Wind field correction method and system applied to single-station remote sensing device

Abstract

Disclosed in the present invention are a wind field correction method and system applied to a single-station remote sensing device. The method comprises the following specific steps: S1, acquiring wind field data by means of a data information acquisition apparatus, and obtaining radial wind speed information of a current acquisition point by means of a data processing module; S2, on the basis of the hypothesis of an uniform wind field, using a classical wind speed reconstruction algorithm for inversion to obtain a three-dimensional wind speed vector of the acquisition point; S3, establishing an error model; and S4, on the basis of an error factor β calculated by means of the error model, performing data correction on the three-dimensional wind speed vector of the acquisition point obtained by means of inversion in S2. Compared with the prior art, the present invention provides a high-precision correction method, which can be operated on low-end hardware, and can also be directly installed and operated on hardware of a measurement device, so that measurement results can be corrected in real time, no additional complexity can be brought to users, and costs are reduced; and the present invention is suitable for solving complex system problems when accurate boundary conditions cannot be obtained.

Description

A wind field correction method and system for single-station remote sensing equipment Technical Field

[0001] This invention relates to the field of wind field correction, and specifically to a wind field correction method and system applied to a single-station remote sensing device. Background Technology

[0002] A single-station remote sensing device (RSD) transmits multiple signals into space, which interact with aerosol particles in the atmosphere. A portion of the backscattered signal is received by the RSD's receiving system and processed to extract wind speed information represented by the atmospheric aerosol particles. Due to the Doppler effect, the particle motion component in the signal transmission-reception direction causes a frequency shift in the scattered signal; therefore, the analyzed wind speed is only the projection component of the wind vector along the radial direction of the signal. To evaluate the complete three-dimensional wind speed vector, radial wind speed information must be provided at least three different line-of-sight directions.

[0003] For single-station remote sensing devices, since all signal transmission points are the same, signals from different directions are transmitted to different locations in space. The fundamental assumption for wind field inversion with all single-station remote sensing devices is that the measured wind field is identical at each line-of-sight location at a given distance from the transmission point. Based on this assumption, using the geometric relationship of projecting the same vector along different directions, a system equation containing at least three unknowns can be generated, thus allowing the inversion of the three-dimensional wind speed vector using the measured radial velocity. As shown in the following equation, the measured radial wind velocity... and the corresponding projection angle It can inversely calculate the three-dimensional wind speed vector. :

[0004] .

[0005] This wind field inversion method requires the assumption that the wind field at which radial wind velocities measured in multiple directions are located has good homogeneity. This is also a necessary prerequisite for ensuring the accuracy of the wind field inversion results obtained by this method.

[0006] The method of obtaining wind speed through a single-station remote sensing measurement device and performing wind field inversion requires the assumption that the wind field where the radial wind speed measured in multiple directions is located has good uniformity; this is also a necessary prerequisite to ensure the accuracy of the wind field inversion results of this method.

[0007] The assumption of uniformity in wind field reconstruction is valid in simple sites with flat terrain and wind fields conforming to atmospheric boundary layer theory. However, when there is non-uniformity in airflow from one direction to another, wind speed reconstruction becomes flawed and introduces additional measurement errors. This often occurs in mountainous areas or complex terrain with significant obstacles, resulting in an average wind speed error typically between 5% and 10%. This measurement error severely hinders the application of single-station remote sensing devices for wind speed measurement in complex terrain.

[0008] Existing technologies generally use models to quantify and correct measurement errors caused by complex terrain. These models typically rely on physical model simulation techniques of wind field areas measured by remote sensing devices. The most common approach in the industry utilizes the Reynolds-averaged Navier-Stokes (RANS) equations under computational fluid dynamics simulation models, combined with boundary conditions introduced by the atmospheric boundary layer (ABL) model, and considers some relevant factors driving wind field flow.

[0009] 1. Changes in terrain undulation, i.e., terrain complexity;

[0010] 2. The extent of terrain coverage, including vegetation cover and ground roughness;

[0011] 3. Atmospheric stability, such as the thermal effect of wind fields driven by diurnal and seasonal cycles.

[0012] In the aforementioned solution employing a traditional pure physical fluid simulation model, the system simulation calculation involves theoretically estimating the wind speed profile at the boundary inlet based on the terrain coverage area defined in the simulation, and then using this as a preset boundary condition for further simulation calculations. However, the estimated wind speed profile differs somewhat from the actual one, leading to a decrease in the accuracy of data correction.

