Power generation regulation and control method and device of wind power plant, electronic equipment, medium and product

Abstract

The invention relates to the technical field of wind power generation, and discloses a power generation regulation and control method and device of a wind power plant, electronic equipment, a storage medium and a program product, and the power generation regulation and control method of the wind power plant comprises the steps: constructing a reference field domain cooperative operation state according to historical wind power operation data; according to the real-time wind power operation data and the reference field domain cooperative operation state, identifying a fault fan in the target wind power plant to obtain a target fault fan; according to the power output value and the disturbance data of the target fault fan, determining a station-level total power vacancy corresponding to the target fault fan; and distributing station-level total power vacancy for each target fan according to the power increasing control parameter of each target fan in the healthy fan set until the station-level total power vacancy distribution is finished. According to the method, the vacancy power is determined by identifying the fault fan, so that the healthy fan bears the vacancy power, and the coordination of the multiple fans and the accuracy of power generation are improved.

Description

Technical Field

[0001] This invention relates to the field of wind power generation technology, specifically to power generation control methods, devices, electronic equipment, media, and products for wind farms. Background Technology

[0002] With the acceleration of the global energy transition, wind power, as an important component of clean energy, is experiencing a continuous expansion in installed capacity and scale. Large wind farms typically consist of dozens or even hundreds of wind turbines, forming a complex aero-mechanical-electrical coupling system. In such high-density wind turbine clusters, as the wind flows through the entire wind farm, the operation of upstream turbines generates a significant wake effect on downstream turbines, leading to a decrease in incoming wind speed and an increase in turbulence intensity. Therefore, it is necessary to regulate the wind turbines of the wind farm according to the total power generation.

[0003] In related technologies, wind farm operation control methods involve adjusting the pitch angle and generator torque of a single wind turbine to capture maximum wind energy at a specific wind speed. Based on data acquisition and monitoring control systems, alarms or shutdowns are triggered by setting fixed thresholds for turbine operating parameters. However, this method fails to coordinate the interaction between turbines at the system level, focusing solely on individual turbine power generation. This exacerbates wake losses across the entire wind farm, creating negative effects and impacting the stability and accuracy of wind power generation. Summary of the Invention

[0004] This invention provides a method, device, electronic equipment, storage medium, and program product for wind farm power generation control, in order to solve the problem that the wind farm operation control method in related technologies exacerbates the wake loss of the entire wind farm, resulting in negative effects and affecting the stability and accuracy of wind power generation.

[0005] In a first aspect, the present invention provides a method for regulating the power generation of a wind farm, comprising: acquiring historical wind power operation data of a target wind farm; constructing a benchmark field coordinated operation state based on the historical wind power operation data; the benchmark field coordinated operation state being used to characterize the balance relationship between instantaneous energy output and structural load distribution of the target wind farm under various benchmark wind resource inputs; acquiring real-time wind power operation data of the target wind farm; identifying faulty wind turbines in the target wind farm based on the real-time wind power operation data and the benchmark field coordinated operation state, thereby obtaining the target faulty wind turbines; and determining the power output value of the target faulty wind turbines. Based on disturbance data, determine the total power deficit at the site level corresponding to the target faulty wind turbine; the total power deficit at the site level corresponding to the target faulty wind turbine is used to characterize the power shortage of the target wind farm caused by the fault of the target wind turbine; obtain the set of healthy wind turbines in the target wind farm, and allocate the total power deficit at the site level to each target wind turbine according to the power boosting control parameters of each target wind turbine in the set of healthy wind turbines, until the allocation of the total power deficit at the site level is completed, so as to regulate the power generation of the target wind farm; the power boosting control parameters are used to characterize the power boosting capability, power boosting control constraints, and power boosting economic efficiency of the healthy wind turbines.

[0006] This invention constructs a baseline field collaborative operating state based on historical wind power operation data, quantitatively characterizing the energy output and structural load balance relationship of wind farms under different wind resource conditions, providing a precise comparison benchmark for subsequent fault identification. Based on real-time wind power operation data and the baseline field collaborative operating state, this invention identifies faulty wind turbines in the target wind farm, obtaining the target faulty turbine. By comparing the differences between real-time data and the baseline state, automatic and accurate identification of faulty wind turbines is achieved. Based on the power output value and disturbance data of the target faulty wind turbine, this invention determines the corresponding total power deficit at the farm level. Through the power output value and disturbance data of the target faulty wind turbine, the farm-level power deficit caused by the fault is accurately quantified, providing a clear quantitative target for subsequent power compensation allocation. This invention acquires a set of healthy wind turbines from a target wind farm. Based on the power boosting control parameters of each target turbine in the healthy turbine set, it allocates a total power deficit at the farm level to each target turbine until the farm-level total power deficit allocation is completed. This allows for power generation regulation of the target wind farm. Power deficit allocation based on the power boosting control parameters of the healthy turbines ensures that the total power deficit is completely covered while also guaranteeing the operational safety of individual turbines and the economic efficiency of overall regulation. Compared with related technologies, this invention considers the mutual influence between turbines, achieves rapid compensation for power deficits, and improves the stability and accuracy of wind power generation.

[0007] In one optional implementation, a baseline field collaborative operation state is constructed based on historical wind power operation data, including: determining the wake influence weight between every two wind turbines in the target wind farm based on historical wind power operation data; the wake influence weight is used to characterize the aerodynamic influence between every two wind turbines caused by the wake; constructing a directed aerodynamic correlation diagram of the target wind farm based on the wake influence weight; the directed aerodynamic correlation diagram is used to represent the aerodynamic influence between multiple wind turbines in the target wind farm; performing cluster analysis on historical wind power operation data to obtain multiple wind condition clusters, performing probability distribution analysis on each wind condition cluster to obtain the baseline power vector and baseline load vector corresponding to each wind condition cluster; and constructing the baseline field collaborative operation state based on the directed aerodynamic correlation diagram, the baseline power vector, and the baseline load vector.

[0008] In one optional implementation, based on real-time wind power operation data and a reference field cooperative operating state, the faulty wind turbines in the target wind farm are identified to obtain the target faulty wind turbines. This includes: matching the real-time wind power operation data with the reference field cooperative operating state to obtain the target directed aerodynamic correlation diagram, target reference power vector, and target reference load vector corresponding to the real-time wind power operation data; determining the real-time power vector and real-time load vector corresponding to the real-time wind power operation data; and determining whether the target wind farm has experienced a field cooperative state failure based on the target reference power vector, target reference load vector, real-time power vector, and real-time load vector. When the target wind farm experiences a field-coordinated imbalance, a power flow matrix is ​​constructed based on the target directed aerodynamic correlation diagram and the power generation of multiple wind turbines in the target wind farm. A power residual vector is obtained based on the difference between the real-time power vector of each wind turbine and the target baseline power vector. The power flow matrix and the power residual vector are input into a preset optimization objective function for optimization to obtain a disturbance potential vector. Wind turbines with disturbance potentials greater than a preset noise threshold are selected as candidate faulty wind turbines. The preset optimization objective function is used to obtain a sparse disturbance vector through norm constraints. The candidate faulty wind turbines are verified, and the target faulty wind turbine is obtained based on the verification results.

[0009] In one optional implementation, determining whether the target wind farm has experienced field-wide coordinated imbalance based on the target reference power vector, the target reference load vector, the real-time power vector, and the real-time load vector includes: determining the total power deviation of the entire field based on the real-time power vector and the target reference power vector; determining the load distribution difference based on the distance between the real-time load vector and the target reference load vector; and determining that the target wind farm has experienced field-wide coordinated imbalance if the total power deviation of the entire field is greater than a preset total power deviation threshold or the load distribution difference is greater than a preset load distribution difference threshold.

[0010] In one optional implementation, the candidate faulty fan is verified, and the target faulty fan is obtained based on the verification results. This includes: acquiring audio data of the candidate faulty fan and extracting voiceprint features from the audio data; the voiceprint features are used to characterize the mechanical operating state of the candidate faulty fan; acquiring historical audio data of the candidate faulty fan and determining the operating range of the feature values ​​based on the historical audio data; when the voiceprint features are not within the operating range of the feature values, determining the cross-correlation coefficient between the voiceprint features and the electrical power deviation of the candidate faulty fan, verifying the candidate faulty fan based on the cross-correlation coefficient, and obtaining the target faulty fan based on the verification results.

