A wind power active power automatic control method for a wind farm monitoring system

Abstract

The present application relates to the field of wind farm monitoring, and discloses a wind power active power automatic control method for a wind farm monitoring system, which is used for enhancing the support capability of the wind farm to the power grid and guaranteeing the safe and stable operation of the wind farm. The wind power active power automatic control method for the wind farm monitoring system comprises the following steps: collecting various data through a sensor network to form a global real-time state data set, equating the wind farm to a virtual synchronous machine model, calculating an equivalent parameter set, a maximum available power upper limit value and an individual operation boundary constraint set, taking a superior instruction as a total target, combining related parameter sets to generate a global optimization power setting value vector through multi-target collaborative optimization calculation, selecting an active control mode according to the situation of the power grid, and adjusting a generated control instruction set. The present application can also automatically generate a wind farm active control efficiency evaluation report, and realize efficient, stable and intelligent control of the wind farm active power.

Description

Technical Field

[0001] This invention relates to the field of wind farm monitoring, and more particularly to an automatic control method for wind power active power in a wind farm monitoring system. Background Technology

[0002] With the continuous growth of global energy demand and increasing emphasis on environmental protection, the development and utilization of renewable energy has become a research hotspot in the energy field. Wind power, as a clean and renewable energy source, has experienced rapid development in recent years, with the scale and installed capacity of wind farms constantly expanding. However, the intermittent and fluctuating nature of wind power poses a significant challenge to the stable operation of the power grid. How to achieve efficient and stable control of the active power of wind farms and improve their support capacity for the power grid has become a key issue that urgently needs to be addressed in the current wind power sector.

[0003] In wind farm monitoring systems, automatic active power control is a core component for ensuring the safe and stable operation of wind farms and achieving grid-friendly integration. Traditional active power control methods for wind farms are mainly based on simple fixed power allocation strategies. While these methods can regulate active power to some extent, they have many limitations when facing complex and ever-changing grid conditions and the internal operating states of wind farms.

[0004] Traditional data acquisition methods primarily rely on conventional Supervisory Control and Data Acquisition (SCADA) systems, which have low sampling rates and struggle to capture instantaneous changes in grid frequency and wind farm operating status. Furthermore, the data sources are limited, lacking the acquisition of critical information such as feedforward wind speed at the wind farm inlet, and the absence of a unified time synchronization mechanism between different data sources results in poor data fusion performance, failing to provide accurate and comprehensive real-time status information for active power control. Existing technologies often simplify wind farms as a simple superposition of multiple independent wind turbines, failing to fully consider the overall dynamic characteristics of the wind farm. When calculating the equivalent parameters of a wind farm, there is a lack of accurate mapping of dynamic characteristics such as the overall kinetic energy, virtual inertia, and damping coefficient. This results in the inability to effectively simulate the wind farm's response to grid frequency changes during active power control, making it difficult to achieve frequency support for the grid. Traditional control methods typically focus solely on meeting the active power setting commands from higher levels, allocating power with a single objective. They consider only reducing grid losses within the wind farm or balancing turbine loads, neglecting other important objectives such as reserving frequency regulation capacity. This singular optimization objective fails to comprehensively consider the diverse operational needs of a wind farm, resulting in poor control performance and difficulty in achieving optimal overall operation of the wind farm.

[0005] Therefore, we propose an automatic control method for wind power active power in wind farm monitoring systems to solve the above problems. Summary of the Invention

[0006] This invention provides an automatic control method for active power of wind power in a wind farm monitoring system, which enhances the wind farm's support capability for the power grid and ensures the safe and stable operation of the wind farm.

[0007] The first aspect of this invention provides an automatic control method for active power of wind power in a wind farm monitoring system. The method includes: collecting grid dispatch instructions, system frequency, grid connection point voltage, active power, reactive power, rotational speed status, and ambient wind speed data at each wind turbine outlet to form a global real-time status dataset for the wind farm; based on the global real-time status dataset, equating the entire wind farm to a virtual synchronous machine model, deriving a set of virtual synchronous machine equivalent parameters, and synchronously calculating the upper limit of the maximum generateable power of the entire farm under the current wind conditions and the individual operating boundary constraint set of each wind turbine; and using the received upper-level active power setting instruction as the overall target... Based on the virtual synchronous machine equivalent parameter set and the individual operating boundary constraint set, a collaborative optimization calculation is performed with multiple optimization objectives, including reducing grid losses within the wind farm, balancing unit loads, and reserving frequency regulation capacity, to generate a global optimized power setpoint vector. According to the real-time frequency deviation of the power grid or a preset operating strategy, an active control mode is selected from multiple preset control modes, and the global optimized power setpoint vector is adjusted based on the selected mode to generate a control command set. The control command set is then issued to execute active power regulation actions, and the new global real-time status data of the wind farm generated after execution is used as feedback and input into the global real-time status data set of the wind farm, thereby forming a closed-loop automatic control system.

[0008] Optionally, in the first implementation of the first aspect of the present invention, the method includes: receiving an active power setting instruction from an upper-level energy management system to form a first input data stream; collecting instantaneous system frequency values ​​and instantaneous three-phase voltage values ​​at the grid connection point to form a second input data stream; periodically collecting the active power output values, reactive power output values, and wind turbine speed values ​​of each unit to form a third input data stream; collecting the feedforward wind speed at the wind farm inlet and the real-time wind speed at each unit to form a fourth input data stream; generating a multi-dimensional wind farm operation status data snapshot based on the first, second, third, and fourth input data streams; dynamically adjusting the sampling frequency configuration parameters of the high-precision frequency and voltage measurement unit and the intelligent measurement and control device according to the rate of change of system frequency in the second input data stream; automatically increasing the sampling rate when the frequency changes drastically; performing data validity verification and noise filtering on the fused multi-dimensional wind farm operation status data snapshot; removing abnormal data points and supplementing smooth data; and outputting a global real-time status dataset of the wind farm.