[0013] Furthermore, widely used fluid simulation models (such as RANS) require significant computational costs and time to obtain satisfactory results. Besides cost, the demands on external computing time and energy consumption also make it difficult for remote sensing devices to perform complex fluid simulations. After receiving data from remote sensing devices, users must outsource professional data post-processing to perform the aforementioned wind speed corrections. This not only prolongs the testing project cycle but also degrades the user experience during wind speed measurement.

[0014] Another drawback of the above solution is that when evaluating the simulated radial velocity at a given measurement height based on simulation results, it directly samples the wind speed in the simulated wind field at the corresponding radial distance. However, in actual measurements, radial velocity is obtained by processing the backscattered signal from a certain volume of atmospheric aerosol particles (also known as the probe volume) using a remote sensing device. The shape and extent of this probe volume depend on the signal propagation speed, the duration of signal reception, and the characteristics of the transmitting system. The single-point sampling method used in previous solutions differs from the actual measurement process, directly reducing the accuracy of the corrected wind speed. Summary of the Invention

[0015] The purpose of this invention is to address the shortcomings of existing traditional pure physical fluid simulation models, which result in discrepancies between the estimated wind speed profile and the actual wind speed, leading to a decrease in the accuracy of data correction. Furthermore, widely used fluid simulation models (such as RANS) require significant computational costs and time to obtain satisfactory results, which not only prolongs the testing project cycle but also reduces the user experience during wind speed measurement. To address these shortcomings, this invention proposes a wind field correction method and system applicable to single-station remote sensing equipment.

[0016] To achieve the above objectives, the present invention adopts the following technical solution:

[0017] A correction method for wind speed measurement using a single-station remote sensing device includes the following specific steps:

[0018] S1. Collect wind field data through a data information acquisition device, and obtain radial wind speed information at the current acquisition point through a data processing module;

[0019] S2. Based on the assumption of a uniform wind field, the three-dimensional wind speed vector of the collection point is obtained by using the classic wind speed reconstruction algorithm.

[0020] S3. Establish an error model;

[0021] S4. Based on the error factor β calculated by the error model, the three-dimensional wind speed vector of the acquisition point obtained by inversion in S2 is corrected.

[0022] As a further preferred embodiment of the present invention, the establishment of the error model includes:

[0023] S31. Obtain geographic data information of the collection points;

[0024] S32. Set the initial conditions and system equations for the simulation, and construct the wind speed flow field model;

[0025] S33. When the wind speed flow field model simulates radial velocity along the line of sight, the wind field simulation model combined with the distance weighting function is used to simulate the wind field and obtain the simulation results: The simulation results include: simulated radial wind speed and simulated real wind vector.

[0026] S34. Based on the simulated radial wind speed, the reconstructed simulated three-dimensional wind vector is calculated using the classic wind speed reconstruction algorithm.

[0027] S35. Compare the simulated real wind vector with the reconstructed simulated three-dimensional wind vector, and establish an error model.

[0028] As a further preferred embodiment of the present invention, the initial simulation conditions are set to obtain the three-dimensional wind speed vector of the acquisition point in S2 through inversion;

[0029] The wind speed flow field model adopts a variational data assimilation simulation model.

[0030] As a further preferred embodiment of the present invention, the variational data assimilation simulation model is configured as a four-dimensional variational data assimilation method, a three-dimensional variational data assimilation method, or an incremental variational data assimilation method.

[0031] As a further preferred embodiment of the present invention, the error model is set as follows:

[0032] ,

[0033] Where β is the error factor.

[0034] As a further preferred embodiment of the present invention, by means of

[0035] ,

[0036] The corrected wind vector is obtained, and the data correction is completed.

[0037] A correction system for wind speed measurement using a single-station remote sensing device includes a data acquisition device, a data processing module, a geographic information database module, an error analysis model, a data correction module, and a data output module.

[0038] Among them, the data information acquisition module is used to collect signals for transmission, sample wind speed at the collection points in the wind field, and send the collected data to the data information processing module; the data information acquisition device includes a remote sensing measurement device.