[0011] In one optional implementation, the total power deficit at the plant level corresponding to the target faulty wind turbine is determined based on the power output value and disturbance data of the target faulty wind turbine, including: determining the average deviation based on the average difference between the reference power value and the power output value of the target faulty wind turbine within a preset time window; determining the cascading power loss based on the product of the power flow matrix elements, disturbance potential, and propagation efficiency factor in the disturbance data of the target faulty wind turbine; and obtaining the total power deficit at the plant level corresponding to the target faulty wind turbine based on the sum of the average deviation and the cascading power loss.

[0012] In one optional implementation, a set of healthy wind turbines in the target wind farm is obtained. Based on the power enhancement control parameters of each target wind turbine in the healthy wind turbine set, a total power deficit at the farm level is allocated to each target wind turbine until the farm-level total power deficit allocation is completed. This includes: determining a set of healthy wind turbines based on the wind turbines in the target wind farm excluding the target faulty wind turbine; inputting the power enhancement cost function in the power enhancement control parameters of each healthy wind turbine in the healthy wind turbine set into a preset optimization objective function based on total compensation power constraints and individual power constraints to obtain a preset power enhancement amount for each healthy wind turbine; the preset optimization objective function is a function that minimizes the total power enhancement cost of the healthy wind turbines; obtaining a product result by multiplying the preset power enhancement amount of each healthy wind turbine by the unit power enhancement cost coefficient; obtaining the power enhancement potential value of each healthy wind turbine by quotienting the power enhancement efficiency in the power enhancement control parameters and the product result; and allocating the farm-level total power deficit to multiple healthy wind turbines according to the power enhancement potential value until the farm-level total power deficit allocation is completed.

[0013] In one optional implementation, the total power deficit at the plant level is allocated to multiple healthy wind turbines according to their power enhancement potential values ​​until the allocation of the total power deficit at the plant level is completed. This includes: selecting a target healthy wind turbine with the highest current power enhancement potential value from the set of healthy wind turbines and allocating the corresponding target power to the target healthy wind turbine; removing the target healthy wind turbine from the set of healthy wind turbines, subtracting the target power from the total power deficit at the plant level, and returning to the step of selecting the target healthy wind turbine with the highest current power enhancement potential value from the set of healthy wind turbines, until the allocation of the total power deficit at the plant level is completed.

[0014] In one optional implementation, the power generation control method for a wind farm further includes: storing the target faulty wind turbine and its fault propagation path to obtain a disturbance event knowledge base; performing statistical analysis on the target faulty wind turbine and its fault propagation path based on the disturbance event knowledge base to determine the expansion trend of the target faulty wind turbine; and generating predictive maintenance instructions based on the expansion trend to maintain the target faulty wind turbine.

[0015] Secondly, the present invention provides a power generation control device for a wind farm, comprising: a baseline operating state construction module, used to acquire historical wind power operation data of a target wind farm, and construct a baseline field coordinated operating state based on the historical wind power operation data; the baseline field coordinated operating state is used to characterize the balance relationship between instantaneous energy output and structural load distribution of the target wind farm under various baseline wind resource inputs; a faulty wind turbine identification module, used to acquire real-time wind power operation data of the target wind farm, and identify faulty wind turbines in the target wind farm based on the real-time wind power operation data and the baseline field coordinated operating state, thereby obtaining the target faulty wind turbine; and a power deficit determination module, used to determine the power deficit... Based on the power output value and disturbance data of the target faulty wind turbine, the total power deficit at the farm level corresponding to the target faulty wind turbine is determined. The total power deficit at the farm level corresponding to the target faulty wind turbine is used to characterize the power shortage of the target wind farm caused by the fault of the target wind turbine. The power allocation module is used to obtain the set of healthy wind turbines in the target wind farm, and allocate the total power deficit at the farm level to each target wind turbine according to the power boost control parameters of each target wind turbine in the healthy wind turbine set, until the allocation of the total power deficit at the farm level is completed, so as to regulate the power generation of the target wind farm. The power boost control parameters are used to characterize the power boost capability, power boost control constraints and power boost economic efficiency of the healthy wind turbines.

[0016] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the wind farm power generation control method of the first aspect or any corresponding embodiment described above.

[0017] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute the wind farm power generation control method of the first aspect or any corresponding embodiment described above.

[0018] Fifthly, the present invention provides a computer program product, including computer instructions, which are used to cause a computer to execute the power generation control method of a wind farm according to the first aspect or any corresponding embodiment described above. Attached Figure Description

[0019] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of an application scenario according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the first method for regulating the power generation of a wind farm according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the second process of the power generation control method for a wind farm according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the third process of the wind farm power generation control method according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the fourth process of the wind farm power generation control method according to an embodiment of the present invention; Figure 6 This is a schematic diagram of the fifth process of the wind farm power generation control method according to an embodiment of the present invention; Figure 7 This is a structural block diagram of a power generation control device for a wind farm according to an embodiment of the present invention; Figure 8 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0023] As an optional application scenario of this invention, such as Figure 1 As shown, the power generation control system of this wind farm may include at least one terminal device and at least one server. Figure 1The system is illustrated in the example, which includes a computer 101, a mobile terminal 102, and a server 103, and the terminal devices such as the computer 101 and the mobile terminal 102 are connected to the server 103 through a network 110.

[0024] Specifically, the terminal device can be a smartphone, tablet, laptop, PDA, desktop computer, game console, smart TV, smart wearable device, in-vehicle terminal, VR (Virtual Reality) device, AR (Augmented Reality) device, etc. Server 103 can be a standalone physical server, a server cluster, a distributed system, or a cloud server providing cloud services. Network 110 can be a wired or wireless network, examples of which include, but are not limited to, the Internet, corporate intranet, local area network, wide area network, mobile communication network, and combinations thereof.

[0025] With the acceleration of the global energy transition, wind power, as an important component of clean energy, is experiencing a continuous expansion in installed capacity and scale. Currently, large wind farms typically consist of dozens or even hundreds of wind turbines, forming a complex aero-mechanical-electrical coupling system. In such high-density wind turbine clusters, as the wind flows through the entire wind farm, the operation of upstream turbines generates a significant wake effect on downstream turbines, leading to a decrease in the incoming wind speed and an increase in turbulence intensity. This mutual interference not only causes a loss in total power generation but also causes dynamic unevenness in the load on the turbine structure, accelerates fatigue damage to key components, and seriously affects the economic benefits and operational safety of the wind farm throughout its entire life cycle.

[0026] To address these challenges, relevant wind farm operation control technologies mainly focus on two levels: firstly, optimal control of individual wind turbines, adjusting pitch angle and generator torque to capture maximum wind energy at specific wind speeds; and secondly, monitoring and alarm systems based on data acquisition and monitoring control systems, triggering alarms or shutdowns by setting fixed thresholds for turbine operating parameters. However, these methods have significant limitations. Individual turbine optimization control strategies cannot coordinate the interaction between turbines at the system level, potentially exacerbating overall wake losses in pursuit of higher individual turbine power generation, resulting in negative effects. Furthermore, SCADA (Supervisory Control and Data Acquisition) alarm mechanisms based on fixed thresholds are lagging, typically only acting after a fault has occurred or its impact has escalated, making it difficult to accurately pinpoint the source of complex disturbances and quantify the impact of localized faults on overall power generation performance.

[0027] Furthermore, when a wind turbine in a wind farm shuts down due to a malfunction or experiences performance degradation, the relevant technologies lack an effective real-time compensation and collaborative recovery mechanism. Operators often have no choice but to passively accept the loss of power generation or rely on experience for manual power dispatch, resulting in slow and inefficient responses. Simultaneously, for systemic disturbance patterns caused by specific turbines and propagating throughout the turbine cluster, the relevant technologies lack effective means of detection and analysis, making it difficult to achieve predictive maintenance upgrades from reactive repairs to proactive early warnings.

[0028] This invention provides a method for regulating the power generation of a wind farm. By identifying faulty wind turbines and determining the power deficit, healthy wind turbines are made to take on the power deficit, thereby improving the coordination of multiple wind turbines and the accuracy of power generation.

[0029] According to an embodiment of the present invention, a method for regulating the power generation of a wind farm is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0030] This embodiment provides a method for controlling the power generation of a wind farm, which can be used with computer equipment. Figure 2 This is a first flowchart of a wind farm power generation control method according to an embodiment of the present invention, as follows: Figure 2 As shown, the process includes the following steps: Step S201: Obtain historical wind power operation data of the target wind farm, and construct a benchmark field collaborative operation state based on the historical wind power operation data; the benchmark field collaborative operation state is used to characterize the balance relationship between instantaneous energy output and structural load distribution of the target wind farm under various benchmark wind resource inputs.

