Engineering wake flow model dynamic correction method and system based on actually measured wind speed of wind power plant

Abstract

The invention discloses an engineering wake flow model dynamic correction method and system based on the actually measured wind speed of a wind power plant, and the method comprises the steps: constructing a dynamic wake flow model according to the time delay characteristic of wake flow propagation, and building an optimization model; performing first-order Taylor expansion on a dynamic single wake flow model and a dynamic superposition model in the dynamic wake flow model to obtain an increment form of the dynamic wake flow model, and forming a discrete state space matrix of a historical wind speed predicted value; and according to the discrete state space matrix of the historical wind speed predicted value, converting the optimization model into a discrete form, setting constraint conditions of a wake flow expansion coefficient and a superposition correction coefficient, converting the constraint conditions into a standard quadratic programming problem, and solving to obtain an optimal model parameter so as to realize dynamic real-time correction of the wake flow model. According to the method, the parameters of the engineering wake flow model are dynamically optimized according to the historical wind speed measured value, so that the calculation precision of the engineering wake flow model is improved, and meanwhile, the wake flow calculation speed and timeliness are guaranteed.

Description

Technical Field

[0001] The present invention belongs to the technical field of wind power generation, and in particular relates to a method and system for dynamically correcting an engineering wake model based on the measured wind speed of a wind farm. Background Art

[0002] As the scale of offshore wind farms continues to grow and the number of turbines increases, the wake effect in the field becomes more and more obvious, which has an impact on the economic operation of offshore wind farms that cannot be ignored. The existence of the wake effect leads to complex aerodynamic coupling characteristics between upstream and downstream wind turbines, which must be considered when optimizing the operation of offshore wind farms.

[0003] In order to characterize the dynamic characteristics of the wake model, mathematical and physical methods are often used to model it to obtain the wake model. According to the different methods used, the wake model is divided into two categories: one is the numerical wake model obtained by computational fluid dynamics method, and the other is the engineering wake model expressed by mathematical formula. The numerical wake model can accurately characterize the dynamic characteristics of the wake effect, but it requires a lot of computing time and computing resources and is not suitable for engineering applications. The engineering wake model can quickly calculate the wake distribution, but the calculation accuracy is limited. It is difficult for existing wake models to balance the calculation speed and calculation accuracy, which poses a challenge to the optimization control of large-scale wind farm operations considering the wake effect.

[0004] At present, some studies have optimized the parameters of the engineering wake model based on statistical methods or data-driven methods to improve the calculation accuracy on the basis of meeting the extremely fast calculation speed of the engineering wake model. However, statistical methods often require the establishment of complex optimization models, which themselves require a long calculation time to optimize parameters, and cannot use real-time measured wind speed data to achieve dynamic optimization. Data-driven methods do not have explicit models and are difficult to apply in wind farm optimization control.

[0005] Therefore, in the current field of engineering wake model correction, it is necessary to further consider balancing calculation speed and accuracy, using real-time measured wind speed data for dynamic correction, etc., so as to provide a more suitable method and system for the dynamic correction of engineering wake models based on wind farm wind speed measured data. Summary of the invention

[0006] Technical problem to be solved by the present invention: In view of the above-mentioned problems in the prior art, a method and system for dynamically correcting an engineering wake model based on the actually measured wind speed of a wind farm are provided. The present invention aims to improve the calculation accuracy of the engineering wake model by dynamically optimizing the parameters of the engineering wake model according to the actually measured values ​​of historical wind speeds, while ensuring the speed and timeliness of the wake calculation.

[0007] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0008] A method for dynamically correcting an engineering wake model based on measured wind speed in a wind farm comprises:

[0009] S1. According to the time delay characteristics of wake propagation, the traditional Jensen wake model and the energy conservation wake superposition model are introduced with the wake delay time to establish a dynamic single wake model and a dynamic superposition model, and the superposition correction coefficient α is introduced into the dynamic superposition model to obtain a dynamic wake model for wind speed correction constituted by the dynamic superposition model and the dynamic single wake model;

[0010] S2. Based on the dynamic wake model, the historical wind speed forecast value is calculated through the historical measured values ​​of the inflow wind speed, the historical thrust coefficient, and the wake expansion coefficient and superposition correction coefficient of the Jensen wake model at the previous moment. The wake expansion coefficient and superposition correction coefficient of the engineering wake model at the current moment are corrected using the first N historical wind speed measured values ​​and historical wind speed forecast values ​​of the organic groups in the wind farm, and the optimization model is established;

[0011] S3, performing first-order Taylor expansion on the dynamic single wake model and the dynamic superposition model in the dynamic wake model to obtain the incremental form of the dynamic wake model and form a discrete state space matrix of the historical wind speed prediction value;

[0012] S4. According to the discrete state space matrix of the historical wind speed prediction value, the optimization model is converted into a discrete form, and the constraints of the wake expansion coefficient and the superposition correction coefficient are set to convert them into a standard quadratic programming problem. The standard quadratic programming problem is solved to obtain the optimal model parameters to realize dynamic real-time correction of the Jensen wake model.

[0013] Optionally, the function expressions for establishing the dynamic single wake model and the dynamic superposition model in step S1 are:

[0014]

[0015] In the above formula, is the predicted value of the wake wind speed caused by wind turbine i to wind turbine n at time t0, is the predicted value of the inflow wind speed of the upstream wind turbine i at time t0, C t,i (t0) is the thrust coefficient of wind turbine i at time t0, k i (t0) is the expansion coefficient of the wake generated by wind turbine i at time t0, d in is the geographical distance between the upstream wind turbine i and the downstream wind turbine n along the wind direction, R is the impeller radius of the wind turbine, is the impeller coverage area of ​​wind turbine n affected by the wake effect caused by wind turbine i, A0 is the swept area of ​​the wind turbine impeller, is the wind turbine n at t0+τ 1,n Wind speed at the time, v ∞ (t0) is the inflow wind speed of the most upstream wind turbine at time t0, α n (t0) is the superposition correction coefficient introduced at time t0, is wind turbine i at t0+τ 1,n The wind speed at the moment, is the ratio of wind turbine i to wind turbine n at t0+τ 1,n The predicted value of the wake wind speed at time τ 1,n is the wake delay time between the most upstream wind turbine 1 and the downstream wind turbine n, and τ 1,n =d in / v ∞ , v ∞ is the inflow wind speed of the most upstream wind turbine, and Γ is the set of upstream turbines that cause wake impact on wind turbine n.

