CN117993172A - Method and system for restoring free wind speed of wind power plant operation wind speed in complex terrain - Google Patents

Abstract

The invention relates to a method and a system for restoring free wind speed of wind power plant operation wind speed in complex terrain, wherein the method comprises the following steps: selecting a data series of running wind speed, wind direction and power generation active power of each wind turbine generator in a wind power plant, a data series of wind speed, wind direction, air temperature and air pressure of a wind power prediction tower, and a data series of numerical weather forecast wind speed and wind direction; dividing a wind area of the wind farm according to wind directions based on wind farm operation data; selecting a representative machine position representing the wind direction of the wind power plant based on the divided wind areas; constructing a wind direction structure of the wind power plant according to the representative wind direction of the fan, constructing a wake loss matrix of the wind direction structure of the wind power plant, and performing consistency check and correction; and restoring the running wind speed of each wind turbine generator to be free wind speed according to the wind direction structure of the wind power plant and the corresponding wake loss matrix. The wind power generation system is high in operability, the running wind speed wake loss of each machine position of the wind power plant can be accurately calculated, and the accuracy of restoring the running wind speed of each wind turbine generator into the free wind speed is improved.

Description

Method and system for restoring free wind speed of wind power plant operation wind speed in complex terrain

Technical Field

The invention belongs to the technical field of wind power generation, and particularly relates to a method for restoring free wind speed of running wind speed of a wind farm in complex terrains.

Background

The wind turbine generator with small single machine capacity of the original wind power plant is transformed, upgraded and removed, and the high-efficiency wind turbine generator with large single machine capacity is installed at a machine position with good wind energy resource condition selected from the machine positions of the removed wind turbine generator.

The wind energy resource analysis of the wind power plant is improved by 'high-pressure low', and the traditional method still adopts wind energy resource wind measurement data of a small number of wind measurement towers before wind power plant construction to calculate the whole wind power plant wind energy resource analysis method. According to the method, due to the fact that the wind measuring tower is insufficient in representativeness, accuracy of analyzing wind energy resources is low, and wind power plant operation data are not fully utilized.

In addition, in wind power prediction of the wind power plant, wind power prediction of the wind power plant is performed by adopting a mode of combining wind power plant numerical weather prediction with wind power prediction towers, the wind power prediction towers of the wind power plant in complex terrains are low in representativeness, wind measurement data of the wind power prediction towers are affected by wake flow of the wind power plant, the wind power prediction accuracy of the wind power plant in complex terrains is low, and if the running wind speed of each unit can be restored to free wind speed, each unit replaces the wind power prediction tower, and the wind power prediction accuracy of the wind power plant is greatly improved.

The wind speed and wind direction data of each wind turbine generator set of the wind power plant and wind measurement data of a wind power prediction tower are affected by running wake of the wind power plant, the wind speed is obviously reduced, the wind power plant cannot be directly used, and the running wind speed is required to be converted into free wind speed.

The wake flow is related to the arrangement of the wind power plant and the direction or wind direction of each machine position, N machine positions of the wind power plant are divided into 16 sector directions, the wind direction combination number of N wind turbines is 16 ^N, and the workload and the calculated amount are extremely large. The method for directly calculating the average wake flow of 16 sectors of each computer position and then restoring the running wind speed of each unit to the free wind speed is simple but has low accuracy, and no feasible method exists at present.

The improvement and upgrading of the large upper pressure of the wind power plant and the wind power prediction are all urgent to find a feasible method for restoring the running wind speed to the free wind speed of each unit so as to improve the accuracy of wind energy resource analysis and wind power prediction of the wind power plant.

Disclosure of Invention

The invention aims to provide a method and a system for restoring free wind speed of running wind speed of a wind farm in complex terrain, which are based on the construction of a wind direction combined structure, restore an average wind speed series of each unit running for 10 minutes in the past of the wind farm to a free wind speed series which is not influenced by wake flow, accurately analyze wind energy resources and wind power prediction of each unit for 'high-pressure low' -transformation and upgrading, accurately predict wind power of the wind farm by replacing a wind power prediction tower with each unit, and provide a solution.

The invention provides a method for restoring free wind speed of wind power plant operation wind speed in complex terrain, which comprises the following steps:

Step 1, selecting a data series of running wind speed, wind direction and power generation active power of each wind turbine generator in a wind power plant, a data series of wind speed, wind direction, air temperature and air pressure of a wind power prediction tower, a data series of numerical weather prediction wind speed and wind direction, a month-by-month actual running power curve of each wind turbine generator, running data of the former four wind turbine generators in the same period for one year, and actual measurement 1 of the wind power plant: 2000 topography;

the three are synchronous and complete for one year of operation data;

step 2, dividing the wind power plant into wind areas according to wind directions based on the selected wind power plant operation data;

Step 3, selecting a representative machine position representing the wind direction of the wind power plant based on the divided wind areas;

step 4, constructing a wind power plant wind direction structure according to the representative wind direction of the fan, constructing a wake loss matrix of the wind power plant wind direction structure, and checking and correcting the consistency of the wake loss matrix achievements;

And 5, restoring the running wind speed of each wind turbine generator to be the free wind speed according to the wind direction structure of the wind power plant and the corresponding wake loss matrix.

Further, the step 2 includes:

1) Selecting a wind direction sector numerical matrix formed by a wind turbine generator running wind direction data series, wherein the method comprises the following steps of:

Taking the maximum annual average wind speed of wind turbines of the wind power plant as a marker post unit, and selecting wind direction sector data of each wind turbine of the wind power plant on the condition that the wind speed of the marker post unit is 2.5m/s greater than the starting wind speed; n wind motor sets form a matrix F= (F _ij)_M×N, wherein M represents the number of time series, N represents the number of wind turbine units, F represents the number of wind direction sectors, and F _ij represents the number of wind direction sectors of an ith fan and a jth time period;

2) The selection of the wind area division index comprises the following steps:

based on the phenomenon that wind flows through a wind power plant have hysteresis and are subjected to terrain disturbance, dividing the machine position with the same or similar wind direction and the occurrence frequency of more than 2/3 into a wind area by statistics in a complete year; the wind area dividing method comprises the following steps: the amplitude of the difference between wind direction sectors of the two wind motor sets is 1, and the frequency of the occurrence of less than or equal to 1 is more than 2/3 throughout the year, which belongs to the same wind area;

3) Calculating a wind direction amplitude matrix, comprising:

Defining wind direction amplitude as the absolute value of the difference of wind direction sectors between two units in the same time period;

The wind direction amplitude matrix calculation method comprises the following steps: taking a kth fan as an example, calculating whether other fans are located in the same wind area with the kth fan or not by using a matrix F= (F _ij)_M×N), wherein each column in the matrix represents different wind direction time series of the fans, and the absolute value of the difference between each column of data and the kth column of data is taken, namely a _ij＝|f_ij-f_ik | (i=1, M; j=1, N) to form an A matrix, wherein A= (a _ij)_M×N;

4) The same wind area judgment is carried out, which comprises the following steps:

counting the frequency of occurrence of less than or equal to 1 according to the columns of the matrix A, and forming a wind area by the columns with the occurrence frequency of more than 2/3;

5) The wind area division of the wind power plant is carried out according to the following steps:

