CN115034159A - Power prediction method, device, storage medium and system for offshore wind farm - Google Patents

Power prediction method, device, storage medium and system for offshore wind farm Download PDF

Info

Publication number
CN115034159A
CN115034159A CN202210718925.4A CN202210718925A CN115034159A CN 115034159 A CN115034159 A CN 115034159A CN 202210718925 A CN202210718925 A CN 202210718925A CN 115034159 A CN115034159 A CN 115034159A
Authority
CN
China
Prior art keywords
wind
power
prediction
data
preset
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202210718925.4A
Other languages
Chinese (zh)
Inventor
于珍
杨银国
陆秋瑜
伍双喜
朱誉
向丽玲
杨璧瑜
华威
骆晓明
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangdong Power Grid Co Ltd
Electric Power Dispatch Control Center of Guangdong Power Grid Co Ltd
Original Assignee
Guangdong Power Grid Co Ltd
Electric Power Dispatch Control Center of Guangdong Power Grid Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Guangdong Power Grid Co Ltd, Electric Power Dispatch Control Center of Guangdong Power Grid Co Ltd filed Critical Guangdong Power Grid Co Ltd
Priority to CN202210718925.4A priority Critical patent/CN115034159A/en
Publication of CN115034159A publication Critical patent/CN115034159A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/28—Design optimisation, verification or simulation using fluid dynamics, e.g. using Navier-Stokes equations or computational fluid dynamics [CFD]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Electricity, gas or water supply
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/004—Generation forecast, e.g. methods or systems for forecasting future energy generation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/10—Numerical modelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2113/00—Details relating to the application field
    • G06F2113/08—Fluids
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Abstract

The invention discloses a power prediction method, a power prediction device, a storage medium and a power prediction system for an offshore wind farm. The power prediction device comprises a data acquisition unit, a calculation fitting unit and a power prediction unit. The method and the device for predicting the power, the computer-readable storage medium and the system improve the accuracy of power prediction by carrying out combined modeling through numerical weather prediction and a first flow field simulation technology and integrating wind measurement data of the real-time wind measuring tower and wake flow calculation; furthermore, the method, the device, the computer readable storage medium and the system for predicting the power of the offshore wind farm provided by the invention further improve the accuracy of power prediction by correcting numerical weather forecast data.