[0016] Optionally, in step S2, the dynamic wake model based on the historical inflow wind speed measured value, the historical thrust coefficient and the wake expansion coefficient of the Jensen wake model at the previous moment and the superposition correction coefficient are used to calculate the historical wind speed prediction value, including the N historical inflow wind speed measured values ​​v before time t0. ∞ (t0-1)~v ∞ (t0-N) and its corresponding historical thrust coefficient C t (t0-1)~

[0017] C t (t0-N), N is the size of the historical window, and the wake expansion coefficient and superposition correction coefficient of the Jensen wake model at the previous moment are combined to calculate the previous N historical wind speed forecast values, and any previous T-th historical wind speed forecast value V pre,n The function expression of (t0-T) is:

[0018]

[0019] In the above formula, C t,i (t0-T) is the Tth historical thrust coefficient, k i (t0-T) is the wake expansion coefficient of the upstream wind turbine i, d in is the geographical distance between the upstream wind turbine i and the downstream wind turbine n along the wind direction, R is the impeller radius of the wind turbine, is the impeller coverage area of ​​wind turbine n affected by the wake effect caused by wind turbine i, A0 is the swept area of ​​the wind turbine impeller, v ∞ (t0-T) is the measured value of the previous T-th historical inflow wind speed, α n(t0-T) is the first Tth superposition correction coefficient of the downstream wind turbine n, V pre,i (t0-T+τ i,n ) is the wind turbine i at t0-T+τ 1,n The historical wind speed forecast value at time, is the ratio of wind turbine i to wind turbine n at t0-T+τ 1,n The predicted value of the wake wind speed at time τ 1,n is the wake delay time between the most upstream wind turbine 1 and the downstream wind turbine n.

[0020] Optionally, the optimization model function expression established in step S2 is:

[0021]

[0022] In the above formula, N wt is the number of wind turbines, V pre,i (t0-T+τ i,n ) is the wind turbine i at t0-T+τ 1,n The historical wind speed forecast value at time V meas,i (t0) is the historical wind speed measured value of wind turbine i at time t0, τ 1,n is the wake delay time between the most upstream wind turbine 1 and the downstream wind turbine n.

[0023] Optionally, when performing first-order Taylor expansion on the dynamic single wake model and the dynamic superposition model in the dynamic wake model in step S3, the function expression of performing first-order Taylor expansion on the dynamic single wake model is:

[0024]

[0025] In the above formula, and are the predicted values ​​of the wake wind speed caused by wind turbine i to wind turbine n. The expansion coefficient k of the wake generated by wind turbine i i , the predicted value of the inflow wind speed of upstream wind turbine i The partial derivative of Δk i (t0) and They are k i (t0) and The increment of the dynamic superposition model is a first-order Taylor expansion function expression:

[0026]

[0027] In the above formula, and are the predicted values ​​of the inflow wind speed of upstream wind turbine n. Predicted value of the wake wind speed caused by wind turbine i to wind turbine n Predicted value of inflow wind speed of upstream wind turbine i And the superposition correction factor α introduced n The partial derivative of is the wind turbine n at t0+τ 1,n The wind speed increment at the time, Δα n (t0) is the increment of the superposition correction coefficient at time t0, is the ratio of wind turbine i to wind turbine n at t0-T+τ 1,n The predicted value increment of the wake wind speed caused by time, is wind turbine i at t0+τ 1,n The wind speed increment at that time, is the ratio of wind turbine i to wind turbine n at t0-T+τ 1,,n The predicted value of the wake wind speed at time τ 1,n is the wake delay time between the most upstream wind turbine 1 and the downstream wind turbine n; the function expression of the incremental form of the dynamic wake model is obtained as follows:

[0028]

[0029] In the above formula, is the wind speed prediction increment matrix at time t0, Δx(t0) is the wake model parameter increment matrix at time t0, M(t0) and N(t0) are the coefficient matrices at time t0, and:

[0030]

[0031] Δx(t0)=[Δk(t0),Δα(t0)] T ,

[0032] Δk(t0)=[Δk1(t0),Δk2(t0),…,Δk n-1 (t0)] T ,

[0033] Δα(t0)=[Δα3(t0),Δα4(t0),…,Δα n (t0)] T ,

[0034] M(t0)=[M1(t0),M2(t0),…,M n (t0)] T ,

[0035] N(t0)=[N1(t0),N2(t0),…,N n (t0)]T ,

[0036] In the above formula, They are wind turbines 2~n at t0+τ 1,2 ~t0+τ 1,n The wind speed increment at time t0, Δk(t0) and Δα(t0) are the increments of the wake expansion coefficient and correction coefficient at time t0, Δk1(t0)~Δk n-1 (t0) is the increment of the wake expansion coefficient of the wake effect caused by unit 1 to unit n-1 at time t0, Δα3(t0)~Δα n (t0) is the correction coefficient increment of the wake superposition model of unit 3 to unit n at time t0, M1(t0)~M n (t0) is an element in the coefficient matrix M(t0), N1(t0)~N n (t0) is an element in the coefficient matrix N(t0), and:

[0037]

[0038] In the above formula, M i and N i are the i-th element in M(t0) and N(t0) respectively; the function expression of the discrete state space matrix of the historical wind speed prediction value is finally formed as follows:

[0039]

[0040] In the above formula, is a matrix composed of the wind speed prediction increment matrix at t0-1 to t0-N, I is an N×(n-1)-dimensional unit matrix, N and M are matrices composed of the coefficient matrices at t0-1 to t0-N, Δx is the optimization variable, and:

[0041]

[0042] M=[M(t0-1),M(t0-2),…,M(t0-N)] T ,

[0043] N=diag[N(t0-1),N(t0-2),…,N(t0-N)],

[0044] In the above formula, It is the incremental matrix of wind speed prediction at time t0-1~t0-N, and M(t0-1)~M(t0-N) and N(t0-1)~N(t0-N) are the coefficient matrices at time t0-1~t0-N.

[0045] Optionally, in step S4, the function expression of the discrete form in which the optimization model is converted is:

[0046]

[0047] In the above formula, N is the size of the historical window, Γ is the set of upstream units that cause wake impact on wind turbine n, is wind turbine i at t0-T+τ 1,n The wind speed increment at the time, V pre,i (t0-T+τ i,n ) is the wind turbine i at t0-T+τ 1,n The historical wind speed forecast value at time V meas,i (t0-T) is the historical wind speed measured value of wind turbine i at time t0-T, Δx(t0) and Δx(t0-1) are the optimization variables at time t0 and t0-1 respectively, Q v and Q x are weight parameters respectively.

[0048] Optionally, the function expression of the constraint conditions of the wake expansion coefficient and the stacking correction coefficient set in step S4 is:

[0049]

[0050] In the above formula, k init is the initial value of the wake expansion coefficient, Δk i (t0) is the increment of wake expansion coefficient at time t0, k i (t0-1) is the wake expansion coefficient at time t0-1, Δα i (t0) is the increment of the superposition correction coefficient at time t0, α i (t0-1) is the superposition correction coefficient at time t0-1.

[0051] In addition, the present invention also provides a system for dynamically correcting an engineering wake model based on the measured wind speed of a wind farm, comprising a microprocessor and a memory connected to each other, wherein the microprocessor is programmed or configured to execute the method for dynamically correcting an engineering wake model based on the measured wind speed of a wind farm.

[0052] In addition, the present invention also provides a computer-readable storage medium, in which a computer program or instruction is stored. The computer program or instruction is programmed or configured to execute the dynamic correction method of the engineering wake model based on the measured wind speed of the wind farm through a processor.

[0053] In addition, the present invention also provides a computer program product, including a computer program or instructions, which are programmed or configured to execute the method for dynamically correcting the engineering wake model based on the measured wind speed of the wind farm through a processor.

[0054] Compared with the prior art, the present invention mainly has the following advantages: the present invention realizes dynamic and real-time optimization of engineering wake model parameters by utilizing the historical wind speed measured values ​​of large-scale offshore wind farms, while taking into account the extremely fast calculation speed of the engineering wake model, the calculation accuracy is improved, and the wind speed prediction effect of the engineering wake model is more accurate. The present invention converts the engineering wake model into an incremental form through a first-order Taylor expansion, and then converts the high-dimensional and high-nonlinear optimization model into a standard quadratic programming problem, which can ensure the speed of solving the optimization problem, meet the real-time performance of the wind farm operation optimization control, and ensure the speed and timeliness of the wake calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 Schematic diagram of the basic flow of the method of the embodiment of the present invention.

[0056] Figure 2 FIG. 4 is a structural diagram of a simulation system in a simulation case according to an embodiment of the present invention.

[0057] Figure 3 This is a simulation verification diagram of the engineering wake model correction of an embodiment of the present invention.

[0058] Figure 4 This is a diagram of parameter optimization results of the engineering wake model correction according to an embodiment of the present invention. DETAILED DESCRIPTION

[0059] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0060] like Figure 1 As shown, the method for dynamically correcting the engineering wake model based on the measured wind speed of the wind farm in this embodiment includes:

[0061] S1. According to the time delay characteristics of wake propagation, the traditional Jensen wake model and the energy conservation wake superposition model are introduced with the wake delay time to establish a dynamic single wake model and a dynamic superposition model, and the superposition correction coefficient α is introduced into the dynamic superposition model to obtain a dynamic wake model for wind speed correction constituted by the dynamic superposition model and the dynamic single wake model;

[0062] S2. Based on the dynamic wake model, the historical wind speed forecast value is calculated through the historical measured values ​​of the inflow wind speed, the historical thrust coefficient, and the wake expansion coefficient and superposition correction coefficient of the Jensen wake model at the previous moment. The wake expansion coefficient and superposition correction coefficient of the engineering wake model at the current moment are corrected using the first N historical wind speed measured values ​​and historical wind speed forecast values ​​of the organic groups in the wind farm, and the optimization model is established;

[0063] S3, performing first-order Taylor expansion on the dynamic single wake model and the dynamic superposition model in the dynamic wake model to obtain the incremental form of the dynamic wake model and form a discrete state space matrix of the historical wind speed prediction value;

[0064] S4. According to the discrete state space matrix of the historical wind speed prediction value, the optimization model is converted into a discrete form, and the constraints of the wake expansion coefficient and the superposition correction coefficient are set to convert them into a standard quadratic programming problem. The standard quadratic programming problem is solved to obtain the optimal model parameters to realize dynamic real-time correction of the Jensen wake model.

[0065] According to the average inflow wind speed V ∞ The geographical distance d along the wind direction from the upstream wind turbine i and the downstream wind turbine n in Calculate the wake delay time τ in , the specific function expression is: τ in =d in / v ∞ The function expression for establishing the dynamic single wake model and the dynamic superposition model in step S1 of this embodiment is:

[0066]

[0067] In the above formula, is the predicted value of the wake wind speed caused by wind turbine i to wind turbine n at time t0, is the predicted value of the inflow wind speed of the upstream wind turbine i at time t0, C t,i (t0) is the thrust coefficient of wind turbine i at time t0, k i (t0) is the expansion coefficient of the wake generated by wind turbine i at time t0, d in is the geographical distance between the upstream wind turbine i and the downstream wind turbine n along the wind direction, R is the impeller radius of the wind turbine, is the impeller coverage area of ​​wind turbine n affected by the wake effect caused by wind turbine i, A0 is the swept area of ​​the wind turbine impeller, is the wind turbine n at t0+τ 1,n Wind speed at the time, v ∞ (t0) is the inflow wind speed of the most upstream wind turbine at time t0, α n (t0) is the superposition correction coefficient introduced at time t0, is wind turbine i at t0+τ 1,n The wind speed at the moment, is the ratio of wind turbine i to wind turbine n at t0+τ 1,n The predicted value of the wake wind speed at time τ 1,n is the wake delay time between the most upstream wind turbine 1 and the downstream wind turbine n, and τ1,n =d in / v ∞ , v ∞ is the inflow wind speed of the most upstream wind turbine, and Γ is the set of upstream turbines that have wake effects on wind turbine n. The ^ at the top of the variable character represents the wind speed prediction value calculated by the dynamic wake model, for example represents the predicted value of the wake wind speed caused by wind turbine i to wind turbine n, represents the predicted value of the inflow wind speed of the upstream wind turbine i. The obtained dynamic single wake model and dynamic superposition model together constitute the dynamic wake model for wind speed correction.

[0068] The dynamic wake model based on step S2 of this embodiment calculates the historical wind speed prediction value by using the historical inflow wind speed measured value, the historical thrust coefficient, the wake expansion coefficient of the Jensen wake model at the previous moment, and the superposition correction coefficient, including the N historical inflow wind speed measured values ​​v before time t0. ∞ (t0-1)~v ∞ (t0-N) and its corresponding historical thrust coefficient C t (t0-1)~

[0069] C t (t0-N), N is the size of the historical window, and the wake expansion coefficient and superposition correction coefficient of the Jensen wake model at the previous moment are combined to calculate the previous N historical wind speed forecast values, and any previous T-th historical wind speed forecast value V pre,n The function expression of (t0-T) is:

[0070]

[0071] In the above formula, C t,i (t0-T) is the Tth historical thrust coefficient, k i (t0-T) is the wake expansion coefficient of the upstream wind turbine i, d in is the geographical distance between the upstream wind turbine i and the downstream wind turbine n along the wind direction, R is the impeller radius of the wind turbine, is the impeller coverage area of ​​wind turbine n affected by the wake effect caused by wind turbine i, A0 is the swept area of ​​the wind turbine impeller, v ∞ (t0-T) is the measured value of the previous T-th historical inflow wind speed, α n (t0-T) is the first Tth superposition correction coefficient of the downstream wind turbine n, V pre,i (t0-T+τ i,n ) is the wind turbine i at t0-T+τ 1,n The historical wind speed forecast value at time, is the ratio of wind turbine i to wind turbine n at t0-T+τ1,n The predicted value of the wake wind speed at time τ 1,n is the wake delay time between the most upstream wind turbine 1 and the downstream wind turbine n.

[0072] Based on the dynamic wake model, the historical measured values ​​of inflow wind speed (V ∞ (t0-1)~V ∞ (t0-N)), historical thrust coefficient (C t (t0-1)~C t The historical wind speed forecast value (V pre (t0-1)~V pre (t0-N)). The previous N historical wind speed measured values ​​of the wind farm’s organic units (V meas (t0-1)~V meas The wake expansion coefficient k(t0) and k(t0) of the engineering wake model at the current moment are corrected by using the historical wind speed prediction value to establish an optimization model. The function expression of the optimization model established in step S2 of this embodiment is:

[0073]

[0074] In the above formula, N is the number of historical wind speed measurements used for correction (correction step length), wt is the number of wind turbines, V pre,i (t0-T+τ i,n ) is the wind turbine i at t0-T+τ 1,n The historical wind speed forecast value at time V meas,i (t0) is the historical wind speed measured value of wind turbine i at time t0, τ 1,n is the wake delay time between the most upstream wind turbine 1 and the downstream wind turbine n.

[0075] In step S3 of this embodiment, when the first-order Taylor expansion is performed on the dynamic single wake model and the dynamic superposition model in the dynamic wake model, the function expression of the first-order Taylor expansion of the dynamic single wake model is:

[0076]

[0077] In the above formula, and are the predicted values ​​of the wake wind speed caused by wind turbine i to wind turbine n. The expansion coefficient k of the wake generated by wind turbine i i , the predicted value of the inflow wind speed of upstream wind turbine i The partial derivative of Δki (t0) and They are k i (t0) and increment.

[0078] In step S3 of this embodiment, the function expression of the first-order Taylor expansion of the dynamic superposition model is:

[0079]

[0080] In the above formula, and are the predicted values ​​of the inflow wind speed of upstream wind turbine n. Predicted value of the wake wind speed caused by wind turbine i to wind turbine n Predicted value of inflow wind speed of upstream wind turbine i And the superposition correction factor α introduced n The partial derivative of is the wind turbine n at t0+τ 1,n The wind speed increment at the time, Δα n (t0) is the increment of the superposition correction coefficient at time t0, is the ratio of wind turbine i to wind turbine n at t0-T+τ 1,n The predicted value increment of the wake wind speed caused by time, is wind turbine i at t0+τ 1,n The wind speed increment at that time, is the ratio of wind turbine i to wind turbine n at t0-T+τ 1,n The predicted value of the wake wind speed at time τ 1,n is the wake delay time between the most upstream wind turbine 1 and the downstream wind turbine n;

[0081] The incremental form of the dynamic wake model can be obtained by the above two first-order Taylor expansion function expressions. In this embodiment, the incremental form of the dynamic wake model is obtained as follows:

[0082]

[0083] In the above formula, is the wind speed prediction increment matrix at time t0, Δx(t0) is the wake model parameter increment matrix at time t0, M(t0) and N(t0) are the coefficient matrices at time t0, and:

[0084]

[0085] Δx(t0)=[Δk(t0),Δα(t0)] T ,

[0086] Δk(t0)=[Δk1(t0),Δk2(t0),…,Δk n-1 (t0)] T ,

[0087] Δα(t0)=[Δα3(t0),Δα4(t0),…,Δα n (t0)] T ,

[0088] M(t0)=[M1(t0),M2(t0),…,M n (t0)] T ,

[0089] N(t0)=[N1(t0),N2(t0),…,N n (t0)] T ,

[0090] In the above formula, They are wind turbines 2~n at t0+τ 1,2 ~t0+τ 1,n The wind speed increment at time t0, Δk(t0) and Δα(t0) are the increments of the wake expansion coefficient and correction coefficient at time t0, Δk1(t0)~Δk n-1 (t0) is the increment of the wake expansion coefficient of the wake effect caused by unit 1 to unit n-1 at time t0, Δα3(t0)~Δα n (t0) is the correction coefficient increment of the wake superposition model of unit 3 to unit n at time t0, M1(t0)~M n (t0) is an element in the coefficient matrix M(t0), N1(t0)~N n (t0) is an element in the coefficient matrix N(t0), and:

[0091]

[0092] In the above formula, M i and N i are the i-th elements in M(t0) and N(t0) respectively;

[0093] Using N historical wind speed measurements to correct the model, the discrete state space matrix of the historical wind speed forecast values ​​from t0-1 to t0-N is written according to the above formula. The function expression of the discrete state space matrix of the historical wind speed forecast values ​​is finally formed as follows:

[0094]

[0095] In the above formula, is a matrix composed of the wind speed prediction increment matrix at t0-1 to t0-N, I is an N×(n-1)-dimensional unit matrix, N and M are matrices composed of the coefficient matrices at t0-1 to t0-N, Δx is the optimization variable, and:

[0096]

[0097] M=[M(t0-1),M(t0-2),…,M(t0-N)] T ,

[0098] N=diag[N(t0-1),N(t0-2),…,N(t0-N)],

[0099] In the above formula, It is the incremental matrix of wind speed prediction at time t0-1~t0-N, and M(t0-1)~M(t0-N) and N(t0-1)~N(t0-N) are the coefficient matrices at time t0-1~t0-N.

[0100] The optimization model can be converted into a discrete form based on the discrete state space matrix. In step S4 of this embodiment, the function expression of the discrete form into which the optimization model is converted is:

[0101]

[0102] In the above formula, N is the size of the historical window, Γ is the set of upstream units that cause wake impact on wind turbine n, is wind turbine i at t0-T+τ 1,n The wind speed increment at the time, V pre,i (t0-T+τ i,n ) is the wind turbine i at t0-T+τ 1,n The historical wind speed forecast value at time V meas,i (t0-T) is the historical wind speed measured value of wind turbine i at time t0-T, Δx(t0) and Δx(t0-1) are the optimization variables at time t0 and t0-1 respectively, Q v and Q x are weight parameters respectively.

[0103] In order to ensure the rationality of the optimization variable Δx, the constraints of the wake expansion coefficient k and the stacking correction coefficient α are set. Specifically, the function expressions of the constraints of the wake expansion coefficient and the stacking correction coefficient set in step S4 of this embodiment are:

[0104]

[0105] In the above formula, k init is the initial value of the wake expansion coefficient, Δk i(t0) is the increment of wake expansion coefficient at time t0, k i (t0-1) is the wake expansion coefficient at time t0-1, Δα i (t0) is the increment of the superposition correction coefficient at time t0, α i (t0-1) is the superposition correction coefficient at time t0-1. Based on the discrete form of the optimization model and the constraints of the optimization variables, the optimization model is converted into a standard quadratic programming problem, and the optimal wake expansion coefficient k and superposition correction coefficient α are solved to achieve dynamic real-time correction of the engineering wake model.

[0106] In order to verify the effectiveness of the method for dynamically correcting the engineering wake model based on the wind speed measured in the wind farm in this embodiment, a dynamic correction simulation system for the engineering wake model based on the wind speed measured data of the wind farm was built by using SimWindFarm software and MATLAB / Simulink software in this embodiment. Figure 2 As shown in the figure. This simulation system regards SimWindFarm as a real offshore wind farm, including 6 NREL 5MW wind turbines arranged along the wind direction, with a spacing of 630 meters between each wind turbine. The output inflow wind speed of each turbine is used as the actual wind speed data of the wind farm.

[0107] In order to compare the wake model correction effects under different average wind speeds and different turbulence intensities, low wind speed (10m / s) and low turbulence (3%), low wind speed and high turbulence (10%), high wind speed (12m / s) and low turbulence, and high wind speed and high turbulence environments are used as simulation conditions. The effect of the dynamic correction of the engineering wake model can be reflected by the measured wind speed values ​​of the wind farm and the calculated values ​​of the engineering wake model. Taking wind turbines 1 to 6 as the research objects, the comparison results of the measured wind speed values ​​of the wind farm, the calculated wind speed values ​​of the corrected engineering wake model, and the calculated wind speed values ​​of the engineering wake model before correction are shown in the following figure. Figure 3 , Table 1 and Table 2, the parameters of the modified engineering wake model are as follows Figure 4 As shown, Figure 4 (a) is the correction result of the wake expansion coefficient, k1~k5 are the wake expansion coefficients of wind turbines 1~5; Figure 4 (b) is the correction result of the superposition correction coefficient, and a3~a6 are the superposition correction coefficients of wind turbines 3~6.

[0108] Table 1: Root mean square error between the calculated and measured wind speeds of wind turbines 1 to 6

[0109] Correction step length Fan 2 Fan 3 Fan 4 Fan 5 Fan 6 average value Calculation time 30 0.114 0.0129 0.0744 0.178 0.126 0.101 0.0136 60 0.105 0.0562 0.0137 0.0639 0.0126 0.05 0.0264 120 0.0338 0.048 0.0232 0.0145 0.0183 0.0276 0.0736 300 0.031 0.0167 0.0204 0.0166 0.0234 0.0216 0.443 600 0.0213 0.0158 0.0252 0.0183 0.0286 0.0218 2.528 No modification 0.398 0.442 0.392 0.331 0.297 0.372 /

[0110] In Table 1, the units of correction step size and calculation time are both seconds.

[0111] Table 2: Average value of the root mean square error between the calculated wind speed and the measured wind speed for all wind turbines

[0112] environment 30 60 120 300 600 No modification Low wind speed and low turbulence 0.102 0.0479 0.0236 0.0146 0.0149 0.372 Low wind speed and high turbulence 0.101 0.0502 0.0276 0.0216 0.0218 0.372 High wind speed and low turbulence 0.236 0.145 0.0919 0.0628 0.0614 0.459 High wind speed and high turbulence 0.365 0.309 0.227 0.159 0.157 0.463

[0113] In Table 2, 30, 60, 120, 300 and 600 are the correction step sizes.

[0114] pass Figure 3 and Figure 4 It can be seen that the wind speed calculation accuracy of the engineering wake model corrected by the dynamic correction method of the engineering wake model based on the measured wind speed of the wind farm in this embodiment is higher. Under different correction step sizes, the root mean square error between the calculated wind speed values ​​and the measured wind speed values ​​of wind turbines 1 to 6 is shown in Table 1. It can be seen from Table 1 that the smaller the correction step size, the shorter the calculation time, but the worse the correction effect. Under four different environments and different correction step sizes, the average value of the root mean square error between the calculated wind speed values ​​and the measured wind speed values ​​of all wind turbines in the wind farm is shown in Table 2. It can be seen that higher average wind speed and turbulence intensity mean greater natural wind speed fluctuations, so low wind speed and low turbulence environments have better correction effects. In addition, the larger the correction step size, the better the correction effect.

[0115] In summary, the method for dynamically correcting the engineering wake model based on the measured wind speed of the wind farm in this embodiment includes introducing a wake delay time into the traditional Jensen wake model and the energy conservation wake superposition model according to the time delay characteristics of wake propagation, establishing a dynamic single wake model and a dynamic superposition model, and in order to further improve the correction accuracy of the engineering wake model, introducing a superposition correction coefficient α into the dynamic superposition model, which together with the dynamic single wake model constitutes a dynamic wake model for wind speed correction; based on the established dynamic wake model, the measured value of the historical inflow wind speed (V ∞ (t0-1)~V ∞ (t0-N)), historical thrust coefficient (C t (t0-1)~C t The historical wind speed prediction values ​​(Vpre(t0-1)~Vpre(t0-N)) are calculated by using the wake expansion coefficient k(t0-1) and the correction coefficient α(t0-1) of the engineering wake model at the previous moment, and the previous N historical wind speed measured values ​​(Vpre(t0-1)~Vpre(t0-N)) of the wind farm organic groups are used. meas (t0-1)~V meas(t0-N)) and the historical wind speed prediction value are used to correct the wake expansion coefficient k(t0) and of the engineering wake model at the current moment, and an optimization model is established; a first-order Taylor expansion is performed on the established dynamic single wake model, and a first-order Taylor expansion is performed on the established dynamic superposition model at the same time, to obtain an incremental form of the dynamic wake model, and a discrete state space matrix of the historical wind speed prediction value is formed; according to the discrete state space matrix of the obtained historical wind speed prediction value, the obtained optimization model is converted into a discrete form, and the constraints of the wake expansion coefficient k and the superposition correction coefficient α are set, and the problem is converted into a standard quadratic programming problem, and the optimal wake model parameters are solved to realize the dynamic real-time correction of the engineering wake model. The dynamic correction method of the engineering wake model based on the measured wind speed of the wind farm in this embodiment has the advantages of high correction accuracy and high real-time performance.

[0116] In addition, the present embodiment also provides a dynamic correction system for an engineering wake model based on the measured wind speed of a wind farm, comprising a microprocessor and a memory connected to each other, wherein the microprocessor is programmed or configured to execute the dynamic correction method for the engineering wake model based on the measured wind speed of the wind farm. Corresponding to the above correction method, the dynamic correction system for an engineering wake model based on the measured wind speed of the wind farm in the present embodiment also has the advantages described in the dynamic correction method for the engineering wake model based on the measured wind speed of the wind farm. In addition, the present embodiment also provides a dynamic correction system for an engineering wake model based on the measured wind speed of a wind farm, comprising a data acquisition and storage system and a solver, wherein a computer program is loaded on the solver, and the computer program executes the dynamic correction method for the engineering wake model based on the measured wind speed of the wind farm as described above when the solver is running. Corresponding to the above correction method, the dynamic correction system for an engineering wake model based on the measured wind speed of the wind farm in the present embodiment also has the advantages described in the dynamic correction method for the engineering wake model based on the measured wind speed of the wind farm.

[0117] In addition, the present embodiment further provides a computer-readable storage medium, in which a computer program or instruction is stored, and the computer program or instruction is programmed or configured to execute the method for dynamically correcting an engineering wake model based on the measured wind speed of a wind farm through a processor. In addition, the present embodiment further provides a computer program product, including a computer program or instruction, and the computer program or instruction is programmed or configured to execute the method for dynamically correcting an engineering wake model based on the measured wind speed of a wind farm through a processor.

[0118] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, may be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the functions in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 These computer program instructions can also be stored in a computer-readable memory that can guide a computer or other programmable data processing device to work in a specific way, so that the instructions stored in the computer-readable memory produce a product including an instruction device, which implements the functions specified in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide for implementing the process in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0119] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A method for dynamically correcting an engineering wake model based on the measured wind speed of a wind farm, characterized in that: include: S1. According to the time delay characteristics of wake propagation, the traditional Jensen wake model and the energy conservation wake superposition model are introduced with the wake delay time to establish a dynamic single wake model and a dynamic superposition model, and the superposition correction coefficient α is introduced into the dynamic superposition model to obtain a dynamic wake model for wind speed correction constituted by the dynamic superposition model and the dynamic single wake model; S2. Based on the dynamic wake model, the historical wind speed forecast value is calculated through the historical measured values ​​of the inflow wind speed, the historical thrust coefficient, and the wake expansion coefficient and superposition correction coefficient of the Jensen wake model at the previous moment. The wake expansion coefficient and superposition correction coefficient of the engineering wake model at the current moment are corrected using the first N historical wind speed measured values ​​and historical wind speed forecast values ​​of the organic groups in the wind farm, and the optimization model is established; S3, performing first-order Taylor expansion on the dynamic single wake model and the dynamic superposition model in the dynamic wake model to obtain the incremental form of the dynamic wake model and form a discrete state space matrix of the historical wind speed prediction value; S4. According to the discrete state space matrix of the historical wind speed prediction value, the optimization model is converted into a discrete form, and the constraints of the wake expansion coefficient and the superposition correction coefficient are set to convert them into a standard quadratic programming problem. The standard quadratic programming problem is solved to obtain the optimal model parameters to realize dynamic real-time correction of the Jensen wake model.

2. The method for dynamically correcting an engineering wake model based on the measured wind speed of a wind farm according to claim 1, characterized in that: The function expressions for establishing the dynamic single wake model and the dynamic superposition model in step S1 are: In the above formula, is the predicted value of the wake wind speed caused by wind turbine i to wind turbine n at time t0, is the predicted value of the inflow wind speed of the upstream wind turbine i at time t0, C t,i (t0) is the thrust coefficient of wind turbine i at time t0, k i (t0) is the expansion coefficient of the wake generated by wind turbine i at time t0, d in is the geographical distance between the upstream wind turbine i and the downstream wind turbine n along the wind direction, R is the impeller radius of the wind turbine, is the impeller coverage area of ​​wind turbine n affected by the wake effect caused by wind turbine i, A0 is the swept area of ​​the wind turbine impeller, is the wind turbine n at t0+τ 1,n Wind speed at the time, v ∞ (t0) is the inflow wind speed of the most upstream wind turbine at time t0, α n (t0) is the superposition correction coefficient introduced at time t0, is wind turbine i at t0+τ 1,n The wind speed at the moment, is the ratio of wind turbine i to wind turbine n at t0+τ 1,n The predicted value of the wake wind speed at time τ 1,n is the wake delay time between the most upstream wind turbine 1 and the downstream wind turbine n, and τ 1,n =d in / v ∞ , v ∞ is the inflow wind speed of the most upstream wind turbine, and Γ is the set of upstream turbines that cause wake impact on wind turbine n.

3. The method for dynamically correcting an engineering wake model based on the measured wind speed of a wind farm according to claim 2 is characterized in that: In step S2, the dynamic wake model is based on the historical wind speed measured value, the historical thrust coefficient and the wake expansion coefficient of the Jensen wake model at the previous moment, and the superposition correction coefficient. The historical wind speed forecast value is calculated by using the N historical inflow wind speed measured values ​​v before time t0. ∞ (t0-1)~v ∞ (t0-N) and its corresponding historical thrust coefficient C t (t0-1)~C t (t0-N), N is the size of the historical window, and the wake expansion coefficient and superposition correction coefficient of the Jensen wake model at the previous moment are combined to calculate the previous N historical wind speed forecast values, and any previous T-th historical wind speed forecast value V pre,n The function expression of (t0-T) is: In the above formula, C t,i (t0-T) is the Tth historical thrust coefficient, k i (t0-T) is the wake expansion coefficient of the upstream wind turbine i, d in is the geographical distance between the upstream wind turbine i and the downstream wind turbine n along the wind direction, R is the impeller radius of the wind turbine, is the impeller coverage area of ​​wind turbine n affected by the wake effect caused by wind turbine i, A0 is the swept area of ​​the wind turbine impeller, v ∞ (t0-T) is the measured value of the previous T-th historical inflow wind speed, α n (t0-T) is the first Tth superposition correction coefficient of the downstream wind turbine n, V pre,i (t0-T+τ i,n ) is the wind turbine i at t0-T+τ 1,n The historical wind speed forecast value at time, is the ratio of wind turbine i to wind turbine n at t0-T+τ 1,n The predicted value of the wake wind speed at time τ 1,n is the wake delay time between the most upstream wind turbine 1 and the downstream wind turbine n.

4. The method for dynamically correcting an engineering wake model based on the measured wind speed of a wind farm according to claim 3 is characterized in that: The optimization model function expression established in step S2 is: In the above formula, N wt is the number of wind turbines, V pre,i (t0-T+τ i,n ) is the wind turbine i at t0-T+τ 1,n The historical wind speed forecast value at time V meas,i (t0) is the historical wind speed measured value of wind turbine i at time t0, τ 1,n is the wake delay time between the most upstream wind turbine 1 and the downstream wind turbine n.