Step one: selecting wind data of N wind motor groups of a wind power plant running in a whole year, converting the annual wind direction series of the N wind motor groups into wind direction sector series, synchronously selecting the wind direction series of N wind motor groups by using a marker post unit with the wind speed of more than 2.5M/s, and forming an M multiplied by N wind direction matrix F= (F _ij)_M×N (i=1, M; j=1, N);

Step two: judging whether the 1 st row of fans in the matrix and other fans belong to the same air area; the absolute value of the difference between each column of data and the 1 st column of data in the matrix is a _ij＝|f_ij-f_i1 | (i=1, M; j=1, N), so that a wind direction amplitude A matrix, A= (a _ij)_M×N;

Step three: for the wind direction amplitude A matrix, counting the occurrence frequency of less than or equal to 1 according to the columns, forming a wind area by the wind turbine generator positions corresponding to the columns with the occurrence frequency of more than 2/3, wherein the number of the wind turbine generator sets forming the wind area is S ₁, and the wind area at least comprises the wind turbine generator set with the sequence number of 1 in the matrix;

Step four: deducting the number of the fan units in the divided wind area and corresponding machine positions, and forming an M multiplied by N1 wind direction matrix F1= (F _ij)_M×N1;

Step five: dividing N machine positions of the wind power plant into H wind areas by analogy according to the second step and the third step, wherein the number of the fan positions contained in each wind area is S ₁、S₂、S₃、…、S_H;

6) Consistency checking is carried out on wind area division of a wind power plant, and the method comprises the following steps:

Selecting according to the wind area division index, wherein the same wind area with the frequency of less than or equal to 1 which is more than 2/3 appears in the whole year according to the amplitude of the difference between wind direction sectors of two wind motor sets as 1;

The wind direction amplitude formed by the division of the wind areas is only judged whether the 1 st row of fans and other rows of fans in the matrix are in the same wind area, then the machine position of the same wind area is deducted, and then whether other machine positions are in the same wind area is judged;

checking whether other machine positions in the same wind area belong to the same wind area as other wind area machine positions, and if so, merging the two wind areas;

7) And dividing the wind power plant wind area into L wind areas according to wind area dividing consistency check, wherein each wind area contains machine positions QS ₁、QS₂、QS₃、…、QS_L.

Further, the step 3 includes:

1) Selecting a representative wind zone representative of a wind farm main wind direction, comprising:

Dividing wind areas of a wind power plant into achievements, wherein the number of machine bits QS ₁、QS₂、QS₃、…、QS_L in each wind area is ordered according to the size, and the wind area with the largest number of machine bits of the wind turbine generator is preferentially selected as a representative wind area; if the number of the selected wind areas can not reach more than half, selecting the wind area with the most number of the selected wind areas and the most number of the selected wind areas to form a combined wind area as a representative wind area;

2) Representing the machine position selection judgment index, comprising:

selecting a wind direction sector of a wind turbine generator set and a numerical weather forecast wind direction sector, wherein the amplitude of the difference between the wind direction sector and the numerical weather forecast wind direction sector is 1, calculating the occurrence frequency of which the annual amplitude is less than or equal to 1, and selecting the machine position with the highest annual occurrence frequency as the representative machine position;

3) The representative machine position selection steps are as follows:

Step one: wind direction data of the wind turbine generator in the representative wind area form a wind direction matrix; the machine position of an NN typhoon motor group in a representative wind area is changed into a sector wind direction in a whole year wind direction series M time periods to form a wind direction matrix FM= (f _ij)_M×NN; the numerical weather forecast wind direction fy data series M time periods are changed into the sector wind direction to form fy _i series i=1, M;

Step two: calculating a wind direction amplitude matrix of the fan and the numerical weather forecast; taking the absolute value of the difference between the kth column data in the wind direction matrix FM and the ith column data in the numerical weather forecast in the ith period f _ik and the numerical weather forecast in the ith period fy _i, and calculating a wind direction amplitude matrix B= (B _ij)_M×NN) according to a calculation formula B _ij＝|f_ik-fy_i |;

Step three: and counting the occurrence frequency which is less than or equal to 1 according to the wind direction amplitude matrix B. The 1 st column shows the frequency P1, the 2 nd column shows the frequencies P2 and …, the K column shows the frequencies PK and …, the NN column shows the frequency PNN, and the fan position with the largest frequency is selected as the representative machine position.

Further, the wind power plant wind direction structure wake loss matrix construction method in the step 4 comprises the following steps:

Step one: selecting a wind speed and wind direction data series according to the wind direction of the representative fan; for N wind power units in a wind power plant, each unit operates a data series of average wind speed and wind direction for 10 minutes in a complete year, and when a representative fan generates a certain wind direction, the data series of average wind speed and wind direction for 10 minutes of N fans in the wind power plant are selected;

step two: obtaining a wake loss matrix of a wind direction structure of a wind farm through calculation, wherein the wake loss matrix comprises the following components:

Taking each fan as a wind measuring tower, taking the data series selected in the step one as input, and calculating the average wake loss of 16 wind direction sectors of N wind turbines of the wind farm by a complex terrain wind farm wake model generating capacity calculation method to obtain a wake loss matrix W= (wl _ij)_N×16, forming N rows and 16 columns of average wake loss matrixes, wherein wl represents the wake loss, wl _ij represents the average wake loss of the ith wind turbine in j wind direction sectors;

Step three: according to the wind direction structure wake loss matrix calculation method of the wind farm in the second step, when wind directions 0, 1, 2, …, 14 and 15 sectors appear on the representative wind machine, average wake loss matrixes of 16 wind direction sectors of the N sets of the wind farm are W0, W1, …, wk, …, W14 and W15.

Further, the wake loss matrix achievement consistency checking and correcting method in the step 4 comprises the following steps:

If the absolute value of the difference between wake losses of the same unit before and after the sector wind direction is calculated within a given error range, the wake loss results are judged to be consistent, the error range is set to be 0.5%, and the wake loss inspection judgment of the unit a in the sector of the wind direction is assumed: w _{Front part ab}-W _{Rear part (S) ab}|×p≤0.5％,W _{Front part ab} represents the last calculated wake loss, W _{Rear part (S) ab} represents the current calculated wake loss, and p represents the frequency of occurrence of the a-unit in the b-wind direction sector.

Further, the wake loss matrix achievement consistency check and correction specifically comprises the following steps:

Step one: selecting a wind speed and wind direction data series according to the wind direction of the representative fan; for N wind turbines of a wind power plant, each wind turbine runs data on average wind speed and wind direction for 10 minutes in the whole year, when M wind directions appear on a representative fan, a data series of average wind speed and wind direction for 10 minutes of N wind turbines of the wind power plant is selected, and a wake loss matrix W _M＝(wl_ij)_N×16 is calculated;

Step two: l wind areas divided by wind power are selected, a wind direction sector with the largest frequency of occurrence of a machine position is selected in each wind area, and consistency test is carried out; except for representing fans, selecting a1, a2, a3, and an aL fan, wherein when the wind direction sector of the representing fan is M, the number of sectors with the largest occurrence frequency of each selected machine position is M1, M2, M3, … and ML respectively;

Step three: when a No. 1 fan generates a wind direction sector M1, a wind power plant N fans are selected for 10min to obtain an average wind speed and wind direction data series;

Step four: according to the data series selected in the step three, a wind direction sector frequency matrix of each fan is calculated in a statistics mode, P= (P _ij)_N×16, (i=1, N; j=0, 15) P represents the frequency of the occurrence of the wind direction of the j-th sector, and P _ij represents the frequency of the occurrence of the wind direction of the j-th sector of the i-th unit;

step five: wind direction structure wake loss matrix calculation:

