Wind farm control method, device, equipment and medium for reducing the impact of prediction errors

Abstract

The present disclosure relates to a wind farm control method, device, equipment and medium for reducing the impact of prediction errors, including: obtaining current measured impact data and current simulated impact data of the target wind farm in the current time period, wherein the current measured impact data and the current simulated impact data both include wind condition data; using a pre-trained control parameter prediction model to process the current measured impact data and the current simulated impact data to obtain target control parameters of the target wind farm in a future time period after the current time period, wherein the target control parameters are operating condition data when the target wind farm outputs maximum power in the future time period; based on the target control parameters, controlling the operation of each unit in the target wind farm. Through the above-mentioned method, the problem of errors in the control strategy caused by the process of determining the predicted wind conditions based on the actual wind conditions is avoided, the elimination effect of the wake effect is improved, and the practical application value of the wind farm wake control technology is enhanced.

Description

Technical Field

[0001] The present disclosure relates to the field of renewable energy power generation technology, and in particular to a wind farm control method, device, equipment and medium for reducing the impact of prediction errors. Background Art

[0002] The units in a wind farm are power generation devices that convert wind energy in the air into electrical energy. While the upstream units obtain wind energy from the wind, a wake area with reduced wind speed is formed downstream, generating a wake effect. The wake effect will cause the wind speed of the downstream units to decrease, ultimately reducing the overall power generation of the wind farm.

[0003] In order to eliminate the wake effect, the related art generally uses the measured wind conditions of the wind farm in the current time period to determine the predicted wind conditions, and then determines the control parameters of the wind farm based on the predicted wind conditions, so that the operation of each unit in the wind farm is controlled based on the control parameters. However, the measured wind conditions do not match the control parameters determined using the predicted wind conditions, so that when the measured wind conditions of the wind farm and the control parameters corresponding to the predicted wind conditions are used for power generation, the power generation loss caused by the wake effect cannot be effectively reduced, and the problem of power generation loss caused by the wake effect cannot be solved. Summary of the invention

[0004] In order to solve the above technical problem or at least partially solve the above technical problem, the present disclosure provides a wind farm control method, device, equipment and medium for reducing the influence of prediction errors.

[0005] In a first aspect, the present disclosure provides a wind farm control method for reducing the impact of prediction errors, the method comprising:

[0006] Acquire current measured impact data and current simulated impact data of the target wind farm in a current time period, wherein both the current measured impact data and the current simulated impact data include wind condition data;

[0007] The current measured impact data and the current simulated impact data are processed by using a pre-trained control parameter prediction model to obtain a target control parameter of the target wind farm in a future time period after the current time period, wherein the target control parameter is an operating condition data of the target wind farm when the target wind farm outputs a maximum power generation power in the future time period;

[0008] Based on the target control parameter, the operation of each unit in the target wind farm is controlled.

[0009] In a second aspect, the present disclosure provides a wind farm control device for reducing the impact of prediction errors, the device comprising:

[0010] An acquisition module, used to acquire current measured impact data and current simulated impact data of a target wind farm in a current time period, wherein both the current measured impact data and the current simulated impact data include wind condition data;

[0011] a determination module, configured to process the current measured impact data and the current simulated impact data using a pre-trained control parameter prediction model to obtain a target control parameter of the target wind farm in a future time period after the current time period, wherein the target control parameter is an operating condition data of the target wind farm when the target wind farm outputs a maximum power generation power in the future time period;

[0012] An operation control module is used to control the operation of each unit in the target wind farm based on the target control parameter.

[0013] In a third aspect, an embodiment of the present disclosure further provides a device, the device comprising:

[0014] one or more processors;

[0015] a storage device for storing one or more programs,

[0016] When one or more programs are executed by one or more processors, the one or more processors implement the method provided in the first aspect.

[0017] In a fourth aspect, an embodiment of the present disclosure further provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method provided in the first aspect is implemented.