[0009] Optionally, in a second implementation of the first aspect of the present invention, the method includes: dynamically calculating the total kinetic energy of the entire wind farm based on the real-time rotational speed and active power values ​​of each wind turbine generator set from the global real-time state dataset of the wind farm, and mapping the total kinetic energy to an equivalent virtual inertia parameter that varies with time; obtaining an equivalent damping coefficient parameter based on the historical power response data and rotational speed fluctuation data of each wind turbine generator set from the global real-time state dataset of the wind farm, and combining the equivalent virtual inertia parameter and the equivalent damping coefficient parameter to form a virtual synchronous machine equivalent parameter set; and calculating the maximum possible wind speed for each wind turbine generator set under the current wind speed based on the feedforward wind speed data and the real-time wind speed data at each turbine generator set from the global real-time state dataset of the wind farm, combined with the pre-stored aerodynamic power characteristic curves of the wind turbine generator sets. Capture power: sum the maximum possible capture power of all units and subtract the estimated network loss value calculated based on real-time power flow to obtain the upper limit of the maximum power output of the entire field under the current wind conditions; Based on the real-time operating data of each unit from the global real-time status dataset of the wind farm, combined with the pre-stored unit mechanical load model and thermal capacity model, calculate the adjustable upper and lower limits of active power of each unit under the current operating state to form the dynamic power adjustment range of each unit; Based on the power flow and voltage data of the collection lines from the global real-time status dataset of the wind farm, combined with the power grid safety operation regulations, calculate the cluster power adjustment constraints aimed at preventing line overload, integrate the dynamic power adjustment range of all units and all cluster power adjustment constraints to generate the individual operating boundary constraint set of each wind turbine unit.

[0010] Optionally, in a third implementation of the first aspect of the present invention, the method includes: using the upper-level active power setting command from the power grid dispatch communication interface as the total power constraint of the optimization problem, and based on the equivalent virtual inertia parameters and equivalent damping coefficient parameters from the virtual synchronous machine equivalent parameter set, constructing a frequency support optimization objective; using the dynamic power adjustment range of each unit and the cluster power adjustment constraint from the individual operating boundary constraint set as boundary conditions, constructing a feasible solution space for the optimization problem, and comprehensively establishing a mathematical model for a multi-objective optimization problem; calculating the sensitivity coefficient of the impact of active power changes of each unit on the overall network loss based on the power flow and voltage data of the collection lines from the global real-time status dataset of the wind farm, and generating a network loss sensitivity matrix; and calculating the load balance index between the groups and its impact on the power generation based on the load status data of each unit from the global real-time status dataset of the wind farm. The sensitivity of load allocation is assessed, and a load balancing sensitivity vector is generated. Based on the real-time frequency fluctuations of the power grid, the weight allocation ratios of the three optimization objectives—reducing network losses, balancing loads, and reserving capacity—are dynamically adjusted to generate a dynamic multi-objective weight coefficient set. Using the network loss sensitivity matrix and the load balancing sensitivity vector as search direction guides, the mathematical model of the multi-objective optimization problem is iteratively solved within the neighborhood of the current operating point, based on the feasible solution space defined by the individual operating boundary constraint set. The Pareto optimal solution is output, yielding a global optimized power setpoint vector. This global optimized power setpoint vector is then substituted into a preset power flow calculation model for verification to ensure that no power grid safety constraints are violated. If the verification fails, the dynamic multi-objective weight coefficient set is adjusted and the solution is re-solved until a safe and feasible optimization result is obtained. The verified global optimized power setpoint vector is then output.

[0011] Optionally, in the fourth implementation of the first aspect of the present invention, the method includes: real-time monitoring of system frequency data from the wind farm grid connection point measurement unit, calculating the deviation value and deviation change rate between the current frequency and the rated frequency, and automatically selecting an optimal active control mode through a mode selection logic table based on preset frequency deviation thresholds and change rate thresholds, combined with operation strategy instructions from the grid dispatch communication interface, and outputting a mode decision signal; receiving the global optimized power setpoint vector from the multi-objective optimization and power allocation decision steps as a basic adjustment instruction, and calling the corresponding control algorithm to correct the basic adjustment instruction according to the control mode type indicated by the mode decision signal, generating a preliminary control instruction set; comparing and verifying the preliminary control instruction set with the adjustable upper and lower limits of each unit's power from the individual operating boundary constraint set, detecting whether the preliminary control instruction set meets the transmission capacity constraint and voltage stability constraint of the collection line, and outputting a control instruction set that has passed safety verification; adding a timestamp, instruction sequence number, and target unit number information to each instruction in the control instruction set, encapsulating the instruction set into a data packet in a standard communication protocol format, adding a data check code, and generating a control instruction set that can be directly sent to the underlying wind turbine controller.

[0012] Optionally, in the fifth implementation of the first aspect of the present invention, the corresponding control algorithm is invoked to modify the basic adjustment command: when the frequency regulation mode is selected, the additional active power adjustment is calculated based on the frequency deviation value and the equivalent parameter set of the virtual synchronous machine; when the margin mode is selected, the reserve capacity of the entire field is increased at the cost of reducing the current total output; when the safety mode is selected, the commands of units that may exceed the limits are preferentially adjusted based on the individual operating boundary constraint set.