Each fan is used as a wind measuring tower, the data series selected in the fourth step are used as input, the average wake loss of 16 wind direction sectors of the N wind power generation sets of the wind power plant is calculated based on a complex terrain wind power plant wake loss model generating capacity calculation method, so that a wake loss matrix WW= (wwl _ij)_N×16 is formed into N rows and 16 columns of average wake loss matrixes, wherein wwl represents the wake loss, wwl _ij represents the average wake loss of the ith wind power generation set in j wind direction sectors;

step six: consistency check of the wake loss achievement of the wind direction sector number M1 of the a1 fan:

(1) In the WM matrix, only keeping wake loss of the wind direction sector number M1 for the a1 fan, and giving 0 value to wake loss of other wind direction sectors, wherein W _M＝(wl_ij)_N×16 is changed into WF _M＝(wf_ij)_N×16;

(2) Calculating a wake loss deviation matrix; forming a wake loss deviation matrix WC= (wcij) _N×16 on the WF _M matrix and the WW matrix according to the calculation formula WC _ij＝|wf_ij-wwl_ij|×p_ij;

(3) If each element WC _ij of the wake loss deviation matrix WC is less than or equal to 0.5%, the wake loss result consistency check meets the requirement, and step eight is entered, otherwise, the a1 fan wind direction structure is subdivided;

step seven: subdividing a1 fan wind direction structure, correcting wake loss deviation matrix, including:

On the basis of the first step, selecting wind direction sectors with the frequency of more than 10% of wind direction occurrence for an a1 fan, wherein the wind direction sectors are M _{a1 1}、M_{a1 2}、...、M_{a1 L-1} respectively, and other sectors form a combined sector M _{a1 L};

Respectively selecting a10 min average wind speed and wind direction data series of N fans of a wind power plant according to the number of the sectors of the wind direction of the a1 fan as M _{a1 1}、M_{a1 2}、...、M_{a1 L};

Respectively calculating average wake loss matrixes of 16 wind direction sectors of N units of the wind power plant according to the number M _{a1 1}、M_{a1 2}、...、M_{a1 L} of wind direction sectors of the a1 fan to obtain a wake loss matrix W _{a1 1},W_{a1 2}、...、W_{a1 L}, and replacing wake loss results W _M＝(wl_ij)_N×16 representing the fan M sectors with the results;

Step eight: according to a similar method, carrying out wake loss result consistency check on fans a2, a3, the first and the second, and when the wake loss matrix result consistency check is not satisfied, subdividing a wind direction structure and correcting results;

Step nine: the results of the 16 wind direction sectors representing the fans are checked and corrected at one time according to the steps.

Further, the step 5 includes:

step one: converting the degree wind directions of the wind direction series of N wind units of the wind power plant into sector wind directions;

Step two: when representing that a fan has a wind direction sector k, selecting the average wind speed, wind direction data and active power series of N fans of the wind power plant for 10 minutes;

Step three: selecting wake loss matrixes W _k＝(wl_ij)_N×16 of 16 wind direction sectors of the N wind motor groups of the wind farm when the fan generates a wind direction k sector, and adopting a correction result if wake correction results of a subdivided wind area structure exist;

step four: for the selected data time series of the average wind speed, wind direction and active power of 10min of running wind of each wind turbine, when the active power is greater than 0 in the mth time period, checking a wake loss matrix according to the machine set number and the wind direction, for example, checking wake loss wl _Lc of the L wind turbine with the wind direction of the c-th sector in the wake loss matrix Wk according to a calculation formula, wherein the wind direction of the L wind turbine is the c-th sector in the m time period of the L wind turbine with the wind speed of V _{Transport and transport ml} Restoring to a free wind speed V _{Self-supporting ml} which is not influenced by wake flow;

step five: when the representative fan wind direction sector k is 0, 1, 2, …, 14 and 15 respectively, the running wind speed, wind direction and active power series of each wind turbine are selected respectively, and the running wind speed is reduced to the free wind speed according to the second step to the fourth step.

The invention also provides a system for restoring the free wind speed of the complex terrain wind farm operation wind speed, which is characterized by comprising a wind speed restoring module, wherein the wind speed restoring module is used for executing the method for restoring the free wind speed of the complex terrain wind farm operation wind speed according to any one of claims 1 to 7.

The invention also provides a non-transitory computer readable storage medium storing computer instructions which, when executed by a processor, implement a method of restoring free wind speed from complex terrain wind farm operating wind speed as claimed in any of claims 1 to 7.

The invention also provides an electronic device, comprising:

a memory and a processor, the memory and the processor being communicatively connected to each other, the memory having stored therein computer instructions, the processor executing the computer instructions to perform a method of restoring free wind speed from operating wind speeds of a complex terrain wind farm as claimed in any of claims 1 to 7.

By means of the scheme, the method and the system for restoring the free wind speed through the wind power plant operation wind speed in the complex terrain are high in operability, the wake loss of the wind power plant operation wind speed in each machine position is accurately calculated based on the wind direction combined structure of each machine position of the wind power plant under the main wind direction, and the accuracy of restoring the wind power plant operation wind speed to the free wind speed is improved. Based on the wind direction combined structure of each machine position of the wind power plant under the main wind direction, the demonstration is sufficient, the analysis is correct, and the result is reasonable. The wind power generation system has the advantages that the annual running wind speed of each wind turbine generator is reduced to the free wind speed, wind energy resource analysis is accurately carried out on 'high-pressure low' -transformation and upgrading of a wind power plant, and wind power of the wind power plant is accurately predicted by replacing a wind power prediction tower with each wind turbine generator. The method is suitable for 'high-pressure-on-low' transformation upgrading and wind power prediction of the wind farm under the condition of complex mountainous terrain in domestic and foreign wind power industries, and has strong applicability.

The foregoing description is only an overview of the present invention, and is intended to provide a better understanding of the present invention, as it is embodied in the following description, with reference to the preferred embodiments of the present invention and the accompanying drawings.

Drawings

FIG. 1 is a flow chart of a method of the present invention for restoring free wind speed from wind farm operating wind speed over complex terrain;

Fig. 2 is a schematic diagram of the structure of the electronic device of the present invention.

Detailed Description

The following describes in further detail the embodiments of the present invention with reference to the drawings and examples. The following examples are illustrative of the invention and are not intended to limit the scope of the invention.

Referring to fig. 1, the embodiment provides a method for restoring free wind speed of a wind farm with complex terrain, which includes:

step S1, selecting a wind power station running wind speed, wind direction and power generation active power data series, a wind power prediction tower wind speed, wind direction, air temperature and air pressure data series, a numerical weather prediction wind speed and wind direction data series, and running data of the three at the same time and in a complete year;

s2, dividing a wind farm into wind areas according to wind directions based on selected wind farm operation data;

Step S3, selecting a representative machine position representing the wind direction of the wind power plant based on the divided wind areas;

S4, constructing a wind power plant wind direction structure according to the representative wind direction of the fan, constructing a wake loss matrix of the wind power plant wind direction structure, and checking and correcting the consistency of the wake loss matrix achievements;

and S5, restoring the running wind speed of each wind turbine generator to be free wind speed according to the wind direction structure of the wind power plant and the corresponding wake loss matrix.