Description

Power prediction method, device, storage medium and system for offshore wind farm
Technical Field
The invention relates to the technical field of power prediction of offshore wind farms, in particular to a power prediction method, a power prediction device, a computer readable storage medium and a power prediction system of an offshore wind farm.
Background
Compared with onshore wind power, the prediction difficulty of offshore wind power is greatly increased. Firstly, the land wind power plant is generally wide in distribution area, and the fluctuation of wind power tends to be moderate along with the increase of the scale of the wind power plant; the offshore wind power station with the same installed capacity has a development trend of high concentration, so that power fluctuation can reach a very remarkable level; and besides, meteorological elements such as temperature, air pressure and wind speed on the sea are easy to change suddenly, so that the large fluctuation of wind field output in a short time is aggravated, and climbing events occur, and the events characterized by serious change of wind power generation are directly related to the low predictability of wind power, so that the overall performance of a wind power prediction statistical model is seriously influenced.
In the prior art, the power of an offshore wind farm is generally predicted by acquiring an offshore wind farm environment and physically modeling the offshore wind farm according to environmental parameters and an environmental model.
However, the prior art still has the following defects: the high specific heat capacity of seawater, the offshore wind flow heat effect and the amplified wake effect make the physical modeling operation of an offshore wind farm very complicated; due to the difference of geographical environments, the flexible generalization of physical modeling has certain defects and is not suitable for the short-term power prediction in the sea; on the basis, different from onshore wind power, the development of NWP (Numerical Weather Prediction) information for offshore wind power plants in China is late, the accuracy is not high, the Prediction error is increased, and therefore inaccurate Numerical Weather Prediction information is also one of main sources of the Prediction error.
Accordingly, there is a need for a method, apparatus, computer readable storage medium, and system for power prediction for offshore wind farms that overcome the above-mentioned deficiencies in the prior art.
Disclosure of Invention
The invention provides a power prediction method, a power prediction device, a computer readable storage medium and a power prediction system for an offshore wind farm, which are used for improving the accuracy of power prediction of the offshore wind farm.
An embodiment of the invention provides a power prediction method for an offshore wind farm, which comprises the following steps:
acquiring an environment data group, generated energy data, real-time anemometer tower data and numerical weather forecast data of an offshore wind farm, and establishing and calculating a first flow field model result of the offshore wind farm according to the environment data group, a preset actuating line wind generator set model and a preset large vortex turbulence model;
calculating a real-time wind resource simulation observation data set and a fan point location wind speed set according to a preset wake condition set, the real-time anemometer tower data and the first flow field model result, and fitting a first statistical relationship between the environment data set and the generated energy data according to the numerical weather forecast data;
and according to the real-time wind resource simulation observation data group, the fan point position wind speed group and the first statistical relationship, performing power prediction on the offshore wind farm to obtain a prediction result.
As an improvement of the above scheme, establishing and calculating a first flow field model result of the offshore wind farm according to the environment data set, a preset actuation linear wind turbine set model and a preset large vortex turbulence model, specifically including:
establishing a first flow field model of the offshore wind farm according to a preset actuating line wind generating set model and a preset large vortex turbulence model;
and calculating a first flow field model result of the offshore wind farm according to the environment data group and the first flow field model.
As an improvement of the above scheme, calculating a real-time wind resource simulation observation data set and a wind speed set of a wind turbine site location according to a preset wake condition set, the real-time wind measuring tower data and the first flow field model result, specifically including:
calculating a wind resource simulation observation data set of each height of each point of a fan in the offshore wind farm according to the first flow field model result and the real-time wind measuring tower data;
and calculating the calculated wind speed of each point of the fan under the wake condition group according to the wind resource simulation observation data group and a preset wake condition group, and storing the wind speed into a fan point position wind speed group.
As an improvement of the above scheme, performing power prediction on the offshore wind farm according to the real-time wind resource simulation observation data set, the wind speed at the wind turbine point location set, and the first statistical relationship to obtain a prediction result specifically includes:
according to the real-time wind resource simulation observation data set, the fan point position wind speed set and the first statistical relationship, based on a CFD power downscaling scale, performing short-term wind power prediction on the offshore wind farm to obtain a short-term prediction result;
according to the real-time wind resource simulation observation data set, the fan point location wind speed set and the first statistical relationship, wind power set prediction is carried out on the offshore wind farm according to multiple prediction methods to obtain a set prediction result;
and outputting the short-term prediction result and the set prediction result as prediction results.
As an improvement of the above scheme, according to the real-time wind resource simulation observation data set, the fan point location wind speed set, and the first statistical relationship, based on a CFD power downscaling scale, performing short-term wind power prediction on the offshore wind farm to obtain a short-term prediction result, specifically including:
correcting the numerical weather forecast data by a preset mesoscale wind speed forecast correcting method according to the real-time wind resource simulation observation data group and the fan point position wind speed group;
selecting corresponding wind power plant micro-scale wind field distribution from a preset micro-scale wind field basic database according to the corrected numerical weather forecast data, and analyzing and calculating the free incoming flow wind speed of each fan;
calculating to obtain the simulated generating power of each fan according to the first statistical relationship, a preset wind turbine generator simulated generating power correction method and the free incoming flow wind speed;
and according to a preset operation maintenance plan of the wind power plant, removing the unit which does not operate and the time period during which the unit does not operate to obtain the whole-plant generated power prediction of the wind power plant.
As an improvement of the above scheme, calculating a calculated wind speed of each point of the wind turbine output under the wake condition group according to the wind resource simulation observation data group and a preset wake condition group, specifically includes:
calculating the calculated wind speed of each point of the fan according to a preset Larsen wake flow calculation model and a wind speed data set; the wind resource simulation observation data set comprises a wind speed data set.
As an improvement of the above, the plurality of prediction methods include: a neural network prediction method based on a tabu algorithm, a power time sequence prediction method, a particle swarm algorithm and a Kalman filtering correction power prediction method.
The invention correspondingly provides a power prediction device of an offshore wind farm, which comprises a data acquisition unit, a calculation fitting unit and a power prediction unit, wherein,
the data acquisition unit is used for acquiring an environment data set, generated energy data, real-time anemometer tower data and numerical weather forecast data of the offshore wind farm, and establishing and calculating a first flow field model result of the offshore wind farm according to the environment data set, a preset actuating line wind power set model and a preset large vortex turbulence model;
the calculation fitting unit is used for calculating a real-time wind resource simulation observation data set and a fan point location wind speed set according to a preset wake flow condition set, the real-time anemometer tower data and the first flow field model result, and fitting a first statistical relationship between the environment data set and the generated energy data according to the numerical weather forecast data;
the power prediction unit is used for performing power prediction on the offshore wind farm according to the real-time wind resource simulation observation data set, the fan point location wind speed set and the first statistical relationship to obtain a prediction result.
As an improvement of the above, the data acquisition unit is further configured to:
establishing a first flow field model of the offshore wind farm according to a preset actuating line wind generating set model and a preset large vortex turbulence model;
and calculating a first flow field model result of the offshore wind farm according to the environment data group and the first flow field model.
As an improvement of the above, the calculation fitting unit is further configured to:
calculating a wind resource simulation observation data set of each height of each point of a fan in the offshore wind farm according to the first flow field model result and the real-time wind measuring tower data;
and calculating the calculated wind speed of each point of the fan under the wake condition group according to the wind resource simulation observation data group and a preset wake condition group, and storing the wind speed into a fan point position wind speed group.