[0013] Optionally, in the sixth implementation of the first aspect of the present invention, the method includes: distributing the control command set to the underlying controllers of each target wind turbine generator unit through the internal communication network of the wind farm; receiving command reception confirmation signals and execution start signals returned by each underlying controller in real time; generating a command execution state matrix; within a preset time window after command execution, collecting execution effect data such as pitch angle, generator torque, and actual output active power; collecting adjusted total active power, frequency characteristics, and voltage change data of the entire field; performing time synchronization and data validity verification on the collected execution effect data; and generating a command execution effect dataset; and including the actual output active power in the command execution effect dataset. The system compares the expected values ​​of the commands with those in the control command set, calculates the command tracking error index, and generates a control performance evaluation report. Based on the command tracking error index and the control performance evaluation report, it dynamically adjusts the virtual synchronous machine equivalent parameter set and control algorithm parameters to form an updated control parameter set. It integrates the actual operating data of each unit in the command execution effect dataset into the wind farm's global real-time status dataset, replacing the original historical data. The updated control parameter set is fed back to the dynamic equivalent and operating boundary calculation steps and the multi-objective optimization and power allocation decision steps as calculation parameters for the next control cycle. Through continuous data updates and parameter adjustments, an adaptive closed-loop automatic control system is formed.

[0014] Optionally, in the seventh implementation of the first aspect of the present invention, the system health status online assessment and early warning steps are further included: based on the vibration, temperature and electrical parameters of each unit from the global real-time status dataset of the wind farm, the remaining service life assessment index of each component is calculated, and a system health status assessment report and early fault warning signal are generated according to the preset early warning threshold; the equipment health index from the system health status assessment report is received, and combined with the grid frequency characteristics from the global real-time status dataset of the wind farm, when a decline in equipment health is detected, the control mode decision logic is automatically adjusted to generate an adaptive control mode strategy; the voltage drop / rise signal and frequency abnormal fluctuation signal from the measurement unit at the wind farm grid connection point are monitored in real time, and when extreme operating conditions exceeding the safety threshold are detected, an emergency power control command is generated based on the virtual synchronous machine equivalent parameter set and the individual operating boundary constraint set; the operating data from each step are periodically collected, including the command execution effect dataset, control performance assessment report and system health status assessment report, and the wind farm active power control efficiency assessment report is automatically generated based on the preset assessment index system, computer group utilization rate, power generation efficiency index and grid support contribution.

[0015] The mechanism of this invention is as follows: Based on a multi-objective real-time optimization power allocation method, multiple optimization objectives such as reducing network loss, balancing load, and reserving frequency modulation capacity are considered in a coordinated manner, and an optimization algorithm guided by sensitivity analysis is used to solve the problem. A complete control system with self-learning capability is formed through closed-loop feedback control. Beneficial effects: Based on real-time data, the equivalent virtual inertia parameter of the total kinetic energy mapping of the entire wind farm is dynamically calculated, and the equivalent damping coefficient parameter is calculated using the least squares parameter identification method. This forms a set of equivalent parameters for the virtual synchronous machine, enabling the model to more accurately reflect the dynamic characteristics of the wind farm. By comprehensively considering factors such as unit mechanical load, heat capacity, power flow of the collector line, and voltage data, a set of individual operating boundary constraints for each unit is generated, providing reasonable boundary conditions for optimization calculations and ensuring the safe operation of the units. A sensitivity-guided gradient descent optimization algorithm is adopted, which uses the network loss sensitivity matrix and load balancing sensitivity vector as search direction guides to iteratively solve the multi-objective optimization problem in the neighborhood of the current running point, outputs the Pareto optimal solution, and obtains the global optimization power setpoint vector, thereby improving the efficiency and accuracy of optimization calculation. Based on the real-time frequency deviation of the power grid or the preset operation strategy, the optimal active control mode is selected from multiple preset control modes, and the mode decision signal is output to clarify the control mode type and parameter set. This enables the wind farm to flexibly adjust the control strategy according to different operating conditions. The preliminary control command set is compared and verified with the adjustable upper and lower limits of the power of each unit to check whether the transmission capacity constraints of the collection line and the voltage stability constraints are met. Over-limit commands are corrected by proportional reduction or priority reduction strategies to ensure that the commands are safe and feasible. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of an embodiment of the automatic control method for wind power active power in a wind farm monitoring system according to the present invention; Figure 2 This is a schematic diagram of an embodiment of an automatic control device for wind power active power in a wind farm monitoring system according to the present invention. Detailed Implementation

[0017] This invention provides an automatic control method for active power of wind power in a wind farm monitoring system, which enhances the wind farm's support capability for the power grid and ensures the safe and stable operation of the wind farm. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0018] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 One embodiment of the automatic control method for wind power active power in a wind farm monitoring system according to the present invention includes: 101. Steps for global status perception and data collection of wind farm: Through the sensor network deployed at the grid connection point, collection lines and each wind turbine, data such as grid dispatch instructions, system frequency, grid connection point voltage, active power, reactive power, speed status and ambient wind speed of each wind turbine outlet are collected to form a global real-time status dataset of the wind farm. It is understood that the executing entity of this invention can be an automatic control device for active power of wind power used in a wind farm monitoring system, or it can be a terminal or a server; the specific implementation is not limited here. This embodiment of the invention will be described using a server as an example.