The following provides a more detailed description of the details of the various steps of the present invention:

1. And selecting wind farm operation data.

Selecting a wind power plant running wind speed, a wind direction, a power generation active power data series (interval time is 10 min), wind power prediction tower wind speed, wind direction, air temperature and air pressure data (interval time is 15 min), numerical weather forecast (NWP) wind speed and wind direction data (interval time is 15 min), and a month-by-month actual running power curve of each wind power plant, wherein the running data of the wind power plant is complete and one year in the same period, and the actual measurement of the wind power plant is 1:2000 topography; the three are integrated with one year of operation data. The running wind speed and the wind direction of the wind turbine are respectively measured data of wind speed and wind direction instruments on a cabin of the wind turbine. The operation of the wind turbine generator is influenced by the wake flow of the fan, and the measured wind speed is defined as the 'operation wind' speed, and is simply referred to as the operation wind speed.

2. The wind power plant divides wind areas according to wind directions.

(1) The basic principle of dividing wind areas according to wind directions. When the airflow flows through the mountain wind farm, the airflow influenced by the mountain topography basically flows along the topography following the general rule of the airflow flow of the complex topography, and stripping and compression can be generated when the airflow flows due to the change of the topography. When the air flow flows through the complicated mountain wind power plant, the wind direction structure of the wind power plant can change slightly under the influence of the terrain factors such as the slope direction, the gradient, the height and the ground roughness of the mountain, but the wind direction main body structure of each machine position of the wind power plant has little change, so that the wind power plant can be divided into different wind areas according to the wind direction.

(2) The degrees wind direction is converted into a wind direction sector. And (3) converting the average degree wind direction (0-360 degrees) of the running wind of each wind turbine generator set for 10 minutes into 16 wind direction sectors (see the table 1 for converting the degree wind direction into a wind direction sector table).

Table 1 degree wind direction to wind direction sector table

(3) And selecting an operation wind direction data series of the wind turbine to construct a wind direction sector matrix. The wind power plant wind turbine generator system annual average wind speed is the marker post unit, and wind direction sector data of each wind power plant wind direction sector data are selected based on the fact that the marker post unit wind speed is 2.5m/s greater than the wind power plant starting wind speed. Assuming that a certain wind farm has N wind turbines, an average wind direction series of N wind turbines for 10min forms a matrix F= (F _ij)_M×N, wherein M represents the time series number, N represents the wind turbine number, F represents the wind direction sector number, and F _ij represents the wind direction sector number of the ith fan and the jth time slot.

(4) And selecting wind area dividing indexes. Considering that wind flows flowing through a wind power plant have hysteresis and are subject to terrain disturbance, and dividing the wind power plant into a wind area with the same or similar wind direction and the occurrence frequency of more than 2/3 of the machine position by statistics in a complete year. The specific dividing method of the wind area comprises the following steps: the amplitude of the difference between the wind direction sectors of the two wind motor sets is 1, and the frequency of the wind motor sets which is less than or equal to 1 in the whole year is more than 2/3, and the wind motor sets belong to the same wind area.

(5) Calculating wind direction amplitude matrix

Wind direction amplitude is defined as the absolute value of the difference in wind direction sector between two groups of wind motors over the same time period.

The wind direction amplitude matrix calculation method comprises the following steps: the matrix F= (F _ij)_M×N, taking the kth fan as an example, calculating whether other fans and the kth fan are located in the same wind area, wherein each column in the matrix represents different wind direction time series of the fans, and the absolute value of the difference between each column of data and the kth column of data is taken, namely a _ij＝|f_ij-f_ik | (i=1, M; j=1, N) forms an A matrix, and A= (a _ij)_M×N).

(6) Same wind area judgment

And counting the frequency of occurrence of less than or equal to 1 according to the columns of the matrix A, and forming a wind area by the columns with the occurrence frequency of more than 2/3.

(7) Wind power plant wind area dividing step

Step one: and (3) selecting wind data of the N wind motor groups of the wind power plant, converting the annual wind direction series of the N wind motor groups into wind direction sector series, synchronously selecting the wind direction series of the N wind motor groups by using a marker post unit with the wind speed of more than 2.5M/s, and forming an M multiplied by N wind direction matrix F= (F _ij)_M×N (i=1, M; j=1, N) in a series of M time periods.

Step two: judging whether the 1 st row of fans in the matrix and other fans belong to the same air area. The absolute value of the difference between each column of data and the 1 st column of data in the matrix is a _ij＝|f_ij-f_i1 | (i=1, m; j=1, n;) to form a wind direction amplitude A matrix, and A= (a _ij)_M×N).

Step three: for the wind direction amplitude A matrix, counting the occurrence frequency of less than or equal to 1 according to the columns, forming a wind area by the wind turbine generator positions (abbreviated as wind turbine positions) corresponding to the columns with the occurrence frequency of more than 2/3, wherein the number of the wind turbine generator positions forming the wind area is S1, and the wind area at least comprises the wind turbine generators with the sequence number of 1 in the matrix.

Step four: and subtracting the number of the divided wind areas and the corresponding positions, and forming an M multiplied by N1 wind direction matrix F1= (F _ij)_M×N1) for the M time periods of the wind direction series of the rest N1 wind motor sets.

Step five: and (3) dividing N machine positions of the wind power plant into H wind areas by analogy according to the second step and the third step, wherein the number of the fan positions contained in each wind area is S ₁、S₂、S₃、…、S_H.

(8) Wind farm wind zone division consistency check

The wind area is selected according to the wind area dividing index, the amplitude of the difference between the wind direction sectors of the two wind motor sets is 1, and the frequency of the occurrence of the wind area which is less than or equal to 1 in the whole year is more than 2/3.

The wind area division only judges whether the 1 st row of fans and other rows of fans in the matrix are in the same wind area, then deducts the same wind area machine position, and then judges whether other machine positions are in the same wind area.

The wind area division consistency check is to check whether other machine positions in the same wind area belong to the same wind area as other wind area machine positions, and if the machine positions belong to the same wind area, the two wind areas are combined.

(9) And dividing the wind power plant wind area into L wind areas according to wind area dividing consistency check, wherein each wind area contains machine positions QS ₁、QS₂、QS₃、…、QS_L.

3. And selecting a representative machine position representing the wind direction of the wind power plant.

(1) Representing the machine position selection requirement. When the airflow flows through the wind power plant in the complicated mountain land, the main airflow of the wind power plant is basically kept relatively stable under the influence of factors such as topography, and the main airflow direction of most fans of the wind power plant is not greatly changed. The wind direction data representing the wind farm area adopts numerical weather forecast data for wind power prediction of the wind farm. And selecting the wind turbine position which can represent the wind directions of a plurality of wind power stations and has the best consistency with the wind direction of the numerical weather forecast as the representative wind position.

(2) A representative wind zone representative of a wind farm main wind direction is selected. Dividing the wind areas of the wind power plant into achievements, wherein the number of machine bits QS ₁、QS₂、QS₃、…、QS_L in each wind area is ordered according to the size, and the wind area with the largest number of machine bits of the fans is selected as a representative wind area. If the number of the selected wind areas can not reach more than half, selecting the wind area with the most number of the selected wind areas and the most number of the selected wind areas to form a combined wind area as a representative wind area.

