power grid connection, used to implement a regional power grid AVC collaborative control method for wind power grid connection, including: The regional power grid AVC master station module is used to collect data from the entire power grid, perform dynamic zoning and global multi-objective reactive power optimization, and issue regional reactive power general instructions. The wind farm cluster logic coordination module consists of dynamically generated logic coordination units, which are used to receive the reactive power total command for the partition and finely allocate it to each reactive power source within the partition. The single-site local execution module, including the wind farm AVC substation, SVG, and energy storage device, is used to execute regulation commands and achieve millisecond-level voltage correction.

[0012] Furthermore, the logical coordination units in the wind farm cluster logical coordination module are generated, reconstructed, and updated in real time according to the dynamic partitioning results.

[0013] Furthermore, the dynamic partitioning is based on a comprehensive coupling index that integrates voltage sensitivity and electrical distance, and is implemented using an improved fuzzy C-means clustering algorithm.

[0014] The beneficial effects of this invention are: (1) The present invention adopts a dynamic voltage regulation partition and logical coordination layer linkage architecture. The logical coordination unit is generated, reconstructed and updated in real time according to the grid operation status and wind power output changes. It does not rely on fixed physical sites and can flexibly adapt to the complex working conditions of grid topology changes and random power output fluctuations in high-proportion wind power grid connection scenarios. It completely solves the problems of voltage regulation command conflict, voltage oscillation and voltage regulation interference in multi-wind farm clusters from the architecture level. The partition decoupling effect is significant and the system's anti-disturbance robustness is greatly enhanced.

[0015] (2) This invention proposes a comprehensive coupling index that integrates voltage sensitivity and electrical distance, and combines improved fuzzy C-means clustering to achieve adaptive dynamic partitioning. Compared with traditional single index and fixed partitioning methods, the partitioning boundary is more reasonable and the coupling distinction is more accurate. It can significantly reduce the probability of regional grid voltage exceeding the limit and improve the voltage stability and qualification level of wind power grid connection point.

[0016] (3) This invention establishes a hierarchical and zoned collaborative allocation strategy for multi-reactive resources such as wind turbines, SVG, and energy storage. Based on voltage regulation sensitivity, reactive power margin and equipment response characteristics, it refines the allocation instructions to achieve complementary advantages of steady-state coarse adjustment and rapid fine adjustment, fully taps the adjustment potential of various reactive equipment, and effectively reduces the active power loss of the entire network while ensuring voltage qualification, thereby improving the economic operation level of the power grid. Attached Figure Description

[0017] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0018] Embodiments of the present invention are described in detail below. Examples of these 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. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0019] This embodiment uses a 220kV / 110kV high-proportion wind power grid in a certain region as an example. The grid nodes refer to key electrical connection points in the power system that require voltage monitoring or participation in reactive power regulation, including but not limited to wind farm grid connection points, substation busbars, wind farm collection station busbars, energy storage power station grid connection points, and regional grid hub nodes. In this embodiment, the following five core key nodes are selected for calculation demonstration: Node 1: 220kV substation busbar 1; Node 2: Grid connection point of wind farm A, with an installed capacity of 250MW; Node 3: Grid connection point of wind farm B, with an installed capacity of 250MW; Node 4: 220kV substation, busbar 2; Node 5: Energy storage device grid connection point, located on the bus side of the collection station 1, equipped with an energy storage device with a capacity of ±20MW / ±20Mvar connected through the PCS converter, with an adjustable reactive power range of -20Mvar to +20Mvar.

[0020] The wind farm uses doubly-fed induction generators (DFIGs). The reactive power regulation range of a single unit is -0.3 pu to 0.3 pu, where pu refers to the per-unit value based on the rated capacity of the turbine. Collection station 1 is equipped with a ±50 Mvar SVG, and collection station 2 is also equipped with a ±50 Mvar SVG. The response time is ≤20 ms. The allowable deviation of the node voltage is ±5% of the rated voltage, and the voltage at the wind power grid connection point must be stable within the range of 0.98 to 1.02 pu.