[0019] Specifically, the steps for synchronous acquisition of multi-source heterogeneous data are as follows: the active power setting command from the upper-level energy management system is received through the protocol converter installed on the power grid dispatch communication interface, forming the first input data stream; By installing a high-precision frequency and voltage measurement unit at the grid connection point of the wind farm, the instantaneous values ​​of the system frequency and the three-phase voltage at the grid connection point are collected simultaneously at a sampling rate higher than that of conventional SCADA systems, forming a second input data stream; The intelligent measurement and control device installed at the controller output of each wind turbine generator periodically collects the active power output value, reactive power output value and wind turbine speed value of each unit to form a third input data stream. By using a lidar array deployed in the wind farm and an ultrasonic anemometer installed on the nacelle, the feedforward wind speed at the wind farm entrance and the real-time wind speed at each turbine are collected to form a fourth input data stream. Timestamp alignment and data fusion steps: Mark each data point in the first, second, third, and fourth input data streams with a unified, precise, synchronized clock timestamp; Based on a unified timestamp, the four data streams are aligned and fused in the time dimension to generate a multi-dimensional wind farm operation status data snapshot with a time stamp. Dynamic sensing configuration and data preprocessing steps: Based on the rate of change of system frequency in the second input data stream, dynamically adjust the sampling frequency configuration parameters of the high-precision frequency and voltage measurement unit and the intelligent measurement and control device, and automatically increase the sampling rate when the frequency changes drastically. The data validity is verified and noise is filtered on the fused multidimensional wind farm operation status data snapshot. Abnormal data points are removed and smoothed data is added. Finally, the global real-time status dataset of the wind farm is output and sent to the dynamic equivalent and operation boundary calculation step as the sole data source for its calculation. 102. Dynamic equivalent and operational boundary calculation steps: Based on the global real-time status dataset of the wind farm, the entire wind farm is equivalent to a virtual synchronous machine model. The set of virtual synchronous machine equivalent parameters characterizing the dynamic characteristics of the entire field is calculated, and the upper limit of the maximum power generation of the entire field under the current wind conditions and the individual operational boundary constraint set of each wind turbine are calculated simultaneously. Specifically, the dynamic calculation steps for the equivalent parameters of the virtual synchronous machine are as follows: Based on the real-time speed and active power values ​​of each wind turbine generator set from the global real-time state dataset of the wind farm, the total kinetic energy of the entire wind farm is dynamically calculated, and this total kinetic energy is mapped to an equivalent virtual inertia parameter that varies with time. Based on the historical power response data and speed fluctuation data of each wind turbine generator set from the global real-time state dataset of the wind farm, an equivalent damping coefficient parameter is calculated using the least squares parameter identification method. The equivalent virtual inertia parameter and the equivalent damping coefficient parameter are combined to form the equivalent parameter set of the virtual synchronous machine. This parameter set will be sent to the multi-objective optimization and power allocation decision steps to construct an optimization objective that includes the virtual inertial response. The overall power potential assessment steps are as follows: Based on the feedforward wind speed data from the global real-time status dataset of the wind farm and the real-time wind speed data at each unit, combined with the pre-stored aerodynamic power characteristic curves of the wind turbine units, the maximum possible capture power of each computer unit under the current wind speed is calculated. The maximum possible capture power of all units is accumulated, and the estimated network loss value calculated based on the real-time power flow is subtracted to obtain the upper limit of the maximum power that can be generated in the entire field under the current wind conditions. This value will be sent to the multi-objective optimization and power allocation decision steps as a key constraint for optimization calculation. Operational boundary constraint generation steps: Based on the real-time operation data of each unit from the global real-time status dataset of the wind farm, combined with the pre-stored unit mechanical load model and thermal capacity model, the upper and lower limits of the adjustable active power of each unit under the current operating state are calculated to form the dynamic power adjustment range of each unit; Based on the power flow and voltage data of the collection lines from the global real-time status dataset of the wind farm, combined with the power grid safety operation regulations, cluster power adjustment constraints aimed at preventing line overload are calculated; The dynamic power adjustment range of all units and all cluster power adjustment constraints are integrated to generate the individual operational boundary constraint set of each wind turbine unit. This constraint set will be sent to the multi-objective optimization and power allocation decision step as the core boundary condition for optimization calculation; 103. Multi-objective optimization and power allocation decision steps: Taking the received active power setting command from the superior as the overall objective, and based on the virtual synchronous machine equivalent parameter set and individual operating boundary constraint set, the multi-objective optimization calculation is carried out to reduce network losses in the field, balance unit load, and reserve frequency regulation capacity, generating a global optimized power setting value vector, which contains the refined active power setting values ​​allocated to each wind turbine cluster or unit; Specifically, the steps for constructing the multi-objective optimization problem are as follows: The active power setting command from the grid dispatch communication interface is used as the total power constraint for the optimization problem; based on the equivalent virtual inertia parameters and equivalent damping coefficient parameters from the virtual synchronous machine equivalent parameter set, a frequency support optimization objective including virtual inertial response capability is constructed; the dynamic power adjustment range of each unit and the cluster power adjustment constraint from the individual operating boundary constraint set are used as boundary conditions to construct the feasible solution space of the optimization problem; and a comprehensive mathematical model for the multi-objective optimization problem is established, simultaneously pursuing reduced network losses, balanced unit loads, and reserved frequency regulation capacity. Real-time sensitivity analysis and weight allocation steps: Based on the power flow and voltage data of the collection lines from the global real-time status dataset of the wind farm, calculate the sensitivity coefficient of the impact of the active power change of each unit on the overall network loss, and generate a network loss sensitivity matrix; Based on the load status data of each unit from the global real-time status dataset of the wind farm, calculate the load balance index between units and its sensitivity to power allocation, and generate a load balance sensitivity vector; According to the real-time frequency fluctuation of the power grid, dynamically adjust the weight allocation ratio of the three optimization objectives of reducing network loss, balancing load, and reserving capacity, and generate a dynamic multi-objective weight coefficient set; Collaborative optimization solution steps: A sensitivity-guided gradient descent optimization algorithm is adopted, using the network loss sensitivity matrix and load balancing sensitivity vector as search direction guides; within the neighborhood of the current running point, the mathematical model of the multi-objective optimization problem is iteratively solved based on the feasible solution space defined by the individual running boundary constraint set; the Pareto optimal solution that simultaneously satisfies multiple optimization objectives is output, i.e., the global optimized power setpoint vector, which will be sent to the control mode decision and command distribution step as its input original power setpoint; Optimization result verification and correction steps: Substitute the global optimized power setpoint vector into the preset power flow calculation model for verification to ensure that no power grid safety constraints are violated; if the verification fails, adjust the dynamic multi-objective weight coefficient set and solve again until a safe and feasible optimization result is obtained; finally, output the verified global optimized power setpoint vector to the subsequent steps. 104. Control mode decision and command distribution steps: Based on the real-time frequency deviation of the power grid or the preset operation strategy, select an active control mode from multiple preset control modes, and make a final adjustment to the global optimized power setpoint vector based on the selected mode to generate a set of control commands. Specifically, the multi-mode decision-making and selection steps are as follows: real-time monitoring of system frequency data from the wind farm grid connection point measurement unit; calculation of the deviation value and rate of change of the current frequency from the rated frequency; based on the preset frequency deviation threshold and rate of change threshold, combined with the operation strategy instructions from the grid dispatch communication interface, an optimal active control mode is automatically selected through the mode selection logic table; outputting a mode decision signal, which clearly indicates the type of control mode to be adopted and its corresponding set of control parameters. Mode Adaptive Command Adjustment Steps: Receive the global optimized power setpoint vector from the multi-objective optimization and power allocation decision steps as the basic adjustment command; Based on the control mode type indicated by the mode decision signal, call the corresponding control algorithm to correct the basic adjustment command: When the frequency modulation mode is selected, calculate the additional active power adjustment amount based on the frequency deviation value and the equivalent parameter set of the virtual synchronizer; When the margin mode is selected, the total reserve capacity of the entire site is increased at the cost of reducing the current total output; when the safety mode is selected, the instructions of units that may exceed the limits are preferentially adjusted based on the individual operating boundary constraint set; and a preliminary control instruction set that has been adjusted in a pattern-based manner is generated. Command safety verification and coordination steps: Compare and verify the preliminary control command set with the adjustable upper and lower limits of each unit's power from the individual operating boundary constraint set; check whether the preliminary control command set meets the transmission capacity constraints of the collector line and voltage stability constraints; if any limit violations are found, coordinate and correct the commands using a proportional reduction or priority reduction strategy to ensure that all commands are within the safe operating range; output the control command set that has passed safety verification. Command encapsulation and distribution preparation steps: Add timestamp, command sequence number and target unit number information to each command in the control command set; encapsulate the command set into a data packet in the standard communication protocol format and add a data check code; generate a control command set that can be directly sent to the underlying wind turbine controller. This command set will be sent to the execution and feedback adjustment steps as the direct basis for its control execution. 105. Execution and Feedback Adjustment Steps: The control command set is sent to the underlying controller of the corresponding wind turbine generator set to control its pitch system and converter to perform active power adjustment actions. At the same time, the new global real-time status data of the wind farm generated after execution is used as feedback and input into the global real-time status data center of the wind farm, thereby forming a closed-loop automatic control system.