(3) Representing the machine position selection judgment index. And counting 10min average wind direction series and contemporaneous numerical weather forecast 15min average wind direction series of the wind turbine generator in a representative wind area of the wind power plant according to 30min intervals, and selecting a fan position which is close to the numerical weather forecast wind direction and has highest occurrence frequency. And selecting the wind direction sector of the wind turbine generator set and the numerical weather forecast wind direction sector, wherein the amplitude of the difference between the wind direction sector and the numerical weather forecast wind direction sector is 1, calculating the occurrence frequency of which the annual amplitude is less than or equal to 1, and selecting the machine position with the highest annual occurrence frequency as the representative machine position.

(4) The representative machine position selection steps are as follows

Step one: wind direction data of wind turbines in the representative wind area form a wind direction matrix. The NN typhoon motor group in the representative wind area is arranged in M time periods of the annual wind direction series, the degree wind direction is changed into a wind direction sector to form a wind direction matrix FM= (f _ij)_M×NN. The numerical weather forecast wind direction f _y data series is arranged in M time periods, the degree wind direction is changed into a wind direction sector to form f _yi series i=1, M.

Step two: calculating a wind direction amplitude matrix of the fan and the numerical weather forecast; taking the absolute value of the difference between the kth column data in the wind direction matrix FM and the ith column data in the numerical weather forecast in the ith period f _ik and the numerical weather forecast in the ith period fy _i, and calculating a wind direction amplitude matrix B= (B _ij)_M×NN) according to a calculation formula B _ij＝|f_ik-f_yi |;

Step three: and counting the occurrence frequency which is less than or equal to 1 according to the wind direction amplitude matrix B. The 1 st column shows the frequency P1, the 2 nd column shows the frequencies P2 and …, the K column shows the frequencies PK and …, the NN column shows the frequency PNN, and the fan position with the largest frequency is selected as the representative machine position.

4. And constructing a wind power plant wind direction structure according to the representative wind direction of the fan, and constructing a wake loss matrix of the wind power plant wind direction structure.

1) Wake loss matrix for constructing wind direction structure of wind farm

The wind direction of the representative fan represents the direction of the dominant airflow of the wind power plant, a wind direction structure of a main body of the wind power plant is constructed according to the wind direction of the representative fan, wake loss of each machine position of the wind power plant is calculated, and a wake loss matrix of the wind direction structure of the wind power plant is constructed, wherein the steps are as follows:

Step one: and selecting a wind speed and wind direction data series according to the wind direction of the representative fan. And selecting an average wind speed and wind direction operation data series of each unit of the wind power plant for 10 minutes in the whole year, and when the representative wind direction k (k=0, 1,2, … and 15) of the wind turbines appears, selecting the average wind speed and wind direction data series of N wind turbines of the wind power plant for 10 minutes.

Step two: a wake loss matrix calculation method of a wind power plant wind direction structure. And (3) taking each fan as a wind measuring tower, selecting generating capacity calculation software suitable for a wake model of the wind farm with complex topography, inputting the data series selected in the step one, calculating the average wake loss of 16 wind direction sectors of N wind turbines of the wind farm, and forming a wake loss matrix W= (wl _ij)_N×16, obtaining N rows and 16 columns of average wake loss matrixes, wherein wl represents the wake loss, wl _ij represents the average wake loss of the ith wind turbine in j wind direction sectors.

Step three: according to the second calculation method, average wake loss matrixes W0, W1, …, wk, …, W14 and W15 of 16 wind direction sectors of the N units of the wind farm are calculated when wind directions 0, 1, 2, …, 14 and 15 sectors of the wind turbine appear respectively.

2) Wake loss matrix achievement consistency check and correction for wind power plant wind direction structure

And constructing a wind power plant main body wind direction structure according to the representative wind direction, and constructing wake loss matrix achievements, wherein whether the accuracy of the achievements meets the requirement or not is checked by carrying out achievements consistency. The wind direction structure is further subdivided, the accuracy of wake loss results can be improved, if the wind direction structure is subdivided, the results of the front and rear times are within a given error range, the consistency check of the results is met, and if the requirements are not met, the results are corrected.

The method comprises the steps of calculating absolute values of wake loss differences between front and back times in the wind direction of a sector where the same unit is located within a given error range according to wake loss result consistency detection judging indexes, judging that results are basically consistent, setting the temporary error range to be 0.5%, eliminating small probability events, and assuming that a unit is used for detecting and judging wake loss of a sector in the b wind direction: w _{Front part ab}-W _{Rear part (S) ab}|×p≤0.5％,W _{Front part ab} represents the last calculated wake loss, W _{Rear part (S) ab} represents the current calculated wake loss, and p represents the frequency of occurrence of the a-unit in the b-wind direction sector. . The steps of wake loss matrix result consistency check and result correction are as follows:

Step one: and selecting a wind speed and wind direction data series according to the wind direction of the representative fan. And (3) for N wind turbines in the wind power plant, each wind turbine operates data on average wind speed and wind direction for 10min in the whole year, when M wind directions appear on the representative fan, the data series of average wind speed and wind direction for 10min of N wind turbines in the wind power plant are selected under the condition, and a wake loss matrix W _M＝(wl_ij)_N×16 is calculated.

Step two: and (3) carrying out consistency test on L wind areas divided by wind power, wherein a wind direction sector with the largest frequency of occurrence of a machine position is selected in each wind area. Except for representative fans, fans a1, a2, a3 and an aL are selected, and when the representative fan wind direction sector is M, the number of sectors with the largest frequency of occurrence of each selected machine position is M1, M2, M3, … and ML respectively.

Step three: when the wind direction sector M1 appears in the fan a1, the condition selects the average wind speed and wind direction data series of the N fans of the wind power plant for 10 min.

Step four: according to the data series selected in the third step, a wind direction sector frequency matrix of each fan is calculated statistically, P= (P _ij)_N×16, (i=1, N; j=0, 15) P represents the frequency of the occurrence of the wind direction of the j-th sector, and P _ij represents the frequency of the occurrence of the wind direction of the j-th sector of the i-th unit

Step five: and calculating a wake loss matrix of the wind direction structure. And each fan is used as a wind measuring tower, the data series selected in the step four are used as input, generating capacity calculation software suitable for a wake loss model of the wind farm in complex terrain is selected, the average wake loss of 16 wind direction sectors of the N wind power generation sets of the wind farm is calculated, and a wake loss matrix WW= (wwl _ij)_N×16 is formed to form an N-row 16-column average wake loss matrix, wherein wwl represents the wake loss, wwl _ij represents the average wake loss of the ith wind power generation set in j wind direction sectors.

Step six: and (5) checking consistency of wake loss results of the wind direction sector number M1 of the a1 fan.

(1) In the WM matrix, only wake loss of the wind direction sector number M1 is reserved for the a1 fan, and each of the other wind direction sector wake loss is given a value of 0, and W _M＝(wl_ij)_N×16 becomes WF _M＝(wf_ij)_N×16.

(2) And calculating a wake loss deviation matrix. The WF _M matrix and WW matrix are calculated as WC _ij＝|wf_ij-wwl_ij|×p_ij to form a wake loss bias matrix wc= (wcij) _N×16.

(3) If each element WC _ij of the wake loss deviation matrix WC is less than or equal to 0.5%, the wake loss result consistency check meets the requirements, and step eight is entered, otherwise the a1 fan wind direction structure is subdivided.

Step seven: subdividing a1 fan wind direction structure and correcting wake loss deviation matrix.

On the basis of the first step, wind direction sectors with the frequency of wind direction more than 10% are selected for the a1 fan, wherein the wind direction sectors are M _{a1 1}、M_{a1 2}、...、M_{a1 L-1} respectively, and other sectors form a combined sector M _{a1 L}.