As an improvement of the above, the power prediction unit is further configured to:
according to the real-time wind resource simulation observation data set, the fan point position wind speed set and the first statistical relationship, based on a CFD power downscaling scale, performing short-term wind power prediction on the offshore wind farm to obtain a short-term prediction result;
according to the real-time wind resource simulation observation data set, the fan point position wind speed set and the first statistical relationship, wind power aggregate prediction is carried out on the offshore wind farm according to multiple prediction methods to obtain an aggregate prediction result;
and outputting the short-term prediction result and the set prediction result as prediction results.
As an improvement of the above, the power prediction unit is further configured to:
correcting the numerical weather forecast data by a preset mesoscale wind speed forecast correcting method according to the real-time wind resource simulation observation data group and the fan point position wind speed group;
selecting corresponding micro-scale wind field distribution of the wind power plant from a preset micro-scale wind field basic database according to the corrected numerical weather forecast data, and analyzing and calculating the free incoming flow wind speed of each fan;
calculating to obtain the simulated generating power of each fan according to the first statistical relationship, a preset wind turbine generator simulated generating power correction method and the free incoming flow wind speed;
and removing the unit which does not operate and the time period during which the unit does not operate according to a preset wind power plant operation maintenance plan to obtain the whole wind power plant generated power prediction.
As an improvement of the above solution, the calculation fitting unit is further configured to:
calculating the calculated wind speed of each point of the fan according to a preset Larsen wake flow calculation model and a wind speed data set; the wind resource simulation observation data group comprises a wind speed data group.
Another embodiment of the present invention provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program runs, the apparatus on which the computer-readable storage medium is located is controlled to execute the power prediction method for offshore wind farm as described above.
Another embodiment of the present invention provides a power prediction system for an offshore wind farm, the power prediction system comprising a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, the processor implementing the power prediction method for an offshore wind farm as described above when executing the computer program.
Compared with the prior art, the technical scheme has the following beneficial effects:
the invention provides a power prediction method, a power prediction device, a computer readable storage medium and a power prediction system for an offshore wind farm, wherein numerical weather forecast and a first flow field simulation technology are used for carrying out combined modeling, real-time anemometer tower anemometry data and wake flow calculation are integrated, and the power prediction method, the power prediction device, the computer readable storage medium and the power prediction system improve the accuracy of power prediction.
Furthermore, the method, the device, the computer readable storage medium and the system for predicting the power of the offshore wind farm provided by the invention further improve the accuracy of power prediction by correcting numerical weather forecast data.
Drawings
FIG. 1 is a schematic flow chart of a power prediction method for an offshore wind farm according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of a power prediction apparatus of an offshore wind farm according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Detailed description of the preferred embodiment
The embodiment of the invention firstly describes a power prediction method of an offshore wind farm. Fig. 1 is a schematic flowchart of a power prediction method for an offshore wind farm according to an embodiment of the present invention.
As shown in fig. 1, the power prediction method includes:
and S1, acquiring an environment data group, generating capacity data, real-time anemometer tower data and numerical weather forecast data of the offshore wind farm, and establishing and calculating a first flow field model result of the offshore wind farm according to the environment data group, a preset actuating line wind generating set model and a preset large vortex turbulence model.
Each wind farm is relatively unique in that the geographic location, terrain and meteorological conditions are all different for each wind farm. Accurate wind farm power forecasting relies on accurate mesoscale data as well as accurate simulation of wind farm topography, roughness, wake and other characteristics. To achieve this accurate simulation, meteorological parameters such as wind speed are transmitted to a wind tower site based on the results of a probabilistic statistical model, an Artificial Neural Network (ANN) and a numerical weather forecast (NWP), and these results are spatially migrated to each wind turbine site based on a CFD (Computational Fluid Dynamics) model and a wake model.
Since the terrain and roughness information has a considerable influence on the accuracy of the result, high-quality terrain elevation and roughness data are very necessary for accurately estimating the wind speed of the wind turbine point. In the embodiment of the invention, the CFD model of the wind power plant under specific terrain, roughness and meteorological conditions is generated firstly, so that the model is maximally close to the objective condition, and the prediction accuracy is improved. The CFD model adopts a computational fluid mechanics simulation method, replaces a traditional actual blade geometric complex model with an optimized actuating line wind turbine generator model, and combines a large vortex turbulence model, thereby bringing the following technical advantages: (1) the adopted double-pass turbulence model has higher precision and more accurate simulation of complex terrain; (2) an isolated solver, a coupling solver and a parallel solver are provided, the functions are more complete, and the solving speed and the convergence are ensured; (3) the optimal wind turbine arrangement can be automatically obtained on the premise of considering IEC (International electrotechnical Commission) wind turbine specification.
The wind power plant CFD model is mainly used for simulating a CFD flow field, the outlet boundary of a calculation domain is set as a pressure boundary condition, the boundary condition of the lower surface is set as a non-slip wall surface boundary condition, and the wind direction and the wind speed at a given height are used as inlet boundary conditions and are set as a wind speed contour line. The method comprises the steps of determining a wind power plant model by solving a Navier-Stokes equation of Reynolds average, selecting an RNGk-e turbulence model for flow field simulation according to engineering practice and applicability of different turbulence models for smooth calculation of a relatively complex offshore wind power plant by adopting a turbulence model of a dual equation in a turbulence closing scheme, solving the model by iterative calculation due to the fact that the flow equation is nonlinear, and gradually iterating and leading a final convergence result from an assumed initial condition.
In this step, the establishing of the terrain model and the wind field CFD calculation specifically includes: after acquiring the fan information and the layout information of the wind farm, transmitting a time sequence file from a reference wind measuring point location to each fan point location, calculating the mesoscale data to the real-time wind measuring tower point location in the wind farm through a statistical model from the mesoscale data, and then calculating the mesoscale data to each fan point location from the real-time wind measuring tower point location; wherein the layout information of fan information and wind farm includes: the coordinates of the fan, the height of the hub and the point position of a real-time anemometer tower in the field are determined.
If there is a barrier around the offshore wind farm, it is contemplated to add a barrier model. The wind field module usually adopts 16 sectors, and for a specific wind power plant, 36 sectors can be adopted to improve the accuracy and ensure that accurate simulation results can be obtained in all directions.
In one embodiment, the establishing and calculating a first flow field model result of the offshore wind farm according to the environmental data set, a preset actuation line wind power set model and a preset large vortex turbulence model specifically includes: establishing a first flow field model of the offshore wind farm according to a preset actuating line wind generating set model and a preset large vortex turbulence model; and calculating a first flow field model result of the offshore wind farm according to the environment data group and the first flow field model.
And S2, calculating a real-time wind resource simulation observation data set and a fan point location wind speed set according to a preset wake flow condition set, the real-time anemometer tower data and the first flow field model result, and fitting a first statistical relationship between the environment data set and the generated energy data according to the numerical weather forecast data.
The statistical method avoids a middle complex physical modeling step, uses information such as high-precision numerical simulation technology NWP forecast and the like as input data, and searches a potential statistical relationship between meteorological information and the power generation amount of the wind power plant through a large amount of analysis. Therefore, a statistical method NWP and a CFD method are jointly used for modeling. In addition, the NWP data is corrected and modeled in consideration of the problems of the offshore wind power multi-climbing event and the NWP inaccurate prediction.