[0020] Specifically, the command distribution and execution monitoring steps are as follows: the control command set from the control mode decision and command distribution steps is distributed to the underlying controller of each target wind turbine through the internal communication network of the wind farm; the command reception confirmation signal and execution start signal returned by each underlying controller are received in real time, and a command execution status matrix is ​​generated; the actual action response of each unit's pitch system and converter is monitored to ensure that it performs active power regulation in accordance with the command requirements. Execution effect data acquisition and preprocessing steps: Within a preset time window after the command is executed, execution effect data such as pitch angle, generator torque, and actual output active power are collected by sensors deployed on each wind turbine; the total active power, frequency characteristics, and voltage change data of the entire field after adjustment are collected by the wind farm grid connection point measurement unit; the collected execution effect data are synchronized in time and the data validity is verified to generate a command execution effect dataset; Control performance evaluation and parameter adjustment steps: Compare the actual output active power in the instruction execution effect dataset with the instruction expectation value in the control instruction set, and calculate the instruction tracking error index; analyze the changes in the full-field frequency characteristics during the adjustment process, evaluate the inertial response effect of the virtual synchronous machine control, and generate a control performance evaluation report; based on the instruction tracking error index and the control performance evaluation report, dynamically adjust the virtual synchronous machine equivalent parameter set and control algorithm parameters to form an updated control parameter set; Closed-loop feedback and data fusion steps: The actual operating data of each unit in the instruction execution effect dataset is merged into the global real-time status dataset of the wind farm, replacing the original historical data; the updated control parameter set is fed back to the dynamic equivalent and operating boundary calculation steps and the multi-objective optimization and power allocation decision steps as the calculation parameters for the next control cycle; through continuous data updates and parameter adjustments, a closed-loop automatic control system with adaptive capabilities is formed.