And respectively selecting an average wind speed and wind direction data series of N fans of the wind power plant for 10 minutes according to the number of the sectors of the wind direction of the a1 fan as M _{a1 1}、M_{a1 2}、...、M_{a1 L}.

And respectively calculating the average wake loss matrixes of the 16 wind direction sectors of the N units of the wind power plant according to the number M _{a1 1}、M_{a1 2}、...、M_{a1 L} of the wind direction sectors of the a1 fan to obtain a wake loss matrix W _{a1 1},W_{a1 2}、...、W_{a1 L}, wherein the wake loss matrixes replace wake loss results W _M＝(wl_ij)_N×16 of the sectors representing the fan M.

Step eight: and according to a similar method, carrying out wake loss result consistency check on fans a2, a3, the first and the second, and when the wake loss matrix result consistency check is not satisfied, subdividing a wind direction structure and correcting results.

Step nine: the results of the 16 wind direction sectors representing the fans are checked and corrected at one time according to the steps.

5. And restoring the running wind speed of each wind turbine to be free wind speed.

And restoring the running data of the N wind power sets of the wind power plant for 10 minutes into the average wind speed and the wind direction which are not influenced by wake loss, wherein the steps are as follows:

step one: converting the degree wind directions of the wind direction series of N wind units of the wind power plant into sector wind directions;

Step two: when representing that a fan has a wind direction sector k, selecting the average wind speed, wind direction data and active power series of N fans of the wind power plant for 10 minutes;

Step three: selecting wake loss matrixes W _k＝(wl_ij)_N×16 of 16 wind direction sectors of the N wind motor groups of the wind farm when the wind direction k sectors appear on the representative wind machine (if wake correction results of a subdivided wind area structure exist, adopting the correction results);

Step four: for the selected data time series of the average wind speed, wind direction and active power of each wind turbine running for 10min, when the active power is greater than 0 in the mth time period, checking a wake loss matrix according to the machine set number and the wind direction, for example, checking wake loss wl _Lc of the L-th wind turbine with the wind direction of the c-th sector in the wake loss matrix Wk (if a correction result is adopted, corresponding correction result is adopted) according to a calculation formula, wherein the wind direction of the L-th wind turbine is the c-th sector in the mth time period of the L-th wind turbine with the wind speed of V _{Transport and transport ml}, and the like Restoring the wind speed V _{Self-supporting ml} to a free wind speed which is not influenced by wake flow (if wake flow correction results of the subdivided wind zone structure exist, restoring the free wind speed to the running wind speed by adopting the mode of the original subdivided wind zone structure);

step five: when the representative fan wind direction sector k is 0, 1, 2, …, 14 and 15 respectively, the running wind speed, wind direction and active power series of each wind turbine are selected respectively, and the running wind speed is reduced to the free wind speed according to the second step to the fourth step.

The wind direction combination structure of each machine position under the main wind direction of the wind power plant is basically stable, wind direction structures of each machine position of the wind power plant are defined according to wind directions of the wind power plant, the same wind region represents the wind direction structure of each machine position in the wind region, different wind regions represent the wind direction combination structure of the wind power plant, the wind region representing the main wind direction of the wind power plant is selected, the machine position closest to the numerical weather forecast wind direction (representing the wind direction of the wind power plant area) is selected as the machine position representing the wind direction by means of the numerical weather forecast historical data of the wind power plant, and the wind direction structure of each machine position of the wind power plant and the corresponding wake loss matrix are constructed according to the wind direction of the representing the wind power plant, so that the running wind speed of each wind power plant is reduced to be free wind speed. The method solves the key problem of wind direction combination arrangement of all fans of the wind power plant, and provides a method for judging wind direction consistency wind area division indexes of the wind power plant, a method for dividing wind areas according to wind directions, a method for representing machine position selection of wind directions of the wind power plant and a method for constructing wind direction structures of the wind power plant according to the wind directions of the representing fans. The method solves the wake loss problem corresponding to the arrangement of wind directions of all fans of the wind power plant, and provides a method for constructing wind direction structures of the wind power plant, constructing corresponding wake loss matrixes and checking and correcting the consistency of results of the wake loss matrixes. The method for restoring the running wind speed of each wind turbine generator to be the free wind speed according to the wind direction structure of the wind power plant and the corresponding wake loss matrix is provided.

The method for restoring the running wind speed of the wind power plant in the complex terrain has strong operability, and the wake loss of the running wind speed of each machine position of the wind power plant is accurately calculated based on the wind direction combined structure of each machine position of the wind power plant under the main wind direction, so that the accuracy of restoring the running wind speed of each wind power set into the free wind speed is improved. Based on the wind direction combined structure of each machine position of the wind power plant under the main wind direction, the demonstration is sufficient, the analysis is correct, and the result is reasonable. The method for restoring the annual running wind speed of each wind turbine generator into the free wind speed provides a feasible solution for accurately analyzing wind energy resources by modifying and upgrading the wind power plant with high pressure and low pressure and accurately predicting the wind power of the wind power plant by replacing the wind power prediction tower with each wind turbine generator. The method is suitable for 'high-pressure-on-low' transformation upgrading and wind power prediction of the wind farm under the condition of complex mountainous terrain in domestic and foreign wind power industries, and has strong applicability.

The invention also provides a system for restoring the free wind speed of the complex terrain wind farm operation wind speed, which comprises a wind speed restoring module, wherein the wind speed restoring module is used for executing the method for restoring the free wind speed of the complex terrain wind farm operation wind speed.

The invention also provides a non-transitory computer readable storage medium storing computer instructions which when executed by a processor implement a method for restoring the running wind speed of the wind farm with complex terrain to a free wind speed.

Referring to fig. 2, an electronic device includes:

The wind power plant wind power generation system comprises a memory 201 and a processor 202, wherein the memory 201 and the processor 202 are in communication connection, computer instructions are stored in the memory 201, and the processor 202 executes the computer instructions, so that a method for restoring the running wind speed of the wind power plant with the complex terrain to the free wind speed is executed.

The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, and it should be noted that it is possible for those skilled in the art to make several improvements and modifications without departing from the technical principle of the present invention, and these improvements and modifications should also be regarded as the protection scope of the present invention.

Claims (10)

1. A method for restoring free wind speed from wind farm operation wind speed in complex terrain, comprising:

Step 1, selecting a data series of running wind speed, wind direction and power generation active power of each wind turbine generator in a wind power plant, a data series of wind speed, wind direction, air temperature and air pressure of a wind power prediction tower, a data series of numerical weather prediction wind speed and wind direction, a month-by-month actual running power curve of each wind turbine generator, running data of the former four wind turbine generators in the same period for one year, and actual measurement 1 of the wind power plant: 2000 topography;

step 2, dividing the wind power plant into wind areas according to wind directions based on the selected wind power plant operation data;

Step 3, selecting a representative machine position representing the wind direction of the wind power plant based on the divided wind areas;

step 4, constructing a wind power plant wind direction structure according to the representative wind direction of the fan, constructing a wake loss matrix of the wind power plant wind direction structure, and checking and correcting the consistency of the wake loss matrix achievements;

And 5, restoring the running wind speed of each wind turbine generator to be the free wind speed according to the wind direction structure of the wind power plant and the corresponding wake loss matrix.