[0031] The target wind farm is the wind farm to be regulated for power generation; the historical wind power operation data is the historical wind power resource data of the past full year. The historical wind power operation data includes historical wind resource data and high-frequency data of historical operating parameters of each wind turbine in the target wind farm during the corresponding time period. For example, the historical wind resource data includes, but is not limited to, inlet wind speed, wind direction, environmental turbulence intensity, wind shear index, air density, etc., and the high-frequency data of historical operating parameters includes, but is not limited to, generator active power, pitch angle, generator speed, yaw position, and physical quantities reflecting structural loads.

[0032] In some optional implementations, after obtaining the historical wind power operation data of the target wind farm, the historical wind power operation data is preprocessed, including: data cleaning, data alignment, data normalization, and the data used to construct the baseline field collaborative operation state is the preprocessed historical wind power operation data.

[0033] In some alternative implementations, the reference wind resource input is a typical wind condition used as a reference to construct a stable reference operating state; the instantaneous energy output is the total power generation of the target wind farm at a certain moment; and the structural load distribution is the stress distribution borne by each structure during operation.

[0034] For example, the existing SCADA system of the target wind farm was upgraded, with 35 additional nacelle acoustic monitoring units deployed. Wind resource data was provided by two vertical profile lidar units and four auxiliary anemometer towers. The data was aggregated to a real-time database and a time-series database deployed in the central control center via industrial Ethernet and the OPCUA protocol.

[0035] Step S202: Obtain real-time wind power operation data of the target wind farm. Based on the real-time wind power operation data and the baseline field coordinated operation status, identify the faulty wind turbines in the target wind farm to obtain the target faulty wind turbines.

[0036] Among them, real-time wind power operation data is real-time wind power resource data, including the wind resource data at the current moment and the operating parameters of all wind turbines in the field, which are obtained through the wind farm monitoring and data acquisition system; faulty wind turbines are wind turbines with abnormal operating status and deviating from the baseline coordinated state; target faulty wind turbines are the faulty wind turbines that are finally identified after identification.

[0037] In some alternative implementations, performance disturbances that are difficult to locate quickly include blade leading-edge corrosion leading to a continuous decline in aerodynamic efficiency, slight deviations in the yaw system encoder causing wind misalignment, and reduced cooling system efficiency leading to generator temperature rise and power limitation.

[0038] Step S203: Determine the total power deficit at the farm level corresponding to the target faulty wind turbine based on the power output value and disturbance data of the target faulty wind turbine; the total power deficit at the farm level corresponding to the target faulty wind turbine is used to characterize the power shortage of the target wind farm caused by the fault of the target faulty wind turbine.

[0039] The power output value is the actual electrical power currently generated by the target faulty wind turbine; the disturbance data is the disturbance data caused by the target faulty wind turbine. For example, the disturbance data includes power flow matrix elements, disturbance potential, and propagation efficiency factor.

[0040] In some alternative implementations, the total power deficit at the site level is the total power deficit of the target wind farm caused by the failure of the target faulty wind turbine.

[0041] Step S204: Obtain the set of healthy wind turbines in the target wind farm. Based on the power enhancement control parameters of each target wind turbine in the set of healthy wind turbines, allocate the total power deficit at the farm level to each target wind turbine until the allocation of the total power deficit at the farm level is completed, so as to regulate the power generation of the target wind farm. The power enhancement control parameters are used to characterize the power enhancement capability, power enhancement control constraints and power enhancement economic efficiency of the healthy wind turbines.

[0042] Among them, the healthy wind turbine set is the set of all wind turbines in the wind farm that are currently operating normally and without faults; the power increase control parameters are a set of parameters that characterize the power increase capability, power increase control constraints and power increase economic efficiency of the healthy wind turbines. For example, the power increase control parameters include the maximum allowable temporary power increase value for each wind turbine, the power increase cost function of the healthy wind turbines, and the power increase efficiency of the healthy wind turbines.

[0043] The wind farm power generation control method provided in this embodiment constructs a benchmark field collaborative operation state based on historical wind power operation data. It quantifies the energy output and structural load balance relationship of the wind farm under different wind resource conditions, providing a precise comparison benchmark for subsequent fault identification. Based on real-time wind power operation data and the benchmark field collaborative operation state, this invention identifies faulty wind turbines in the target wind farm, obtaining the target faulty turbine. By comparing the differences between real-time data and the benchmark state, it achieves automatic and accurate identification of the faulty turbine. Based on the power output value and disturbance data of the target faulty turbine, this invention determines the corresponding total power deficit at the farm level. Through the power output value and disturbance data of the target faulty turbine, it accurately quantifies the farm level power deficit caused by the fault, providing a clear quantitative target for subsequent power compensation allocation. This invention acquires a set of healthy wind turbines from a target wind farm. Based on the power boosting control parameters of each target turbine in the healthy turbine set, it allocates a total power deficit at the farm level to each target turbine until the farm-level total power deficit allocation is completed. This allows for power generation regulation of the target wind farm. Power deficit allocation based on the power boosting control parameters of the healthy turbines ensures that the total power deficit is completely covered while also guaranteeing the operational safety of individual turbines and the economic efficiency of overall regulation. Compared with related technologies, this invention considers the mutual influence between turbines, achieves rapid compensation for power deficits, and improves the stability and accuracy of wind power generation.

[0044] This embodiment provides a method for controlling the power generation of a wind farm, which can be used with computer equipment. Figure 3 This is a second flowchart of a wind farm power generation control method according to an embodiment of the present invention, as follows: Figure 3 As shown, the process includes the following steps: Step S301: Obtain historical wind power operation data of the target wind farm, and construct a benchmark field collaborative operation state based on the historical wind power operation data; the benchmark field collaborative operation state is used to characterize the balance relationship between instantaneous energy output and structural load distribution of the target wind farm under various benchmark wind resource inputs.

[0045] Specifically, step S301 includes: Step S3011: Based on historical wind power operation data, determine the wake influence weight between every two wind turbines in the target wind farm; the wake influence weight is used to characterize the aerodynamic influence caused by the wake between every two wind turbines.

[0046] Based on the actual geographical coordinates of the wind farm and the distribution of the main wind directions within the region, a computational fluid dynamics model or a validated high-precision empirical wake model is used to determine the wake influence weight between any two wind turbines in the target wind farm. For example, the formula for determining the wake influence weight is as follows:

[0047] in, For wind turbine i With wind turbine j The weight of the wake effect between them For free-flow wind speed, For downstream wind turbines j At actual wind speed, Free-flow wind speed Downstream wind turbines j The difference between the actual wind speed and the wind speed.

[0048] Step S3012: Construct a directed aerodynamic correlation diagram of the target wind farm based on the wake influence weight; the directed aerodynamic correlation diagram is used to analyze the aerodynamic influence between multiple wind turbines in the target wind farm.

[0049] Among them, the weighting based on wake influence Construct a directed aerodynamic correlation graph of a wind farm. In the directed aerodynamic correlation diagram For a set of nodes, nodes , indicating a fan, Depends on the edge set, directed edge Indicates wind turbine i For the fan j There are significant aerodynamic effects.

[0050] Step S3013: Perform cluster analysis on historical wind power operation data to obtain multiple wind condition clusters, perform probability distribution analysis on each wind condition cluster to obtain the reference power vector and reference load vector corresponding to each wind condition cluster.

[0051] In this process, historical wind power operation data is clustered according to wind conditions to obtain multiple wind condition clusters. Specifically, the KMeans or DBSCAN algorithm is used to divide the historical wind power operation data into several typical wind condition clusters based on wind speed and wind direction as the main features.

[0052] In some alternative implementations, for each typical wind condition cluster, the wind turbine power vector corresponding to all samples within that wind condition cluster is calculated. and load vector The joint probability distribution is specifically fitted using a multivariate Gaussian distribution or a Gaussian mixture model, with the expected value of this distribution used as the baseline power vector for this wind condition. and reference load vector .

[0053] Step S3014: Construct the reference field cooperative operating state based on the directed aerodynamic correlation diagram, the reference power vector, and the reference load vector.

[0054] Among them, all typical wind conditions are clustered together , The corresponding directed aerodynamic correlation diagrams are stored together to form a queryable reference field cooperative operating state. The baseline field collaborative operation state represents the optimal or normal operating state that the wind turbine cluster can achieve under typical historical wind conditions.

[0055] Step S302: Obtain real-time wind power operation data of the target wind farm. Based on the real-time wind power operation data and the baseline field coordinated operation status, identify the faulty wind turbines in the target wind farm to obtain the target faulty wind turbines.

[0056] Specifically, step S302 includes: Step S3021: Based on the real-time wind power operation data, match is performed in the reference field cooperative operation state to obtain the target directional aerodynamic correlation diagram, target reference power vector, and target reference load vector corresponding to the real-time wind power operation data.