[0018] Compared with the prior art, the technical solution provided by the embodiments of the present disclosure has the following advantages:

[0019] A wind farm control method, device and medium for reducing the impact of prediction errors in the disclosed embodiments obtains the current measured impact data and the current simulated impact data of the target wind farm in the current time period, wherein the current measured impact data and the current simulated impact data both include wind condition data; uses a pre-trained control parameter prediction model to process the current measured impact data and the current simulated impact data to obtain the target control parameters of the target wind farm in the future time period after the current time period, wherein the target control parameters are the operating condition data when the target wind farm outputs the maximum power generation in the future time period; based on the target control parameters, the operation of each unit in the target wind farm is controlled. Through the above-mentioned method, the current measured impact data including wind condition data, the current simulated impact data and the pre-trained control parameter prediction model can be used to directly determine the target control parameters for controlling the operation of each unit in the wind farm, thereby avoiding the problem of errors in the control strategy caused by the process of determining the predicted wind condition based on the actual wind condition, improving the elimination effect of the wake effect, and ultimately improving the practical application value of the wind farm wake control technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0021] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0022] Figure 1 A schematic flow chart of a wind farm control method for reducing the impact of prediction errors provided by an embodiment of the present disclosure;

[0023] Figure 2 A flow chart of a method for controlling a parameter prediction model provided by an embodiment of the present disclosure;

[0024] Figure 3 A schematic diagram of the structure of a wind farm control device for reducing the impact of prediction errors provided by an embodiment of the present disclosure;

[0025] Figure 4 A schematic diagram of the structure of a wind farm control device for reducing the impact of prediction errors provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0026] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features in the embodiments can be combined with each other without conflict.

[0027] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.

[0028] The embodiments of the present disclosure provide a wind farm control method, device and medium capable of eliminating wake effects and reducing the impact of prediction errors.

[0029] Next, first combine Figure 1 to Figure 2 A wind farm control method for reducing the impact of prediction errors provided by an embodiment of the present disclosure is described.

[0030] Figure 1 A schematic flow chart of a wind farm control method for reducing the impact of prediction errors provided by an embodiment of the present disclosure is shown.

[0031] In the disclosed embodiment, Figure 1 The wind farm control method for reducing the impact of prediction errors shown can be executed by a wind farm control device for reducing the impact of prediction errors. The wind farm control device for reducing the impact of prediction errors can be an electronic device or a server. The electronic device can include but is not limited to fixed terminals such as smart phones, laptops, desktop computers, etc. The server can be a cloud server or a server cluster or other device with storage and computing functions. The embodiments of the present disclosure are explained in detail with the electronic device as the execution subject.

[0032] like Figure 1 As shown, the wind farm control method for reducing the impact of prediction errors may include the following steps.

[0033] S110 . Obtain current measured impact data and current simulated impact data of the target wind farm in a current time period, wherein both the current measured impact data and the current simulated impact data include wind condition data.

[0034] In actual application, when it is desired to improve the elimination effect of the wake effect, real-time data and simulation data that affect the working conditions in the target wind farm are collected to determine the control parameters of the target wind farm using the real-time data and simulation data.

[0035] The target wind farm may be any large wind farm that needs to eliminate the wake effect.

[0036] The currently measured impact data refers to the actual impact data of the target wind farm in the current time period. Optionally, the currently measured impact data includes one or more of the following combinations: currently measured wind condition data, currently measured temperature data, currently measured humidity data, and currently measured air pressure data.

[0037] The current simulation impact data is simulation data of the target wind farm operating in the current time period. Optionally, the current simulation impact data includes one or more of the following combinations: current simulation wind condition data, current wake impact relationship between each unit, current simulation temperature data, current simulation humidity data, and current simulation air pressure data.

[0038] Specifically, the method for determining the wake influence relationship includes but is not limited to the following methods:

[0039] Obtain the power generation influence relationship of each unit in the target wind farm, and use the power generation influence relationship as the wake influence relationship;

[0040] Obtaining the relative position relationship of each unit in the target wind farm under the current measured wind conditions in the current measured impact data, and using the relative position relationship as the wake impact relationship;

[0041] The impeller area of ​​each unit in the target wind farm in the wake area of ​​other units is obtained, and the impeller area is used as the wake influence relationship.