[0021] In this embodiment, the multiple reactive power resources include: doubly-fed wind turbines (wind farm A, wind farm B), SVG (collector station 1, collector station 2), and energy storage devices. See also Figure 1 The following steps detail the collaborative control process for these resources.

[0022] S1: Construct a three-tier dynamic logic architecture

[0023] According to the method of the present invention, a three-layer dynamic logical architecture is first constructed. This architecture is a purely logical functional division and is not bound to a fixed physical site.

[0024] Regional power grid AVC master station layer: deployed in the power grid dispatch and control center. In this embodiment, an industrial server is used to run the data acquisition and monitoring (SCADA) system and AVC application software.

[0025] Wind farm cluster logical coordination layer: This layer is not a fixed physical device, but a collection of logical functional units generated in real time from the subsequent dynamic partitioning results. In this embodiment, two logical coordination units will be generated in the end, each unit corresponding to a voltage control partition, responsible for receiving the total reactive power command of the partition issued by the master station and distributing it downward.

[0026] The local execution layer at a single wind farm site includes AVC substations, SVG (gathering station), and energy storage devices for each wind farm. In this embodiment, wind farm A and wind farm B are each equipped with an AVC substation, gathering station 1 is equipped with an SVG controller, and the energy storage device is equipped with an energy storage device controller.

[0027] S2: Dynamic Partitioning and Logical Coordination Unit Generation

[0028] 2.1 Basic Operational Data Collection

[0029] The main station layer collects real-time power grid operation data across the entire region, including: voltage amplitude at each node, active / reactive power flow on lines, real-time output of wind farms, power grid topology, and rated capacity and current adjustable range of each reactive power source. This embodiment demonstrates calculations based on data from a typical operating condition.

[0030] 2.2 Voltage Sensitivity Calculation

[0031] Voltage sensitivity Indicates at node When a unit reactive power change (e.g., 1 Mvar) is injected at a node, The voltage amplitude change (in kV / Mvar). This embodiment calculates a 5×5 voltage sensitivity matrix based on a linearized power flow model (i.e., the Jacobian matrix sub-block in Newton-Raphson power flow calculation).

[0032] The first in the matrix Line 1 The column element is .For example, This means that injecting 1Mvar of reactive power at Node5 (the grid connection point of the energy storage device) increases the voltage of Node1 by about 0.0011kV. The energy storage has a voltage regulation capability for the hub node that is comparable to that of a wind farm.

[0033] 2.3 Electrical Distance Calculation

[0034] Electrical distance is based on the impedance matrix of power grid nodes. Calculation, where For the self-impedance of node i, Let J be the self-impedance of node j. Let be the mutual impedance between node i and node j. Electrical distance. The smaller the value, the tighter the electrical connection between nodes. This embodiment calculates the electrical distance matrix as follows:

[0035] It can be seen that the electrical distance between Node5 and Node1 is only 0.11, and the distances between Node5 and Node2 and Node3 are 0.13 and 0.15 respectively, indicating that the energy storage device is closely connected to the collection station 1 and the two wind farms, and is suitable to be included in the same zone.

[0036] 2.4 Calculation of Comprehensive Coupling Degree

[0037] Take weighting coefficients , This value can be adjusted based on the scale of the power grid and the differences in physical dimensions between voltage sensitivity and electrical distance, with the principle being to ensure that the two contributions are on the same order of magnitude. The comprehensive coupling degree matrix is ​​calculated as follows:

[0038] In the matrix The larger the value, the more likely it is to be a node. and The more suitable it is to be assigned to the same partition. Observation shows that the coupling values ​​between Node1, Node2, Node3, and Node5 (0.3577, 0.3514, 0.3982, 0.3551, 0.3489, etc.) are significantly higher than their coupling with Node4 (0.2760, 0.2667, 0.2705, 0.2751). Therefore, there is a physical basis to divide the first four nodes into one partition and Node4 into a separate partition.

