Wind farm wake prediction method based on improved physical information neural network

Abstract

The present invention discloses a wind farm wake prediction method based on an improved physical information neural network. The method constructs a basic physical information neural network that integrates a laser wind radar module, a two-dimensional Navier-Stokes equation, and an actuator disk model. The basic physical information neural network is trained using a cosine annealing algorithm and a dynamic loss function weight, and a stepwise time-saving strategy is introduced to construct an improved physical information neural network. A training dataset is constructed based on line-of-sight wind speed data from sparse measurement points of the laser wind radar, and the improved physical information neural network is trained to obtain a wind farm wake prediction model. Spatial position data and time data are input to obtain predicted wind farm wake wind speed and air pressure. The improved physical information neural network in the present invention combines measurement data with physical laws and, by adopting a precision improvement strategy and a stepwise time-saving strategy, achieves accurate and efficient prediction of the entire wind farm wake.

Description

Technical Field

[0001] The present invention relates to wind power generation and deep learning technology, and in particular to a wind farm wake prediction method based on an improved physical information neural network. Background Art

[0002] As one of the most important renewable energy sources, wind power generation technology has been rapidly developing worldwide. With the expansion of wind farms and the development of larger wind turbines, wind energy losses due to the wake effect have become increasingly serious. Therefore, in order to mitigate the impact of the wake effect and improve the efficiency of wind turbines in utilizing wind energy, accurate and efficient wake prediction is of great significance for the optimized design and control of wind farms. The wake effect refers to the turbulent area formed behind the upstream wind turbine after it extracts wind energy from the windward side. This results in a decrease in wind speed and increased turbulence received by the downstream wind turbine, thereby reducing the wind turbine's power generation efficiency and increasing fatigue losses.

[0003] Currently, traditional wake prediction models based on deep learning technology require a large amount of data for model training. However, in practical applications, it is often difficult to obtain the required amount of data, which hinders its widespread application in engineering practice.

[0004] LiDAR technology has been widely used in the wind power sector due to its advantages, including high measurement accuracy, high temporal and spatial resolution, wide detection range, and fast response speed. However, LiDAR can only measure the apparent wind speed and direction at sparse locations along the laser beam, and cannot provide detailed information about the wind turbine wake.

[0005] Therefore, there is an urgent need for a technical solution that can use the line-of-sight wind speed data of sparse measurement points of laser wind radar to achieve accurate and efficient prediction of wind farm wakes. Summary of the Invention

[0006] Purpose of the invention: The purpose of the present invention is to solve the deficiencies in the prior art and to provide a wind farm wake prediction method based on an improved physical information neural network.

[0007] Technical Solution: The present invention provides a wind farm wake prediction method based on an improved physical information neural network. A wind farm wake prediction model is constructed. The input data of the wind farm wake prediction model includes spatial position data and time data, and the output data of the wind farm wake prediction model is the wind farm wake wind speed. The wind farm wake prediction model includes an improved physical information neural network and integrates a laser wind measurement radar module, a two-dimensional Navier-Stokes equation, and an actuator disk model. The laser wind measurement radar module is combined with sparse measurement data of the laser wind measurement radar, and the two-dimensional Navier-Stokes equation and the actuator disk model serve as physical law constraints. The specific method is as follows:

[0008] Step 1: For the basic physical information neural network, a cosine annealing algorithm and a dynamic loss function weighting strategy are used to integrate the laser wind radar module, the two-dimensional Navier-Stokes equations, and the actuator disk model to construct an improved physical information neural network.

[0009] Step 2: Construct a training data set and introduce a step-by-step time saving strategy to train the improved physical information neural network obtained in step 1 to obtain a wind farm wake prediction model. During the training process, the laser wind radar module calculates the line-of-sight wind speed u LoS The training dataset includes the line-of-sight wind speed data from sparse measurement points of the laser wind radar.

[0010] Step 3: Input spatial position data and time data into the trained wind farm wake prediction model, that is, input the space-time coordinates (x, y, t), and obtain the predicted wind farm wake directional wind speed u, spanwise wind speed v, and air pressure p.

[0011] The loss function of the laser wind radar module in the present invention Essentially, it refers to the comparison between the predicted apparent wind speed and the measured apparent wind speed; one of the goals in the neural network training process is to minimize the difference between the predicted apparent wind speed and the measured apparent wind speed, and its ultimate goal is to achieve the total loss function minimize.

[0012] Furthermore, the two-dimensional Navier-Stokes equation is expressed as:

[0013]

[0014] Where, represents partial derivative operation; u represents the streamwise wind speed; v represents the spanwise wind speed; t represents the prediction time; x represents the streamwise position; y represents the spanwise position; ρ represents the air density; p represents the air pressure; υ represents the kinematic viscosity of the air;

[0015] The expression of the actuator disc model is:

[0016]

[0017] Where x f Indicates the flow position of the center of the actuator disc; y f is the spanwise position of the actuator disc center; f(x f ,y f ) means that the coordinates of the center position are (x f ,y f ) of the actuator disk; N elem is the total number of discrete elements of the actuator disk; f irepresents the volume force at the discrete element of the i-th actuator disk; ε is the smoothing parameter of the Gaussian kernel; exp(·) represents the exponential function with the mathematical constant e as the base; is the flow direction position of the discrete element of the i-th actuator disk; represents the spanwise position of the discrete element of the i-th actuator disk.

[0018] Furthermore, through the loss function The laser wind radar module, two-dimensional Navier-Stokes equations and actuator disk model are integrated into the basic physical information neural network, and the loss function The formula is:

[0019]

[0020] Where w1 represents the loss function weight of the laser wind radar module; is the loss function of the laser wind radar module; w2 represents the loss function weight of the Navier-Stokes equation; is the loss function of the Navier-Stokes equation; w3 represents the loss function weight of the actuator disk model; is the loss function of the actuator disk model; finally, the neural network training is performed with the goal of minimizing the loss function value.

[0021] Furthermore, the loss function is weighted by the dynamic loss function strategy. Make adaptive adjustments so that the loss function The three weights in are dynamically updated through gradient ascent, and the expressions are as follows:

[0022]

[0023] Where, and Represent the loss function weights of the laser wind radar module in the k+1th and kth training rounds respectively; is the learning rate when the weight w1 is updated; Indicates the partial derivative of the loss function with respect to the weight w1; is the total loss function; Θ k is the training parameter of the neural network in the kth training round; and Represent the loss function weights of the Navier-Stokes equation at the k+1th and kth training rounds respectively; is the learning rate when the weight w2 is updated; Indicates the partial derivative of the loss function with respect to the weight w2; and Represent the loss function weights of the actuator disk model at the k+1th and kth training rounds respectively; Indicates the partial derivative of the loss function with respect to the weight w3.

[0024] Furthermore, the use of the cosine annealing algorithm is a warm restart stochastic gradient descent method. The cosine annealing algorithm is used to gradually reduce the learning rate in each warm start cycle during the training of the improved physical information neural network, thereby improving the performance of the improved physical information neural network. The formula is expressed as:

[0025]

[0026] Where η j represents the learning rate of the jth training round; η max is the maximum learning rate; η min is the minimum learning rate; T cur Indicates the number of rounds the neural network has trained since the warm restart; T e It indicates the number of training times required between two adjacent warm restarts. Its initial value is T0. After each warm restart, the coefficient T mult Expand, that is, T e+1 =T mult ·T e .

[0027] Furthermore, the specific method of introducing a gradual time-saving strategy to train the improved physical information neural network in step 2 is as follows: the entire training data set is decomposed into multiple sub-training sets according to the time domain, and each sub-training set is gradually trained by the same improved physical information neural network according to the time sequence; except for the sub-training set used in the initial training, the subsequent sub-training sets will be combined with the previous sub-training set data to train the improved physical information neural network; after each step of training is completed, the improved physical information neural network will use the training parameter information as the initial condition for the next step of training.

[0028] Furthermore, the visual wind speed u predicted by the laser wind radar module during training is LoS The calculation formula is:

[0029] u LoS =u train cos(θ)-v train sin(θ)(10)

[0030] Where u LoS is the predicted wind speed in the direction of the laser wind radar; u train is the flow direction wind speed output during the neural network training process; v train is the spanwise wind speed output during the neural network training process; θ is the azimuth angle of the laser wind radar measurement beam.

[0031] Beneficial effects: The present invention introduces two precision improvement strategies, namely the cosine annealing algorithm and the dynamic loss function weight, and a gradual time saving strategy, into the basic physical information neural network to form an improved physical information neural network, thereby achieving more accurate and more efficient prediction of wind farm wakes; in addition, the present invention integrates the wind speed data constraints of sparse measurement points of the laser wind radar, the two-dimensional Navier-Stokes equation constraints, and the actuator disk model constraints, and can achieve accurate and efficient prediction of wind farm wakes by combining measurement data with physical laws. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a flow chart of the overall method of the present invention.

[0033] Figure 2 This is a layout diagram of a wind farm calculation example in an embodiment of the present invention.

[0034] Figure 3 1 is a diagram of a neural network structure in an embodiment of the present invention.

[0035] Figure 4 3 is a comparison diagram of the actual wake and the wake predicted by the neural network in an embodiment of the present invention. DETAILED DESCRIPTION

[0036] The technical solution of the present invention is described in detail below, but the protection scope of the present invention is not limited to the embodiments.

[0037] like Figure 1 As shown, the present invention provides a wind farm wake prediction method based on an improved physical information neural network, constructing a wind farm wake prediction model. The input data of the wind farm wake prediction model includes spatial position data and time data, and the output data of the wind farm wake prediction model is the wind farm wake wind speed. The wind farm wake prediction model includes an improved physical information neural network and integrates a laser wind measurement radar module, a two-dimensional Navier-Stokes equation, and an actuator disk model. The laser wind measurement radar module combines sparse measurement data of the laser wind measurement radar, and the two-dimensional Navier-Stokes equation and the actuator disk model serve as physical law constraints. The specific method is as follows:

[0038] Step 1: For the basic physical information neural network, a cosine annealing algorithm and a dynamic loss function weighting strategy are used to integrate the laser wind radar module, the two-dimensional Navier-Stokes equations, and the actuator disk model to construct an improved physical information neural network.

[0039] Step 2: Construct a training data set and introduce a step-by-step time saving strategy to train the improved physical information neural network obtained in step 1 to obtain a wind farm wake prediction model. During the training process, the laser wind radar module calculates the line-of-sight wind speed uLoS The training dataset includes the line-of-sight wind speed data from sparse measurement points of the laser wind radar.

[0040] Step 3: Input spatial position data and time data (space-time coordinates (x, y, t)) into the trained wind farm wake prediction model to obtain the predicted wind speed u, wind speed v and air pressure p of the wind farm wake.

[0041] The spatial position data input to the wind farm wake prediction model in the present invention includes the two-dimensional coordinates (x, y) of the wake, such as Figure 2 As shown in the horizontal and vertical coordinates in ; the time data input to the wind farm wake prediction model includes time t. For example, if a wake of 400s needs to be predicted, the time data t is from 1s to 400s; finally, the spatial position data and time data are input together into the wind farm wake prediction model in the form of space-time coordinates (x, y, t), and then the wake flow direction wind speed u, span direction wind speed v and air pressure p of the corresponding coordinate point (x, y) at time t are directly output to the wind farm wake prediction model; the air pressure p is output because there is a term for finding the partial derivative of the air pressure p in the two-dimensional Navier-Stokes equation.

[0042] The two-dimensional Navier-Stokes equation in this embodiment is expressed as:

[0043]

[0044] Where, represents the partial derivative operation; u is the streamwise wind speed; v is the spanwise wind speed; t is the prediction time; x is the streamwise position; y is the spanwise position; ρ is the air density; p is the air pressure; and υ is the air kinematic viscosity. The expression of the actuator disk model is:

[0045]

[0046] Where x f Indicates the flow position of the center of the actuator disc; y f is the spanwise position of the actuator disc center; f(x f ,y f ) means that the coordinates of the center position are (x f ,y f ) of the actuator disk; N elem is the total number of discrete elements of the actuator disk; f i represents the volume force at the discrete element of the i-th actuator disk; ε is the smoothing parameter of the Gaussian kernel; exp(·) represents the exponential function with the mathematical constant e as the base; is the flow direction position of the discrete element of the i-th actuator disk; represents the spanwise position of the discrete element of the i-th actuator disk.

[0047] Step 1: Through the loss function The laser wind radar module, two-dimensional Navier-Stokes equations and actuator disk model are integrated into the basic physical information neural network, and the loss function The formula is:

[0048]

[0049] Where w1 represents the loss function weight of the laser wind radar module; is the loss function of the laser wind radar module; w2 represents the loss function weight of the Navier-Stokes equation; is the loss function of the Navier-Stokes equation; w3 represents the loss function weight of the actuator disk model; is the loss function of the actuator disk model.

[0050] In order to improve the prediction accuracy, the loss function is further adjusted by the dynamic loss function weight strategy. Make adaptive adjustments so that the loss function The three weights in are dynamically updated through gradient ascent, and the expressions are as follows:

[0051]

[0052] Where, and Represent the loss function weights of the laser wind radar module in the k+1th and kth training rounds respectively; is the learning rate when the weight w1 is updated; Indicates the partial derivative of the loss function with respect to the weight w1; is the total loss function; Θ k is the training parameter of the neural network in the kth training round; and Represent the loss function weights of the Navier-Stokes equation at the k+1th and kth training rounds respectively; is the learning rate when the weight w2 is updated; Indicates the partial derivative of the loss function with respect to the weight w2; and Represent the loss function weights of the actuator disk model at the k+1th and kth training rounds respectively; Indicates the partial derivative of the loss function with respect to the weight w3.

[0053] To further optimize the learning rate, the cosine annealing algorithm is used to gradually reduce the learning rate during each warm start cycle of the improved physical information neural network training. The expression is:

[0054]

[0055] Where η j represents the learning rate of the jth training round; η max is the maximum learning rate; η min is the minimum learning rate; T cur Indicates the number of rounds the neural network has trained since the warm restart; T e It indicates the number of training times required between two adjacent warm restarts. Its initial value is T0. After each warm restart, the coefficient T mult Expand, that is, T e+1 =T mult ·T e .

[0056] In order to reduce the training time and make the neural network model more efficient, step 2 also introduces a gradual time-saving strategy to train the improved physical information neural network. The specific method is: the entire training data set is decomposed into multiple sub-training sets according to the time domain, and each sub-training set is gradually trained by the same improved physical information neural network according to the time sequence; except for the sub-training set used in the initial training, the subsequent sub-training sets will be combined with the previous sub-training set data to train the improved physical information neural network; after each step of training is completed, the improved physical information neural network will use the training parameter information as the initial condition for the next step of training.

[0057] The visual wind speed u predicted by the laser wind radar module during the training process of this embodiment is LoS The calculation formula is:

[0058] u LoS =u train cos(θ)-v train sin(θ)

[0059] Where u LoS is the predicted wind speed in the direction of the laser wind radar; u train is the flow direction wind speed output during the neural network training process; v train is the spanwise wind speed output during the neural network training process; θ is the azimuth angle of the laser wind radar measurement beam.

[0060] After completing the above steps, the streamwise and spanwise wind speeds of the wake field are obtained through simulation calculation (regarded as the real streamwise and spanwise wind speed data). Then, the required laser wind radar data is extracted from the obtained simulation data and a data set is constructed. Then, a wind farm wake prediction model is obtained through neural network training. Finally, the streamwise and spanwise wind speeds of the wake are predicted.

[0061] When evaluating the prediction accuracy, the wake vortex flow direction wind speed and span direction wind speed predicted by the trained wake prediction model are compared with the flow direction wind speed and span direction wind speed (real wind speed data) of the wake field obtained by simulation calculation (the difference between the two).

[0062] The relevant prediction accuracy evaluation indicators are:

[0063]

[0064] Where u MRMSE represents the average value of the root mean square error of the wind speed; T is the total prediction time, N pred Indicates the number of predicted spatial location points; (x k ,y k ,t) is the spatial position data and time data of the prediction point; u pred (x k ,y k ,t) represents the neural network's response to the spatial position point (x k ,y k ) is the predicted value of the stream wind speed at time t; u true (x k ,y k ,t) represents the spatial position point (x k ,y k ) the true value of the streamwise wind speed at time t; v MRMSE represents the average value of the root mean square error of the spanwise wind speed; v pred (x k ,y k ,t) represents the neural network's response to the spatial position point (x k ,y k ) is the predicted spanwise wind speed at time t; v true (x k ,y k ,t) represents the spatial position point (x k ,y k ) is the true value of the spanwise wind speed at time t.

[0065] To verify the effectiveness of the technical solution of the present invention, this embodiment uses a wind farm containing six NREL 5MW wind turbines as the research object and applies the technical solution of the present invention. In the wind farm of this embodiment, the diameter of the wind turbine rotor is D = 126.4m, the flow spacing of the wind turbines is 5D, and the span spacing is 3D. Figure 2 shown.

[0066] This example uses the FAST.Farm medium-fidelity multi-physics engineering simulation tool, and performs a 700-second simulation with an inflow of 10 m / s and 6% turbulence. The laser wind radar is configured with a measurement range of 550 m and a 30-m interval between adjacent measurement points. The laser beam scans the plane at hub height at a speed of 3.5° / s within a 42° angle. Figure 2 The black dots in the middle are schematic diagrams of sparse measurement points of the laser wind radar. To ensure that the wake is fully formed and stable, the wind speed data at the hub height plane is saved from the last 400 seconds of simulation as the true wind speed value. Based on the laser wind radar configuration, the line-of-sight wind speed data of the laser wind radar sparse measurement points is extracted from the saved wind speed data to construct a neural network training dataset. The training dataset is used to train the neural network based on the step-by-step time saving strategy. Figure 3 The neural network model shown is trained.

[0067] The technical solution of the present invention is compared with the existing physical information neural network to improve the wind farm wake prediction accuracy and training speed (see Table 1 for the results), and the comparison between the real wake results and the wake results predicted by the method of the present invention is given, as shown in the figure. Figure 4 As shown. Among them, Figure 4 (a) in the figure is the wind farm wake obtained by simulation calculation; Figure 4 (b) in the figure shows the wake field map predicted by the basic physical information neural network on the left, and the right figure shows the error map of the wake field predicted by the basic physical information neural network. Figure 4 In (c), the left picture is the wake field map predicted by the improved physical information neural network, and the right picture is the wake field error map predicted by the improved physical information neural network.

[0068] The above results show that the method proposed in the present invention can achieve accurate and efficient prediction of wind farm wakes based on sparse wind speed data of laser wind radar measurement points.

[0069] Table 1 Comparison of prediction accuracy and training time of neural networks

[0070]

[0071] In summary, the improved physical information neural network method in the present invention combines the sparse measurement data of the laser wind radar with physical laws, and achieves accurate and efficient prediction of the entire wind farm wake by adopting a precision improvement strategy and a gradual time saving strategy.

Claims

1. A wind farm wake prediction method based on an improved physical information neural network, characterized in that: A wind farm wake prediction model is constructed. The input data of the wind farm wake prediction model includes spatial position data and time data, and the output data of the wind farm wake prediction model is the wind speed of the wind farm wake. The wind farm wake prediction model includes an improved physical information neural network and integrates a laser wind measurement radar module, a two-dimensional Navier-Stokes equation, and an actuator disk model. The laser wind measurement radar module combines sparse measurement data of the laser wind measurement radar, and the two-dimensional Navier-Stokes equation and the actuator disk model as physical law constraints. The specific method is as follows: Step 1: For the basic physical information neural network, a cosine annealing algorithm and a dynamic loss function weighting strategy are used to integrate the laser wind radar module, the two-dimensional Navier-Stokes equations, and the actuator disk model to construct an improved physical information neural network. Step 2: Construct a training data set and introduce a step-by-step time saving strategy to train the improved physical information neural network obtained in step 1 to obtain a wind farm wake prediction model. During the training process, the laser wind radar module predicts and calculates the line-of-sight wind speed u. LoS The training dataset includes the line-of-sight wind speed data from sparse measurement points of the laser wind radar. Step 3: Input spatial position data and time data into the trained wind farm wake prediction model, that is, input the space-time coordinates (x, y, t), and obtain the predicted wind farm wake directional wind speed u, spanwise wind speed v, and air pressure p.

2. The wind farm wake prediction method based on the improved physical information neural network according to claim 1 is characterized in that: The two-dimensional Navier-Stokes equation is expressed as: Where, represents the partial derivative operation; u is the wind speed in the stream direction; v is the wind speed in the span direction; t is the predicted time; x is the position in the stream direction; y is the position in the span direction; ρ is the air density; p is the air pressure; υ is the kinematic viscosity of the air; The expression of the actuator disc model is: Where x f Indicates the flow position of the center of the actuator disc; y f is the spanwise position of the actuator disc center; f(x f ,y f ) means that the coordinates of the center position are (x f ,y f ) of the actuator disk; N elem is the total number of discrete elements of the actuator disk; f i represents the volume force at the discrete element of the i-th actuator disk; ε is the smoothing parameter of the Gaussian kernel; exp(·) represents the exponential function with the mathematical constant e as the base; is the flow direction position of the discrete element of the i-th actuator disk; represents the spanwise position of the discrete element of the i-th actuator disk.

3. The wind farm wake prediction method based on the improved physical information neural network according to claim 1 is characterized in that: Through the loss function The laser wind radar module, two-dimensional Navier-Stokes equations and actuator disk model are integrated into the basic physical information neural network, and the loss function The formula is: Where w1 represents the loss function weight of the laser wind radar module; is the loss function of the laser wind radar module; w2 represents the loss function weight of the Navier-Stokes equation; is the loss function of the Navier-Stokes equation; w3 represents the loss function weight of the actuator disk model; is the loss function of the actuator disk model.

4. The wind farm wake prediction method based on the improved physical information neural network according to claim 3 is characterized in that: The loss function is weighted by the dynamic loss function Make adaptive adjustments so that the loss function The three weights in are dynamically updated through gradient ascent, and the expressions are as follows: Where, and Represent the loss function weights of the laser wind radar module in the k+1th and kth training rounds respectively; is the learning rate when the weight w1 is updated; Indicates the partial derivative of the loss function with respect to the weight w1; is the total loss function; Θ k is the training parameter of the neural network in the kth training round; and Represent the loss function weights of the Navier-Stokes equation at the k+1th and kth training rounds respectively; is the learning rate when the weight w2 is updated; Indicates the partial derivative of the loss function with respect to the weight w2; and Represent the loss function weights of the actuator disk model at the k+1th and kth training rounds respectively; Indicates the partial derivative of the loss function with respect to the weight w3.

5. The wind farm wake prediction method based on improved physical information neural network according to claim 1 or 3, characterized in that: The cosine annealing algorithm is used to gradually reduce the learning rate in each warm start cycle during the training of the improved physical information neural network. The expression is: Where η j represents the learning rate of the jth training round; η max is the maximum learning rate; η min is the minimum learning rate; T cur Indicates the number of rounds the neural network has trained since the warm restart; T e It indicates the number of training times required between two adjacent warm restarts. Its initial value is T0. After each warm restart, the coefficient T mult Expand, that is, T e+l =T mult ·T e .

6. The wind farm wake prediction method based on improved physical information neural network according to claim 1 is characterized in that: The specific method of introducing the step-by-step time saving strategy to train the improved physical information neural network in step 2 is as follows: the entire training data set is decomposed into multiple sub-training sets according to the time domain, and each sub-training set is gradually trained by the same improved physical information neural network according to the time sequence; except for the sub-training set used in the initial training, the subsequent sub-training sets will be combined with the previous sub-training set data to train the improved physical information neural network; After each step of training is completed, the improved physical information neural network uses the training parameter information as the initial condition for the next step of training.

7. The wind farm wake prediction method based on improved physical information neural network according to claim 1 is characterized in that: The visual wind speed u predicted by the laser wind radar module during training LoS The calculation formula is: you LoS =u train coS(θ)-v train sin(θ) Where u LoS is the predicted wind speed in the direction of the laser wind radar; u train is the flow direction wind speed output during the neural network training process; v train is the spanwise wind speed output during the neural network training process; θ is the azimuth angle of the laser wind radar measurement beam.

Images