[0042] Among them, the power generation influence relationship can be understood as the constraint relationship that affects the power generation power of each unit.

[0043] The relative position relationship may be the relative coordinate values ​​of each generator set.

[0044] Among them, the impeller area can be used to evaluate the impact of each unit on the wake area of ​​other units.

[0045] S120. Use a pre-trained control parameter prediction model to process the current measured impact data and the current simulated impact data to obtain the target control parameters of the target wind farm in a future time period after the current time period, wherein the target control parameters are the operating condition data of the target wind farm when it outputs the maximum power generation power in the future time period.

[0046] In order to avoid the problem that the wake effect cannot be effectively solved due to the mismatch between the measured wind conditions and the control parameters determined by the predicted wind conditions, in this embodiment, the current measured impact data and the current simulation impact data are input into a pre-trained control parameter prediction model to determine the target control parameters when the target wind farm outputs the maximum power generation using the control parameter prediction model.

[0047] Among them, the specific implementation methods of S120 include but are not limited to the following methods: based on the spatiotemporal feature extraction network in the control parameter prediction model, extract the current measured spatiotemporal features from the current measured impact data, and extract the current simulation spatiotemporal features from the current simulation impact data; based on the attention mechanism in the control parameter prediction model, perform attention processing on the current measured spatiotemporal features and the current simulation spatiotemporal features respectively; based on the fully connected network in the control parameter prediction model, fit the current measured spatiotemporal features after attention processing and the current simulation spatiotemporal features after attention processing, and use the obtained fitting data as the target control parameters.

[0048] Among them, the current measured spatiotemporal features include measured features in the time dimension and the space dimension, and the current simulated spatiotemporal features include predicted features in the time dimension and the space dimension.

[0049] The target control parameter refers to the predicted operating condition when all units in the target wind farm output maximum power. Optionally, the target control parameter includes but is not limited to the yaw angle, rotation speed, and pitch angle of each unit.

[0050] The maximum power generation power may be determined based on historical data of a reference wind farm corresponding to the target wind farm.

[0051] Therefore, the spatiotemporal feature extraction network, attention mechanism and fully connected network in the control parameter prediction model are used to process the current measured impact data and the current simulation impact data, so as to accurately determine the operating condition data of each unit when it outputs maximum power in the future time period.

[0052] S130: Based on the target control parameters, control the operation of each unit in the target wind farm.

[0053] It can be understood that since the target control parameters are the operating data that make each unit output the maximum power in the future time period, when the target control parameters are used to control the operation of each unit in the target wind farm, the target wind farm can output the maximum power, thereby effectively solving the problem of power generation loss caused by the wake effect.

[0054] In the disclosed embodiment, the current measured impact data and the current simulated impact data of the target wind farm in the current time period are obtained, wherein the current measured impact data and the current simulated impact data both include wind condition data; the current measured impact data and the current simulated impact data are processed using a pre-trained control parameter prediction model to obtain target control parameters of the target wind farm in a future time period after the current time period, wherein the target control parameters are operating condition data when the target wind farm outputs the maximum power generation in the future time period; based on the target control parameters, the operation of each unit in the target wind farm is controlled. Through the above-mentioned method, the current measured impact data including wind condition data, the current simulated impact data and the pre-trained control parameter prediction model can be used to directly determine the target control parameters for controlling the operation of each unit in the wind farm, thereby avoiding the problem of errors in the control strategy caused by the process of determining the predicted wind conditions based on the actual wind conditions, improving the effect of eliminating the wake effect, and ultimately improving the practical application value of the wind farm wake control technology.

[0055] In another embodiment of the present disclosure, a training method of a control parameter prediction model is explained in detail.

[0056] Figure 2 A flow chart of a method for training a control parameter prediction model provided by an embodiment of the present disclosure is shown.