In one embodiment, calculating a real-time wind resource simulation observation data set and a wind speed set of a wind turbine site location according to a preset wake condition set, the real-time anemometer tower data and the first flow field model result specifically includes: calculating a wind resource simulation observation data set of each height of each point of a fan in the offshore wind farm according to the first flow field model result and the real-time wind measuring tower data; and calculating the calculated wind speed of each point of the fan under the wake condition group according to the wind resource simulation observation data group and a preset wake condition group, and storing the wind speed into a fan point position wind speed group.
In an embodiment, calculating a calculated wind speed of each point of the wind turbine output under the wake condition group according to the wind resource simulation observation data group and a preset wake condition group specifically includes: calculating the calculated wind speed of each point of the fan according to a preset Larsen wake flow calculation model and a wind speed data set; the wind resource simulation observation data group comprises a wind speed data group.
And S3, performing power prediction on the offshore wind farm according to the real-time wind resource simulation observation data set, the fan point location wind speed set and the first statistical relationship to obtain a prediction result.
Considering that wind field distribution of a wind field area is a result of movement of a weather system driven by local terrain, the movement of the weather system can be forecasted by adopting a mesoscale numerical mode, and because the local terrain is stable, wind field distribution under various weather background conditions can be simulated in advance by adopting a CFD (computational fluid dynamics) technology, a microscale wind field basic database is established, and then wind power forecast of each wind turbine is given out by selecting wind field distribution of a corresponding wind power plant directly according to the mesoscale numerical weather forecast.
Meanwhile, because the data used by each prediction model and method are not completely the same, the combined prediction method can utilize the useful information of various single prediction methods to the maximum extent, and the prediction accuracy of the system is increased to a certain extent. Specifically, different prediction models and methods are combined, so that diversified model structures are obtained in a proper weighted average mode according to information provided by multiple prediction methods, and the prediction accuracy can be further improved by forming a combined prediction model. The method generally uses absolute error as a criterion to calculate the weight coefficients of the combined prediction method.
In one embodiment, the power prediction of the offshore wind farm according to the real-time wind resource simulation observation data set, the wind speed group of the wind turbine site location, and the first statistical relationship to obtain a prediction result specifically includes: according to the real-time wind resource simulation observation data set, the fan point position wind speed set and the first statistical relationship, based on a CFD power downscaling scale, performing short-term wind power prediction on the offshore wind farm to obtain a short-term prediction result; according to the real-time wind resource simulation observation data set, the fan point position wind speed set and the first statistical relationship, wind power aggregate prediction is carried out on the offshore wind farm according to multiple prediction methods to obtain an aggregate prediction result; and outputting the short-term prediction result and the set prediction result as prediction results.
In one embodiment, according to the real-time wind resource simulation observation data set, the wind speed set at the wind turbine point location, and the first statistical relationship, based on a CFD power down scale, performing short-term wind power prediction on the offshore wind farm to obtain a short-term prediction result, specifically including: correcting the numerical weather forecast data by a preset mesoscale wind speed forecast correcting method according to the real-time wind resource simulation observation data group and the fan point position wind speed group; selecting corresponding wind power plant micro-scale wind field distribution from a preset micro-scale wind field basic database according to the corrected numerical weather forecast data, and analyzing and calculating the free incoming flow wind speed of each fan; calculating to obtain the simulated generating power of each fan according to the first statistical relationship, a preset wind turbine generator simulated generating power correction method and the free incoming flow wind speed; and according to a preset operation maintenance plan of the wind power plant, removing the unit which does not operate and the time period during which the unit does not operate to obtain the whole-plant generated power prediction of the wind power plant.
In one embodiment, the plurality of prediction methods includes: a neural network prediction method based on a tabu algorithm, a power time sequence prediction method, a particle swarm algorithm and a Kalman filtering correction power prediction method.
The neural network prediction method based on the tabu algorithm comprises the following steps:
tabu Search (TS) is an extension of local neighborhood Search, a global stepwise optimization algorithm, and is a simulation of human intelligence process. The TS algorithm avoids circuitous searching by introducing a flexible storage structure and corresponding taboo criteria, and privileged some good states forbidden through scofflaw criteria, thereby ensuring diversified effective exploration to finally realize global optimization. The basic idea is as follows: assuming that a solution neighborhood is given, firstly, an initial local solution x is found out in the neighborhood as a current solution, the current solution is made to be an optimal solution, then the current solution x is used as a starting point, the optimal solution x ' is searched in the solution neighborhood, in order to avoid circuitous search caused by the fact that x ' is the same as x, a tabu table for memorizing recent operation is arranged, if the current search operation is the operation recorded in the tabu table, the search operation is forbidden, and if not, x ' is used for replacing x as the current solution.
The BP neural network is a multi-layer feedforward network trained according to an error back propagation algorithm and comprises two processes: the forward propagation process of information and the backward propagation process of errors. And the information forward propagation process is that input information is input into the network from the input layer and is transmitted to the intermediate layer, the intermediate layer performs conversion processing on the information and then transmits the processed result to the output layer, and the actual output information of the network is obtained at the output layer. And the error back propagation process is that the error passes through the output layer in a gradient decreasing mode, the weight of each layer is corrected, and the error is back-propagated one by one according to the sequence of the middle layer and the input layer. Both processes are repeated until the error between the network output and the desired output is reduced to an acceptable level. The principle used in error back propagation is the "negative gradient descent" theory.
The neural network has strong nonlinear fitting capability, can map any complex nonlinear relation, has simple learning rule, is convenient for computer realization, has strong robustness, memory capability, nonlinear mapping capability and strong self-learning capability, but is easy to fall into 'over-learning' in the learning process. The basic principle of the application of the artificial neural network based on the tabu search algorithm is that the tabu search algorithm with a memory function is used for optimally learning the weight of the neural network, and the global optimal solution is obtained by using the global search capability of the tabu search algorithm, so that the training is prevented from falling into local minimum.
For further description, the process of optimizing the neural network by the tabu search algorithm will be described by way of example:
suppose thatThe error function of a BP neural network is f ═ f (W) h ,W o ,θ h ,θ o ) Wherein W is h 、W o 、θ h 、θ o Respectively representing the connection weight of the input layer and the hidden layer, the connection weight of the hidden layer and the output layer, the threshold value of the hidden layer and the threshold value of the output layer. The optimization of the network is minf (W) h ,W o ,θ h ,θ o ) The process of (2). For convenience of description, Δ is used herein to denote (W) h ,W o ,θ h ,θ o )。
First, Δ is initialized (specifically, each component of Δ is assigned a smaller random number, denoted as Δ initial ) (ii) a Then, with Δ best Represents the best solution, Δ, of the search guide so far now Represents the current solution, let Δ best ＝Δ initial ，Δ now ＝Δ initial And will be now Storing in a tabu table; then, Δ is generated initial A neighborhood solution of new Calculating f (Δ) new ) And f (Δ) best ) (ii) a If f (. DELTA. new ) If the continuous times (the preset times) are not changed, stopping the algorithm and outputting the result, otherwise, continuing the following steps; if f (Δ) new )<f(Δ best ) Then a is best ＝Δ new ，Δ now ＝Δ new ，Δ new Entering a tabu table, and moving the sequence of the points stored in the table backwards, namely updating the tabu table; if f (Δ) new )≥f(Δ best ) Determine Δ new Whether or not it is contraindicated, if Δ new In a given neighborhood of a certain memory point in the tabu table, Δ is represented new If forbidden, a neighborhood solution vector delta is regenerated new (ii) a If Δ new Not being contraindicated, then now ＝Δ new Meanwhile, updating a tabu table; finally, Δ is generated now A neighborhood solution of new Turning to the fourth step; when the training is finished, obtaining a trained weight vector W h 、W o And a threshold vector θ h 、θ o 。
Time series analysis is an effective parameterized time domain analysis method for processing dynamic data. A differential Autoregressive Moving Average model (ARIMA- -Autoregressive Integrated Moving Average) is used for prediction and control by analyzing time series data. The invention adopts a power time series prediction technology, establishes a time series mathematical model according to historical record data of power, describes the statistical regularity of the random variable variation process of the power by using the mathematical model on one hand, and establishes a mathematical expression of power prediction on the basis of the mathematical model on the other hand to predict the future power.