[0057] Specifically, based on the wind speed and direction in the real-time wind power operation data, in the reference field collaborative operation state, the most similar reference wind conditions are retrieved through nearest neighbor search or similarity matching, and the target oriented aerodynamic correlation graph, target reference power vector, and target reference load vector corresponding to the most similar reference wind conditions are read.

[0058] Step S3022: Based on the real-time wind power operation data, determine the real-time power vector and real-time load vector corresponding to the real-time wind power operation data. Based on the target reference power vector, target reference load vector, real-time power vector, and real-time load vector, determine whether the target wind farm has experienced field cooperative state imbalance.

[0059] Specifically, based on real-time wind power operation data, the real-time power vector and real-time load vector corresponding to the real-time wind power operation data are determined. This includes: performing cluster analysis on the real-time wind power operation data to obtain multiple real-time wind condition clusters; and performing probability distribution analysis on each real-time wind condition cluster to obtain the real-time power vector corresponding to each real-time wind condition cluster. and real-time load vector .

[0060] In some optional implementations, an imbalance determination is made by calculating the total power deviation and load distribution difference of the target wind farm and comparing them with preset total power deviation thresholds and preset load distribution difference thresholds.

[0061] In some optional implementations, step S3022 above includes: Step a1: Determine the total power deviation across the entire field based on the real-time power vector and the target reference power vector.

[0062] For example, the formula for determining the total power deviation of the entire field is:

[0063] in, This represents the total power deviation across the entire field. As the reference power vector, This is the real-time power vector.

[0064] In some alternative implementations, the total power deviation across the entire field is used to reflect the overall change in the total power generation capacity of the entire field.

[0065] Step a2: Determine the load distribution difference based on the distance between the real-time load vector and the target reference load vector.

[0066] For example, the formula for determining the load distribution difference is:

[0067] in, Due to differences in load distribution, For real-time load vectors, This is the reference load vector.

[0068] In some alternative implementations, load distribution differences are used to reflect whether the uniformity of load distribution among wind turbines has changed. Sometimes, even if the total power remains constant, a deterioration in load distribution may indicate potential risks.

[0069] Step a3: If the total power deviation of the entire field is greater than the preset total power deviation threshold, or the load distribution difference is greater than the preset load distribution difference threshold, then it is determined that the target wind farm has experienced a field cooperative state imbalance.

[0070] Among them, the preset total power deviation threshold Difference threshold between load distribution and preset threshold It can be set based on the statistical fluctuation characteristics of historical data or operation and maintenance experience.

[0071] For example, when or If the target wind farm is found to have a field-coordinated state imbalance, the subsequent diagnostic process will be triggered immediately.

[0072] In some alternative implementations, when and If the target wind farm does not experience a field-coordinated state imbalance, the subsequent process of determining the faulty wind turbine will not be carried out.

[0073] Step S3023: When the target wind farm experiences field cooperative state imbalance, a power flow matrix is ​​constructed based on the target directional aerodynamic correlation diagram and the power generation of multiple wind turbines in the target wind farm. The power residual vector is obtained based on the difference between the real-time power vector of each wind turbine and the target reference power vector.

[0074] Specifically, based on the target directed aerodynamic correlation graph, the elements of the wind farm topological adjacency matrix are determined, i.e., whether there are direct and significant aerodynamic correlations between wind turbines. Based on the elements of the wind farm topological adjacency matrix and the power generation of multiple wind turbines, a power flow matrix is ​​constructed. F For example, the power flow matrix F elements in It can be represented as:

[0075] in, These are elements in the power flow matrix. These are the elements of the wind field topological adjacency matrix. For wind turbine i Power generation capacity, For wind turbine j Power generation capacity.

[0076] In some alternative implementations, Indicates wind turbine For the fan The intensity of the influence of the standardized power flow.

[0077] In some alternative implementations, the formula for determining the power residual vector is:

[0078] in, The power residual vector, As the reference power vector, This is the real-time power vector.

[0079] Step S3024: Input the power flow matrix and power residual vector into the preset optimization objective function for optimization to obtain the disturbance potential vector. The wind turbine corresponding to the target disturbance potential in the disturbance potential vector that is greater than the preset noise threshold is taken as the candidate faulty wind turbine. The preset optimization objective function is used to obtain a sparse disturbance vector through norm constraints.

[0080] The preset optimization objective function can be expressed as:

[0081] in, Represents the perturbation potential vector Take the minimum value. The power residual vector, F For the power flow matrix, This is a regularization parameter used to control the sparsity of the solutions. It is the disturbance potential vector of each wind turbine, and its non-zero elements indicate possible disturbance sources (i.e., candidate faulty wind turbines).

[0082] In some alternative implementations, optimization algorithms such as coordinate descent or near-end gradient descent are used to solve for the perturbation potential vector.

[0083] In some alternative implementations, the obtained perturbation potential vector is... The elements in the array are sorted in descending order of absolute value, and the top-ranked elements are selected. Position and its Fans with noise levels greater than a preset noise threshold are considered candidate faulty fans.

[0084] In some alternative implementations, in the directed aerodynamic correlation diagram Starting from the marked candidate faulty wind turbines, perform a depth-first or breadth-first traversal along the directed edges to depict the propagation path of the suspected disturbances affected by them.

[0085] Step S3025: Verify the candidate faulty fan and obtain the target faulty fan based on the verification results.

[0086] In some optional implementations, step S3025 above includes: Step b1: Obtain audio data of candidate faulty fans and extract voiceprint features from the audio data; voiceprint features are used to characterize the mechanical operating state of candidate faulty fans.

[0087] Among these measures, instructions are sent to candidate faulty wind turbines to activate the high-frequency acoustic sensors pre-installed in their nacelles or tower bases to collect 3060 seconds of audio data.

[0088] In some optional implementations, the acquired audio data is preprocessed, including noise reduction and frame windowing, and the data from which voiceprint features are subsequently extracted is the preprocessed audio data.

[0089] In some alternative implementations, acoustic signature features characterizing the mechanical operating state of the wind turbine are extracted from each frame of audio data. Exemplarily, these acoustic signature features include: Mel frequency cepstral coefficients, sound pressure level spectrum, spectral centroid, and spectral roll-off point.

[0090] Step b2: Obtain historical audio data of candidate faulty wind turbines, and determine the operating range of characteristic values ​​based on the historical audio data.

[0091] This involves acquiring historical audio data of candidate faulty wind turbines under normal conditions, extracting historical voiceprint feature data, training a baseline model using a Gaussian Mixture Model (GMM), or calculating the normal range of historical voiceprint feature data to obtain the feature value operating range.

[0092] Step b3: When the acoustic signature is not within the characteristic value operating range, determine the cross-correlation coefficient between the acoustic signature and the electrical power deviation of the candidate faulty fan, verify the candidate faulty fan based on the cross-correlation coefficient, and obtain the target faulty fan based on the verification results.

[0093] In some alternative implementations, the voiceprint features are compared with a benchmark model or the range of feature values. For example, the negative log-likelihood of the voiceprint features relative to the benchmark model is calculated, or the Mahalanobis distance between the voiceprint features and the range of feature values ​​is calculated. The larger the distance, the higher the degree of voiceprint anomaly. Therefore, when the voiceprint features are not within the range of feature values, it indicates that the degree of voiceprint anomaly is higher.

[0094] In some alternative implementations, the exponential time series corresponding to the acoustic signature features is aligned with the time series corresponding to the electrical power deviation of the candidate faulty fan.

[0095] In some optional implementations, the cross-correlation coefficient between the exponential time series corresponding to the acoustic signature features and the time series corresponding to the electrical power deviation of the candidate faulty wind turbine is calculated, and their correlation under different time lags is analyzed.

[0096] Specifically, if the verification results show that the abnormal acoustic signature appears earlier than or simultaneously with the significant decrease in electrical power, and the trends of the two at key time points are highly consistent, then this constitutes strong double verification evidence, confirming the candidate faulty fan as the target faulty fan. It also indicates that the abnormal mechanical condition is the cause of the electrical performance degradation, thus confirming that the candidate faulty fan is the actual source of the disturbance.

[0097] Step S303: Based on the power output value and disturbance data of the target faulty wind turbine, determine the total power deficit at the farm level corresponding to the target faulty wind turbine; the total power deficit at the farm level corresponding to the target faulty wind turbine is used to characterize the power shortage in the target wind farm caused by the fault of the target wind turbine. For details, please refer to... Figure 2 Step S203 of the illustrated embodiment will not be described again here.