[0057] like Figure 2 As shown, the training method of the control parameter prediction model includes the following steps.

[0058] S210 . Obtain historical measured impact data and historical simulation impact data of the reference wind farm in a historical time period, and measured control parameters of the reference wind farm in a future time period after the historical time period.

[0059] The reference wind farm is any wind farm whose impact data is close to that of the target wind farm.

[0060] The historical time period may be a time period such as 3 days, 5 days, or 7 days before the current time.

[0061] The historical measured impact data refers to the actual impact data of the reference wind farm in the historical time period. Optionally, the historical measured impact data includes one or more of the following combinations: historical measured wind condition data, historical measured temperature data, historical measured humidity data, and historical measured air pressure data.

[0062] The historical simulation impact data is simulation data of the operation of the reference wind farm in the historical time period. Optionally, the historical simulation impact data includes one or more of the following combinations: historical simulation wind condition data, historical wake impact relationship between each unit, historical simulation temperature data, historical simulation humidity data, and historical simulation air pressure data.

[0063] Specifically, the method for determining the historical wake influence relationship includes but is not limited to the following methods:

[0064] Obtain the power generation influence relationship of each unit in the reference wind farm, and use the power generation influence relationship as the historical wake influence relationship;

[0065] Obtain the relative position relationship of each unit in the reference wind farm under the historical measured wind conditions in the historical measured impact data, and use the relative position relationship as the historical wake impact relationship;

[0066] The impeller area of ​​each unit in the reference wind farm in the wake area of ​​other units is obtained, and the impeller area is used as the historical wake influence relationship.

[0067] The measured control parameters refer to the measured working conditions of each unit in the reference wind farm in a historical period of time. Optionally, the measured control parameters include but are not limited to the yaw angle, rotation speed, and pitch angle of each unit.

[0068] S220: Obtain the measured unit power of each unit in the reference wind farm within a historical period, wherein the sum of the measured unit powers of each wind turbine is the measured power of the wind farm.

[0069] The measured unit power is the actual power generated by each unit in the reference wind farm under the historical measured impact data and according to the measured control parameters. Then the measured unit power generated by each unit in the reference wind farm is added to obtain the wind farm measured power.

[0070] S230, using historical measured impact data, historical simulation impact data, measured control parameters, wind farm measured power and maximum power generation, iteratively train the preset neural network to obtain a control parameter prediction model.

[0071] In this embodiment, specific implementations of S230 include but are not limited to the following:

[0072] S2301, using a preset neural network to process historical measured impact data and historical simulation impact data to obtain prediction control parameters when the reference wind farm outputs the measured power of the wind farm;

[0073] S2302: If the measured power of the wind farm is not equal to the maximum power generation, iteratively train the preset neural network based on the measured control parameters and the predicted control parameters until the measured power of the wind farm output by the reference wind farm is equal to the maximum power generation, thereby obtaining a control parameter prediction model.

[0074] Among them, the preset neural network may include a spatiotemporal feature extraction network, an attention mechanism, and a fully connected network. Accordingly, S2301 may specifically include the following steps: based on the spatiotemporal feature extraction network in the preset neural network, extract historical measured spatiotemporal features from historical measured impact data, and extract historical simulation spatiotemporal features from historical simulation impact data; based on the attention mechanism in the preset neural network, perform attention processing on the historical measured spatiotemporal features and the historical simulation spatiotemporal features respectively; based on the fully connected network in the preset neural network, fit the historical measured spatiotemporal features after attention processing and the historical simulation spatiotemporal features after attention processing, and use the obtained fitting data as the prediction control parameters.

[0075] Among them, the historical measured spatiotemporal features include measured features in the time dimension and the space dimension, and the historical simulated spatiotemporal features include predicted features in the time dimension and the space dimension.

[0076] It can be understood that if the measured power of the wind farm is not equal to the maximum power generation power, it means that the wake effect of the target wind farm has not been effectively solved. In this case, the preset neural network is iteratively trained based on the measured control parameters and the predicted control parameters until the measured power of the wind farm output by the reference wind farm is equal to the maximum power generation power, and the control parameter prediction model is obtained. At this time, the control parameter prediction model can be used to output the maximum power generation power of the reference wind farm.