ARIMA consists of three parts: auto-regressive term autoregressive (ar), differential term integrated (i), and moving average term Moving Average (MA). ARIMA is based on ARMA, whose mathematical expression is:
Figure BDA0003710341020000121
wherein, X t Represents a time series, i.e. a power series;
Figure BDA0003710341020000122
representing the autoregressive term AR, a j Is a constant number, X t-j Is the observed value at time t-j,
Figure BDA0003710341020000123
the linear combination of the past observed values is obtained; b is a mixture of k Is a constant value e t Is a white noise sequence and is a white noise sequence,
Figure BDA0003710341020000124
the moving average term MA representing a white noise sequence. Therefore, the AR model describes the memory of the system to the past self-state, and the MA model describes the memory of the system to the past self-state and the noise entering the system. The value of a time series at a certain time can be represented by a linear combination of p historical observed values and q-term moving average of a white noise series, and the time series is the ARMA (p, q) process.
In power prediction, falseLet the sample function of power X (t) be x (t), and the sampling value (history of power) be x 1 ,x 2 ,…,x t …, in a specific implementation, assume that a finite sequence of a finite number of sample values is x 1 ,x 2 ,…,x N . Time series prediction is based on the finite sequence to infer the nature of the original sequence that generated the finite sequence. It is difficult to exactly find the original sequence, but instead of the original sequence, a prediction model can be found that substantially conforms to the limited sequence. This process requires pattern recognition and parameter estimation. The model identification is to determine which of the AR, MA, and ARMA the required prediction model belongs to or does not belong to. Parameter estimation, i.e. after model identification, the unknown parameters in the model are calculated according to a suitable method. Model identification and parameter estimation are performed based on a finite-sequence equation to infer the nature of the original sequence equation, which cannot be completely accurate, and therefore a check is required to determine whether the stochastic model is appropriate.
The discrimination basis of the model identification is the autocorrelation function and the partial correlation function of the sequence. After model identification, the class, structure and order of the model are preliminarily determined, and then unknown parameters in the model need to be estimated. The methods for parameter estimation are many, and the most applied methods are mainly moment estimation and least square estimation. After model identification and parameter estimation, the next problem is to determine whether the model is proper, and if the model meets the requirements after inspection, prediction can be carried out; conversely, if there is evidence of a severe misfit, the model may need to be modified or re-identified until the requirements are met. If the random model preliminarily determined by the model identification and the parameter estimation fails to pass the test, that is, if the model is determined to be inappropriate by the test, the following processing can be performed: and evaluating the statistical significance of the parameter estimation value by using the standard deviation of the parameter estimation. Generally, it is specified that the term of the parameter as a coefficient should be deleted from the model if the absolute value of the parameter point estimate is less than two times the standard deviation, which is the statistical significance difference or its significance. Meanwhile, if an autoregressive term or a moving average term needs to be added into the model, the model is added into the preliminarily determined model, but the added new model needs to be tested to have higher applicability than the originally preliminarily determined model, otherwise, the addition of the parameter term is meaningless; according to the information provided by residual analysis residual, the model is modified properly; if the preliminarily determined model is not modified, the model can be identified again; when the preliminarily determined model is considered to be inappropriate through model inspection, after the model is modified or re-identified to obtain a new model, the model inspection is also required to determine the reasonableness of the model.
For further description, a particle swarm algorithm will be described by way of example.
The particle swarm optimization is an effective global optimization algorithm, starts to simulate the foraging process of a bird swarm, is an algorithm optimized by utilizing swarm intelligence, and realizes swarm intelligent optimization search through cooperation and competition among individuals in a swarm. In theory, each optimization problem is considered to be a flock of birds which forage in the air, the so-called "particles" are each foraging "bird" flying in the air, i.e. a solution to the optimization problem, and the optimal solution is the "food" which the flock of birds will ultimately find.
In the particle swarm algorithm, the algorithm first generates an initial solution, i.e., randomly initializes N particle composition groups z ═ { z ] in a D-dimensional feasible solution space 1 ,z 2 ,…,z N Two vectors, position and velocity, i.e. z, for each particle i ＝{z i1 ,z i2 ,…,z iD And v i ＝{v i1 ,v i2 ,…,v iD And then calculating the fitness value of the particles according to the objective function, and performing iterative search in an S-dimensional solution space. In this process, the particles search through the individuals themselves for the optimal solution, i.e., "individual extremum" P id And the best solution, i.e. "extremum" P, to which the population is seeking gd To update its position and speed. Each particle updates its position and velocity according to:
v id (t+1)＝ωv id (t)+c 1 r 1 [p id -z id (t)]+c 2 r 2 [p gd -z id (t)]；
z id (t+1)＝z id (t)+v id (t+1)；
in the formula, v id (t +1) denotes the velocity of the ith particle in the d-dimension in t +1 iterations, ω is the inertial weight, c 1 ,c 2 Is an acceleration constant, r 1 ，r 2 Is a random number between 0 and 1.
The inertial weight ω represents the maintenance of the current velocity, and ω adjusts the global and local search capabilities of the particle. The larger inertia weight enables the particles to have stronger global search capability, and the smaller inertia weight enables the particles to have stronger local search capability. Learning factor c 1 ，c 2 The method respectively represents the self-learning ability and the social learning ability of the particles, namely the maximum step length of the particles which are adjusted to fly to the individual optimum and the group optimum is determined, the influence of the group experience and the individual experience of the particles on the particles is determined, and the information exchange between the individual and between the individual and the group is reflected.
For further description, the power prediction method modified by kalman filtering will be described by way of example.
When the statistical method is applied to wind power plant power prediction, the NWP is the data basis for prediction, so the accuracy of data provided by the NWP directly determines the reliability of a final power prediction result. Generally, the capacity of NWP processing secondary grid phenomenon is not enough to make up for the defect of physical parameter initialization, so that a certain degree of system error exists in meteorological mode output, and the final prediction precision is inevitably influenced by the error introduced into a wind power metering model.
Various statistical methods evolved from MOS can be used to eliminate such systematic errors. However, the establishment of MOS equations requires a lot of history, which is difficult to accumulate, and the parameters of dynamically updated equations are not accessible to MOS. The Kalman filtering algorithms are different, a large amount of historical data is not needed, and the prediction equation coefficients can be dynamically corrected only through error feedback, so that the method has important practical significance for improving the accuracy of meteorological mode output.
The embodiment of the invention also adopts a wind power plant power prediction model modified by a Kalman filtering algorithm. Historical meteorological data provided by NWP (non-meteorological data) is modified by a Kalman filtering algorithm to form a training set of a BP (Back propagation) neural network, power sequences of all fans acquired from a wind turbine set and a supervisory control System (SCADA) are used as a target set of the BP neural network, and a nonlinear mapping relation between the meteorological data and power output is obtained after full training, namely a BP network prediction model. Similarly, the future meteorological data corrected by the Kalman filtering algorithm is trained by the neural network to obtain the final predicted power output.
When the wind speed time series output by the NWP is corrected by the Kalman filter, the prediction error of the wind speed is taken as a function of the NWP wind speed output data, and the error is estimated. Suppose v t Is the output wind speed of the NWP model at time t, y t Is the prediction error at time t, y t Can use a about v t Is described by a third order polynomial of:
y t ＝x 0,t +x 1,t v t +x 2,t v t 2 +x 3,t v t 3 +q t ；
in the formula, x i,t (i ═ 0,1,2,3) are coefficients estimated using a kalman filter; q. q.s t For the gaussian nonlinear system error generated in the previous step, the coefficient matrix to be estimated is used as the state vector, i.e. x t ＝[x 0,t x 1,t x 2,t x 3,t ] T By y t As observation vector, observation matrix H t ＝[1 v t v t 2 v t 3 ]Based on the unit matrix taken as the system matrix, the system equation and the measurement equation are as follows:
x t ＝x t-1 +w t ；
y t ＝H t x t +q t 。
in one embodiment, the power prediction method further comprises: and optimizing the CFD model of the wind power plant in a preset time period. Specifically, after the wind farm CFD model of the reference wind farm in the area is established, the wind farm CFD model needs to be operated for a period of time, and then the wind farm CFD model is optimized according to comprehensive observation data of the wind farm in a certain period of time.