2. The method for restoring free wind speed in wind farm operation according to claim 1, wherein the step2 comprises:

1) Selecting a wind direction sector numerical matrix formed by a wind turbine generator running wind direction data series, wherein the method comprises the following steps of:

Taking the maximum annual average wind speed of wind turbines of the wind power plant as a marker post unit, and selecting wind direction sector data of each wind turbine of the wind power plant on the condition that the wind speed of the marker post unit is 2.5m/s greater than the starting wind speed; n wind motor sets form a matrix F= (F _ij)_M×N, wherein M represents the number of time series, N represents the number of wind turbine units, F represents the number of wind direction sectors, and F _ij represents the number of wind direction sectors of an ith fan and a jth time period;

2) The selection of the wind area division index comprises the following steps:

based on the phenomenon that wind flows through a wind power plant have hysteresis and are subjected to terrain disturbance, dividing the machine position with the same or similar wind direction and the occurrence frequency of more than 2/3 into a wind area by statistics in a complete year; the wind area dividing method comprises the following steps: the amplitude of the difference between wind direction sectors of the two wind motor sets is 1, and the frequency of the occurrence of less than or equal to 1 is more than 2/3 throughout the year, which belongs to the same wind area;

3) Calculating a wind direction amplitude matrix, comprising:

Defining wind direction amplitude as the absolute value of the difference of wind direction sectors between two units in the same time period;

The wind direction amplitude matrix calculation method comprises the following steps: taking a kth fan as an example, calculating whether other fans are located in the same wind area with the kth fan or not by using a matrix F= (F _ij)_M×N), wherein each column in the matrix represents different wind direction time series of the fans, and the absolute value of the difference between each column of data and the kth column of data is taken, namely a _ij＝|f_ij-f_ik | (i=1, M; j=1, N) to form an A matrix, wherein A= (a _ij)_M×N;

4) The same wind area judgment is carried out, which comprises the following steps:

counting the frequency of occurrence of less than or equal to 1 according to the columns of the matrix A, and forming a wind area by the columns with the occurrence frequency of more than 2/3;

5) The wind area division of the wind power plant is carried out according to the following steps:

Step one: selecting wind data of N wind motor groups of a wind power plant running in a whole year, converting the annual wind direction series of the N wind motor groups into wind direction sector series, synchronously selecting the wind direction series of N wind motor groups by using a marker post unit with the wind speed of more than 2.5M/s, and forming an M multiplied by N wind direction matrix F= (F _ij)_M×N (i=1, M; j=1, N);

Step two: judging whether the 1 st row of fans in the matrix and other fans belong to the same air area; the absolute value of the difference between each column of data and the 1 st column of data in the matrix is a _ij＝|f_ij-f_i1 | (i=1, M; j=1, N), so that a wind direction amplitude A matrix, A= (a _ij)_M×N;

Step three: for the wind direction amplitude A matrix, counting the occurrence frequency of less than or equal to 1 according to the columns, forming a wind area by the wind turbine generator positions corresponding to the columns with the occurrence frequency of more than 2/3, wherein the number of the wind turbine generator sets forming the wind area is S ₁, and the wind area at least comprises the wind turbine generator set with the sequence number of 1 in the matrix;

Step four: deducting the number of the fan units in the divided wind area and corresponding machine positions, and forming an M multiplied by N1 wind direction matrix F1= (F _ij)_M×N1;

Step five: dividing N machine positions of the wind power plant into H wind areas by analogy according to the second step and the third step, wherein the number of the fan positions contained in each wind area is S ₁、S₂、S₃、…、S_H;

6) Consistency checking is carried out on wind area division of a wind power plant, and the method comprises the following steps:

Selecting according to the wind area division index, wherein the same wind area with the frequency of less than or equal to 1 which is more than 2/3 appears in the whole year according to the amplitude of the difference between wind direction sectors of two wind motor sets as 1;

The wind direction amplitude formed by the division of the wind areas is only judged whether the 1 st row of fans and other rows of fans in the matrix are in the same wind area, then the machine position of the same wind area is deducted, and then whether other machine positions are in the same wind area is judged;

checking whether other machine positions in the same wind area belong to the same wind area as other wind area machine positions, and if so, merging the two wind areas;

7) And dividing the wind power plant wind area into L wind areas according to wind area dividing consistency check, wherein each wind area contains machine positions QS ₁、QS₂、QS₃、…、QS_L.

3. The method for restoring free wind speed to operational wind speed of a wind farm with complex terrain according to claim 2, wherein the step 3 comprises:

1) Selecting a representative wind zone representative of a wind farm main wind direction, comprising:

Dividing wind areas of a wind power plant into achievements, wherein the number of machine bits QS ₁、QS₂、QS₃、…、QS_L in each wind area is ordered according to the size, and the wind area with the largest number of machine bits of the wind turbine generator is preferentially selected as a representative wind area; if the number of the selected wind areas can not reach more than half, selecting the wind area with the most number of the selected wind areas and the most number of the selected wind areas to form a combined wind area as a representative wind area;

2) Representing the machine position selection judgment index, comprising:

selecting a wind direction sector of a wind turbine generator set and a numerical weather forecast wind direction sector, wherein the amplitude of the difference between the wind direction sector and the numerical weather forecast wind direction sector is 1, calculating the occurrence frequency of which the annual amplitude is less than or equal to 1, and selecting the machine position with the highest annual occurrence frequency as the representative machine position;

3) The representative machine position selection steps are as follows:

Step one: wind direction data of the wind turbine generator in the representative wind area form a wind direction matrix; the machine position of an NN typhoon motor group in a representative wind area is changed into a sector wind direction in a whole year wind direction series M time periods to form a wind direction matrix FM= (f _ij)_M×NN; the numerical weather forecast wind direction fy data series M time periods are changed into the sector wind direction to form fy _i series i=1, M;

Step two: calculating a wind direction amplitude matrix of the fan and the numerical weather forecast; taking the absolute value of the difference between the kth column data in the wind direction matrix FM and the ith column data in the numerical weather forecast in the ith period f _ik and the numerical weather forecast in the ith period fy _i, and calculating a wind direction amplitude matrix B= (B _ij)_M×NN) according to a calculation formula B _ij＝|f_ik-fy_i |;

Step three: and counting the occurrence frequency which is less than or equal to 1 according to the wind direction amplitude matrix B. The 1 st column shows the frequency P1, the 2 nd column shows the frequencies P2 and …, the K column shows the frequencies PK and …, the NN column shows the frequency PNN, and the fan position with the largest frequency is selected as the representative machine position.

4. A method for restoring free wind speed from operational wind speed in a wind farm of complex terrain according to claim 3, wherein the wind farm wind direction structure wake loss matrix building method in step 4 comprises:

Step one: selecting a wind speed and wind direction data series according to the wind direction of the representative fan; for N wind power units in a wind power plant, each unit operates a data series of average wind speed and wind direction for 10 minutes in a complete year, and when a representative fan generates a certain wind direction, the data series of average wind speed and wind direction for 10 minutes of N fans in the wind power plant are selected;

step two: obtaining a wake loss matrix of a wind direction structure of a wind farm through calculation, wherein the wake loss matrix comprises the following components:

Taking each fan as a wind measuring tower, taking the data series selected in the step one as input, and calculating the average wake loss of 16 wind direction sectors of N wind turbines of the wind farm by a complex terrain wind farm wake model generating capacity calculation method to obtain a wake loss matrix W= (wl _ij)_N×16, forming N rows and 16 columns of average wake loss matrixes, wherein wl represents the wake loss, wl _ij represents the average wake loss of the ith wind turbine in j wind direction sectors;

Step three: according to the wind direction structure wake loss matrix calculation method of the wind farm in the second step, when wind directions 0, 1, 2, …, 14 and 15 sectors appear on the representative wind machine, average wake loss matrixes of 16 wind direction sectors of the N sets of the wind farm are W0, W1, …, wk, …, W14 and W15.