[0077] Therefore, by utilizing the spatiotemporal feature extraction network, attention mechanism and fully connected network of the preset neural network, the historical measured impact data and the historical simulation impact data are processed, so as to accurately determine the prediction control parameters when the reference wind farm outputs the measured power of the wind farm, and when the measured power of the wind farm is not equal to the maximum power generation power, the preset neural network is iteratively trained to train a control parameter prediction model that can output the maximum power generation power of the target wind farm.

[0078] In this embodiment, the specific method for determining the maximum power generation power includes but is not limited to the following methods: obtaining the measured wind conditions of the reference wind farm in a future time period after the historical time period, and obtaining the measured control parameters corresponding to the measured wind conditions; using a power simulation model to simulate and calculate the measured control parameters under the measured wind conditions to obtain multiple simulated powers; selecting the simulated power with the largest power value from the multiple simulated powers as the maximum power generation power.

[0079] Among them, the future time period after the historical time period is equal to the future time period after the current time period.

[0080] The power simulation model can process the measured wind conditions and the measured control parameters to calculate multiple simulated powers of the reference wind farm. Furthermore, the multiple simulated powers are sorted and the simulated power with the largest power value is selected to obtain the maximum generated power.

[0081] Therefore, a plurality of simulated powers can be determined by using the power simulation model, and then the power with the largest power value can be selected from the plurality of simulated powers, thereby accurately determining the maximum generated power.

[0082] The embodiment of the present disclosure also provides a wind farm control device for reducing the impact of prediction errors, which is used to implement the above-mentioned wind farm control method for reducing the impact of prediction errors. Figure 3 In the embodiment of the present disclosure, the wind farm control device for reducing the impact of prediction errors may be an electronic device. The electronic device may include a mobile terminal, a tablet computer, or other device with a communication function.

[0083] Figure 3 A schematic structural diagram of a wind farm control device for reducing the impact of prediction errors provided by an embodiment of the present disclosure is shown.

[0084] like Figure 3 As shown, the wind farm control device 300 for reducing the impact of prediction errors may include:

[0085] An acquisition module 310 is used to acquire current measured impact data and current simulated impact data of a target wind farm in a current time period, wherein both the current measured impact data and the current simulated impact data include wind condition data;

[0086] A determination module 320 is used to process the current measured impact data and the current simulation impact data using a pre-trained control parameter prediction model to obtain a target control parameter of the target wind farm in a future time period after the current time period, wherein the target control parameter is an operating condition data of the target wind farm when the target wind farm outputs a maximum power generation power in the future time period;

[0087] The operation control module 330 is used to control the operation of each unit in the target wind farm based on the target control parameter.

[0088] In the disclosed embodiment, the current measured impact data and the current simulated impact data of the target wind farm in the current time period are obtained, wherein the current measured impact data and the current simulated impact data both include wind condition data; the current measured impact data and the current simulated impact data are processed using a pre-trained control parameter prediction model to obtain target control parameters of the target wind farm in a future time period after the current time period, wherein the target control parameters are operating condition data when the target wind farm outputs the maximum power generation in the future time period; based on the target control parameters, the operation of each unit in the target wind farm is controlled. Through the above-mentioned method, the current measured impact data including wind condition data, the current simulated impact data and the pre-trained control parameter prediction model can be used to directly determine the target control parameters for controlling the operation of each unit in the wind farm, thereby avoiding the problem of errors in the control strategy caused by the process of determining the predicted wind conditions based on the actual wind conditions, improving the effect of eliminating the wake effect, and ultimately improving the practical application value of the wind farm wake control technology.