The embodiment of the invention describes a power prediction method of an offshore wind farm, which carries out combined modeling through numerical weather forecast and a first flow field simulation technology, integrates wind measurement data of a real-time wind measuring tower and wake flow calculation, and improves the accuracy of power prediction by the power prediction method, the power prediction device, a computer readable storage medium and a system; furthermore, the power prediction method for the offshore wind farm described in the embodiment of the invention further improves the accuracy of power prediction by correcting numerical weather forecast data.
Detailed description of the invention
Besides, the embodiment of the invention also discloses a power prediction device of the offshore wind farm. Fig. 2 is a schematic structural diagram of a power prediction apparatus of an offshore wind farm according to an embodiment of the present invention.
As shown in fig. 2, the power prediction apparatus includes a data acquisition unit 11, a calculation fitting unit 12, and a power prediction unit 13.
The data acquisition unit 11 is used for acquiring an environment data set, generated energy data, real-time anemometer tower data and numerical weather forecast data of the offshore wind farm, and establishing and calculating a first flow field model result of the offshore wind farm according to the environment data set, a preset actuation linear wind turbine set model and a preset large vortex turbulence model.
The calculation fitting unit 12 is configured to calculate a real-time wind resource simulation observation data set and a wind speed set of a wind turbine site location according to a preset wake condition set, the real-time anemometer tower data, and the first flow field model result, and fit a first statistical relationship between the environment data set and the generated energy data according to the numerical weather forecast data.
The power prediction unit 13 is configured to perform power prediction on the offshore wind farm according to the real-time wind resource simulation observation data set, the wind speed at the wind turbine point location set, and the first statistical relationship, so as to obtain a prediction result.
In one embodiment, the data acquisition unit 11 is further configured to: establishing a first flow field model of the offshore wind farm according to a preset actuating line wind generating set model and a preset large vortex turbulence model; and calculating a first flow field model result of the offshore wind farm according to the environment data group and the first flow field model.
In one embodiment, the calculation fitting unit 12 is further configured to: calculating a wind resource simulation observation data set of each height of each point of a fan in the offshore wind farm according to the first flow field model result and the real-time wind measuring tower data; and calculating the calculated wind speed of each point of the fan under the wake condition group according to the wind resource simulation observation data group and a preset wake condition group, and storing the wind speed into a fan point position wind speed group.
In one embodiment, the power prediction unit 13 is further configured to: according to the real-time wind resource simulation observation data set, the fan point position wind speed set and the first statistical relationship, based on a CFD power downscaling scale, performing short-term wind power prediction on the offshore wind farm to obtain a short-term prediction result; according to the real-time wind resource simulation observation data set, the fan point position wind speed set and the first statistical relationship, wind power aggregate prediction is carried out on the offshore wind farm according to multiple prediction methods to obtain an aggregate prediction result; and outputting the short-term prediction result and the set prediction result as prediction results.
In one embodiment, the power prediction unit 13 is further configured to: correcting the numerical weather forecast data by a preset mesoscale wind speed forecast correcting method according to the real-time wind resource simulation observation data group and the fan point location wind speed group; selecting corresponding micro-scale wind field distribution of the wind power plant from a preset micro-scale wind field basic database according to the corrected numerical weather forecast data, and analyzing and calculating the free incoming flow wind speed of each fan; calculating to obtain the simulated generating power of each fan according to the first statistical relationship, a preset wind turbine generator simulated generating power correction method and the free incoming flow wind speed; and removing the unit which does not operate and the time period during which the unit does not operate according to a preset wind power plant operation maintenance plan to obtain the whole wind power plant generated power prediction.
In one embodiment, the calculation fitting unit 12 is further configured to: calculating the calculated wind speed of each point of the fan according to a preset Larsen wake flow calculation model and a wind speed data set; the wind resource simulation observation data set comprises a wind speed data set.
Wherein, the unit integrated by the power prediction device can be stored in a computer readable storage medium if the unit is realized in the form of software functional unit and sold or used as an independent product. The computer readable storage medium comprises a stored computer program, wherein the apparatus on which the computer readable storage medium is based is controlled when the computer program is run to perform the method for power prediction of an offshore wind farm as described above.
Based on such understanding, all or part of the flow in the method according to the above embodiments may be implemented by a computer program, which may be stored in a computer readable storage medium and used by a processor to implement the steps of the above embodiments of the method. Wherein the computer program comprises computer program code, which may be in the form of source code, object code, an executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, usb disk, removable hard disk, magnetic disk, optical disk, computer Memory, Read-Only Memory (ROM), Random Access Memory (RAM), electrical carrier wave signals, telecommunications signals, software distribution medium, and the like. It should be noted that the computer-readable medium may contain suitable additions or subtractions depending on the requirements of legislation and patent practice in jurisdictions, for example, in some jurisdictions, computer-readable media may not include electrical carrier signals or telecommunication signals in accordance with legislation and patent practice.
It should be noted that the above-described device embodiments are merely illustrative, where the units described as separate parts may or may not be physically separate, and the parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. In addition, in the drawings of the embodiment of the device provided by the invention, the connection relationship between the units indicates that the units have communication connection therebetween, and the connection relationship can be specifically implemented as one or more communication buses or signal lines. One of ordinary skill in the art can understand and implement it without inventive effort.
The embodiment of the invention describes a power prediction device and a computer readable storage medium of an offshore wind farm, which are used for carrying out combined modeling through numerical weather forecast and a first flow field simulation technology, and integrating real-time anemometer tower anemometry data and wake flow calculation, and the power prediction device and the computer readable storage medium improve the accuracy of power prediction; furthermore, the power prediction device and the computer-readable storage medium for the offshore wind farm described in the embodiments of the present invention further improve the accuracy of power prediction by correcting the numerical weather forecast data.
Detailed description of the preferred embodiment
Besides the method and the device, the embodiment of the invention also describes a power prediction system of the offshore wind farm.
The power prediction system comprises a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, the processor implementing the power prediction method of the offshore wind farm as described above when executing the computer program.
The Processor may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic, discrete hardware components, etc. The general purpose processor may be a microprocessor or the processor may be any conventional processor or the like that is the control center for the device and that connects the various parts of the overall device using various interfaces and lines.
The memory may be used to store the computer programs and/or modules, and the processor may implement the various functions of the apparatus by running or executing the computer programs and/or modules stored in the memory, as well as by invoking data stored in the memory. The memory may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required by at least one function (such as a sound playing function, an image playing function, etc.), and the like; the storage data area may store data (such as audio data, a phonebook, etc.) created according to the use of the cellular phone, and the like. In addition, the memory may include high speed random access memory, and may also include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), at least one magnetic disk storage device, a Flash memory device, or other volatile solid state storage device.
The embodiment of the invention describes a power prediction system of an offshore wind farm, which carries out combined modeling through numerical weather forecast and a first flow field simulation technology, integrates wind measurement data of a real-time wind measuring tower and wake flow calculation, and improves the accuracy of power prediction by the power prediction method, the power prediction device, a computer readable storage medium and the power prediction system; furthermore, the power prediction system of the offshore wind farm described in the embodiment of the invention further corrects the numerical weather forecast data, so that the accuracy of power prediction is further improved.
While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention.