[0021] 106. Online assessment and early warning steps for system health status: Based on the vibration, temperature, and electrical parameters of each unit from the global real-time status dataset of the wind farm, a fault prediction method based on a physical model is used to calculate the remaining service life assessment index of each component; according to the preset early warning threshold, a system health status assessment report and early fault warning signals are generated. Control mode adaptive optimization steps: Receive equipment health indicators from the system health status assessment report, and combine them with grid frequency characteristics from the global real-time status dataset of the wind farm; when a decline in equipment health is detected, automatically adjust the control mode decision logic to generate a control mode adaptive strategy; feed the control mode adaptive strategy back to the control mode decision and command distribution steps as a new input condition for its mode selection. Extreme operating condition emergency coordination and control steps: Real-time monitoring of voltage sag / surge signals and abnormal frequency fluctuation signals from the wind farm grid connection point measurement unit; When an extreme operating condition exceeding the safety threshold is detected, an emergency power control command is generated based on the virtual synchronous machine equivalent parameter set and individual operating boundary constraint set; Beyond conventional optimization allocation logic, priority is given to ensuring grid stability, forming an emergency control command set and issuing it for execution; Operational performance comprehensive evaluation and report generation steps: Periodically collect operational data from each step, including instruction execution effect datasets, control performance evaluation reports, and system health status evaluation reports; based on a preset evaluation index system, comprehensive indicators such as computer group utilization, power generation efficiency index, and grid support contribution; automatically generate wind farm active power control performance evaluation reports, and feed back key indicators to the multi-objective optimization and power allocation decision steps for adaptive adjustment of the optimization objective weights.

[0022] In this embodiment of the invention, a multi-source heterogeneous data synchronous acquisition method is adopted to obtain data from multiple dimensions such as power grid scheduling, grid connection point, wind turbine controller and wind farm environment. Through timestamp alignment and fusion technology, a multi-dimensional wind farm operation status data snapshot with time stamp is generated, which can comprehensively and accurately grasp the real-time operation status of the wind farm, provide a reliable data foundation for subsequent dynamic equivalence, optimization decision-making, etc., and improve the accuracy and effectiveness of control. The entire wind farm is modeled as a virtual synchronous machine, and the equivalent parameter set is dynamically calculated. At the same time, factors such as feedforward wind speed, real-time wind speed, unit mechanical load model and heat capacity model are comprehensively considered to calculate the upper limit of the maximum power generation of the entire farm and the individual operating boundary constraint set of each unit. This can accurately reflect the dynamic characteristics of the wind farm, provide reasonable constraints for multi-objective optimization, make the power allocation more scientific and reasonable, give full play to the power generation potential of the wind farm, and ensure the safe operation of the units. The system performs collaborative optimization calculations with multiple optimization objectives, including reducing network losses within the wind farm, balancing unit loads, and reserving frequency regulation capacity. Based on real-time sensitivity analysis and dynamic weight allocation, it employs a sensitivity-guided gradient descent optimization algorithm to solve the multi-objective optimization problem. Under the premise of meeting the active power setting instructions from the higher level, it comprehensively considers multiple optimization objectives to achieve global optimization allocation of active power in the wind farm, improve the operating efficiency and economic benefits of the wind farm, and enhance its support capability for the power grid. Multiple preset control modes are designed, such as frequency regulation mode, margin mode, and safety mode. The control mode is automatically selected according to the real-time frequency deviation of the power grid or the preset operation strategy. The global optimized power setpoint vector is finally adjusted. The control strategy can be flexibly adjusted according to different operating conditions and power grid needs, thereby improving the wind farm's adaptability and response to the power grid and ensuring the safe and stable operation of the power grid. The execution effect data is fed back to the global real-time status dataset, and the equivalent parameter set of the virtual synchronous machine and the control algorithm parameters are dynamically adjusted to form an adaptive closed-loop automatic control system. This system can monitor the control effect in real time, automatically adjust the control parameters according to the actual situation, improve the stability and robustness of the control system, and enable the wind farm to maintain good operating performance under different operating conditions. The fault prediction method based on physical models calculates the remaining service life assessment index of each component, realizing online assessment and early warning of system health status. At the same time, when extreme operating conditions are detected, emergency power control commands can be generated based on the equivalent parameter set of virtual synchronous machine and the individual operating boundary constraint set. This can help to detect potential equipment failures in advance, take timely measures for maintenance and repair, and extend the service life of equipment. Under extreme operating conditions, it can respond quickly, ensure grid stability, and improve the reliability and safety of wind farms. By periodically collecting operational data, including comprehensive indicators such as computer group utilization, power generation efficiency index, and grid support contribution, the system automatically generates active power control efficiency assessment reports for wind farms and feeds key indicators back to the multi-objective optimization and power allocation decision-making process. This enables a comprehensive evaluation of the wind farm's operational efficiency, providing a basis for operation and management decisions. Furthermore, by feeding back key indicators, the system adaptively adjusts the weights of optimization objectives, further improving the wind farm's operational efficiency and economic benefits.

[0023] Figure 2 This is a schematic diagram of the structure of an automatic wind power control device 200 for a wind farm monitoring system, provided by an embodiment of the present invention. The automatic wind power control device 200 for the wind farm monitoring system can vary significantly due to different configurations or performance characteristics. It may include one or more central processing units (CPUs) 210 (e.g., one or more processors) and a memory 220, and one or more storage media 230 (e.g., one or more mass storage devices) storing application programs 233 or data 232. The memory 220 and storage media 230 can be temporary or persistent storage. The program stored in the storage media 230 may include one or more modules (not shown in the diagram), each module including a series of instruction operations on the automatic wind power control device 200 for the wind farm monitoring system. Furthermore, the processor 210 may be configured to communicate with the storage media 230 and execute the series of instruction operations in the storage media 230 on the automatic wind power control device 200 for the wind farm monitoring system.