[0039] The data information processing module is used to retrieve the radial wind speed at the acquisition point and obtain the measured wind vector through the wind field reconstruction algorithm. The measured wind vector and radial wind speed are then sent to the error analysis model.

[0040] The error analysis model retrieves relevant geographic information about the wind field from the geographic information database module, combines it with measured wind vectors and radial wind speeds to perform error analysis, estimate and quantify the error, and send the results to the data correction module.

[0041] The data correction module receives the measured wind vector and radial wind speed sent by the data information processing module, compares and analyzes them with the error estimate and quantification obtained from the error analysis model, corrects them, and sends the corrected measured wind vector to the data output module.

[0042] The data output module is used to output the corrected measured wind vector.

[0043] As a further preferred embodiment of the present invention, the data acquisition device is configured as a pulsed lidar, a continuous wave lidar, an acoustic radar, or an electromagnetic radar.

[0044] As a further preferred embodiment of the present invention, the data information acquisition device and the data information processing module are mounted on a cabin-type installation device, a floating installation device, or a vehicle-mounted installation device.

[0045] As a further preferred embodiment of the present invention, the error analysis model, data correction module, and data output module are installed on the main equipment of the remote sensing device, on an external device, or on a cloud platform.

[0046] The wind field correction method and system proposed in this invention for single-station remote sensing equipment have the following advantages compared with the prior art:

[0047] 1. This invention processes measured data using a variational data assimilation method, which can significantly improve the simulation speed and accuracy of wind field models and accurately estimate measurement errors;

[0048] 2. This invention optimizes the wind speed reconstruction error introduced by the uniformity assumption, thereby improving the accuracy of wind speed measurement;

[0049] 3. In the process of establishing the model, the present invention incorporates the real-time measurement data of the device into the wind field simulation, and uses the actual wind field data as the boundary conditions of the simulation, which effectively improves the accuracy of the wind speed correction results.

[0050] 4. By adding a distance weighting function during the model building process, this invention further improves the accuracy of the correction method. Attached Figure Description

[0051] Figure 1 is a schematic diagram of the measurement of the ground-based single-station vertical wind profile remote sensing device at four different measurement heights;

[0052] Figure 2 is a schematic diagram of wind field reconstruction by a single-station vertical wind profile pulse lidar on a simple ground surface.

[0053] Figure 3 is a schematic diagram of wind field reconstruction by a ground-based single-station vertical wind profile pulse lidar under complex terrain.

[0054] Figure 4 is a data processing flowchart of the wind field correction method for a single-station remote sensing device involved in this invention;

[0055] Figure 5 shows a first embodiment of the wind field correction system of the single-station remote sensing equipment involved in this invention;

[0056] Figure 6 illustrates a second embodiment of the wind field correction system for a single-station remote sensing device involved in this invention.

[0057] Figure 7 shows the third embodiment of the wind field correction system of the single-station remote sensing equipment involved in this invention.

[0058] The meanings of the labels in the figure are as follows: 1. Remote sensing device; 2. Signal transmitted along the line of sight; 3. Computational grid for wind field simulation in complex sites; 4. Simple shape; 5. Actual measurement point; 6. Volume of probe cylinder; 7. Actual wind field; 8. Radial velocity along the line of sight; 9. Original measured wind vector. Detailed Implementation

[0059] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0060] This invention provides a high-precision calibration method that can run on low-end hardware; it can also be directly installed and run on the hardware of the measuring device, thereby enabling real-time calibration of measurement results without introducing additional complexity to the user and reducing costs.

[0061] This invention is particularly suitable for dealing with complex system problems where accurate boundary conditions cannot be obtained, such as wind field simulation under complex terrain.

[0062] Unlike the traditional method of directly sampling simulation results at a selected measurement distance, this invention uses a weighted summation method that samples all simulation results from the RSD installation location to the maximum measurement distance in the line-of-sight direction. This method is more faithful to the physical process of RSD measurement in real life because it takes into account the physical characteristics related to the atmospheric sounding volume.

[0063] Example 1: The system includes a data acquisition device, a data processing module, a geographic information database module, an error analysis model, a data correction module, and a data output module.

[0064] The data acquisition module is used to collect signals for transmission, sample wind speeds at collection points in the wind field, and send the collected data to the data processing module; the data acquisition device includes a remote sensing measurement device.