Claims (10)

1. A power prediction method for an offshore wind farm, characterized in that the power prediction method comprises:
acquiring an environment data group, generated energy data, real-time anemometer tower data and numerical weather forecast data of an offshore wind farm, and establishing and calculating a first flow field model result of the offshore wind farm according to the environment data group, a preset actuating line wind generator set model and a preset large vortex turbulence model;
calculating a real-time wind resource simulation observation data set and a fan point location wind speed set according to a preset wake condition set, the real-time anemometer tower data and the first flow field model result, and fitting a first statistical relationship between the environment data set and the generated energy data according to the numerical weather forecast data;
and according to the real-time wind resource simulation observation data group, the fan point position wind speed group and the first statistical relationship, performing power prediction on the offshore wind farm to obtain a prediction result.
2. The method for power prediction of an offshore wind farm according to claim 1, wherein the establishing and calculating of the first flow field model result of the offshore wind farm according to the environmental data set, a preset actuation line wind power set model and a preset large vortex turbulence model specifically comprises:
establishing a first flow field model of the offshore wind farm according to a preset actuating line wind generating set model and a preset large vortex turbulence model;
and calculating a first flow field model result of the offshore wind farm according to the environment data group and the first flow field model.
3. The method for predicting power of an offshore wind farm according to claim 2, wherein a real-time wind resource simulation observation data set and a wind speed set of a wind turbine site location are calculated according to a preset wake condition set, the real-time anemometer tower data and the first flow field model result, and specifically comprises:
calculating a wind resource simulation observation data set of each height of each point of a fan in the offshore wind farm according to the first flow field model result and the real-time wind measuring tower data;
and calculating the calculated wind speed of each point of the fan under the wake condition group according to the wind resource simulation observation data group and a preset wake condition group, and storing the wind speed into a fan point position wind speed group.
4. The method for predicting power of an offshore wind farm according to claim 3, wherein the predicting power of the offshore wind farm according to the real-time wind resource simulation observation data set, the wind speed set of the wind turbine site location and the first statistical relationship to obtain a prediction result specifically comprises:
according to the real-time wind resource simulation observation data set, the fan point position wind speed set and the first statistical relationship, based on a CFD power downscaling scale, performing short-term wind power prediction on the offshore wind farm to obtain a short-term prediction result;
according to the real-time wind resource simulation observation data set, the fan point position wind speed set and the first statistical relationship, wind power aggregate prediction is carried out on the offshore wind farm according to multiple prediction methods to obtain an aggregate prediction result;
and outputting the short-term prediction result and the set prediction result as prediction results.
5. The method for power prediction of an offshore wind farm according to claim 4, wherein wind power short-term prediction is performed on the offshore wind farm based on CFD power down scale according to the real-time wind resource simulation observation data set, the wind speed set of the wind turbine site and the first statistical relationship to obtain a short-term prediction result, specifically comprising:
correcting the numerical weather forecast data by a preset mesoscale wind speed forecast correcting method according to the real-time wind resource simulation observation data group and the fan point position wind speed group;
selecting corresponding micro-scale wind field distribution of the wind power plant from a preset micro-scale wind field basic database according to the corrected numerical weather forecast data, and analyzing and calculating the free incoming flow wind speed of each fan;
calculating to obtain the simulated generating power of each fan according to the first statistical relationship, a preset wind turbine generator simulated generating power correction method and the free incoming flow wind speed;
and removing the unit which does not operate and the time period during which the unit does not operate according to a preset wind power plant operation maintenance plan to obtain the whole wind power plant generated power prediction.
6. The method for predicting power of an offshore wind farm according to claim 5, wherein the step of calculating the calculated wind speed of each point of the wind turbine under the wake condition group according to the wind resource simulation observation data group and a preset wake condition group specifically comprises:
calculating the calculated wind speed of each point of the fan according to a preset Larsen wake flow calculation model and a wind speed data set; the wind resource simulation observation data group comprises a wind speed data group.
7. Method for power prediction of an offshore wind farm according to any of the claims 4-6, characterized in that said plurality of prediction methods comprises: a neural network prediction method based on a tabu algorithm, a power time sequence prediction method, a particle swarm algorithm and a Kalman filtering correction power prediction method.
8. A power prediction device for an offshore wind farm, characterized in that the power prediction device comprises a data acquisition unit, a calculation fitting unit and a power prediction unit, wherein,
the data acquisition unit is used for acquiring an environment data group, generated energy data, real-time anemometer tower data and numerical weather forecast data of the offshore wind farm, and establishing and calculating a first flow field model result of the offshore wind farm according to the environment data group, a preset actuating line wind generating set model and a preset large vortex turbulence model;
the calculation fitting unit is used for calculating a real-time wind resource simulation observation data set and a fan point location wind speed set according to a preset wake flow condition set, the real-time anemometer tower data and the first flow field model result, and fitting a first statistical relationship between the environment data set and the generated energy data according to the numerical weather forecast data;
the power prediction unit is used for performing power prediction on the offshore wind farm according to the real-time wind resource simulation observation data set, the fan point location wind speed set and the first statistical relationship to obtain a prediction result.
9. A computer-readable storage medium, comprising a stored computer program, wherein the computer program, when executed, controls a device on which the computer-readable storage medium is located to perform a method of power prediction for an offshore wind farm according to any one of claims 1 to 7.
10. A power prediction system for an offshore wind farm, characterized in that the power prediction system comprises a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, the processor when executing the computer program implementing a power prediction method for an offshore wind farm according to any one of claims 1 to 7.
CN202210718925.4A 2022-06-23 2022-06-23 Power prediction method, device, storage medium and system for offshore wind farm Pending CN115034159A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210718925.4A CN115034159A (en) 2022-06-23 2022-06-23 Power prediction method, device, storage medium and system for offshore wind farm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210718925.4A CN115034159A (en) 2022-06-23 2022-06-23 Power prediction method, device, storage medium and system for offshore wind farm

