CN111950780A - Wind power plant short-term power prediction method - Google Patents
Wind power plant short-term power prediction method Download PDFInfo
- Publication number
- CN111950780A CN111950780A CN202010755411.7A CN202010755411A CN111950780A CN 111950780 A CN111950780 A CN 111950780A CN 202010755411 A CN202010755411 A CN 202010755411A CN 111950780 A CN111950780 A CN 111950780A
- Authority
- CN
- China
- Prior art keywords
- meteorological
- parameters
- wind
- power plant
- fans
- 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
Links
- 238000000034 method Methods 0.000 title claims abstract description 51
- 238000012937 correction Methods 0.000 claims abstract description 48
- 238000005259 measurement Methods 0.000 claims description 31
- 238000010248 power generation Methods 0.000 claims description 27
- 238000013528 artificial neural network Methods 0.000 claims description 21
- 238000012549 training Methods 0.000 claims description 17
- 230000003044 adaptive effect Effects 0.000 claims description 10
- 230000000694 effects Effects 0.000 abstract description 6
- 230000005684 electric field Effects 0.000 abstract description 6
- 238000004364 calculation method Methods 0.000 description 7
- 230000005611 electricity Effects 0.000 description 4
- 238000009434 installation Methods 0.000 description 3
- 230000002093 peripheral effect Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
Images
Classifications
-
- 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
- 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
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- 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
- G06N3/045—Combinations of networks
-
- 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
- G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
-
- 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
-
- 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—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- 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
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A30/00—Adapting or protecting infrastructure or their operation
-
- 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
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Economics (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Data Mining & Analysis (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Development Economics (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Primary Health Care (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
A wind farm short-term power prediction method comprises the following steps: acquiring real-time forecast meteorological parameters of the location of a wind power plant, wherein the meteorological parameters comprise wind speed, wind direction, temperature and humidity; correcting the real-time forecast meteorological parameters according to preset correction models corresponding to all fans in the wind power plant to obtain real-time meteorological parameters corresponding to all fans; and determining the predicted generating power of each fan in the wind farm by taking the real-time meteorological parameters of each fan in the wind farm as input parameters based on a preset generating power prediction BP network model. The method can solve the influence of actual meteorological element deviation among fans on electric field power prediction caused by terrain difference and wake effect in an electric field area, and can effectively respond to the requirements of a dispatching department.
Description
Technical Field
The invention relates to the technical field of wind power generation, in particular to a wind power plant short-term power prediction method considering terrain and wake flow influence characteristics.
Background
Wind power generation is used as a novel renewable energy source with the widest development prospect, and the scale and the technology of the wind power generation are rapidly developed in China. However, the indirection and fluctuation of wind power generation are the largest brake elbows limiting the development of the wind power generation, and after large-scale grid connection, the wind power generation will have great impact on the aspects of prediction, scheduling, stable operation, optimization control and the like of a power grid. The accurate wind power prediction method can help the wind power plant to effectively respond to the requirements of a dispatching department, reduce the influence of wind power generation on a power grid, realize the maximum utilization of wind energy and improve the operation benefit of the wind power plant.
Wind power prediction mainly depends on numerical weather forecast, and the wind power plant and surrounding terrain are complex and prone to causing deviation of the numerical weather forecast. Meanwhile, the numerical weather forecast is regional generally, the fans of the wind power plant are scattered and affected by the terrain and the positions of the fans, the actual meteorological elements of each fan are different from the numerical weather forecast, and the influence of wake flow on the wind speed, the wind direction and the like during the operation of the fans causes large deviation of a prediction result.
Disclosure of Invention
Objects of the invention
The invention aims to provide a method for predicting the short-term power of a wind power plant, which is used for solving the influence of actual meteorological element deviation among fans on electric power prediction caused by terrain difference and wake effect in an electric field area and can effectively respond to the requirements of a dispatching department.
(II) technical scheme
In order to solve the above problem, a first aspect of the present invention provides a method for predicting short-term power of a wind farm, which at least includes the following steps:
acquiring real-time forecast meteorological parameters of the location of a wind power plant, wherein the meteorological parameters comprise wind speed, wind direction, temperature and humidity;
correcting the real-time forecast meteorological parameters according to preset correction models corresponding to all fans in the wind power plant to obtain real-time meteorological parameters corresponding to all fans;
and determining the predicted generating power of each fan in the wind farm by taking the real-time meteorological parameters of each fan in the wind farm as input parameters based on a preset generating power prediction BP network model.
Specifically, in this application, wind-powered electricity generation field includes a plurality of fans, and each fan is from taking meteorological detection device, and wind-powered electricity generation field is equipped with meteorological collection equipment such as anemometry tower simultaneously for gather meteorological parameters such as wind speed, wind direction, temperature and humidity.
Specifically, the correcting the real-time forecast meteorological parameters according to the preset correction model corresponding to each fan in the wind power plant to obtain the real-time meteorological parameters corresponding to each fan includes:
acquiring actual measurement meteorological parameter historical records acquired by all fans in the wind power plant;
comparing actual measurement meteorological parameters acquired by all fans at the same time in history with actual measurement meteorological parameters acquired by meteorological acquisition equipment of a wind power plant at the same time, and dividing the fans in the wind power plant into a standard pole fan set and a non-standard pole fan set according to a comparison result;
correcting real-time forecast meteorological parameters according to a preset correction model corresponding to the benchmark fan unit to obtain real-time meteorological parameters corresponding to all fans in the benchmark fan unit;
and correcting the real-time meteorological parameters corresponding to the fans in the benchmark fan unit according to the preset correction model corresponding to each non-benchmark fan in the non-benchmark fan unit to obtain the real-time meteorological parameters corresponding to each fan in the non-benchmark fan unit.
In a specific embodiment, a plurality of fans with the smallest difference between the measured meteorological parameters of the fans and the measured meteorological parameters collected by the meteorological collection equipment are used as the benchmark fans. In other embodiments, the wind farm workers may also divide the wind farm according to the actual situation of the installation position of the wind turbine of the wind farm.
The method converts the nonlinear relation between the actual meteorological data of the wind power plant and the meteorological station forecast meteorological data into a forecast data self-correction BP network model containing the wind power plant which influences the forecasting accuracy and geographic information such as peripheral terrain, roughness and the like.
Specifically, the preset correction model corresponding to the benchmark fan unit is obtained by the following method:
acquiring a historical record of meteorological data of the location of a wind power plant, wherein the meteorological data comprises forecast meteorological parameters of a local weather forecast and corresponding actual measurement meteorological parameters acquired by meteorological acquisition equipment of the wind power plant;
and according to the historical records, taking the forecasted meteorological parameters as input parameters, taking the actually measured meteorological parameters collected by the meteorological collection equipment of the wind power plant at the same time as output parameters, and training the adaptive BP neural network algorithm model to obtain a preset correction model corresponding to the benchmark fan unit.
Specifically, the preset correction model corresponding to the non-standard bar fan unit is obtained by the following method:
the preset correction model corresponding to each non-standard pole fan in the non-standard pole fan set is obtained by the following method:
and training the adaptive BP neural network algorithm model by taking the measured meteorological parameters of all the fans in the benchmark fan set as input parameters and the measured meteorological parameters of the non-benchmark fans in the same time as output parameters according to the historical records to obtain the preset correction models corresponding to all the non-benchmark fans.
Further, the wind farm comprises a plurality of meteorological acquisition devices, and the method for predicting the short-term power of the wind farm further comprises the following steps:
dividing the wind power plant into at least two sub-regions according to the position of the meteorological acquisition equipment of the wind power plant;
comparing the actual measurement meteorological parameters acquired by the fans in the subareas at the same time in history with the actual measurement meteorological parameters acquired by the meteorological acquisition equipment of the subareas at the same time;
and dividing the fans in the sub-areas into a benchmark fan group and a non-benchmark fan group according to the comparison result.
Specifically, the wind farm comprises a plurality of meteorological collection devices, and the wind farm is divided into at least two sub-areas according to the position of the meteorological collection devices of the wind farm, wherein:
the method for acquiring real-time forecast meteorological parameters of the location of the wind power plant comprises the following steps:
acquiring initial forecast meteorological parameters of the location of a wind power plant, wherein the initial forecast meteorological parameters comprise the initial forecast meteorological parameters of the wind power plant and the initial meteorological forecast parameters provided by meteorological stations around the wind power plant;
correcting the initial forecast gas phase parameters according to the preset correction model of each subregion to obtain real-time forecast meteorological parameters corresponding to each subregion;
correspondingly, the correcting the real-time forecast meteorological parameters according to the preset correction model corresponding to each fan in the wind power plant comprises the following steps:
and correcting the real-time forecast meteorological parameters of the corresponding sub-areas according to the preset correction models corresponding to the fans in the sub-areas of the wind power plant.
Specifically, the preset correction model of each sub-region is obtained by the following method:
acquiring a historical record of meteorological data of the location of a wind power plant, wherein the meteorological data comprises forecast meteorological parameters of local weather forecast and corresponding actual measurement meteorological parameters acquired by meteorological acquisition equipment of each subregion;
and according to the historical records, taking the forecasted meteorological parameters as input parameters, taking the actually measured meteorological parameters acquired by the sub-region meteorological acquisition equipment at the same time as output parameters, and training the adaptive BP neural network algorithm model to obtain a preset correction model of each sub-region.
Specifically, the preset generated power prediction BP network model is obtained by the following method:
acquiring a power generation power historical record corresponding to an actually measured meteorological parameter historical record acquired by each fan in the wind power plant;
and training the self-adaptive BP neural network algorithm model by taking the actual measurement meteorological parameters collected by the fans as input parameters and the corresponding generated power as output parameters according to the historical records to obtain a preset generated power prediction BP network model.
Specifically, after the predicted generated power of each wind turbine in the wind farm is determined, the method further includes:
and determining the total generated power of the wind power plant according to the predicted generated power of each fan based on a preset total power prediction model.
Specifically, the preset total power prediction model is obtained by:
acquiring a total generating power historical record of the wind power plant;
and training the self-adaptive BP neural network algorithm model by taking the generated power of each fan as an input parameter and taking the total generated power of the corresponding wind power plant at the same time as an output parameter according to the historical record to obtain a preset total power prediction model.
The method realizes the correction of the numerical weather forecast precision and accounts for the electric field loss, and further improves the accuracy of the power prediction of the wind power plant.
(III) advantageous effects
The technical scheme of the invention has the following beneficial technical effects:
the wind power plant power prediction method is used for solving the influence of actual meteorological element deviation among fans on wind power prediction caused by terrain difference and wake effect in an electric field area and effectively responding to the requirements of a dispatching department.
Drawings
FIG. 1 is a flow chart of a method for predicting short-term power of a wind farm provided by the present application;
fig. 2 is a flow chart diagram of a short-term power prediction method for a wind farm according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the accompanying drawings in conjunction with the following detailed description. It should be understood that the description is intended to be exemplary only, and is not intended to limit the scope of the present invention. Moreover, in the following description, descriptions of well-known structures and techniques are omitted so as to not unnecessarily obscure the concepts of the present invention.
The application provides a method for predicting short-term power of a wind power plant, which at least comprises the following steps:
step 101: acquiring real-time forecast meteorological parameters of the location of a wind power plant, wherein the meteorological parameters comprise wind speed, wind direction, temperature and humidity;
step 102: correcting the real-time forecast meteorological parameters according to preset correction models corresponding to all fans in the wind power plant to obtain real-time meteorological parameters corresponding to all fans;
step 103: and determining the predicted generating power of each fan in the wind farm by taking the real-time meteorological parameters of each fan in the wind farm as input parameters based on a preset generating power prediction BP network model.
According to the method, the forecast meteorological parameters are corrected according to the correction models of the fans, the actual meteorological element deviation among the fans caused by the wake effect is fully considered, and the accuracy of the prediction result is ensured.
Specifically, in this application, wind-powered electricity generation field includes a plurality of fans, and each fan is from taking meteorological detection device, and wind-powered electricity generation field is equipped with meteorological collection equipment such as anemometry tower simultaneously for gather meteorological parameters such as wind speed, wind direction, temperature and humidity. The forecasted meteorological parameters are provided by the wind farm and/or by surrounding weather stations.
Specifically, the correcting the real-time forecast meteorological parameters according to the preset correction model corresponding to each fan in the wind power plant to obtain the real-time meteorological parameters corresponding to each fan includes:
acquiring actual measurement meteorological parameter historical records acquired by all fans in the wind power plant;
comparing actual measurement meteorological parameters acquired by all fans at the same time in history with actual measurement meteorological parameters acquired by meteorological acquisition equipment of a wind power plant at the same time, and dividing the fans in the wind power plant into a standard pole fan set and a non-standard pole fan set according to a comparison result;
correcting real-time forecast meteorological parameters according to a preset correction model corresponding to the benchmark fan unit to obtain real-time meteorological parameters corresponding to all fans in the benchmark fan unit;
and correcting the real-time meteorological parameters corresponding to the fans in the benchmark fan unit according to the preset correction model corresponding to each non-benchmark fan in the non-benchmark fan unit to obtain the real-time meteorological parameters corresponding to each fan in the non-benchmark fan unit.
In a specific embodiment, a plurality of fans with the smallest difference between the measured meteorological parameters of the fans and the measured meteorological parameters collected by the meteorological collection equipment are used as the benchmark fans. In other embodiments, the wind farm workers may also divide the wind farm according to the actual situation of the installation position of the wind turbine of the wind farm.
The method converts the nonlinear relation between the actual meteorological data of the wind power plant and the meteorological station forecast meteorological data into a forecast data self-correction BP network model containing the wind power plant which influences the forecasting accuracy and geographic information such as peripheral terrain, roughness and the like.
Specifically, the preset correction model corresponding to the benchmark fan unit is obtained by the following method:
acquiring a historical record of meteorological data of the location of a wind power plant, wherein the meteorological data comprises forecast meteorological parameters of a local weather forecast and corresponding actual measurement meteorological parameters acquired by meteorological acquisition equipment of the wind power plant;
according to the historical records, the forecasted meteorological parameters are used as input parameters, the measured meteorological parameters collected by the meteorological collection equipment of the wind power plant at the same time are used as output parameters, the self-adaptive BP neural network algorithm model is trained to obtain a BP network model 1, and the BP network model 1 is used as a preset correction model corresponding to the benchmark fan set.
Specifically, the preset correction model corresponding to the non-standard pole fan is obtained by the following method:
according to the historical records, the measured meteorological parameters of all fans in the benchmark fan set are used as input parameters, the measured meteorological parameters of the non-benchmark fans in the same time are used as output parameters to train the self-adaptive BP neural network algorithm model, a BP network model 2 is obtained, and the BP network model 2 is used as a preset correction model corresponding to all the non-benchmark fans.
Further, the wind farm comprises a plurality of meteorological acquisition devices, and the method for predicting the short-term power of the wind farm further comprises the following steps:
dividing the wind power plant into at least two sub-regions according to the position of the meteorological acquisition equipment of the wind power plant;
comparing the actual measurement meteorological parameters acquired by the fans in the subareas at the same time in history with the actual measurement meteorological parameters acquired by the meteorological acquisition equipment of the subareas at the same time;
and dividing the fans in the sub-areas into a benchmark fan group and a non-benchmark fan group according to the comparison result.
In an embodiment, the wind farm comprises a plurality of meteorological collection devices, the wind farm being divided into at least two sub-areas according to its position, wherein:
the method for acquiring real-time forecast meteorological parameters of the location of the wind power plant comprises the following steps:
acquiring initial forecast meteorological parameters of the location of a wind power plant, wherein the initial forecast meteorological parameters comprise the initial forecast meteorological parameters of the wind power plant and the initial meteorological forecast parameters provided by meteorological stations around the wind power plant;
correcting the initial forecast gas phase parameters according to the preset correction model of each subregion to obtain real-time forecast meteorological parameters corresponding to each subregion;
correspondingly, the correcting the real-time forecast meteorological parameters according to the preset correction model corresponding to each fan in the wind power plant comprises the following steps:
and correcting the real-time forecast meteorological parameters of the corresponding sub-areas according to the preset correction models corresponding to the fans in the sub-areas of the wind power plant.
Specifically, the preset correction model of each sub-region is obtained by the following method:
acquiring a historical record of meteorological data of the location of a wind power plant, wherein the meteorological data comprises forecast meteorological parameters of local weather forecast and corresponding actual measurement meteorological parameters acquired by meteorological acquisition equipment of each subregion;
and according to the historical records, taking the forecasted meteorological parameters as input parameters, taking the actually measured meteorological parameters acquired by the sub-region meteorological acquisition equipment at the same time as output parameters, and training the adaptive BP neural network algorithm model to obtain a preset correction model of each sub-region.
The method further considers the actual meteorological element deviation of each regional part caused by the terrain difference, and further improves the accuracy and reliability of the wind power prediction result.
Specifically, the preset generated power prediction BP network model is obtained by the following method:
acquiring a power generation power historical record corresponding to an actually measured meteorological parameter historical record acquired by each fan in the wind power plant;
according to the historical records, the actual measurement meteorological parameters collected by the fans are used as input parameters, the corresponding generated power is used as output parameters to train the self-adaptive BP neural network algorithm model to obtain a BP network model 3, and the BP network model 3 is used as a preset generated power prediction BP network model.
Specifically, after the predicted generated power of each wind turbine in the wind farm is determined, the method further includes:
and determining the total generated power of the wind power plant according to the predicted generated power of each fan based on a preset total power prediction model.
Specifically, the preset total power prediction model is obtained by:
acquiring a total generating power historical record of the wind power plant;
according to the historical records, the generated power of each fan is used as an input parameter, the total generated power of the corresponding wind power plant at the same time is used as an output parameter, the self-adaptive BP neural network algorithm model is trained to obtain a BP network model 4, and the BP network model 4 is used as a preset total power prediction model.
The method realizes the correction of the numerical weather forecast precision and accounts for the electric field loss, and further improves the accuracy of the power prediction of the wind power plant.
In a specific embodiment of the present application, a method for predicting short-term power of a wind farm specifically includes the following steps:
step 1: establishing a basic learning step size self-adaptive adjustment BP neural network algorithm model, which comprises input and output layer parameters, a hidden layer node number calculation formula, a transfer function, an error limit value, a weight initial adjustment rate, a threshold value initial adjustment rate, a learning step size adjustment formula and the like;
step 2: the method comprises the following steps of performing subregion division on a wind power plant through the position of wind power plant meteorological collection equipment (such as a anemometer tower);
and step 3: acquiring historical records of meteorological data of a wind power plant in which a sub-area of the wind power plant is located and surrounding meteorological stations, wherein the meteorological data comprise forecast meteorological parameters of weather forecasts of the wind power plant and the surrounding meteorological stations and corresponding actually-measured meteorological parameters acquired by meteorological acquisition equipment of the sub-area, and the meteorological parameters comprise wind speed, wind direction, temperature and humidity;
step 4, according to the historical records, taking the forecasted meteorological parameters as input parameters, acquiring actual measured meteorological parameters as output parameters by corresponding meteorological acquisition equipment, and training the adaptive BP neural network algorithm model to obtain a forecasted meteorological parameter self-correction BP network model 1;
step 5, acquiring actual measurement meteorological parameter historical records acquired by each fan in the sub-area of the wind power plant, comparing the actual measurement meteorological parameters acquired by each fan at the same time in the history with the actual measurement meteorological parameters acquired by the sub-area meteorological acquisition equipment at the same time, and determining the benchmark fans in the sub-area according to the principle that the actual measurement meteorological parameters acquired by the fans are closest to the actual measurement meteorological parameters acquired by the meteorological acquisition equipment, so that the fans in the sub-area are divided into a benchmark fan group and a non-benchmark fan group;
step 6, according to historical records, taking the actually measured meteorological parameters of all fans in the benchmark fan set as input parameters, taking the actually measured meteorological parameters of non-benchmark fans in the same sub-area at the same time as output parameters to train the self-adaptive BP neural network algorithm model, and obtaining a meteorological data BP network model 2 between each non-benchmark fan and the benchmark fan by converting the mutual influence of wake flows among all the fans;
step 7, acquiring a power generation power historical record corresponding to the actual measurement meteorological parameter historical record acquired by each fan in the wind power plant, taking the actual measurement meteorological parameter acquired by each fan as an input parameter, and taking the corresponding power generation power as an output parameter to train the self-adaptive BP neural network algorithm model to obtain a BP network model 3;
step 8, acquiring a historical record of the total generated power of the wind power plant, taking the generated power of each fan as an input parameter and taking the total generated power of the corresponding wind power plant as an output parameter, and training an adaptive BP neural network algorithm model to obtain a BP network model 4;
and 9, acquiring real-time forecast meteorological parameters of the location of the wind power plant, taking the real-time forecast meteorological parameters as input parameters, determining real-time meteorological parameters corresponding to the benchmark fans according to the BP network model 1, determining real-time meteorological parameters corresponding to non-benchmark fans according to the real-time meteorological parameters corresponding to the benchmark fans based on the BP network model 2, taking the real-time meteorological parameters of the fans in the wind power plant as the input parameters, determining the predicted power generation power of the fans in the wind power plant according to the BP network model 3, and determining the power generation power of the wind power plant according to the predicted power generation power of the fans and the BP network model 4.
Specifically, as shown in fig. 2, in this embodiment, the BP neural network algorithm model in step 1 specifically includes:
hidden layer node number calculation formula
transfer function:
error limit value root mean square relative error
Wherein, Vcal is a predicted value, Vreal is an actual expected value, and N is the number of samples;
the initial weight adjustment rate w is 0.0035; initial threshold adjustment rate b: 0.001;
weight adjustment rate/threshold adjustment rate learning step-size adaptive formula:
In step 2, two anemometer towers and twenty fans are built in the wind power plant, the wind power plant is divided into two sub-areas, namely an area I and an area II, ten fans are arranged in each sub-area, and two peripheral weather stations (a weather station I and a weather station II) are arranged.
As shown in table 1, twelve data (meteorological parameters) in historical weather forecasts of the wind power plant and the surrounding meteorological stations are selected as input layer nodes in step 4, 8 historical collected meteorological data of the first area and the second area are used as output layer nodes, the middle layer nodes are set to be 7 according to a calculation formula, and the area forecast data self-correction BP network model 1 is obtained through training.
TABLE 1
And 5, comparing the historical collected meteorological data curves of the fans with the historical collected meteorological data curves of the meteorological collection equipment of the subzone, and setting three benchmark fans in the first zone and the second zone according to the installation positions of the fans of the wind power plant.
As shown in table 2, twelve historical collected (actually measured) meteorological data of the benchmark fans are selected as input layer nodes in the step 6, historical collected meteorological data of four single non-benchmark fans are used as output layer nodes, the middle layer nodes are set to be 6 according to a calculation formula, a non-benchmark fan meteorological data BP network model 2 is obtained through training, and 7 non-benchmark fan meteorological data BP network models are respectively arranged in each region.
TABLE 2
As shown in table 3, in step 7, historical collected meteorological data of four fans are selected as input layer nodes, historical power generation data (power generation power) is selected as output layer nodes, the middle layer nodes are set to be 3 according to a calculation formula, a BP network model for power generation of each fan is obtained through training, and twenty fans correspond to the BP network model for power generation of twenty fans.
TABLE 3
As shown in table 4, twenty wind turbine historical power generation data are selected as input layer nodes, historical power generation data (total power) of a common connection branch of a wind farm are selected as output layer nodes, the middle layer nodes are set to be 7 according to a calculation formula, and a wind farm calculation and loss actual power generation model is obtained through training.
TABLE 4
And substituting the weather forecast data of the next-day wind power plant and the surrounding weather stations into the regional forecast data self-correction BP network model shown in the table 1 to obtain the accurate weather forecast data of the first region and the second region of the next-day wind power plant.
And substituting the obtained accurate meteorological prediction data of the first area and the second area of the wind power plant in the next day into a meteorological prediction data BP network model of non-standard pole wind turbine meteorological data of each area shown in the table 2 to obtain the meteorological prediction data of each wind turbine in each area.
And substituting the meteorological forecast data of each fan into the power generation BP network model of each fan shown in the table 3 to obtain the independent power generation predicted power of each fan.
And substituting the individual power generation prediction power of each fan into the wind power plant meter and the loss actual power generation model shown in the table 4 to obtain the final power generation prediction data of the next day of the wind power plant.
It is to be understood that the above-described embodiments of the present invention are merely illustrative of or explaining the principles of the invention and are not to be construed as limiting the invention. Therefore, any modification, equivalent replacement, improvement and the like made without departing from the spirit and scope of the present invention should be included in the protection scope of the present invention. Further, it is intended that the appended claims cover all such variations and modifications as fall within the scope and boundaries of the appended claims or the equivalents of such scope and boundaries.
Claims (10)
1. A method for predicting short-term power of a wind power plant is characterized by at least comprising the following steps:
acquiring real-time forecast meteorological parameters of the location of a wind power plant, wherein the meteorological parameters comprise wind speed, wind direction, temperature and humidity;
correcting the real-time forecast meteorological parameters according to preset correction models corresponding to all fans in the wind power plant to obtain real-time meteorological parameters corresponding to all fans;
and determining the predicted generating power of each fan in the wind farm by taking the real-time meteorological parameters of each fan in the wind farm as input parameters based on a preset generating power prediction BP network model.
2. The wind farm short-term power prediction method according to claim 1,
the method for correcting the real-time forecast meteorological parameters according to the preset correction models corresponding to the fans in the wind power plant to obtain the real-time meteorological parameters corresponding to the fans specifically comprises the following steps:
acquiring actual measurement meteorological parameter historical records acquired by all fans in the wind power plant;
comparing actual measurement meteorological parameters acquired by all fans at the same time in history with actual measurement meteorological parameters acquired by meteorological acquisition equipment of a wind power plant at the same time, and dividing the fans in the wind power plant into a standard pole fan set and a non-standard pole fan set according to a comparison result;
correcting real-time forecast meteorological parameters according to a preset correction model corresponding to the benchmark fan unit to obtain real-time meteorological parameters corresponding to all fans in the benchmark fan unit;
and correcting the real-time meteorological parameters corresponding to the fans in the benchmark fan unit according to the preset correction model corresponding to each non-benchmark fan in the non-benchmark fan unit to obtain the real-time meteorological parameters corresponding to each fan in the non-benchmark fan unit.
3. The wind farm short-term power prediction method according to claim 2, characterized in that the preset correction model corresponding to the benchmark fan unit is obtained by the following method:
acquiring a historical record of meteorological data of the location of a wind power plant, wherein the meteorological data comprises forecast meteorological parameters of a local weather forecast and corresponding actual measurement meteorological parameters acquired by meteorological acquisition equipment of the wind power plant;
and according to the historical records, taking the forecasted meteorological parameters as input parameters, taking the actually measured meteorological parameters collected by the meteorological collection equipment of the wind power plant at the same time as output parameters, and training the adaptive BP neural network algorithm model to obtain a preset correction model corresponding to the benchmark fan unit.
4. The wind farm short-term power prediction method according to claim 3, wherein the preset correction model corresponding to each non-standard-bar fan in the non-standard-bar fan set is obtained by the following method:
and training the adaptive BP neural network algorithm model by taking the measured meteorological parameters of all the fans in the benchmark fan set as input parameters and the measured meteorological parameters of the non-benchmark fans in the same time as output parameters according to the historical records to obtain the preset correction models corresponding to all the non-benchmark fans.
5. The wind farm short-term power prediction method according to claim 2, further comprising:
dividing the wind power plant into at least two sub-regions according to the position of the meteorological acquisition equipment of the wind power plant;
comparing the actual measurement meteorological parameters acquired by the fans in the subareas at the same time in history with the actual measurement meteorological parameters acquired by the meteorological acquisition equipment of the subareas at the same time;
and dividing the fans in the sub-areas into a benchmark fan group and a non-benchmark fan group according to the comparison result.
6. The method for predicting the short-term power of the wind farm according to claim 5, wherein the step of acquiring real-time forecast meteorological parameters of the location of the wind farm comprises the following steps:
acquiring initial forecast meteorological parameters of the location of a wind power plant, wherein the initial forecast meteorological parameters comprise the initial forecast meteorological parameters of the wind power plant and the initial meteorological forecast parameters provided by meteorological stations around the wind power plant;
correcting the initial forecast gas phase parameters according to the preset correction model of each subregion to obtain real-time forecast meteorological parameters corresponding to each subregion;
correspondingly, the correcting the real-time forecast meteorological parameters according to the preset correction model corresponding to each fan in the wind power plant comprises the following steps:
and correcting the real-time forecast meteorological parameters of the corresponding sub-areas according to the preset correction models corresponding to the fans in the sub-areas of the wind power plant.
7. The wind farm short-term power prediction method according to claim 6, characterized in that the preset modified model of each sub-area is obtained by the following method:
acquiring a historical record of meteorological data of the location of a wind power plant, wherein the meteorological data comprises forecast meteorological parameters of local weather forecast and corresponding actual measurement meteorological parameters acquired by meteorological acquisition equipment of each subregion;
and according to the historical records, taking the forecasted meteorological parameters as input parameters, taking the actually measured meteorological parameters acquired by the sub-region meteorological acquisition equipment at the same time as output parameters, and training the adaptive BP neural network algorithm model to obtain a preset correction model of each sub-region.
8. The wind farm short-term power prediction method according to claim 1, characterized in that the preset generated power prediction BP network model is obtained by the following method:
acquiring a power generation power historical record corresponding to an actually measured meteorological parameter historical record acquired by each fan in the wind power plant;
and training the self-adaptive BP neural network algorithm model by taking the actual measurement meteorological parameters collected by the fans as input parameters and the corresponding generated power as output parameters according to the historical records to obtain a preset generated power prediction BP network model.
9. The wind farm short-term power prediction method according to claim 1, after determining the predicted generated power of each wind turbine within the wind farm, further comprising:
and determining the total generated power of the wind power plant according to the predicted generated power of each fan based on a preset total power prediction model.
10. The wind farm short-term power prediction method according to claim 9, characterized in that the preset total power prediction model is obtained by:
acquiring a total generating power historical record of the wind power plant;
and training the self-adaptive BP neural network algorithm model by taking the generated power of each fan as an input parameter and taking the total generated power of the corresponding wind power plant at the same time as an output parameter according to the historical record to obtain a preset total power prediction model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010755411.7A CN111950780A (en) | 2020-07-31 | 2020-07-31 | Wind power plant short-term power prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010755411.7A CN111950780A (en) | 2020-07-31 | 2020-07-31 | Wind power plant short-term power prediction method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111950780A true CN111950780A (en) | 2020-11-17 |
Family
ID=73338816
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010755411.7A Pending CN111950780A (en) | 2020-07-31 | 2020-07-31 | Wind power plant short-term power prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111950780A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117713039A (en) * | 2023-11-01 | 2024-03-15 | 宁夏青铜峡市华能雷避窑光伏发电有限公司 | Power plant power generation control method based on regional new energy power generation prediction |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102269124A (en) * | 2011-06-30 | 2011-12-07 | 内蒙古电力勘测设计院 | Ultra-short term wind power station generated power forecasting system |
| CN103001249A (en) * | 2012-11-15 | 2013-03-27 | 中国电力科学研究院 | Method for predicting short-term power of wind farm on the basis of BP (back propagation) neural network |
| CN103679298A (en) * | 2013-12-27 | 2014-03-26 | 中能电力科技开发有限公司 | Method for evaluating wind power plant short-period power prediction precision |
| CN104124715A (en) * | 2014-08-11 | 2014-10-29 | 国家电网公司 | Automatic gain scheduling method for wind power plants distributed as priority level according to blower involvement |
| CN104484704A (en) * | 2014-12-22 | 2015-04-01 | 国家电网公司 | Wind power plant theoretical power determination method based on BP neural network |
| CN109325607A (en) * | 2017-08-01 | 2019-02-12 | 中国电力科学研究院 | A kind of short-term wind power forecast method and system |
-
2020
- 2020-07-31 CN CN202010755411.7A patent/CN111950780A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102269124A (en) * | 2011-06-30 | 2011-12-07 | 内蒙古电力勘测设计院 | Ultra-short term wind power station generated power forecasting system |
| CN103001249A (en) * | 2012-11-15 | 2013-03-27 | 中国电力科学研究院 | Method for predicting short-term power of wind farm on the basis of BP (back propagation) neural network |
| CN103679298A (en) * | 2013-12-27 | 2014-03-26 | 中能电力科技开发有限公司 | Method for evaluating wind power plant short-period power prediction precision |
| CN104124715A (en) * | 2014-08-11 | 2014-10-29 | 国家电网公司 | Automatic gain scheduling method for wind power plants distributed as priority level according to blower involvement |
| CN104484704A (en) * | 2014-12-22 | 2015-04-01 | 国家电网公司 | Wind power plant theoretical power determination method based on BP neural network |
| CN109325607A (en) * | 2017-08-01 | 2019-02-12 | 中国电力科学研究院 | A kind of short-term wind power forecast method and system |
Non-Patent Citations (1)
| Title |
|---|
| 王尤嘉等: "基于特征聚类的区域风电短期功率统计升尺度预测", 《电网技术》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117713039A (en) * | 2023-11-01 | 2024-03-15 | 宁夏青铜峡市华能雷避窑光伏发电有限公司 | Power plant power generation control method based on regional new energy power generation prediction |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| TWI476430B (en) | A wind energy forecasting method with extreme wind speed prediction function | |
| US7523001B2 (en) | Method and apparatus for operating wind turbine generators | |
| US9337656B2 (en) | Method and system for forecasting wind energy | |
| CN107133702B (en) | Full-field power prediction method for wind power plant | |
| CN107292514B (en) | Site selection method and device for anemometer tower in production operation of wind power plant | |
| CN103904682B (en) | A kind of power forecasting method based on scene mixed model | |
| CN101634722A (en) | Method and arrangement for the forecast of wind-resources | |
| CN111488896B (en) | Distribution line time-varying fault probability calculation method based on multi-source data mining | |
| CN101794996A (en) | Real-time predicting method for output of wind electric field | |
| CN113205210B (en) | Wind power plant wind speed and power prediction method, system and equipment for complex terrain and storage medium | |
| CN103225588B (en) | Wind power generation power curve optimization method based on pattern recognition technology | |
| CN105389634A (en) | Combined short-term wind power prediction system and method | |
| Prasad et al. | Some of the design and methodology considerations in wind resource assessment | |
| CN109167387A (en) | Wind field wind power forecasting method | |
| CN101414317A (en) | Equivalent wind speed method for processing wind electric field static power equivalence dispersion problem | |
| CN103244354A (en) | Power curve self-adapting optimizing method for wind generating sets | |
| CN102945509A (en) | Wind electric field power prediction method | |
| WO2011124226A1 (en) | Method and system for forecasting wind energy | |
| CN107422180A (en) | A kind of power prediction system of the photovoltaic plant based on cloud monitoring | |
| CN105046349B (en) | It is a kind of meter and wake effect wind power forecasting method | |
| CN106650977A (en) | Short-term power prediction method used for newly-built wind farm | |
| CN107194141A (en) | A kind of region wind energy resources becomes more meticulous appraisal procedure | |
| CN111342499A (en) | Wind power plant real-time scheduling method based on wind power prediction data | |
| CN111950780A (en) | Wind power plant short-term power prediction method | |
| CN113610285A (en) | Power prediction method for distributed wind power |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20201117 |
|
| RJ01 | Rejection of invention patent application after publication |