[0039] 2.5 Improved Fuzzy C-means Clustering Dynamic Partitioning

[0040] Using the comprehensive coupling matrix To input, perform the following steps: (1) Initial parameter settings: Preset number of partitions (Pre-determined based on the spatial distribution and operation and maintenance experience of wind power clusters, or automatically optimized through clustering effectiveness indicators). Fuzzy weighted index (Typically, 1.5 to 2.5 is used; in this embodiment, 2 is used). The centrally located hub nodes are selected as the initial cluster centers: Node1 (located at the center of partition 1) and Node4 (located at the center of partition 2), with initial cluster center vectors of respectively... ; ; (2) Calculate the electrical characteristic distance according to the formula. : With Node2 ( For example, it is related to the cluster center. Distance:

[0041] With cluster center Distance:

[0042] Obviously Node2 should belong to partition 1.

[0043] With Node5 ( For example, it is related to the cluster center. Distance:

[0044] With cluster center Distance:

[0045] It is evident that Node5 and V1 are very close, and should belong to partition 1.

[0046] (3) Calculate the probability of belonging, according to the formula Calculate the probability of Node2 matching partition 1:

[0047] The probability of partition 2 .

[0048] The probability of Node5 matching partition 1:

[0049] Similar calculations show that the probability of Node1 and Node3 belonging to partition 1 is greater than 0.99, and the probability of Node4 belonging to partition 2 is greater than 0.99.

[0050] (4) Update cluster centers

[0051] The cluster centers are recalculated using the affiliation probability as the weight. For example, the first component of the cluster center in partition 1 after the update. ,in , After two rounds of iterative calculation, the change in node affiliation probability is less than 0.001, indicating that the cluster centers are stable.

[0052] (5) Generating Logical Coordination Units: After the partitioning stabilizes, the final partitioning result is: Partition 1 contains Node 1, Node 2, Node 3, and Node 5; Partition 2 contains Node 4. The regional power grid AVC master station layer dynamically generates a logical coordination unit for each partition. This unit is created in real time with the partitioning result and is bound to all node and device information within the partition. In particular, the logical coordination unit of Partition 1 is bound to four reactive power sources: wind farm A, wind farm B, SVG, and energy storage devices. When wind power output fluctuates significantly (more than 20%) or the power grid topology changes, the above partitioning process is automatically re-executed, the old logical coordination unit is canceled, and a new unit is regenerated.

[0053] S3: Global Multi-Objective Reactive Power Optimization at the Main Station Layer

[0054] The main station layer solves the global multi-objective reactive power optimization model with a period of 5 to 15 minutes (10 minutes in this embodiment).

[0055] Objective function: Minimum node voltage deviation The total number of critical nodes in the power grid that require voltage monitoring (such as wind farm grid connection points, substation busbars, energy storage grid connection points, etc.).

[0056] :node The actual voltage amplitude.

[0057] :node The voltage reference value (usually the rated voltage, such as 1.0 pu).

[0058] Minimum active power loss across the entire network : Active power loss across the entire network.

[0059] : Connecting nodes and nodes The line conductivity reflects the line's ability to conduct electricity.

[0060] :node and nodes The voltage amplitude.

[0061] :node With nodes The voltage phase angle difference between them.

[0062] The cosine of the phase angle difference affects the transmission power of the line.

[0063] The reactive power equipment operates the fewest times. : Reactive power equipment number (such as a certain wind turbine, SVG or energy storage device).

[0064] :equipment In the current optimization cycle (time point) (The unproductive output)

[0065] :equipment In the previous optimization cycle (time point) (The unproductive output)

[0066] Constraints: Node voltage constraints :node The allowable lower voltage limit (0.95 pu in this embodiment, and 0.98 pu at the wind power grid connection point).

[0067] :node The upper limit of the allowable voltage (1.05 pu in this embodiment, and 1.02 pu for the wind power grid connection point).

[0068] Reactive capacity constraints :equipment The lower limit of reactive power output (the maximum value of reactive power absorbed, which is negative).

[0069] :equipment The upper limit of reactive power output (the maximum amount of reactive power generated).

[0070] In this embodiment: the wind turbine is -75Mvar to +75Mvar, the SVG is -50Mvar to +50Mvar, and the energy storage is -20Mvar to +20Mvar.

[0071] Line power flow constraints :line The active power transmitted uplink.