Publications (1)

Publication Number Publication Date
CN115034159A true CN115034159A (en) 2022-09-09

Family

ID=83126880

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210718925.4A Pending CN115034159A (en) 2022-06-23 2022-06-23 Power prediction method, device, storage medium and system for offshore wind farm

Country Status (1)

Country Link
CN (1) CN115034159A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115693666A (en) * 2022-12-30 2023-02-03 中国华能集团清洁能源技术研究院有限公司 Offshore wind farm generated energy determination method and system based on satellite inversion

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115693666A (en) * 2022-12-30 2023-02-03 中国华能集团清洁能源技术研究院有限公司 Offshore wind farm generated energy determination method and system based on satellite inversion

Similar Documents

Publication Publication Date Title
Samadianfard et al. Wind speed prediction using a hybrid model of the multi-layer perceptron and whale optimization algorithm
CN111210082B (en) Precipitation prediction method based on optimized BP neural network algorithm
Hung et al. An artificial neural network model for rainfall forecasting in Bangkok, Thailand
Sudheer et al. Estimating actual evapotranspiration from limited climatic data using neural computing technique
CN109146162B (en) A kind of probability wind speed forecasting method based on integrated Recognition with Recurrent Neural Network
CN112288164B (en) Wind power combined prediction method considering spatial correlation and correcting numerical weather forecast
Ma et al. Ultra-short-term wind generation forecast based on multivariate empirical dynamic modeling
CN110619432A (en) Feature extraction hydrological forecasting method based on deep learning
CN112765912A (en) Evaluation method for social and economic exposure degree of flood disasters based on climate mode set
CN104036328A (en) Self-adaptive wind power prediction system and prediction method
CN111695290A (en) Short-term runoff intelligent forecasting hybrid model method suitable for variable environment
CN108802857A (en) A kind of Meteorology Forecast System based on meteorological data
CN115907436B (en) Quality coupling prediction-based water resource water environment regulation and control method and system
Niu et al. Short-term wind speed hybrid forecasting model based on bias correcting study and its application
CN116595394A (en) Training method of wind speed correction model, wind speed prediction method, wind speed prediction equipment and medium
CN115034159A (en) Power prediction method, device, storage medium and system for offshore wind farm
CN109146196B (en) Residential community district water consumption prediction method
CN116976227B (en) Storm water increasing forecasting method and system based on LSTM machine learning
CN112307672A (en) BP neural network short-term wind power prediction method based on cuckoo algorithm optimization
CN116760006A (en) Wind power prediction method, device, equipment and storage medium
Wang et al. Rainfall-runoff simulation using simulated annealing wavelet bp neural networks
CN107590346B (en) Downscaling correction model based on spatial multi-correlation solution set algorithm
CN115859765A (en) Method, device, equipment and storage medium for predicting city expansion
Bantupalli et al. Wind Speed forecasting using empirical mode decomposition with ANN and ARIMA models
CN114897204A (en) Method and device for predicting short-term wind speed of offshore wind farm

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination