CN114723299A - Offshore wind farm adjustable capacity assessment method considering submarine cable line loss calculation - Google Patents

Offshore wind farm adjustable capacity assessment method considering submarine cable line loss calculation Download PDF

Info

Publication number
CN114723299A
CN114723299A CN202210382206.XA CN202210382206A CN114723299A CN 114723299 A CN114723299 A CN 114723299A CN 202210382206 A CN202210382206 A CN 202210382206A CN 114723299 A CN114723299 A CN 114723299A
Authority
CN
China
Prior art keywords
power
offshore wind
offshore
wind farm
submarine cable
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
CN202210382206.XA
Other languages
Chinese (zh)
Inventor
梅睿
袁超
陈燕擎
唐一铭
钱鹏
杨春
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.)
Jiangsu Fangtian Power Technology Co Ltd
Original Assignee
Jiangsu Fangtian Power Technology 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 Jiangsu Fangtian Power Technology Co Ltd filed Critical Jiangsu Fangtian Power Technology Co Ltd
Priority to CN202210382206.XA priority Critical patent/CN114723299A/en
Publication of CN114723299A publication Critical patent/CN114723299A/en
Pending legal-status Critical Current

Links

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
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
    • G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
    • 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
    • G06N3/08—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
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E10/00—Energy generation through renewable energy sources
    • Y02E10/70—Wind energy
    • Y02E10/76—Power conversion electric or electronic aspects

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Mathematical Physics (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Health & Medical Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Strategic Management (AREA)
  • Software Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computational Mathematics (AREA)
  • Educational Administration (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Development Economics (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Molecular Biology (AREA)
  • Mathematical Analysis (AREA)
  • Biomedical Technology (AREA)
  • Computational Linguistics (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Biophysics (AREA)
  • Tourism & Hospitality (AREA)
  • Databases & Information Systems (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Public Health (AREA)
  • Probability & Statistics with Applications (AREA)
  • Evolutionary Biology (AREA)
  • Algebra (AREA)
  • Water Supply & Treatment (AREA)
  • Computer Vision & Pattern Recognition (AREA)

Abstract

The invention discloses an offshore wind farm adjustable capacity assessment method considering sea cable line loss calculation, which comprises the following steps: 1) inputting historical data of a fan SCADA (supervisory control and data acquisition) of an offshore wind farm and historical data of an operation and maintenance system of the offshore wind farm; 2) carrying out spectral clustering on the wind turbine generator based on SCADA historical data of the offshore wind farm fan; 3) establishing mapping from the clustering power of the wind turbine generator to the power of a current collecting wire of an offshore wind farm; 4) establishing mapping from the power of the current collecting wire of the offshore wind plant to the power loss of the submarine cable; 5) establishing an adjustable capacity evaluation model based on the power of a current collecting wire of an offshore wind plant and the power loss of a submarine cable; 6) and evaluating the adjustable capacity of the wind power plant based on the SCADA real-time information of the wind turbine of the offshore wind power plant.

Description

Offshore wind farm adjustable capacity assessment method considering submarine cable line loss calculation
Technical Field
The invention belongs to the field of offshore wind power generation, and particularly relates to an offshore wind power plant adjustable capacity evaluation method considering sea cable line loss calculation.
Background
Offshore wind power generation is an important new energy utilization mode, and as the permeability of offshore wind power in a power grid is continuously improved, certain impact is brought to the power grid by randomness and fluctuation characteristics of the offshore wind power. A fan of the offshore wind farm generates power and collects the power through a current collecting system, the power is boosted in a booster station and then transmitted to land grid-connected through a high-voltage submarine cable, and power loss is generated in the process. The existing cable loss calculation model does not have a specific calculation standard specially aiming at submarine cables, so that the calculation error of the submarine cable loss of an offshore wind power plant is large, and the adjustable capacity of a grid-connected point cannot be accurately and quickly estimated by Automatic Generation Control (AGC), so that the adjustment precision and the response speed of the AGC are influenced. The method comprises the steps of calculating the line loss of a submarine cable by using collector wire power, establishing mapping from fan power to grid-connected point power based on a neural network, finally evaluating the real-time maximum adjustable capacity of a wind power plant by using the real-time maximum generable capacity of a wind turbine generator, providing basis for power grid dispatching and promoting the consumption of offshore wind power.
Disclosure of Invention
The technical problem is as follows: the method can provide basis for power grid scheduling and promote coordinated development of offshore wind power and a power grid.
The technical scheme is as follows: the method for evaluating the adjustable capacity of the offshore wind farm by considering the calculation of the sea cable line loss comprises the following steps:
1) inputting historical data of a fan SCADA (supervisory control and data acquisition) of an offshore wind farm and historical data of an operation and maintenance system of the offshore wind farm;
2) carrying out spectral clustering on the wind turbine generator based on SCADA historical data of the offshore wind farm fan;
3) establishing mapping from the clustering power of the wind turbine generator to the power of a current collecting wire of an offshore wind farm;
4) establishing mapping from the power of the current-collecting wire of the offshore wind plant to the power loss of the submarine cable;
5) establishing an adjustable capacity evaluation model based on the power of a current collecting wire of an offshore wind plant and the power loss of a submarine cable;
6) and evaluating the adjustable capacity of the wind power plant based on the SCADA real-time information of the wind turbine of the offshore wind power plant.
In the step 1), the input SCADA historical data of the wind turbine comprises historical active power P of all wind turbines of the offshore wind farmwl(ii) a The input historical data of the offshore wind power plant operation and maintenance system comprises the power P of an offshore collecting wireglThe power P of the head end of the submarine cableslPower P at the end of submarine cablemlOffshore booster station power PhlLand centralized control center power Pll(ii) a The input duration of the historical data is 30 days, the time interval is 15 minutes, and the time dimension of each type of time sequence data is that T is 30 multiplied by 24 multiplied by 4 is 2880; historical active power PwlThe dimensionality is a multiplied by T, wherein a is the number of the fans; power P of offshore power collectorglDimension is b multiplied by T, wherein b is the number of the collecting wires; head end power P of submarine cableslAt the end of the sea cable, power PmlPower P of offshore booster stationhlAnd land centralized control center power PllThe dimensions are all 1 × T.
In the step 2), the method for carrying out spectral clustering on the wind turbine generator based on the historical data of the SCADA of the offshore wind farm fan comprises the following steps:
201) calculating the Pearson correlation coefficient of the power sequence phasor combination of any fan i and fan j of the offshore wind farm and recording the Pearson correlation coefficient as pijThus, a similarity matrix P is constructed:
Figure BDA0003591589250000021
202) based on the similarity matrix P, a degree matrix H is calculated:
Figure BDA0003591589250000022
the matrix is a diagonal matrix, element hiIs the sum of the ith row elements in the matrix P;
204) constructing a Laplace matrix L, and carrying out standardization treatment on the Laplace matrix L:
L＝H-P (3)
L′＝H-0.5LH-0.5 (4)
205) determining the number m of spectral clusters, solving the first m minimum eigenvalues of L' and corresponding eigenvectors, normalizing the eigenvectors, and constructing a new matrix Ua×m；
206) To matrix Ua×mThe row vectors are clustered by using K-means to obtain the division C of m clusters of the wind turbine generator1,C2,L CmAnd realizing the spectral clustering of the wind turbine generator.
In the step 3), the method for establishing the mapping from the clustering power of the wind turbine generator to the power of the offshore wind farm current collector is as follows:
301) based on the division results of clusters in the wind turbine generator spectrum clustering, calculating a historical sequence Z of the total power of the wind turbine generators in m clusters:
Figure BDA0003591589250000031
the ith row vector of Z represents the historical time sequence of the total power of the wind generating sets in the ith cluster;
302) based on a BP neural network, establishing a mapping from the clustering power of the wind turbine generator to the power of a collecting wire:
2880 column vectors of a historical sequence Z of the total power of the m clustered wind turbine generators are used as the input of a BP neural network training set, and the power P of an offshore collecting wiregl2880 column vectors of the BP neural network training set are used as the output of the BP neural network training set, and the number of input samples and output samples of the neural network training is 2880;
in a BP neural network, the input vector is X ═ X (X)1,x2,L,xn)T(ii) a Implicit expressionThe layer output is Y ═ Y1,y2,L,ym)TThe output of the output layer is O ═ O1,o2,L ol)TThe expected value is D ═ D1,d2,L,dl)T，ωij、ωjkRespectively are the weight from the input layer to the hidden layer and the weight from the hidden layer to the output layer;
computing hidden layer output
Figure BDA0003591589250000032
Computing output layer output
Figure BDA0003591589250000033
In the above two formulas, f (g) is an activation function of the BP neural network, and the output error E of the network is as follows:
Figure BDA0003591589250000034
substituting the equations (6) and (7) into the error expression (8) to obtain the equation (9)
Figure BDA0003591589250000035
From the equation (9), the output error E of the network is a weight ωij、ωjkThe weight is adjusted to reduce the output error of the network, and when the output error E of the network is smaller than an expected value or the training frequency reaches the standard, the training of the network is finished;
the weight value is updated along the direction of the negative gradient of the output error of the network so as to achieve the purpose of reducing the error,
Figure BDA0003591589250000041
Figure BDA0003591589250000042
eta is the learning rate and is 0-1; iteratively updating the weight value of the network according to the increment calculated by the formula,
ωij(n+1)＝ωij(n)+Δωij (13)
ωjk(n+1)＝ωjk(n)+Δωjk (14)
after the updating is finished, forward transmission of the information is carried out again until the training times are larger than the target value or the error is smaller than the target value; based on the BP neural network training process, establishing a mapping from the clustering power of the wind turbine generator to the power of the collecting wire:
PG＝F1(PZ) (15)
wherein, F1Representing a trained neural network model, PZRepresenting cluster power, P, of a group of wind turbinesGRepresenting the corresponding collector line power.
In the step 4), the mapping method for establishing the power of the current collecting wire of the offshore wind plant to the power loss of the submarine cable is as follows:
401) calculating historical data of power loss of the submarine cable:
Pxl＝Psl-Pml (16)
wherein, PxlHistorical data of power loss of the submarine cable is shown, the dimensionality is 1 x b, and b is the number of collecting wires;
402) collecting power P of electric wire by offshore wind plantglFor input, power loss P of submarine cablexlFor output, establishing a mapping from the power of the current collector of the offshore wind farm to the power loss of the submarine cable based on the BP neural network in the step 302):
PX＝F2(PG) (17)
wherein, PGIs a set of collector line powers, PXRepresenting the corresponding submarine cable power loss.
In the step 5), an adjustable capacity evaluation model based on the power of the current collecting wire of the offshore wind plant and the power loss of the submarine cable is established as follows:
PK＝sum(PG)-F2(PG)-ave(Phl+Pll) (18)
where sum () represents the vector sum and ave () represents the vector mean.
In the step 6), the model for evaluating the adjustable capacity of the wind power plant based on the real-time information of the SCADA of the offshore wind plant fans is as follows:
601) reading the real-time maximum generating capacity of the wind turbine generator in SCADA real-time information of the wind turbine generator in the offshore wind plant, calculating the sum of the maximum generating capacities of the m clusters of the wind turbine generator based on the spectral clustering division result of the m clusters in the step 206), forming an m-dimensional vector, and recording the m-dimensional vector as PZS；
602) The formula for calculating the real-time adjustable capacity of the offshore wind power plant is as follows:
PKS＝sum(F1(PZS))-F2(F1(PZS))-ave(Phl+Pll) (19)
wherein, PKSAnd representing the real-time maximum adjustable capacity of the wind power plant.
Has the advantages that: compared with the prior art, the invention has the following advantages: calculating the line loss of a submarine cable by utilizing the collector wire power, establishing a mapping from the fan power to the power of a grid-connected point based on a neural network, and finally estimating the real-time maximum adjustable capacity of a wind power plant by utilizing the real-time maximum capacity that can be generated by a wind turbine generator, so that a basis is provided for power grid scheduling, and offshore wind power consumption is promoted; by the method, the estimation precision of the adjustable capacity can be effectively improved, the AGC (automatic gain control) performance of the offshore wind farm is improved, and the coordinated development of offshore wind power and a power grid is promoted.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a diagram of a spectral clustering result of a wind turbine;
FIG. 3 is a graph of the results of collector line power calculations;
FIG. 4 is a graph of submarine cable loss calculations;
FIG. 5 is a graph of tunable capacity estimation.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
As shown in fig. 1, an offshore wind farm adjustable capacity evaluation method considering sea cable line loss calculation according to an embodiment of the present invention includes the following steps:
1) inputting historical data of a fan SCADA (supervisory control and data acquisition) of an offshore wind farm and historical data of an operation and maintenance system of the offshore wind farm;
2) carrying out spectral clustering on the wind turbine generator based on SCADA historical data of the offshore wind farm fan;
3) establishing mapping from the clustering power of the wind turbine generator to the power of a current collecting wire of an offshore wind farm;
4) establishing mapping from the power of the current collecting wire of the offshore wind plant to the power loss of the submarine cable;
5) establishing an adjustable capacity evaluation model based on the power of a current collecting wire of an offshore wind plant and the power loss of a submarine cable;
6) and evaluating the adjustable capacity of the wind power plant based on the SCADA real-time information of the wind turbine of the offshore wind power plant.
In the step 1), the input SCADA historical data of the fan comprises historical active power P of all wind turbines of the offshore wind farmwl(ii) a The input historical data of the offshore wind power plant operation and maintenance system comprises the power P of an offshore collecting wireglThe power P of the head end of the submarine cableslPower P at the end of submarine cablemlOffshore booster station power PhlLand centralized control center power Pll(ii) a The input duration of the historical data is 30 days, the time interval is 15 minutes, and the time dimension of each type of time sequence data is that T is 30 multiplied by 24 multiplied by 4 is 2880; historical active power PwlThe dimensionality is a multiplied by T, wherein a is the number of the fans; power P of offshore power collectorglDimension is b multiplied by T, wherein b is the number of the collecting wires; head end power P of submarine cableslAt the end of the sea cable, power PmlPower P of offshore booster stationhlAnd land centralized control center power PllThe dimensions are all 1 × T.
In the step 2), the method for carrying out spectral clustering on the wind turbine generator based on the historical data of the SCADA of the offshore wind farm fan comprises the following steps:
201) calculating the Pearson correlation coefficient of the power sequence phasor combination of any fan i and fan j of the offshore wind farm and recording the Pearson correlation coefficient as pijThus, a similarity matrix P is constructed:
Figure BDA0003591589250000061
202) based on the similarity matrix P, calculating a degree matrix H:
Figure BDA0003591589250000062
the matrix is a diagonal matrix, element hiIs the sum of the ith row elements in the matrix P;
204) constructing a Laplace matrix L, and carrying out standardization treatment on the Laplace matrix L:
L＝H-P (3)
L′＝H-0.5LH-0.5 (4)
205) determining the number m of spectral clusters, solving the first m minimum eigenvalues of L' and corresponding eigenvectors, normalizing the eigenvectors, and constructing a new matrix Ua×m；
206) To matrix Ua×mThe row vectors are clustered by using K-means to obtain the division C of m clusters of the wind turbine generator1,C2,L CmAnd realizing the spectral clustering of the wind turbine generator.
In the step 3), the method for establishing the mapping from the clustering power of the wind turbine generator to the power of the power collecting line of the offshore wind farm is as follows:
301) based on the division results of clusters in the wind turbine generator spectrum clustering, calculating a historical sequence Z of the total power of the wind turbine generators in m clusters:
Figure BDA0003591589250000071
the ith row vector of Z represents the historical time sequence of the total power of the wind generating sets in the ith cluster;
302) establishing a mapping from the clustering power of the wind turbine generator to the power of the collecting wire based on a BP neural network:
2880 column vectors of a historical sequence Z of the total power of the m clustered wind turbine generators are used as the input of a BP neural network training set, and the power P of an offshore collecting wiregl2880 column vectors of the BP neural network training set are used as the output of the BP neural network training set, and the number of input samples and output samples of the neural network training set is 2880;
in a BP neural network, the input vector is X ═ X (X)1,x2,L,xn)T(ii) a The hidden layer output is Y ═ Y1,y2,L,ym)TThe output of the output layer is O ═ O1,o2,L ol)TThe expected value is D ═ D1,d2,L,dl)T，ωij、ωjkRespectively are the weight from the input layer to the hidden layer and the weight from the hidden layer to the output layer;
computing hidden layer output
Figure BDA0003591589250000072
Computing output layer output
Figure BDA0003591589250000073
In the above two formulas, f (g) is an activation function of the BP neural network, and the output error E of the network is as follows:
Figure BDA0003591589250000074
substituting the equations (6) and (7) into the error expression (8) to obtain the equation (9)
Figure BDA0003591589250000081
From the equation (9), the output error E of the network is a weight ωij、ωjkThe weight is adjusted to reduce the output error of the network, and when the output error E of the network is smaller than an expected value or the training frequency reaches the standard, the training of the network is finished;
the weight value is updated along the direction of the negative gradient of the output error of the network so as to achieve the purpose of reducing the error,
Figure BDA0003591589250000082
Figure BDA0003591589250000083
eta is the learning rate and is 0-1; iteratively updating the weight value of the network according to the increment calculated by the formula,
ωij(n+1)＝ωij(n)+Δωij (13)
ωjk(n+1)＝ωjk(n)+Δωjk (14)
after the updating is finished, forward transmission of the information is carried out again until the training times are larger than the target value or the error is smaller than the target value; based on the BP neural network training process, establishing a mapping from the clustering power of the wind turbine generator to the power of the collecting wire:
PG＝F1(PZ) (15)
wherein, F1Representing a trained neural network model, PZRepresenting the cluster power, P, of a group of wind turbinesGRepresenting the corresponding collector line power.
In the step 4), the mapping method for establishing the power of the current collecting wire of the offshore wind plant to the power loss of the submarine cable is as follows:
401) calculating historical data of power loss of the submarine cable:
Pxl＝Psl-Pml (16)
wherein, PxlHistorical data of power loss of the submarine cable is represented, the dimensionality is 1 x b, and b is the number of collecting wires;
402) collecting power P with offshore wind plantglFor input, power loss P of submarine cablexlFor output, establishing a mapping from the power of the current collector of the offshore wind farm to the power loss of the submarine cable based on the BP neural network in the step 302):
PX＝F2(PG) (17)
wherein, PGIs a set of collector line powers, PXRepresenting the corresponding submarine cable power loss.
In the step 5), an adjustable capacity evaluation model based on the power of the current collecting wire of the offshore wind plant and the power loss of the submarine cable is established as follows:
PK＝sum(PG)-F2(PG)-ave(Phl+Pll) (18)
where sum () represents vector summation and ave () represents vector averaging.
In the step 6), the model for evaluating the adjustable capacity of the wind power plant based on the SCADA real-time information of the offshore wind plant fans is as follows:
601) reading the real-time maximum generating capacity of the wind turbine generator in the SCADA real-time information of the offshore wind farm fan, calculating the sum of the maximum generating capacities of the m clusters of the wind turbine generator based on the spectral clustering division result of the m clusters of the wind turbine generator in the step 206), forming an m-dimensional vector, and recording the m-dimensional vector as PZS；
602) The formula for calculating the real-time adjustable capacity of the offshore wind power plant is as follows:
PKS＝sum(F1(PZS))-F2(F1(PZS))-ave(Phl+Pll) (19)
wherein, PKSAnd representing the real-time maximum adjustable capacity of the wind power plant.
Examples
In this embodiment, collected data of 67 wind turbines of a certain offshore wind farm in Jiangsu is used, historical data of the offshore wind farm for 30 days is input, the time interval is 15 minutes, spectral clustering is performed on the wind turbines based on historical power, the clustering result is shown in FIG. 2, and the wind turbines do not represent actual spatial positions in the graph. As can be seen from fig. 2, 67 wind turbines are grouped into 12 classes.
According to the clustering result, the fan power is processed, then the processed fan power is used for mapping the power of the current collector wire, and the calculation result is shown in fig. 3. And (3) directly using the original power sequence to perform mapping comparison without clustering, and taking the average absolute error (MAE) and the Root Mean Square Error (RMSE) as evaluation indexes, wherein the error pair ratio is shown in Table 1:
TABLE 1 comparison of collector line power calculation errors with and without spectral clustering
MAE/MW RMSE/MW
Using spectral clustering 1.63 2.14
Not using spectral clustering 1.96 2.83
It can be seen that the power calculation error of the collector wire adopting spectral clustering is obviously reduced. The calculated collector line power is used to map the submarine cable power loss, and the result is shown in fig. 4. Compared with the calculation of the power loss of the submarine cable by using the power of the fan, the error ratio is shown in table 2.
TABLE 2 comparison of submarine cable power loss calculation errors using collector wire power and fan power
MAE/MW RMSE/MW
Power of current collector 0.18 0.23
Power of fan 0.24 0.31
The result shows that the result of calculating the power loss of the submarine cable by adopting the power of the collector wire is more accurate. Finally, the fan power is summed, and various losses are subtracted to obtain the adjustable capacity, and the result is shown in fig. 5.
The method provided by the invention is adopted for mapping step by step, and compared with the prior method for directly mapping the adjustable capacity by utilizing the fan power, the error ratio is shown in a table 3.
TABLE 3 comparison of the method of the present invention with the adjustable capacity error of direct fan power mapping
Figure BDA0003591589250000101
Through the error comparison, the error of the evaluation method provided by the invention is effectively reduced, and the effectiveness and superiority of the provided method compared with the prior art are verified. The method provided by the invention utilizes the power of the power collecting wire to calculate the power loss of the submarine cable, establishes the mapping from the fan power to the power of the grid-connected point based on the neural network, finally realizes the evaluation of the real-time maximum adjustable capacity of the wind power plant by utilizing the real-time maximum capacity of the wind turbine generator, provides a basis for the power grid dispatching and promotes the offshore wind power consumption; by the method, the estimation precision of the adjustable capacity can be effectively improved, the AGC (automatic gain control) adjustment performance of the offshore wind farm is further improved, and the coordinated development of offshore wind power and a power grid is promoted. The error is greatly reduced, and the effectiveness and the superiority of the method are verified.

Claims (7)

1. An offshore wind farm adjustable capacity assessment method considering sea cable line loss calculation is characterized by comprising the following steps:
1) inputting historical data of a fan SCADA (supervisory control and data acquisition) of an offshore wind farm and historical data of an operation and maintenance system of the offshore wind farm;
2) carrying out spectral clustering on the wind turbine generator based on SCADA historical data of the offshore wind farm fan;
3) establishing mapping from the clustering power of the wind turbine generator to the power of a current collecting wire of an offshore wind farm;
4) establishing mapping from the power of the current collecting wire of the offshore wind plant to the power loss of the submarine cable;
5) establishing an adjustable capacity evaluation model based on the power of a current collecting wire of an offshore wind plant and the power loss of a submarine cable;
6) and evaluating the adjustable capacity of the wind power plant based on the SCADA real-time information of the wind turbine of the offshore wind power plant.
2. The method for evaluating the adjustable capacity of the offshore wind farm by considering the calculation of the sea cable loss according to claim 1, wherein in the step 1), the input SCADA historical data of the wind turbines comprises historical active power P of all the wind turbines of the offshore wind farmwl(ii) a The input historical data of the offshore wind power plant operation and maintenance system comprises offshore collecting wiresPower PglThe power P of the head end of the submarine cableslPower P at the end of submarine cablemlOffshore booster station power PhlLand centralized control center power Pll(ii) a The input duration of the historical data is 30 days, the time interval is 15 minutes, and the time dimension of each type of time sequence data is that T is 30 multiplied by 24 multiplied by 4 is 2880; historical active power PwlThe dimensionality is a multiplied by T, wherein a is the number of the fans; power P of offshore power collectorglThe dimensionality is b multiplied by T, wherein b is the number of the collecting wires; head end power P of submarine cableslAt the end of the sea cable, power PmlPower P of offshore booster stationhlAnd land centralized control center power PllThe dimensions are all 1 × T.
3. The method for evaluating the adjustable capacity of the offshore wind farm in consideration of the sea cable loss calculation according to claim 2, wherein in the step 2), the method for performing spectral clustering on the wind turbines based on SCADA historical data of the offshore wind farm fans comprises the following steps:
201) calculating the Pearson correlation coefficient of the power sequence phasor combination of any fan i and fan j of the offshore wind farm and recording the Pearson correlation coefficient as pijThus, a similarity matrix P is constructed:
Figure FDA0003591589240000011
202) based on the similarity matrix P, a degree matrix H is calculated:
Figure FDA0003591589240000021
the matrix is a diagonal matrix, element hiIs the sum of the ith row elements in the matrix P;
204) constructing a Laplace matrix L, and carrying out standardization treatment on the Laplace matrix L:
L＝H-P (3)
L′＝H-0.5LH-0.5 (4)
205) determining spectral clusteringThe number m of classes, the first m minimum eigenvalues of L' and corresponding eigenvectors are obtained, and a new matrix U is constructed after the eigenvectors are normalizeda×m；
206) To matrix Ua×mThe row vectors are clustered by using K-means to obtain the division C of m clusters of the wind turbine generator1,C2,L CmAnd realizing the spectral clustering of the wind turbine generator.
4. The method for evaluating the adjustable capacity of the offshore wind farm by considering the sea cable loss calculation according to claim 3, wherein in the step 3), the method for establishing the mapping from the cluster power of the wind turbine generator to the current collector power of the offshore wind farm is as follows:
301) based on the division results of clusters in the wind turbine generator spectrum clustering, calculating a historical sequence Z of the total power of the wind turbine generators in m clusters:
Figure FDA0003591589240000022
the ith row vector of Z represents the historical time sequence of the total power of the wind generating sets in the ith cluster;
302) establishing a mapping from the clustering power of the wind turbine generator to the power of the collecting wire based on a BP neural network:
2880 column vectors of a historical sequence Z of the total power of the m clustered wind turbine generators are used as the input of a BP neural network training set, and the power P of an offshore collecting wiregl2880 column vectors of the BP neural network training set are used as the output of the BP neural network training set, and the number of input samples and output samples of the neural network training is 2880;
in a BP neural network, the input vector is X ═ X1,x2,L,xn)T(ii) a The hidden layer output is Y ═ Y1,y2,L,ym)TThe output of the output layer is O ═ O1,o2,L ol)TThe expected value is D ═ D1,d2,L,dl)T，ωij、ωjkWeights from input layer to hidden layer and weights from hidden layer to output layer, respectively；
Computing hidden layer output
Figure FDA0003591589240000031
Computing output layer output
Figure FDA0003591589240000032
In the above two formulas, f (g) is an activation function of the BP neural network, and the output error E of the network is as follows:
Figure FDA0003591589240000033
substituting the equations (6) and (7) into the error expression (8) to obtain the equation (9)
Figure FDA0003591589240000034
From the equation (9), the output error E of the network is a weight ωij、ωjkWhen the output error E of the network is smaller than an expected value or the training times reach a standard, the training of the network is finished;
the weight value is updated along the direction of the negative gradient of the output error of the network so as to achieve the purpose of reducing the error,
Figure FDA0003591589240000035
Figure FDA0003591589240000036
eta is the learning rate and is 0-1; iteratively updating the weight value of the network according to the increment calculated by the formula,
ωij(n+1)＝ωij(n)+Δωij (13)
ωjk(n+1)＝ωjk(n)+Δωjk (14)
after the updating is finished, forward transmission of the information is carried out again until the training times are larger than the target value or the error is smaller than the target value; based on the BP neural network training process, establishing a mapping from the clustering power of the wind turbine generator to the power of the collecting wire:
PG＝F1(PZ) (15)
wherein, F1Representing a trained neural network model, PZRepresenting the cluster power, P, of a group of wind turbinesGRepresenting the corresponding collector line power.
5. The method for evaluating the adjustable capacity of the offshore wind farm by considering the offshore cable loss calculation according to claim 4, wherein in the step 4), the mapping method for establishing the power of the current collecting wire of the offshore wind farm to the power loss of the submarine cable is as follows:
401) calculating historical data of power loss of the submarine cable:
Pxl＝Psl-Pml (16)
wherein, PxlHistorical data of power loss of the submarine cable is shown, the dimensionality is 1 x b, and b is the number of collecting wires;
402) collecting power P of electric wire by offshore wind plantglFor input, power loss P of submarine cablexlFor output, establishing a mapping from the power of the current collector of the offshore wind farm to the power loss of the submarine cable based on the BP neural network in the step 302):
PX＝F2(PG) (17)
wherein, PGIs a set of collector line powers, PXRepresenting the corresponding submarine cable power loss.
6. The method for evaluating the adjustable capacity of the offshore wind farm by considering the offshore cable loss calculation according to claim 5, wherein in the step 5), an adjustable capacity evaluation model based on the power of the current collection line of the offshore wind farm and the power loss of the submarine cable is established as follows:
PK＝sum(PG)-F2(PG)-ave(Phl+Pll) (18)
where sum () represents the vector sum and ave () represents the vector mean.
7. The method for evaluating the adjustable capacity of the offshore wind farm in consideration of the sea cable loss calculation according to claim 6, wherein in the step 6), the model for evaluating the adjustable capacity of the wind farm based on the SCADA real-time information of the offshore wind farm fans is as follows:
601) reading the real-time maximum generating capacity of the wind turbine generator in the SCADA real-time information of the offshore wind farm fan, calculating the sum of the maximum generating capacities of the m clusters of the wind turbine generator based on the spectral clustering division result of the m clusters of the wind turbine generator in the step 206), forming an m-dimensional vector, and recording the m-dimensional vector as PZS；
602) The formula for calculating the real-time adjustable capacity of the offshore wind power plant is as follows:
PKS＝sum(F1(PZS))-F2(F1(PZS))-ave(Phl+Pll) (19)
wherein, PKSAnd representing the real-time maximum adjustable capacity of the wind power plant.
CN202210382206.XA 2022-04-12 2022-04-12 Offshore wind farm adjustable capacity assessment method considering submarine cable line loss calculation Pending CN114723299A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210382206.XA CN114723299A (en) 2022-04-12 2022-04-12 Offshore wind farm adjustable capacity assessment method considering submarine cable line loss calculation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210382206.XA CN114723299A (en) 2022-04-12 2022-04-12 Offshore wind farm adjustable capacity assessment method considering submarine cable line loss calculation