[0098] Step S304: Obtain the set of healthy wind turbines for the target wind farm. Based on the power boost control parameters of each target wind turbine in the healthy wind turbine set, allocate a total power deficit at the farm level to each target wind turbine until the farm level total power deficit allocation is completed, in order to regulate the power generation of the target wind farm. The power boost control parameters are used to characterize the power boost capability, power boost control constraints, and power boost economic efficiency of the healthy wind turbines. For details, please refer to [link to relevant documentation]. Figure 2 Step S204 of the illustrated embodiment will not be described again here.

[0099] The wind farm power generation control method provided in this embodiment achieves a positioning accuracy of over 96% for performance disturbance sources through a dual verification mechanism of graph calculation and acoustic fingerprint. The average positioning time is reduced from several hours in traditional experience analysis to less than 10 minutes, greatly improving the efficiency of operation and maintenance response.

[0100] This embodiment provides a method for controlling the power generation of a wind farm, which can be used with computer equipment. Figure 4 This is a third flowchart of a wind farm power generation control method according to an embodiment of the present invention, as follows: Figure 4 As shown, the process includes the following steps: Step S401: Obtain historical wind power operation data of the target wind farm, and construct a baseline field collaborative operation state based on the historical wind power operation data. The baseline field collaborative operation state is used to characterize the balance relationship between instantaneous energy output and structural load distribution of the target wind farm under various baseline wind resource inputs. For details, please refer to... Figure 3 Step S301 of the illustrated embodiment will not be described again here.

[0101] Step S402: Obtain real-time wind power operation data of the target wind farm. Based on the real-time wind power operation data and the baseline field coordinated operation status, identify the faulty wind turbines in the target wind farm to obtain the target faulty wind turbines. For details, please refer to [link to relevant documentation]. Figure 3 Step S302 of the illustrated embodiment will not be described again here.

[0102] Step S403: Determine the total power deficit at the farm level corresponding to the target faulty wind turbine based on the power output value and disturbance data of the target faulty wind turbine; the total power deficit at the farm level corresponding to the target faulty wind turbine is used to characterize the power shortage of the target wind farm caused by the fault of the target faulty wind turbine.

[0103] Specifically, step S403 includes: Step S4031: Determine the average deviation based on the average difference between the reference power value and the power output value of the target faulty fan within a preset time window.

[0104] For example, the formula for determining the average deviation is:

[0105] in, The average deviation. To analyze the length of the time window, This is the starting moment of the field's cooperative state imbalance. For a moment Target faulty fan The reference power value, For a moment Target faulty fan The power output value.

[0106] In some alternative implementations, the average deviation is used to characterize the average deviation of the power output of the target faulty wind turbine from its reference value during the period of sustained imbalance.

[0107] Step S4032: Determine the cascading power loss based on the product of the power flow matrix elements, the disturbance potential, and the propagation efficiency factor in the disturbance data of the target faulty wind turbine.

[0108] For example, the formula for determining cascading power loss is:

[0109] in, For cascading power loss, These are elements of the power flow matrix. Target faulty fan The disturbance potential. As a propagation efficiency factor, The target set of faulty fans.

[0110] In some alternative implementations, cascading power loss is used to characterize the cascading power loss caused by a target faulty wind turbine to its downstream wind turbines.

[0111] Step S4033: Based on the sum of the average deviation and the cascading power loss, obtain the total power deficit at the station level corresponding to the target faulty wind turbine.

[0112] For example, the formula for determining the total power deficit at the station level is:

[0113] in, This represents a shortfall in total power at the station level. The average deviation. This is due to cascading power loss.

[0114] In some alternative implementations, the total power deficit at the site level is used to characterize the total power deficit at the site level caused by the failure of the target faulty wind turbine.

[0115] Step S404: Obtain the set of healthy wind turbines in the target wind farm. Based on the power boosting control parameters of each target wind turbine in the set of healthy wind turbines, allocate the total power deficit at the farm level to each target wind turbine until the allocation of the total power deficit at the farm level is completed, so as to regulate the power generation of the target wind farm. The power boosting control parameters are used to characterize the power boosting capability, power boosting control constraints, and power boosting economic efficiency of the healthy wind turbines.

[0116] Specifically, step S404 includes: Step S4041: Determine the set of healthy wind turbines based on the wind turbines in the target wind farm, excluding the target faulty wind turbine.

[0117] Specifically, identified target faulty wind turbines, as well as those whose performance has significantly degraded due to severe impacts on the disturbance propagation path, are marked as unhealthy wind turbines. All wind turbines in the target wind farm other than those in unhealthy condition are identified as the set of healthy wind turbines. .

[0118] Step S1042: Based on the total compensation power constraint and the individual power constraint, the power increase cost function in the power increase control parameters of each healthy fan in the healthy fan set is input into the preset optimization objective function to obtain the preset power increase amount of each healthy fan; the preset optimization objective function is a function that minimizes the total power increase cost of the healthy fan.

[0119] Among them, for each health fan in the health fan cluster The power boost control parameters include the maximum permissible temporary power boost value for each fan in the healthy fan cluster. , Power-increasing cost function Health fan power enhancement efficiency Specifically, the maximum permissible temporary power increase value for each fan in a healthy fan hub. The assessment comprehensively considers the wind turbine's current operating point, rotor aerodynamic limits, generator capacity limits, and the performance capabilities of the pitch and yaw systems, while preserving safety margins. Power increase cost function. The calculation formula is:

[0120] in, This is a cost factor that reflects the sensitivity of a healthy fan to additional loads. This represents the maximum permissible temporary power increase for each wind turbine. The unit power increase cost coefficient.

[0121] In some alternative implementations, the power efficiency of the health fan... The evaluation method is as follows: wind turbines located upwind and less affected by wake have a greater direct contribution to the overall power generation increase. High; conversely, increased power from downwind turbines may exacerbate the wake effect on downstream turbines. Lower.

[0122] In some optional implementations, the preset optimization objective function can be expressed as:

[0123] in, The target function can be a predefined optimization function, or it can be a global utility function. As decision variables, This represents a shortfall in total power at the station level. For the power increase cost function, This is the cost coefficient. To be assigned to the The power increase of the health fan. The total number of health fans is the number of health fan sets. Size.

[0124] In some alternative implementations, the total compensation power constraint can be expressed as:

[0125] in, To be assigned to the The power increase of the health fan. This represents a shortfall in total power at the station level. For health-conscious fans.

[0126] In some alternative implementations, the individual power constraint can be expressed as:

[0127] in, To be assigned to the The power increase of the health fan. For the centralized health fan The maximum permissible temporary power increase value for a health fan.

[0128] In some optional implementations, the power increase cost function in the power increase control parameters of each healthy fan in the healthy fan set is input into a preset optimization objective function to obtain a preset power increase amount for each healthy fan. .

[0129] Step S4043: Based on the product of the preset power increase amount of each healthy fan and the unit power increase cost coefficient, obtain the product result. Based on the quotient of the power increase efficiency in the power increase control parameters and the product result, obtain the power increase potential value of each healthy fan.

[0130] For example, the formula for determining the power enhancement potential value is:

[0131] in, For the first The potential for power enhancement of Taiwan's health-promoting fans. For the first The power enhancement efficiency of the Taiwan health fan. To be assigned to the The power increase of the health fan. For the first The unit power increase cost coefficient of the health fan.

[0132] Step S4044: According to the power enhancement potential value, the total power deficit at the site level is allocated to multiple healthy wind turbines until the allocation of the total power deficit at the site level is completed.

[0133] In some optional implementations, step S4044 above includes: Step c1: In the set of healthy fans, select the target healthy fan with the highest current power improvement potential value and allocate the corresponding target power to the target healthy fan.

[0134] Step c2 involves removing the target healthy fan from the healthy fan set, subtracting the target power from the total power deficit at the site level, and returning to the step of selecting the target healthy fan with the highest current power improvement potential value from the healthy fan set, until the allocation of the total power deficit at the site level is completed.

[0135] Among them, in the healthy fan cluster, the target healthy fan with the highest current power improvement potential value is selected, and the power it can provide is allocated to the target healthy fan. Specifically, the formula for allocating the power it can provide to the target healthy fan is:

[0136] in, For this allocation to the first The actual increased power output of the health fan, i.e., the target power, is shown in min, which represents the minimum value. For the centralized health fan The maximum permissible temporary power increase value for a Taiwan-made health fan. This represents the current power deficit.

[0137] In some alternative implementations, the smaller of the maximum additional capacity of the wind turbine and the remaining deficit is taken as the allocation amount for this time, which ensures that the wind turbine capacity is not exceeded and avoids excessive additional capacity.

[0138] In some alternative implementations, subtracting the allocation amount from the remaining deficit can be expressed as:

[0139] in, The remaining power deficit is calculated by subtracting the target power from the total power deficit at the power station level. For this allocation to the first The actual increased power output of the Taiwan health fan, i.e., the target power output. This represents the current power deficit.

[0140] In some alternative implementations, the step of selecting the target healthy wind turbine with the highest current power enhancement potential value from the set of healthy wind turbines is returned until the allocation of the total power deficit at the station level is completed or the current remaining power deficit is less than or equal to 0.

[0141] In some alternative implementations, the final power allocation scheme will be... The commands are converted into specific control instructions and sent to the corresponding healthy wind turbines through the wind farm energy management system. After receiving the instructions, the wind turbines adjust their power setpoints at the controller level to achieve coordinated power increase and jointly compensate for the power deficit at the farm level.

[0142] The wind farm power generation control method provided in this embodiment realizes the transformation from maximizing the power generation of a single unit to maximizing the overall net revenue of the wind farm. During the test period, the overall power generation of the entire farm increased by approximately 2.1% year-on-year. At the same time, through the coordinated optimization of load distribution, the equivalent fatigue load of key components was reduced to varying degrees, extending the service life of the equipment.

[0143] This embodiment provides a method for controlling the power generation of a wind farm, which can be used with computer equipment. Figure 5 This is a fourth flowchart of a wind farm power generation control method according to an embodiment of the present invention, as follows: Figure 5 As shown, the process includes the following steps: Step S501: Store the target faulty fan and the fault propagation path of the target faulty fan to obtain a disturbance event knowledge base.

[0144] The information corresponding to the target faulty fan and the fault propagation path of the target faulty fan includes: the fan number of the disturbance source, the fault type, the time of occurrence, the duration, the disturbance propagation path, the total power loss caused, and the compensation and maintenance measures taken.

[0145] Step S502: Based on the disturbance event knowledge base, perform statistical analysis on the target faulty fan and the fault propagation path of the target faulty fan to determine the expansion trend of the target faulty fan.

[0146] This involves regularly analyzing the historical disturbance event database, focusing on two main trends: the recurrence and reinforcement trends of the same source point, and the expansion trend of the impact range of the same propagation path. Specifically, the recurrence and reinforcement trend of the same source point involves analyzing whether the frequency of a specific wind turbine or a wind turbine in a specific location acting as a disturbance source increases over time, or whether the average power deficit it causes shows an upward trend. The expansion trend of the impact range of the same propagation path involves analyzing whether the number of affected wind turbines or the geographical area of ​​a specific disturbance propagation path expands systematically over time. Control chart methods from statistical process control, or trend testing methods from time series analysis, are used to determine whether the above trends are statistically significant, rather than random fluctuations.

[0147] Step S503: Based on the expansion trend, generate a predictive maintenance instruction to maintain the target faulty wind turbine.

[0148] Once a disturbance source or propagation path is identified as having a regular expansion trend, it is determined that there may be an evolving systemic defect behind it. Based on this determination, specific and operable predictive maintenance instructions are automatically generated.

[0149] In this embodiment of the invention, a historical disturbance event data warehouse is constructed to record the disturbance source, fault type, occurrence time, duration, disturbance propagation path, total power deficit, and compensation and maintenance measures taken for each real disturbance event. The data warehouse is periodically mined and analyzed to identify the recurrence and reinforcement trends of the same source or the expansion trend of the influence range of the same propagation path. Statistical process control or time series trend testing methods are used to determine whether the trend is statistically significant. If a regular expansion trend is identified, predictive maintenance instructions for the real disturbance source and the affected critical links are automatically generated.

[0150] This embodiment provides a method for controlling the power generation of a wind farm, which can be used with computer equipment. Figure 6 This is a fifth flowchart of a wind farm power generation control method according to an embodiment of the present invention, as follows: Figure 6 As shown, the process includes the following steps: Step S601: Obtain historical data and construct a baseline field collaborative operating state.

[0151] Among them, by analyzing and learning from historical normal operation data, a benchmark model of energy output and load distribution that wind farms should follow under different typical wind conditions was established, providing an accurate benchmark for subsequent real-time status assessment.

[0152] Specifically, historical wind resource inputs and wind turbine cluster operating parameters are obtained to construct a baseline field collaborative operation mode.

[0153] Step S602: Real-time perception and field co-state imbalance diagnosis and location of suspected disturbance sources.

[0154] Among these measures, the system continuously monitors the real-time operating status of the wind turbine cluster, quickly identifies field-coordinated imbalance events by comparing them with the baseline state, and uses graph computing technology to accurately locate suspected disturbance sources and their impact paths.

[0155] Specifically, the system senses wind resource input and wind turbine cluster operating parameters in real time, continuously compares the field collaborative operation state with the benchmark field collaborative operation state in real time, and determines that a field collaborative state imbalance has occurred when the total output power or load distribution of the entire field deviates from the benchmark. The system then uses graph calculation to analyze power flow data to locate the suspected disturbance source of the imbalance and the disturbance propagation path caused by the suspected disturbance source.

[0156] Step S603: Verification of the real disturbance source and calculation of power deficit based on dual voiceprint verification.

[0157] Among them, by introducing acoustic sensing data at suspected disturbance sources and performing cross-modal correlation analysis with electrical power data, the disturbance sources can be dually verified and confirmed, and the resulting station-level power deficit can be accurately quantified.

[0158] Specifically, after identifying the field-domain cooperative state imbalance, the acoustic sensors located on the suspected disturbance source are instructed to collect the operating sound patterns. By analyzing the temporal and logical correlation between the abnormal physical characteristics of the sound patterns and the abnormal electrical power, the suspected disturbance source is double-verified to confirm the real disturbance source. The total power deficit at the field level caused by the real performance disturbance event corresponding to the real disturbance source is then calculated.

[0159] Step S604: A game theory-based strategy for collaborative power enhancement compensation of healthy wind turbines.

[0160] After identifying the actual source of disturbance and the power deficit, a non-cooperative game theory algorithm is used to coordinate healthy, unaffected wind turbines within the field to temporarily increase power output in the most economical and effective way, thereby compensating for the total power deficit of the system and maintaining the power level required by the grid dispatch.

[0161] Specifically, a game theory algorithm is used to coordinate healthy wind turbines in the field, calculate the temporary power increase potential value of each wind turbine considering the power increase cost and global contribution efficiency, and drive the healthy wind turbine with the highest potential value to perform collaborative power increase to compensate for the total power shortage at the field level.

[0162] Step S605: Systemic defect identification and predictive maintenance instruction generation.

[0163] Among these methods, by deeply mining historical disturbance event data and analyzing the evolution of disturbance propagation paths, potential and developing systemic defects in wind farms can be identified, and predictive maintenance instructions can be generated to prevent problems before they occur.

[0164] Specifically, by mining data on the propagation path of the actual disturbance source, when the propagation range and impact pattern of the disturbance propagation path show a regular expansion trend, it is determined that there is an evolving systemic defect, and predictive maintenance instructions are generated for the actual disturbance source and the affected critical links.

[0165] This invention first constructs a baseline field-wide collaborative operating state using historical data; then, it monitors and locates imbalances and suspected disturbance sources in the field-wide collaborative state in real time; next, it uses acoustic signatures and power data to confirm the actual disturbance sources and calculate power deficits; subsequently, it uses game theory to coordinate the power increase of healthy wind turbines to compensate for the deficits; finally, it mines historical data to identify systemic defects and generates predictive maintenance instructions, realizing a shift from passive response to proactive early warning, and from single-unit optimization to site-wide collaboration. This invention, by constructing a baseline for field-wide collaborative operating state, combining graph calculation and acoustic signature dual verification to locate actual disturbance sources, and utilizing game theory algorithms to drive the collaborative power increase of healthy wind turbines, achieves rapid compensation for power deficits and early warning of systemic defects.

[0166] This embodiment also provides a power generation control device for a wind farm, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0167] This embodiment provides a power generation control device for a wind farm, such as... Figure 7 As shown, it includes: The baseline operating state construction module 701 is used to acquire historical wind power operation data of the target wind farm and construct a baseline field collaborative operating state based on the historical wind power operation data. The baseline field collaborative operating state is used to characterize the balance relationship between instantaneous energy output and structural load distribution of the target wind farm under various baseline wind resource inputs.

[0168] The faulty wind turbine identification module 702 is used to acquire real-time wind power operation data of the target wind farm, and identify the faulty wind turbines in the target wind farm based on the real-time wind power operation data and the baseline field collaborative operation status, thereby obtaining the target faulty wind turbine.

[0169] The power deficit determination module 703 is used to determine the total power deficit at the farm level corresponding to the target faulty wind turbine based on the power output value and disturbance data of the target faulty wind turbine. The total power deficit at the farm level corresponding to the target faulty wind turbine is used to characterize the power shortage of the target wind farm caused by the fault of the target faulty wind turbine.

[0170] The power allocation module 704 is used to acquire the set of healthy wind turbines in the target wind farm, and allocate the total power deficit at the farm level to each target wind turbine according to the power boosting control parameters of each target wind turbine in the set of healthy wind turbines, until the allocation of the total power deficit at the farm level is completed, so as to regulate the power generation of the target wind farm; the power boosting control parameters are used to characterize the power boosting capability, power boosting control constraints and power boosting economic efficiency of the healthy wind turbines.

[0171] In some optional implementations, the baseline runtime construction module 701 includes: The wake influence weight determination unit is used to determine the wake influence weight between every two wind turbines in the target wind farm based on historical wind power operation data; the wake influence weight is used to characterize the aerodynamic influence caused by the wake between every two wind turbines.

[0172] The aerodynamic correlation diagram determination unit is used to construct a directed aerodynamic correlation diagram of the target wind farm based on the wake influence weight; the directed aerodynamic correlation diagram is used to determine the aerodynamic influence between multiple wind turbines in the target wind farm.

[0173] The clustering analysis unit is used to perform clustering analysis on historical wind power operation data to obtain multiple wind condition clusters. Probability distribution analysis is performed on each wind condition cluster to obtain the reference power vector and reference load vector corresponding to each wind condition cluster.

[0174] The reference operating state construction unit is used to construct the reference field cooperative operating state based on the directed aerodynamic correlation diagram, the reference power vector, and the reference load vector.

[0175] In some alternative implementations, the faulty fan identification module 702 includes: The reference data determination unit is used to perform matching in the reference field cooperative operation state based on real-time wind power operation data to obtain the target directional aerodynamic correlation diagram, target reference power vector, and target reference load vector corresponding to the real-time wind power operation data.

[0176] The cooperative state identification unit is used to determine the real-time power vector and real-time load vector corresponding to the real-time wind power operation data based on the real-time wind power operation data, and to determine whether the target wind farm has experienced field cooperative state imbalance based on the target reference power vector, target reference load vector, real-time power vector, and real-time load vector.

[0177] The residual data determination unit is used to construct a power flow matrix based on the target directional aerodynamic correlation diagram and the power generation of multiple wind turbines in the target wind farm when the target wind farm experiences field cooperative state imbalance. The power residual vector is obtained based on the difference between the real-time power vector of each wind turbine and the target reference power vector.

[0178] The candidate faulty wind turbine determination unit is used to input the power flow matrix and power residual vector into the preset optimization objective function for optimization to obtain the disturbance potential vector. The wind turbines corresponding to the target disturbance potentials in the disturbance potential vector that are greater than the preset noise threshold are used as candidate faulty wind turbines. The preset optimization objective function is used to obtain a sparse disturbance vector through norm constraints.

[0179] The wind turbine fault verification unit is used to verify candidate faulty wind turbines and obtain the target faulty wind turbine based on the verification results.

[0180] In some optional implementations, the cooperative state identification unit includes: The total power deviation determination subunit is used to determine the total power deviation across the entire field based on the real-time power vector and the target reference power vector.

[0181] The load distribution determination sub-unit is used to determine the load distribution difference based on the distance between the real-time load vector and the target reference load vector.

[0182] The cooperative state identification subunit is used to determine that the target wind farm has experienced a field cooperative state imbalance if the total power deviation of the entire field is greater than a preset total power deviation threshold, or the load distribution difference is greater than a preset load distribution difference threshold.

[0183] In some optional implementations, the wind turbine fault verification unit includes: The voiceprint feature determination subunit is used to acquire audio data of candidate faulty fans and extract voiceprint features from the audio data; the voiceprint features are used to characterize the mechanical operating state of candidate faulty fans.

[0184] The operating range determination subunit is used to acquire historical audio data of candidate faulty fans and determine the operating range of characteristic values ​​based on the historical audio data.

[0185] The faulty fan verification subunit is used to determine the cross-correlation coefficient between the acoustic signature and the electrical power deviation of the candidate faulty fan when the acoustic signature is not within the characteristic value operating range. The candidate faulty fan is verified based on the cross-correlation coefficient, and the target faulty fan is obtained based on the verification results.

[0186] In some alternative implementations, the power deficit determination module 703 includes: The deviation determination unit is used to determine the average deviation based on the average difference between the reference power value and the power output value of the target faulty fan within a preset time window.

[0187] The power loss determination unit is used to determine the cascading power loss based on the product of the power flow matrix elements, the disturbance potential, and the propagation efficiency factor in the disturbance data of the target faulty wind turbine.

[0188] The power deficit determination unit is used to obtain the total power deficit at the plant level corresponding to the target faulty wind turbine based on the sum of the average deviation and the cascading power loss.

[0189] In some alternative implementations, the power distribution module 704 includes: The healthy wind turbine determination unit is used to determine the set of healthy wind turbines in the target wind farm, excluding the target faulty wind turbine.

[0190] The power increase determination unit is used to input the power increase cost function in the power increase control parameters of each healthy fan in the healthy fan set into the preset optimization objective function based on the total compensation power constraint and the individual power constraint, so as to obtain the preset power increase amount of each healthy fan; the preset optimization objective function is a function that minimizes the total power increase cost of the healthy fan.

[0191] The potential value determination unit is used to obtain the product result by multiplying the preset power increase amount of each healthy fan by the unit power increase cost coefficient, and to obtain the power increase potential value of each healthy fan by the quotient of the power increase efficiency in the power increase control parameters and the product result.

[0192] The power distribution unit is used to allocate the total power deficit at the site level to multiple healthy wind turbines according to the power enhancement potential value, until the allocation of the total power deficit at the site level is completed.

[0193] In some alternative implementations, the power distribution unit includes: The target power allocation subunit is used to select the target healthy fan with the highest current power improvement potential value from the healthy fan set, and allocate the corresponding target power to the target healthy fan.

[0194] The power allocation subunit is used to remove the target healthy fan from the healthy fan set, subtract the target power from the total power deficit at the site level, and return to the step of selecting the target healthy fan with the highest current power improvement potential value in the healthy fan set, until the total power deficit at the site level is allocated.

[0195] In some alternative implementations, the power generation control method for wind farms further includes: The faulty fan information storage module is used to store the target faulty fan and the fault propagation path of the target faulty fan, and to obtain a disturbance event knowledge base.

[0196] The fault statistics and analysis module is used to perform statistical analysis on the target faulty wind turbine and the fault propagation path of the target faulty wind turbine based on the disturbance event knowledge base, and to determine the expansion trend of the target faulty wind turbine.

[0197] The maintenance instruction generation module is used to generate predictive maintenance instructions based on the expansion trend in order to maintain the target faulty wind turbine.

[0198] The wind farm power generation control device provided in this embodiment of the invention can execute the wind farm power generation control method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the above modules and units are the same as in the corresponding embodiments described above, and will not be repeated here.

[0199] Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0200] The following is a detailed reference. Figure 8 This diagram illustrates a suitable structural schematic for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 801, which can perform various appropriate actions and processes based on a program stored in read-only memory (ROM) 802 or a program loaded from memory 808 into random access memory (RAM) 803. The RAM 803 also stores various programs and data required for the operation of the electronic device. The processor 801, ROM 802, and RAM 803 are interconnected via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0201] Typically, the following devices can be connected to I / O interface 805: input devices 806 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 807 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 808 including, for example, magnetic tapes, hard disks, etc.; and communication devices 809. Communication device 809 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 8 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0202] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 809, or installed from a memory 808, or installed from a ROM 802. When the computer program is executed by the processor 801, it performs the functions defined in the wind farm power generation control method of the embodiments of the present invention.

[0203] Figure 8 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0204] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the wind farm power generation control method shown in the above embodiments is implemented.

[0205] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0206] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for regulating the power generation of a wind farm, characterized in that, The method includes: Historical wind power operation data of the target wind farm is obtained, and a benchmark field collaborative operation state is constructed based on the historical wind power operation data. The benchmark field collaborative operation state is used to characterize the balance relationship between instantaneous energy output and structural load distribution of the target wind farm under various benchmark wind resource inputs. The real-time wind power operation data of the target wind farm is obtained. Based on the real-time wind power operation data and the collaborative operation state of the reference field, the faulty wind turbines in the target wind farm are identified to obtain the target faulty wind turbines. Based on the power output value and disturbance data of the target faulty wind turbine, the total power deficit at the farm level corresponding to the target faulty wind turbine is determined; the total power deficit at the farm level corresponding to the target faulty wind turbine is used to characterize the power shortage of the target wind farm caused by the fault of the target faulty wind turbine; A set of healthy wind turbines for the target wind farm is obtained. Based on the power enhancement control parameters of each target wind turbine in the set of healthy wind turbines, the total power deficit at the farm level is allocated to each target wind turbine until the allocation of the total power deficit at the farm level is completed, so as to regulate the power generation of the target wind farm. The power enhancement control parameters are used to characterize the power enhancement capability, power enhancement control constraints, and power enhancement economic efficiency of the healthy wind turbines.

2. The method according to claim 1, characterized in that, The step of constructing a baseline field collaborative operation state based on the historical wind power operation data includes: Based on the historical wind power operation data, the wake influence weight between every two wind turbines in the target wind farm is determined; the wake influence weight is used to characterize the aerodynamic influence between every two wind turbines caused by the wake. Based on the wake influence weight, a directed aerodynamic correlation diagram of the target wind farm is constructed; the directed aerodynamic correlation diagram is used to analyze the aerodynamic influence between multiple wind turbines in the target wind farm. Cluster analysis is performed on the historical wind power operation data to obtain multiple wind condition clusters. Probability distribution analysis is performed on each wind condition cluster to obtain the reference power vector and reference load vector corresponding to each wind condition cluster. Based on the directional aerodynamic correlation diagram, the reference power vector, and the reference load vector, the reference field cooperative operating state is constructed.

3. The method according to claim 2, characterized in that, The step of identifying faulty wind turbines in the target wind farm based on the real-time wind power operation data and the baseline field cooperative operation state, to obtain the target faulty wind turbines, includes: Based on the real-time wind power operation data, matching is performed in the reference field cooperative operation state to obtain the target directional aerodynamic correlation diagram, target reference power vector, and target reference load vector corresponding to the real-time wind power operation data; Based on the real-time wind power operation data, determine the real-time power vector and real-time load vector corresponding to the real-time wind power operation data. Based on the target reference power vector, the target reference load vector, the real-time power vector, and the real-time load vector, determine whether the target wind farm has experienced field cooperative state imbalance. When the target wind farm experiences a field cooperative state imbalance, a power flow matrix is ​​constructed based on the target directional aerodynamic correlation diagram and the power generation of multiple wind turbines in the target wind farm. The power residual vector is obtained based on the difference between the real-time power vector of each wind turbine and the target reference power vector. The power flow matrix and the power residual vector are input into a preset optimization objective function for optimization to obtain a disturbance potential vector. The wind turbines corresponding to the target disturbance potentials in the disturbance potential vector that are greater than a preset noise threshold are selected as candidate faulty wind turbines. The preset optimization objective function is used to obtain a sparse disturbance vector through norm constraints. The candidate faulty fans are verified, and the target faulty fan is obtained based on the verification results.

4. The method according to claim 3, characterized in that, The step of determining whether the target wind farm has experienced a field-wide cooperative state imbalance based on the target reference power vector, the target reference load vector, the real-time power vector, and the real-time load vector includes: The total power deviation across the entire field is determined based on the real-time power vector and the target reference power vector. The load distribution difference is determined based on the distance between the real-time load vector and the target reference load vector; If the total power deviation of the entire field is greater than a preset total power deviation threshold, or the load distribution difference is greater than a preset load distribution difference threshold, then it is determined that the target wind farm has experienced a field cooperative state imbalance.

5. The method according to claim 3, characterized in that, The step of verifying the candidate faulty wind turbines and obtaining the target faulty wind turbine based on the verification results includes: Audio data of the candidate faulty fan is acquired, and voiceprint features are extracted from the audio data; the voiceprint features are used to characterize the mechanical operating state of the candidate faulty fan. Obtain historical audio data of the candidate faulty wind turbines, and determine the operating range of the characteristic values ​​based on the historical audio data; When the voiceprint feature is not within the operating range of the feature value, the cross-correlation coefficient between the voiceprint feature and the electrical power deviation of the candidate faulty fan is determined, the candidate faulty fan is verified based on the cross-correlation coefficient, and the target faulty fan is obtained based on the verification result.

6. The method according to any one of claims 1 to 5, characterized in that, The step of determining the total power deficit at the plant level corresponding to the target faulty wind turbine based on the power output value and disturbance data of the target faulty wind turbine includes: The average deviation is determined based on the average difference between the reference power value and the power output value of the target faulty fan within a preset time window. The cascading power loss is determined by multiplying the power flow matrix elements, the disturbance potential, and the propagation efficiency factor in the disturbance data of the target faulty wind turbine. The total power deficit at the plant level corresponding to the target faulty wind turbine is obtained by summing the average deviation and the cascading power loss.

7. The method according to any one of claims 1 to 5, characterized in that, The process of obtaining a set of healthy wind turbines for the target wind farm, and allocating the total power deficit at the farm level to each target wind turbine according to the power increase control parameters of each target wind turbine in the set of healthy wind turbines, until the allocation of the total power deficit at the farm level is completed, includes: The set of healthy wind turbines is determined based on the wind turbines in the target wind farm, excluding the target faulty wind turbine. Based on the total compensation power constraint and the individual power constraint, the power increase cost function in the power increase control parameters of each healthy fan in the healthy fan set is input into the preset optimization objective function to obtain the preset power increase amount of each healthy fan; the preset optimization objective function is a function that minimizes the total power increase cost of the healthy fan. The power increase preset amount of each of the healthy fans is multiplied by the unit power increase cost coefficient to obtain the product result. The power increase potential value of each of the healthy fans is obtained by quotient of the power increase efficiency in the power increase control parameters and the product result. According to the power enhancement potential value, the total power deficit at the site level is allocated to multiple healthy wind turbines until the allocation of the total power deficit at the site level is completed.

8. The method according to claim 7, characterized in that, The step of allocating the total power deficit at the power plant level to multiple healthy wind turbines according to the power enhancement potential value, until the allocation of the total power deficit at the power plant level is completed, includes: From the set of healthy fans, select the target healthy fan with the highest current power enhancement potential value and assign the corresponding target power to the target healthy fan; The steps of removing the target healthy fan from the set of healthy fans, subtracting the target power from the total power deficit at the site level, returning to the set of healthy fans, and selecting the target healthy fan with the highest current power improvement potential value are repeated until the allocation of the total power deficit at the site level is completed.

9. The method according to any one of claims 1 to 5, characterized in that, The method further includes: The target faulty fan and its fault propagation path are stored to obtain a disturbance event knowledge base; Based on the disturbance event knowledge base, statistical analysis is performed on the target faulty wind turbine and the fault propagation path of the target faulty wind turbine to determine the expansion trend of the target faulty wind turbine. Based on the expansion trend, predictive maintenance instructions are generated to maintain the target faulty wind turbine.

10. A power generation control device for a wind farm, characterized in that, The device includes: The benchmark operating state construction module is used to acquire historical wind power operation data of the target wind farm and construct a benchmark field collaborative operating state based on the historical wind power operation data; the benchmark field collaborative operating state is used to characterize the balance relationship between instantaneous energy output and structural load distribution of the target wind farm under various benchmark wind resource inputs; The faulty wind turbine identification module is used to acquire real-time wind power operation data of the target wind farm, and identify the faulty wind turbines in the target wind farm based on the real-time wind power operation data and the baseline field cooperative operation state to obtain the target faulty wind turbines; The power deficit determination module is used to determine the total power deficit at the farm level corresponding to the target faulty wind turbine based on the power output value and disturbance data of the target faulty wind turbine; the total power deficit at the farm level corresponding to the target faulty wind turbine is used to characterize the amount of power shortage in the target wind farm caused by the fault of the target faulty wind turbine; The power allocation module is used to acquire a set of healthy wind turbines in the target wind farm, and allocate the total power deficit at the farm level to each target wind turbine according to the power boosting control parameters of each target wind turbine in the set of healthy wind turbines, until the allocation of the total power deficit at the farm level is completed, so as to regulate the power generation of the target wind farm; the power boosting control parameters are used to characterize the power boosting capability, power boosting control constraints, and power boosting economic efficiency of the healthy wind turbines.

11. An electronic device, characterized in that, include: The system includes a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the power generation control method of the wind farm as described in any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the power generation control method of the wind farm as described in any one of claims 1 to 9.

13. A computer program product, characterized in that, It includes computer instructions for causing a computer to execute the power generation control method of the wind farm as described in any one of claims 1 to 9.

Images