5. The method for dynamically correcting an engineering wake model based on the measured wind speed of a wind farm according to claim 1, characterized in that: When the dynamic single wake model and the dynamic superposition model in the dynamic wake model are subjected to first-order Taylor expansion in step S3, the function expression of the first-order Taylor expansion of the dynamic single wake model is: In the above formula, and are the predicted values ​​of the wake wind speed caused by wind turbine i to wind turbine n. The expansion coefficient k of the wake generated by wind turbine i i , the predicted value of the inflow wind speed of upstream wind turbine i The partial derivative of Δk i (t0) and They are k i (t0) and The increment of the dynamic superposition model is a first-order Taylor expansion function expression: In the above formula, and are the predicted values ​​of the inflow wind speed of upstream wind turbine n. Predicted value of the wake wind speed caused by wind turbine i to wind turbine n Predicted value of inflow wind speed of upstream wind turbine i And the superposition correction factor α introduced n The partial derivative of is the wind turbine n at t0+τ 1,n The wind speed increment at the time, Δα n (t0) is the increment of the superposition correction coefficient at time t0, is the ratio of wind turbine i to wind turbine n at t0-T+τ 1,n The predicted value increment of the wake wind speed caused by time, is wind turbine i at t0+τ 1,n The wind speed increment at that time, is the ratio of wind turbine i to wind turbine n at t0-T+τ 1,n The predicted value of the wake wind speed at time τ 1,n is the wake delay time between the most upstream wind turbine 1 and the downstream wind turbine n; the function expression of the incremental form of the dynamic wake model is obtained as follows: In the above formula, is the wind speed prediction increment matrix at time t0, Δx(t0) is the wake model parameter increment matrix at time t0, M(t0) and N(t0) are the coefficient matrices at time t0, and: Δx(t0)=[Δk(t0),Δα(t0)] T , Δk(t0)=[Δk1(t0),Δk2(t0),…,Δk n-1 (t0)] T , Δα(t0)=[Δα3(t0),Δα4(t0),…,Δα n (t0)] T , M(t0)=[M1(t0),M2(t0),…,M n (t0)] T , N(t0)=[N1(t0),N2(t0),…,N n (t0)] T , In the above formula, They are wind turbines 2~n at t0+τ 1,2 ~t0+τ 1,n The wind speed increment at time t0, Δk(t0) and Δα(t0) are the increments of the wake expansion coefficient and correction coefficient at time t0, Δk1(t0)~Δk n-1 (t0) is the increment of the wake expansion coefficient of the wake effect caused by unit 1 to unit n-1 at time t0, Δα3(t0)~Δα n (t0) is the correction coefficient increment of the wake superposition model of unit 3 to unit n at time t0, M1(t0)~M n (t0) is an element in the coefficient matrix M(t0), N1(t0)~N n (t0) is an element in the coefficient matrix N(t0), and: In the above formula, M i and N i are the i-th element in M(t0) and N(t0) respectively; the function expression of the discrete state space matrix of the historical wind speed prediction value is finally formed as follows: In the above formula, is a matrix composed of the wind speed prediction increment matrix at t0-1 to t0-N, I is an N×(n-1)-dimensional unit matrix, N and M are matrices composed of the coefficient matrices at t0-1 to t0-N, Δx is the optimization variable, and: M=[M(t0-1),M(t0-2),…,M(t0-N)] T , N=diag[N(t0-1),N(t0-2),…,N(t0-N)], In the above formula, It is the incremental matrix of wind speed prediction at time t0-1~t0-N, and M(t0-1)~M(t0-N) and N(t0-1)~N(t0-N) are the coefficient matrices at time t0-1~t0-N.

6. The method for dynamically correcting an engineering wake model based on the measured wind speed of a wind farm according to claim 5, characterized in that: The function expression of the discrete form in which the optimization model is converted in step S4 is: In the above formula, N is the size of the historical window, Γ is the set of upstream units that cause wake impact on wind turbine n, is wind turbine i at t0-T+τ 1,n The wind speed increment at time, V pre,i (t0-T+τ i,n ) is the wind turbine i at t0-T+τ 1,n The historical wind speed forecast value at time V meas,i (t0-T) is the historical wind speed measured value of wind turbine i at time t0-T, Δx(t0) and Δx(t0-1) are the optimization variables at time t0 and t0-1 respectively, Q v and Q x are weight parameters respectively.

7. The method for dynamically correcting an engineering wake model based on the measured wind speed of a wind farm according to claim 6, characterized in that: The function expression of the constraint conditions of the wake expansion coefficient and the stacking correction coefficient set in step S4 is: In the above formula, k init is the initial value of the wake expansion coefficient, Δk i (t0) is the increment of wake expansion coefficient at time t0, k i (t0-1) is the wake expansion coefficient at time t0-1, Δα i (t0) is the increment of the superposition correction coefficient at time t0, α i (t0-1) is the superposition correction coefficient at time t0-1.

8. A dynamic correction system for an engineering wake model based on the measured wind speed of a wind farm, comprising a microprocessor and a memory connected to each other, characterized in that: The microprocessor is programmed or configured to execute the method for dynamically correcting an engineering wake model based on the measured wind speed of a wind farm as described in any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program or instruction stored therein, characterized in that: The computer program or instruction is programmed or configured to execute the method for dynamically correcting an engineering wake model based on the measured wind speed of a wind farm as described in any one of claims 1 to 7 through a processor.

10. A computer program product comprising a computer program or instructions, characterized in that The computer program or instruction is programmed or configured to execute the method for dynamically correcting an engineering wake model based on the measured wind speed of a wind farm as described in any one of claims 1 to 7 through a processor.

Images