[0072] The maximum active power transmission capacity allowed for this line.

[0073] Solving the above mixed-integer nonlinear programming model yields the total reactive power command for each partition: Partition 1 absorbs 28 Mvar of reactive power (i.e., Partition 2 outputs 12 Mvar of reactive power ( The master station layer sends instructions to the logical coordination units corresponding to partition 1 and partition 2, respectively.

[0074] S4: Internal allocation of the logical coordination layer of the wind farm cluster

[0075] Taking partition 1 as an example, its logical coordination unit receives instructions. (Reactive power absorption). Reactive power sources within Zone 1 include: SVG (rated capacity 50 Mvar), Wind Farm A (equivalent turbine capacity 250 Mvar, reactive power range -75 to +75 Mvar), Wind Farm B (equivalent turbine capacity 250 Mvar, reactive power range -75 to +75 Mvar), and energy storage devices (rated capacity 20 Mvar, reactive power range -20 to +20 Mvar). Currently, all equipment is operating at 0 Mvar.

[0076] (1) Calculate the voltage regulation sensitivity weight

[0077] Using the central point voltage (Node1 voltage) as a reference, the voltage sensitivity of each device to Node1 is taken from the first row of matrix S: SVG subsite (installed on Node1): ; Wind Farm A (Node 2): ; Wind Farm B (Node 3): ; Energy storage (Node 5): .

[0078] denominator Therefore:

[0079] (2) Calculate the reactive power adjustability margin coefficient

[0080] SVG: ; Wind Farm A: ; Wind farm B: Same as A. ; Energy storage: .

[0081] (3) Constructing the comprehensive allocation coefficient

[0082] Pick (Balancing sensitivity and margin, recommended value is 0.5–0.7), Calculation:

[0083] (4) Normalization

[0084] (5) Calculate the reactive power regulation of a single device

[0085] (6) Over-limit constraint correction

[0086] All adjustments are within the adjustable range of the equipment (SVG: ±50Mvar; wind turbine: ±75Mvar; energy storage: ±20Mvar), requiring no correction. If any equipment exceeds its limit (e.g., the energy storage command exceeds the 20Mvar absorption limit), the excess will be proportionally redistributed to other equipment that has not exceeded their limits until all commands meet the constraints.

[0087] (7) Division of labor based on speed characteristics

[0088] In this embodiment, the logic coordination unit makes a final allocation correction based on the device response speed: Wind farm (response speed in seconds to minutes): Handles steady-state baseline components; allocation instructions are... , ; SVG (response speed in milliseconds): Handles rapid fluctuations and correction components, and the allocation instructions are as follows. In addition, an extra capacity (approximately 3 Mvar) is reserved for instantaneous correction; Energy storage (response speed in milliseconds): also handles rapid fluctuations and correction components, and the allocation command is... Approximately 2 Mvar is reserved for collaborative correction.

[0089] The total absorption remains 28Mvar after the correction. This allocation result has been uploaded to the main site for filing.

[0090] S5: Rapid Response at the Single-Station Local Execution Layer

[0091] Steady-state command execution: The AVC substation of wind farm A receives the command "absorb 6.97 Mvar", and the AVC substation of wind farm B receives the command "absorb 6.76 Mvar". Steady-state reactive power output is achieved by adjusting the wind turbine excitation system. The SVG controller receives the command "absorb 7.19 Mvar", and the energy storage device controller receives the command "absorb 7.08 Mvar". The IGBT and PCS converters are adjusted respectively, with response times ≤20ms, achieving steady-state reactive power absorption.

[0092] Instantaneous fluctuation correction: Assuming that during operation, wind farm A experiences a sudden increase in active power output due to a change in wind speed, causing the voltage at grid connection point Node2 to momentarily exceed the 1.02 pu upper limit. The energy storage device at the local execution layer collaborates with the SVG to perform millisecond-level rapid correction: The SVG controller (deployed on Node1) detected the voltage anomaly through high-frequency sampling (sampling period of 1ms) and increased the reactive power absorption from 7.19Mvar to 12.19Mvar (an increase of 5Mvar) within 2ms. Meanwhile, the energy storage controller (Node5) increases reactive power absorption from 7.08 Mvar to 12.08 Mvar (an increase of 5 Mvar) within 5 ms. The two combined increase absorption by 10Mvar, bringing the Node2 voltage back to the acceptable range of 1.01pu within 1.5 seconds.

[0093] During the correction process, the energy storage device releases a portion of the active power (according to the PCS capacity limit, the active power automatically decreases when reactive power increases), but this does not affect the grid frequency stability.

[0094] Feedback loop: The status of SVG and energy storage equipment (actual reactive power output, remaining margin) is fed back to the logic coordination unit and the master station layer in real time. The master station layer recalculates the steady-state allocation in the next slow timescale optimization cycle (10 minutes later), incorporating the additional reactive power absorption caused by this instantaneous correction (SVG and energy storage each absorbed an additional 5 Mvar) into the next round of optimization. By adjusting the wind turbine steady-state base value or redistributing the power, the equipment is restored to the economically optimal operating point.

[0095] The embodiments described herein are preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, all equivalent changes made in accordance with the structure, shape, and principle of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A regional power grid AVC collaborative control method for wind power grid connection, characterized in that, Includes the following steps: S1: Construct a three-layer dynamic logic architecture; the three-layer dynamic logic architecture includes a regional power grid AVC master station layer, a wind farm cluster logic coordination layer composed of dynamically generated logic coordination units, and a single-station local execution layer, which are connected in sequence; the single-station local execution layer includes wind farm AVC substations, SVG and energy storage devices; S2: In the AVC master station layer of the regional power grid, a comprehensive coupling index is constructed based on voltage sensitivity and electrical distance, and an improved fuzzy C-means clustering algorithm is used to adaptively and dynamically partition the power grid nodes, generating a corresponding logical coordination unit for each partition; the logical coordination unit is dynamically generated, reconstructed and updated according to the partitioning results, and multiple logical coordination units are arranged in parallel to form the logical coordination layer of the wind farm cluster; The improved fuzzy C-means clustering algorithm includes the following sub-steps: Based on the preset number of partitions K in the wind power cluster layout, the hub node with the central electrical location is selected as the initial cluster center; Compute the electrical characteristic distance between nodes and cluster centers ;in, For nodes With the Cluster centers The smaller the electrical characteristic distance between nodes, the higher the fit between the node and the partition; Total number of critical nodes; For nodes With nodes The overall coupling degree; For the first Cluster centers With nodes The equivalent coupling degree; Calculate the node affiliation probability ,satisfy Where m is the fuzzy weighted index; For nodes Belonging to the The probability of each partition; The preset number of partitions; For nodes With the Electrical feature distance between cluster centers; For nodes With the Electrical feature distance between cluster centers; For fuzzy weighted index, take This controls the degree of ambiguity in the partitioning; For the summation index, from 1 to Satisfy normalization constraints ; Update cluster center ;in, For the updated number The first cluster center One component; : No. The set of nodes contained in each partition; For nodes Belongs to the partition The probability of; It is a fuzzy weighted index; For nodes With nodes The overall coupling degree; Iterate until node affiliation is stable, then generate the corresponding logical coordination unit. S3: In the AVC master station layer of the regional power grid, the global multi-objective reactive power optimization model is solved periodically to generate the total reactive power command for each partition and send it to the corresponding logical coordination unit. S4: In each logic coordination unit, receive the general reactive power command for the zone, and allocate reactive power regulation to each wind farm, SVG and energy storage device according to the voltage regulation sensitivity, reactive power adjustability margin and response characteristics of each reactive power source in the zone. S5: In the local execution layer of a single wind farm, the wind farm AVC substation controller, SVG controller and energy storage device controller send the received single-device adjustment commands from the logic coordination unit to the corresponding execution devices.

2. The regional power grid AVC collaborative control method for wind power grid connection according to claim 1, characterized in that, The method for constructing the overall coupling degree described in S2 is as follows: first calculate the voltage sensitivity between nodes. and electrical distance Then, through the weighted fusion formula The calculation yields a comprehensive coupling degree matrix; where, For nodes For nodes Voltage sensitivity, representing the node Changes in injected reactive power cause node Rate of change of voltage; For nodes The voltage amplitude; For nodes Injected reactive power, The symbol is for partial differentials. This indicates that, with other variables remaining constant, the node voltage amplitude For nodes Injected reactive power The partial derivatives; node With nodes The electrical distance between them; the smaller the value, the closer the electrical connection. For nodes Self-impedance; For nodes Self-impedance; For nodes With nodes Mutual impedance between them; For nodes With nodes The overall coupling degree between them; the larger the value, the stronger the coupling and the more suitable they are to be assigned to the same partition. Voltage sensitivity The absolute value; For nodes With nodes Electrical distance between them; , For the weighting coefficients, satisfying This is used to balance the contributions of voltage sensitivity and electrical distance to the overall coupling.

3. The regional power grid AVC collaborative control method for wind power grid connection according to claim 1, characterized in that, The objective functions of the global multi-objective reactive power optimization model described in S3 include: minimizing node voltage deviation, minimizing the total active power loss, and minimizing the number of reactive power equipment operations; the constraints include: node voltage constraints, reactive power capacity constraints, and line power flow constraints.

4. The regional power grid AVC collaborative control method for wind power grid connection according to claim 3, characterized in that, The allocation of reactive power regulation in S4 includes: Calculate the voltage regulation sensitivity weight ;in, For the first in the partition The voltage regulation sensitivity weight of the reactive power source to the central point voltage; For the first The absolute value of the voltage sensitivity of the reactive power source; It is the sum of the absolute values ​​of the voltage sensitivity of all reactive power sources within the partition; Calculate the reactive power adjustability margin factor ;in, For the first The reactive power adjustable margin coefficient of a reactive power source reflects its residual regulation capacity. For the first The upper limit of reactive power output of the Taiwan reactive power source; For the first The lower limit of reactive power output of the Taiwan reactive power source; For the first Rated capacity of the reactive power source; Constructing a comprehensive allocation coefficient ,in, For the first The overall allocation coefficient of the reactive power source; This is a weighting coefficient, ranging from 0.5 to 0.7, used to balance the contributions of sensitivity weight and margin coefficient. Weights for voltage regulation sensitivity; This is the reactive power adjustability margin coefficient; For the comprehensive allocation coefficient Normalization is performed: in, The normalized comprehensive allocation coefficient satisfies ; This is the sum of the comprehensive allocation coefficients for all reactive power sources within the partition; Calculate the reactive power regulation of a single device ;in, No. The reactive power regulation of the reactive power source allocation; The normalized comprehensive allocation coefficient; For total reactive power command of the partition; Limit the adjustment amount that exceeds the limit, and redistribute the difference to the remaining equipment that does not exceed the limit; Assign steady-state base value components to the AVC substations of the wind farm, and assign rapid fluctuation and correction components to the SVG and energy storage devices.

5. A regional power grid AVC collaborative control system for wind power grid connection, used to implement the regional power grid AVC collaborative control method for wind power grid connection as described in any one of claims 1 to 4, characterized in that, include: The regional power grid AVC master station module is used to collect data from the entire power grid, perform dynamic zoning and global multi-objective reactive power optimization, and issue regional reactive power general instructions. The wind farm cluster logic coordination module consists of dynamically generated logic coordination units, which are used to receive the reactive power total command for the partition and finely allocate it to each reactive power source within the partition. The single-site local execution module, including the wind farm AVC substation, SVG, and energy storage device, is used to execute regulation commands and achieve millisecond-level voltage correction.

6. The regional power grid AVC collaborative control system for wind power grid connection according to claim 5, characterized in that, The logical coordination units in the wind farm cluster logical coordination module are generated, reconstructed, and updated in real time according to the dynamic partitioning results.

7. The regional power grid AVC collaborative control system for wind power grid connection according to claim 5, characterized in that, The dynamic partitioning is based on a comprehensive coupling index that integrates voltage sensitivity and electrical distance, and is implemented using an improved fuzzy C-means clustering algorithm.

Images