[0089] In some embodiments, the determination module 320 includes:

[0090] an extraction unit, configured to extract current measured spatiotemporal features from the current measured impact data and current simulated spatiotemporal features from the current simulated impact data based on a spatiotemporal feature extraction network in the control parameter prediction model;

[0091] An attention processing unit, configured to perform attention processing on the current measured spatiotemporal features and the current simulated spatiotemporal features respectively based on the attention mechanism in the control parameter prediction model;

[0092] A fitting unit is used to fit the current measured spatiotemporal features after attention processing and the current simulated spatiotemporal features after attention processing based on the fully connected network in the control parameter prediction model, and use the obtained fitting data as the target control parameter.

[0093] In some embodiments, the currently measured impact data includes one or more combinations of the following: currently measured wind condition data, currently measured temperature data, currently measured humidity data, and currently measured air pressure data;

[0094] The current simulation influence data includes one or more of the following combinations: current simulation wind condition data, current wake influence relationship between each unit, current simulation temperature data, current simulation humidity data and current simulation air pressure data.

[0095] In some embodiments, the device further includes: a wake influence relationship determination module; the wake influence relationship determination module includes:

[0096] A first determining unit is used to obtain the power generation influence relationship of each unit in the target wind farm, and use the power generation influence relationship as the wake influence relationship;

[0097] A second determination unit is used to obtain a relative position relationship of each unit in the target wind farm under the current measured wind condition in the current measured impact data, and use the relative position relationship as the wake impact relationship;

[0098] The third determining unit is used to obtain the impeller area of ​​each unit in the target wind farm that is in the wake area of ​​other units, and use the impeller area as the wake influence relationship.

[0099] In some embodiments, the apparatus further comprises:

[0100] A first acquisition module is used to acquire historical measured impact data and historical simulation impact data of a reference wind farm in a historical time period and measured control parameters of the reference wind farm in a future time period after the historical time period;

[0101] A second acquisition module is used to obtain the measured unit power of each unit in the reference wind farm within the historical time period, wherein the sum of the measured unit power of each wind turbine is the measured power of the wind farm;

[0102] An iterative training module is used to iteratively train the preset neural network using the historical measured impact data, the historical simulation impact data, the measured control parameters, the wind farm measured power and the maximum power generation to obtain the control parameter prediction model.

[0103] In some embodiments, the iterative training module is specifically used to:

[0104] Processing the historical measured impact data and the historical simulation impact data using the preset neural network to obtain prediction control parameters when the reference wind farm outputs the wind farm measured power;

[0105] If the measured power of the wind farm is not equal to the maximum power generation power, the preset neural network is iteratively trained based on the measured control parameters and the predicted control parameters until the measured power of the wind farm output by the reference wind farm is equal to the maximum power generation power, thereby obtaining the control parameter prediction model.

[0106] In some embodiments, the device includes: a maximum power generation determination module; the maximum power generation determination module includes:

[0107] A historical data acquisition unit, used to acquire the measured wind conditions of the reference wind farm in a future time period after the historical time period, and to acquire the measured control parameters corresponding to the measured wind conditions;

[0108] A simulation calculation unit, configured to use a power simulation model to perform simulation calculation on the measured control parameters under the measured wind conditions to obtain a plurality of simulated powers;

[0109] The maximum generated power determination unit is used to select a simulated power with the largest power value from the multiple simulated powers as the maximum generated power.

[0110] It should be noted that Figure 3 The wind farm control device 300 for reducing the impact of prediction errors can be implemented as follows: Figure 1 to Figure 2 The various steps in the method embodiment shown in the figure are implemented Figure 1 to Figure 2 The various processes and effects in the method or system embodiments shown are not described in detail here.

[0111] Figure 4 A schematic structural diagram of a wind farm control device for reducing the impact of prediction errors provided by an embodiment of the present disclosure is shown.

[0112] like Figure 4 As shown, the wind farm control device for reducing the impact of prediction errors may include a processor 401 and a memory 402 storing computer program instructions.