[0024] The wind power active power automatic control device 200 for a wind farm monitoring system may also include one or more power supplies 240, one or more wired or wireless network interfaces 250, one or more input / output interfaces 260, and / or one or more operating systems 231, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that... Figure 2 The structure of the wind power active power automatic control device shown for a wind farm monitoring system does not constitute a limitation on the wind power active power automatic control device for a wind farm monitoring system. It may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0025] The present invention also provides an automatic control device for active power of wind power in a wind farm monitoring system. The automatic control device for active power of wind power in a wind farm monitoring system includes a memory and a processor. The memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the processor performs the steps of the automatic control method for active power of wind power in the wind farm monitoring system described in the above embodiments.

[0026] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the automatic control method for wind power active power of the wind farm monitoring system.

[0027] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0028] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0029] The above-described 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 the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for automatic control of active power of wind power in a wind farm monitoring system, characterized in that, The automatic control method for active power of wind power in a wind farm monitoring system includes: Collect grid dispatch instructions, system frequency, grid connection point voltage, active power, reactive power, speed status of each wind turbine outlet, and ambient wind speed data to form a global real-time status dataset for the wind farm. Based on the global real-time status dataset of the wind farm, the entire wind farm is equivalent to a virtual synchronous machine model, the equivalent parameter set of the virtual synchronous machine is obtained, and the upper limit of the maximum power generation of the entire field under the current wind conditions and the individual operating boundary constraint set of each wind turbine are calculated synchronously. Taking the received active power setting command from the superior as the overall objective, and based on the virtual synchronous machine equivalent parameter set and the individual operating boundary constraint set, collaborative optimization calculations are performed with multiple optimization objectives, including reducing network losses within the field, balancing unit loads, and reserving frequency regulation capacity, to generate a global optimized power setpoint vector; specifically: Using the active power setting command from the power grid dispatch communication interface as the total power constraint for the optimization problem, and based on the equivalent virtual inertia parameters and equivalent damping coefficient parameters from the equivalent parameter set of the virtual synchronous machine, a frequency support optimization objective is constructed. Using the dynamic power adjustment range of each unit and the cluster power adjustment constraint from the individual operation boundary constraint set as boundary conditions, a feasible solution space for the optimization problem is constructed, and a mathematical model for the multi-objective optimization problem is established in a comprehensive manner. Based on the power flow and voltage data of the collection lines from the global real-time status dataset of the wind farm, the sensitivity coefficient of the impact of the active power change of each unit on the network loss of the entire farm is calculated, and the network loss sensitivity matrix is ​​generated. Based on the load status data of each unit from the global real-time status dataset of the wind farm, the load balance index between computer groups and its sensitivity to power allocation are used to generate a load balance sensitivity vector. According to the real-time frequency fluctuation of the power grid, the weight allocation ratio of the three optimization objectives of reducing network loss, balancing load and reserving capacity is dynamically adjusted to generate a dynamic multi-objective weight coefficient set. Using the network loss sensitivity matrix and the load balancing sensitivity vector as search direction guides, within the neighborhood of the current running point, based on the feasible solution space defined by the individual running boundary constraint set, the mathematical model of the multi-objective optimization problem is iteratively solved to output the Pareto optimal solution and obtain the global optimization power setpoint vector. The global optimized power setpoint vector is substituted into the preset power flow calculation model for verification to ensure that no power grid safety constraints are violated. If the verification fails, the dynamic multi-objective weight coefficient set is adjusted and the solution is recalculated until a safe and feasible optimization result is obtained, and the verified global optimized power setpoint vector is output. Based on the real-time frequency deviation of the power grid or the preset operating strategy, an active control mode is selected from multiple preset control modes, and the global optimized power setpoint vector is adjusted based on the selected mode to generate a control instruction set. The control command set is issued to execute active power adjustment actions, and the new global real-time status data of the wind farm generated after execution is used as feedback and input into the global real-time status data set of the wind farm, thereby forming a closed-loop automatic control system.

2. The automatic control method for active power of wind power in a wind farm monitoring system according to claim 1, characterized in that, include: The system receives active power setting instructions from the superior energy management system to form the first input data stream. It also collects the instantaneous values ​​of the system frequency and the three-phase voltage at the grid connection point to form the second input data stream. The system periodically collects the active power output value, reactive power output value, and wind turbine speed value of each unit to form the third input data stream. Finally, it collects the feedforward wind speed at the wind farm entrance and the real-time wind speed at each unit to form the fourth input data stream. A multi-dimensional wind farm operation status data snapshot is generated based on the first input data stream, the second input data stream, the third input data stream, and the fourth input data stream; Based on the rate of change of system frequency in the second input data stream, the sampling frequency configuration parameters of the high-precision frequency and voltage measurement unit and the intelligent measurement and control device are dynamically adjusted. When the frequency changes drastically, the sampling rate is automatically increased. The data validity is verified and noise is filtered on the fused multi-dimensional wind farm operation status data snapshot. Abnormal data points are removed and smoothed data is added. The global real-time status dataset of the wind farm is output.