Publications (1)

Publication Number Publication Date
CN114723299A true CN114723299A (en) 2022-07-08

Family

ID=82243799

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210382206.XA Pending CN114723299A (en) 2022-04-12 2022-04-12 Offshore wind farm adjustable capacity assessment method considering submarine cable line loss calculation

Country Status (1)

Country Link
CN (1) CN114723299A (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107194625A (en) * 2017-07-25 2017-09-22 国家电网公司 Wind power plant based on neutral net abandons wind-powered electricity generation amount appraisal procedure
WO2020016904A1 (en) * 2018-07-16 2020-01-23 Tata Power Solar Systems Ltd. A system and method of reducing power losses in pv installations
CN112886641A (en) * 2021-03-11 2021-06-01 国网电力科学研究院武汉南瑞有限责任公司 Submarine cable power transmission system analysis model and reactive power configuration method
CN114186488A (en) * 2021-12-06 2022-03-15 国网福建省电力有限公司经济技术研究院 Topological structure optimization method and system for large-scale offshore wind farm current collection system

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107194625A (en) * 2017-07-25 2017-09-22 国家电网公司 Wind power plant based on neutral net abandons wind-powered electricity generation amount appraisal procedure
WO2020016904A1 (en) * 2018-07-16 2020-01-23 Tata Power Solar Systems Ltd. A system and method of reducing power losses in pv installations
CN112886641A (en) * 2021-03-11 2021-06-01 国网电力科学研究院武汉南瑞有限责任公司 Submarine cable power transmission system analysis model and reactive power configuration method
CN114186488A (en) * 2021-12-06 2022-03-15 国网福建省电力有限公司经济技术研究院 Topological structure optimization method and system for large-scale offshore wind farm current collection system

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
康慨;卢胜;: "基于WT的风电场电能损耗计算与分析", 电气应用, no. 19, 5 October 2016 (2016-10-05) *

Similar Documents

Publication Publication Date Title
CN109086928A (en) Photovoltaic plant realtime power prediction technique based on SAGA-FCM-LSSVM model
CN107766968A (en) Short-term wind speed forecasting method based on CAPSO RVM built-up patterns
CN111079343B (en) Wind turbine generator effective wind speed estimation method based on width learning
CN110365053B (en) Short-term wind power prediction method based on delay optimization strategy
CN105574615A (en) Spatial correlation and genetic algorithm (GA) based wind power forecast method for wavelet-BP neural network
CN112070311A (en) Day-ahead light power prediction method based on similar day clustering and meteorological factor weighting
CN109412203B (en) Grid-connected capacity limit calculation method for wind power multi-feed-in power system
CN110941909A (en) Method and system for identifying and calculating dominance of parameters of photovoltaic power station and wind power plant
CN110298494A (en) A kind of wind power forecasting method based on Segment Clustering and Combinatorial Optimization
CN107947228A (en) The Stochastic Stability Analysis method of electric system containing wind-powered electricity generation based on Markov theories
CN108022019A (en) Wind power plant dispatching method and system based on Wind turbines classification
CN115578016A (en) Online evaluation method for frequency modulation capability of wind power plant with incomplete model
CN116722540A (en) Cluster dynamic partitioning strategy based on distributed photovoltaic output prediction and clustering algorithm
CN114444784A (en) Multi-target arrangement optimization method and system for wind power plant positions
CN105303319A (en) Wind power plant dynamic clustering equivalence method based on operating data
CN111639852B (en) Real-time evaluation method and system for vibration state of hydroelectric generating set based on wavelet singular value
CN105207255B (en) A kind of power system peak regulation computational methods suitable for wind power output
CN114723299A (en) Offshore wind farm adjustable capacity assessment method considering submarine cable line loss calculation
CN114429233A (en) Cluster algorithm-based station-level optimal power control method and system
CN113780661B (en) Wind power ultra-short-term power prediction method for abrupt wind speed
CN110571794A (en) Transient model equivalent calculation method suitable for doubly-fed wind turbine wind power plant
CN114880863B (en) Self-adaptive frequency division order reduction method for distributed renewable energy cluster impedance aggregation model
CN116306236A (en) Wind power plant dynamic equivalent modeling method and system based on GRU-LSTM-FC combined network
CN116245318A (en) Day-ahead scheduling method and system for power system based on big data filling and clustering
CN115345073A (en) Grid-coordinated genetic algorithm-based wind power plant position optimization method

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