[0113] Specifically, the processor 401 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0114] The memory 402 may include a large capacity memory for information or instructions. By way of example and not limitation, the memory 402 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a tape, or a universal serial bus (USB) drive or a combination of two or more of these. Where appropriate, the memory 402 may include a removable or non-removable (or fixed) medium. Where appropriate, the memory 402 may be inside or outside the integrated gateway device. In a particular embodiment, the memory 402 is a non-volatile solid-state memory. In a particular embodiment, the memory 402 includes a read-only memory (ROM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (Electrically Erasable Programmable ROM, EEPROM), an electrically rewritable ROM (EAROM) or a flash memory, or a combination of two or more of these.

[0115] The processor 401 reads and executes the computer program instructions stored in the memory 402 to perform the steps of the wind farm control method for reducing the impact of prediction errors provided in the embodiment of the present disclosure.

[0116] In one example, the wind farm control device for reducing the impact of prediction errors may further include a transceiver 403 and a bus 404. Figure 4 As shown, the processor 401, the memory 402 and the transceiver 403 are connected via a bus 404 and communicate with each other.

[0117] The bus 404 includes hardware, software, or both. For example, but not limitation, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industrial Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a Memory bus, a MicroChannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable buses or a combination of two or more of these. Where appropriate, the bus 404 may include one or more buses. Although embodiments of the present application describe and illustrate a particular bus, the present application contemplates any suitable bus or interconnect.

[0118] The following is an embodiment of a computer-readable storage medium provided in an embodiment of the present disclosure. The computer-readable storage medium and the wind farm control method for reducing the impact of prediction errors in the above-mentioned embodiments belong to the same inventive concept. For details not described in detail in the embodiment of the computer-readable storage medium, reference can be made to the above-mentioned embodiment of the wind farm control method for reducing the impact of prediction errors.

[0119] This embodiment provides a storage medium including computer executable instructions. When the computer executable instructions are executed by a computer processor, they are used to execute a wind farm control method for reducing the influence of prediction errors.

[0120] Of course, the storage medium containing computer executable instructions provided in the embodiment of the present disclosure is not limited to the above method operations, and the computer executable instructions can also execute related operations in the wind farm control method for reducing the impact of prediction errors provided in any embodiment of the present disclosure.

[0121] Through the above description of the implementation methods, the technical personnel in the relevant field can clearly understand that the present disclosure can be implemented by means of software and necessary general hardware, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present disclosure can be essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for a computer cloud platform (which can be a personal computer, a server, or a network cloud platform, etc.) to execute the wind farm control method for reducing the impact of prediction errors provided by various embodiments of the present disclosure.

[0122] Note that the above are only preferred embodiments of the present disclosure and the technical principles used. Those skilled in the art will understand that the present disclosure is not limited to the specific embodiments herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present disclosure. Therefore, although the present disclosure is described in more detail through the above embodiments, the present disclosure is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present disclosure, and the scope of the present disclosure is determined by the scope of the appended claims.

Claims

1. A wind farm control method for reducing the impact of prediction errors, characterized in that: include: Acquire current measured impact data and current simulated impact data of the target wind farm in a current time period, wherein both the current measured impact data and the current simulated impact data include wind condition data; The current measured impact data and the current simulated impact data are processed by using a pre-trained control parameter prediction model to obtain a target control parameter of the target wind farm in a future time period after the current time period, wherein the target control parameter is an operating condition data of the target wind farm when the target wind farm outputs a maximum power generation power in the future time period; Based on the target control parameter, controlling the operation of each unit in the target wind farm; The using of the pre-trained control parameter prediction model to process the current measured impact data and the current simulated impact data to obtain the target control parameter of the target wind farm in a future time period after the current time period includes: Based on the spatiotemporal feature extraction network in the control parameter prediction model, extracting current measured spatiotemporal features from the current measured impact data, and extracting current simulation spatiotemporal features from the current simulation impact data; Based on the attention mechanism in the control parameter prediction model, the current measured spatiotemporal features and the current simulated spatiotemporal features are respectively subjected to attention processing; Based on the fully connected network in the control parameter prediction model, the current measured spatiotemporal features after attention processing and the current simulated spatiotemporal features after attention processing are fitted, and the obtained fitting data is used as the target control parameter.

2. The method according to claim 1, characterized in that The currently measured impact data includes one or more of the following combinations: currently measured wind condition data, currently measured temperature data, currently measured humidity data, and currently measured air pressure data; The current simulation influence data includes one or more of the following combinations: current simulation wind condition data, current wake influence relationship between each unit, current simulation temperature data, current simulation humidity data and current simulation air pressure data.

3. The method according to claim 2, characterized in that The method for determining the wake influence relationship includes: Acquire the power generation influence relationship of each unit in the target wind farm, and use the power generation influence relationship as the wake influence relationship; Obtaining a relative position relationship of each unit in the target wind farm under the current measured wind condition in the current measured impact data, and using the relative position relationship as the wake impact relationship; The impeller area of ​​each unit in the target wind farm in the wake area of ​​other units is obtained, and the impeller area is used as the wake influence relationship.

4. The method according to claim 1, characterized in that: The method further comprises: Acquire historical measured impact data and historical simulation impact data of a reference wind farm in a historical time period and measured control parameters of the reference wind farm in a future time period after the historical time period; Obtaining the measured unit power of each unit in the reference wind farm within the historical time period, wherein the sum of the measured unit powers of each unit is the measured power of the wind farm; The preset neural network is iteratively trained using the historical measured impact data, the historical simulation impact data, the measured control parameters, the wind farm measured power and the maximum generated power to obtain the control parameter prediction model.

5. The method according to claim 4, characterized in that The method of iteratively training the preset neural network by using the historical measured impact data, the historical simulation impact data, the measured control parameters, the wind farm measured power and the maximum power generation to obtain the control parameter prediction model includes: Processing the historical measured impact data and the historical simulation impact data using the preset neural network to obtain prediction control parameters when the reference wind farm outputs the wind farm measured power; If the measured power of the wind farm is not equal to the maximum power generation power, the preset neural network is iteratively trained based on the measured control parameters and the predicted control parameters until the measured power of the wind farm output by the reference wind farm is equal to the maximum power generation power, thereby obtaining the control parameter prediction model.

6. The method according to claim 4, characterized in that The method for determining the maximum power generation includes: Obtaining measured wind conditions of the reference wind farm in a future time period after the historical time period, and obtaining measured control parameters corresponding to the measured wind conditions; Using a power simulation model, simulating and calculating the measured control parameters under the measured wind conditions to obtain a plurality of simulated powers; The simulated power with the largest power value is selected from the multiple simulated powers as the maximum generated power.

7. A wind farm control device for reducing the impact of prediction errors, characterized in that: include: An acquisition module, used to acquire current measured impact data and current simulated impact data of a target wind farm in a current time period, wherein both the current measured impact data and the current simulated impact data include wind condition data; a determination module, configured to process the current measured impact data and the current simulated impact data using a pre-trained control parameter prediction model to obtain a target control parameter of the target wind farm in a future time period after the current time period, wherein the target control parameter is an operating condition data of the target wind farm when the target wind farm outputs a maximum power generation power in the future time period; An operation control module, used for controlling the operation of each unit in the target wind farm based on the target control parameter; The determining module comprises: an extraction unit, configured to extract current measured spatiotemporal features from the current measured impact data and current simulated spatiotemporal features from the current simulated impact data based on a spatiotemporal feature extraction network in the control parameter prediction model; An attention processing unit, configured to perform attention processing on the current measured spatiotemporal features and the current simulated spatiotemporal features respectively based on the attention mechanism in the control parameter prediction model; A fitting unit is used to fit the current measured spatiotemporal features after attention processing and the current simulated spatiotemporal features after attention processing based on the fully connected network in the control parameter prediction model, and use the obtained fitting data as the target control parameter.

8. A device, characterized in that include: processor; A memory for storing executable instructions; The processor is used to read the executable instructions from the memory and execute the executable instructions to implement the method described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the processor implements the method according to any one of claims 1 to 6.

Images