3. The automatic control method for active power of wind power in a wind farm monitoring system according to claim 2, characterized in that, include: Based on the real-time speed and active power values ​​of each wind turbine generator set from the global real-time status dataset of the wind farm, the total kinetic energy of the entire wind farm is dynamically calculated and mapped to an equivalent virtual inertia parameter that varies with time. Based on the historical power response data and speed fluctuation data of each wind turbine generator set from the global real-time status dataset of the wind farm, the equivalent damping coefficient parameter is obtained. The equivalent virtual inertia parameter and the equivalent damping coefficient parameter are combined to form the virtual synchronous machine equivalent parameter set. Based on the feedforward wind speed data from the global real-time status dataset of the wind farm and the real-time wind speed data at each turbine, combined with the pre-stored aerodynamic power characteristic curves of the wind turbines, the maximum possible capture power of each computer group under the current wind speed is calculated. The maximum possible capture power of all turbines is accumulated, and the estimated network loss value calculated based on the real-time power flow is deducted to obtain the upper limit of the maximum power that can be generated in the entire field under the current wind conditions. Based on the real-time operating data of each unit from the global real-time status dataset of the wind farm, and combined with the pre-stored unit mechanical load model and thermal capacity model, the upper and lower limits of the adjustable active power of each unit under the current operating state are calculated to form the dynamic power adjustment range of each unit. Based on the power flow and voltage data of the collection lines from the global real-time status dataset of the wind farm, and in conjunction with the power grid safety operation regulations, cluster power regulation constraints aimed at preventing line overload are calculated. The dynamic power regulation range of all units and all cluster power regulation constraints are integrated to generate individual operating boundary constraint sets for each wind turbine unit.

4. The automatic control method for active power of wind power in a wind farm monitoring system according to claim 1, characterized in that, include: The system frequency data from the measurement unit at the grid connection point of the wind farm is monitored in real time. The deviation value and the rate of change of the current frequency from the rated frequency are calculated. Based on the preset frequency deviation threshold and the rate of change threshold, combined with the operation strategy instructions from the grid dispatch communication interface, an optimal active control mode is automatically selected through the mode selection logic table, and the mode decision signal is output. The system receives the global optimized power setpoint vector from the multi-objective optimization and power allocation decision steps as the basic adjustment command. Based on the control mode type indicated by the mode decision signal, the system calls the corresponding control algorithm to modify the basic adjustment command and generate a preliminary control command set. The preliminary control instruction set is compared and verified with the adjustable upper and lower limits of the power of each unit from the individual operating boundary constraint set. The preliminary control instruction set is checked to see if it meets the transmission capacity constraint of the collector line and the voltage stability constraint. The control instruction set that has passed the safety verification is output. Add a timestamp, instruction sequence number, and target unit number to each instruction in the control instruction set, encapsulate the instruction set into a data packet in a standard communication protocol format, add a data checksum, and generate a control instruction set that can be directly sent to the underlying wind turbine controller.

5. The automatic control method for active power of wind power in a wind farm monitoring system according to claim 4, wherein the corresponding control algorithm is invoked to correct the basic adjustment command: When frequency modulation mode is selected, the additional active power regulation is calculated based on the frequency deviation value and the equivalent parameter set of the virtual synchronous machine. When the margin mode is selected, the total reserve capacity of the entire field is increased at the cost of reducing the current total output. When the safety mode is selected, the unit instructions that may exceed the limits are preferentially adjusted based on the individual operating boundary constraint set.

6. The automatic control method for active power of wind power in a wind farm monitoring system according to claim 4, characterized in that, include: The control command set is distributed to the underlying controller of each target wind turbine through the internal communication network of the wind farm. The command reception confirmation signal and execution start signal returned by each underlying controller are received in real time, and a command execution status matrix is ​​generated. Within a preset time window after the command is executed, execution effect data such as pitch angle, generator torque, and actual output active power are collected. Data on total active power, frequency characteristics, and voltage changes after adjustment are also collected. The collected execution effect data are synchronized in time and validated for data validity to generate a command execution effect dataset. The actual output active power in the instruction execution effect dataset is compared with the instruction expectation value in the control instruction set to calculate the instruction tracking error index and generate a control performance evaluation report. Based on the instruction tracking error index and control performance evaluation report, the equivalent parameter set of the virtual synchronizer and the control algorithm parameters are dynamically adjusted to form an updated control parameter set. The actual operating data of each unit in the instruction execution effect dataset is integrated into the global real-time status dataset of the wind farm, replacing the original historical data. The updated control parameter set is fed back to the dynamic equivalent and operating boundary calculation steps and the multi-objective optimization and power allocation decision steps as the calculation parameters for the next control cycle. Through continuous data updates and parameter adjustments, a closed-loop automatic control system with adaptive capabilities is formed.

7. The automatic control method for active power of wind power in a wind farm monitoring system according to claim 1, characterized in that, It also includes online assessment and early warning steps for system health status: Based on the vibration, temperature and electrical parameters of each unit from the global real-time status dataset of the wind farm, the remaining service life assessment index of each component is calculated, and a system health status assessment report and early fault warning signals are generated according to the preset warning threshold. The system receives equipment health indicators from the system health status assessment report and combines them with grid frequency characteristics from the global real-time status dataset of the wind farm. When a decline in equipment health is detected, the system automatically adjusts the control mode decision logic and generates an adaptive control mode strategy. Real-time monitoring of voltage sag / surge signals and abnormal frequency fluctuation signals from the measurement unit at the wind farm grid connection point; when extreme operating conditions exceeding the safety threshold are detected, emergency power control commands are generated based on the virtual synchronous machine equivalent parameter set and individual operating boundary constraint set. Periodically collect operational data from each step, including instruction execution effect datasets, control performance evaluation reports, and system health status evaluation reports. Based on a preset evaluation index system, including computer group utilization, power generation efficiency index, and grid support contribution, automatically generate wind farm active power control efficiency evaluation reports.

Images