5. The method for restoring free wind speed to operational wind speed of a wind farm with complex terrain according to claim 4, wherein the wake loss matrix achievement consistency check and correction method in step 4 comprises:

If the absolute value of the difference between wake losses of the same unit before and after the sector wind direction is calculated within a given error range, the wake loss results are judged to be consistent, the error range is set to be 0.5%, and the wake loss inspection judgment of the unit a in the sector of the wind direction is assumed: w _{Front part ab}-W _{Rear part (S) ab}|×p≤0.5％,W _{Front part ab} represents the last calculated wake loss, W _{Rear part (S) ab} represents the current calculated wake loss, and p represents the frequency of occurrence of the a-unit in the b-wind direction sector.

6. The method for restoring free wind speed in wind farm operation in complex terrain according to claim 5, wherein the wake loss matrix achievement consistency check and correction specifically comprises the following steps:

Step one: selecting a wind speed and wind direction data series according to the wind direction of the representative fan; for N wind turbines of a wind power plant, each wind turbine runs data on average wind speed and wind direction for 10 minutes in the whole year, when M wind directions appear on a representative fan, a data series of average wind speed and wind direction for 10 minutes of N wind turbines of the wind power plant is selected, and a wake loss matrix W _M＝(wl_ij)_N×16 is calculated;

Step two: l wind areas divided by wind power are selected, a wind direction sector with the largest frequency of occurrence of a machine position is selected in each wind area, and consistency test is carried out; except for representing fans, selecting a1, a2, a3, and an aL fan, wherein when the wind direction sector of the representing fan is M, the number of sectors with the largest occurrence frequency of each selected machine position is M1, M2, M3, … and ML respectively;

Step three: when a No. 1 fan generates a wind direction sector M1, a wind power plant N fans are selected for 10min to obtain an average wind speed and wind direction data series;

Step four: according to the data series selected in the step three, a wind direction sector frequency matrix of each fan is calculated in a statistics mode, P= (P _ij)_N×16, (i=1, N; j=0, 15) P represents the frequency of the occurrence of the wind direction of the j-th sector, and P _ij represents the frequency of the occurrence of the wind direction of the j-th sector of the i-th unit;

step five: wind direction structure wake loss matrix calculation:

Each fan is used as a wind measuring tower, the data series selected in the fourth step are used as input, the average wake loss of 16 wind direction sectors of the N wind power generation sets of the wind power plant is calculated based on a complex terrain wind power plant wake loss model generating capacity calculation method, so that a wake loss matrix WW= (wwl _ij)_N×16 is formed into N rows and 16 columns of average wake loss matrixes, wherein wwl represents the wake loss, wwl _ij represents the average wake loss of the ith wind power generation set in j wind direction sectors;

step six: consistency check of the wake loss achievement of the wind direction sector number M1 of the a1 fan:

(1) In the WM matrix, only keeping wake loss of the wind direction sector number M1 for the a1 fan, and giving 0 value to wake loss of other wind direction sectors, wherein W _M＝(wl_ij)_N×16 is changed into WF _M＝(wf_ij)_N×16;

(2) Calculating a wake loss deviation matrix; forming a wake loss deviation matrix WC= (WC _ij)_N×16) on the WF _M matrix and the WW matrix according to the calculation formula WC _ij＝|wf_ij-wwl_ij|×p_ij;

(3) If each element WC _ij of the wake loss deviation matrix WC is less than or equal to 0.5%, the wake loss result consistency check meets the requirement, and step eight is entered, otherwise, the a1 fan wind direction structure is subdivided;

step seven: subdividing a1 fan wind direction structure, correcting wake loss deviation matrix, including:

on the basis of the first step, selecting wind direction sectors with the frequency of more than 10% of wind direction occurrence for an a1 fan, wherein the wind direction sectors are M _{a1 1}、M_{a1 2}、...、M_{a1 L-1} respectively, and other sectors form a combined sector M _{a1 L};

Respectively selecting a10 min average wind speed and wind direction data series of N fans of a wind power plant according to the number of the sectors of the wind direction of the a1 fan as M _{a1 1}、M_{a1 2}、...、M_{a1 L};

Respectively calculating average wake loss matrixes of 16 wind direction sectors of N units of the wind power plant according to the number M _{a1 1}、M_{a1 2}、...、M_{a1 L} of wind direction sectors of the a1 fan to obtain a wake loss matrix W _{a1 1},W_{a1 2}、...、W_{a1 L}, and replacing wake loss results W _M＝(wl_ij)_N×16 representing the fan M sectors with the results;

Step eight: according to a similar method, carrying out wake loss result consistency check on fans a2, a3, the first and the second, and when the wake loss matrix result consistency check is not satisfied, subdividing a wind direction structure and correcting results;

Step nine: the results of the 16 wind direction sectors representing the fans are checked and corrected at one time according to the steps.

7. The method for restoring free wind speed to operational wind speed of a wind farm with complex terrain according to claim 6, wherein the step 5 comprises:

step one: converting the degree wind directions of the wind direction series of N wind units of the wind power plant into sector wind directions;

Step two: when representing that a fan has a wind direction sector k, selecting the average wind speed, wind direction data and active power series of N fans of the wind power plant for 10 minutes;

Step three: selecting wake loss matrixes W _k＝(wl_ij)_N×16 of 16 wind direction sectors of the N wind motor groups of the wind farm when the fan generates a wind direction k sector, and adopting a correction result if wake correction results of a subdivided wind area structure exist;

step four: for the selected data time series of the average wind speed, wind direction and active power of 10min of running wind of each wind turbine, when the active power is greater than 0 in the mth time period, checking a wake loss matrix according to the machine set number and the wind direction, for example, checking wake loss wl _Lc of the L wind turbine with the wind direction of the c-th sector in the wake loss matrix Wk according to a calculation formula, wherein the wind direction of the L wind turbine is the c-th sector in the m time period of the L wind turbine with the wind speed of V _{Transport and transport ml} Restoring to a free wind speed V _{Self-supporting ml} which is not influenced by wake flow;

step five: when the representative fan wind direction sector k is 0, 1, 2, …, 14 and 15 respectively, the running wind speed, wind direction and active power series of each wind turbine are selected respectively, and the running wind speed is reduced to the free wind speed according to the second step to the fourth step.

8. A system for restoring free wind speed from complex terrain wind farm operating wind speed, comprising a wind speed restoration module for performing a method for restoring free wind speed from complex terrain wind farm operating wind speed as claimed in any of claims 1 to 7.

9. A non-transitory computer readable storage medium storing computer instructions which, when executed by a processor, implement a method of restoring free wind speed from complex terrain wind farm operating wind speed as claimed in any of claims 1 to 7.

10. An electronic device, comprising:

a memory and a processor, the memory and the processor being communicatively connected to each other, the memory having stored therein computer instructions, the processor executing the computer instructions to perform a method of restoring free wind speed from operating wind speeds of a complex terrain wind farm as claimed in any of claims 1 to 7.