[0065] The data information processing module is used to invert the radial wind speed at the acquisition point and obtain the measured wind vector through the wind field reconstruction algorithm. The measured wind vector and radial wind speed are then sent to the error analysis model.

[0066] The error analysis model retrieves relevant geographic information about the wind field from the geographic information database module, combines it with measured wind vectors and radial wind speeds to perform error analysis, estimate and quantify the error, and then sends the results to the data correction module.

[0067] The data correction module receives the measured wind vector and radial wind speed sent by the data information processing module, compares and analyzes them with the error estimate and quantification obtained from the error analysis model, corrects them, and sends the corrected measured wind vector to the data output module.

[0068] The data output module is used to output the corrected measured wind vector.

[0069] Example 2: S1. Collect wind field data through a data information acquisition device, and obtain radial wind speed information of the current acquisition point through a data processing module.

[0070] A single-station remote sensing device emits multiple signals into space, which interact with aerosol particles in the atmosphere. Part of the backscattered signals are received by the remote sensing device's receiving system and processed to obtain radial wind speed information.

[0071] S2. Based on the assumption of a uniform wind field, the three-dimensional wind speed vector of the collection point is obtained by using the classic wind speed reconstruction algorithm.

[0072] Based on the assumption of a uniform wind field, the classical wind speed reconstruction algorithm is used as shown in the following equation, utilizing multiple measured radial wind speeds. and the corresponding projection angle It can inversely calculate the three-dimensional wind speed vector. :

[0073] .

[0074] S3. Establish an error model.

[0075] When the assumption of wind field uniformity cannot be satisfied under complex terrain, an error model needs to be established to reasonably correct the inverted wind speed.

[0076] S31. Obtain geographic data information of the collection point; geographic data information includes inlet velocity, ground roughness length, atmospheric stability, eddy viscosity, or any other parameters used in the physical model.

[0077] S32. Set the initial conditions and system equations for the simulation, and construct the wind speed flow field model.

[0078] The three-dimensional wind speed vector of the acquisition point obtained by inversion in S2 is used as the initial condition.

[0079] The wind speed flow field model adopts a variational data assimilation simulation model.

[0080] This invention employs the VDA method for windflow modeling. The data assimilation method can be simplified into an optimization problem, namely, finding the optimal solution of the objective function that satisfies the constraints. This objective function can reasonably quantify the difference between simulation data and actual observation data, while the system constraint equations originate from the fundamental physical equations describing the operating laws of the objective system.

[0081] There are two main types of applications in data assimilation, both of which can be used in this invention and are covered by the same form:

[0082] State estimation involves directly adjusting the results of the physical model to align the physical fields with available observations. In this application, the state variable x being optimized can include wind speed, pressure, temperature, or any other physical field calculated by the model.

[0083] Parameter estimation involves calibrating the parameters of the physical model so that it produces results that more closely approximate available observations. In this application, the state variable x being optimized may include a reference inlet velocity, ground roughness length, atmospheric stability, eddy viscosity, or any other parameter used by the physical model.

[0084] To assimilate a single, instantaneously available measurement, the most widely used VDA method is called three-dimensional variational assimilation (3D-Var), where the objective function is typically expressed as:

[0085] ,

[0086] in:

[0087] x is the updated system state variable, i.e., the current estimate of the wind field;

[0088] It is the system's objective function (cost function);

[0089] T is the transpose notation of the matrix;

[0090] This refers to the system background state, i.e., the initial state of the wind field;

[0091] B is the error covariance matrix of the system's background state, i.e., the uncertainty estimate of the purely physical model. For its inverse matrix, regions with lower model uncertainty are given greater weight in the objective function;

[0092] It is the set of available observations, that is, the available measurements obtained from a single-station RSD;

[0093] It is a set of simulated observations. The H operator maps the system state to the observation space. That is, for a single-station RSD, the operator H represents the wind field reconstruction process.

[0094] R is the error covariance matrix of the observations, which is the uncertainty estimate of the single-station RSD measurements. For its inverse matrix, observations with lower uncertainty are given greater weight in the objective function.

[0095] This item This term describes the augmented weighted least squares deviation that describes the state of the physical model system. This describes the augmented weighted least squares bias among the available reconstructed observation data. Minimize the objective function. Essentially, it involves selecting the system state variables that minimize the error between the simulated data and the measured data.

[0096] In setting initial simulation conditions and system simulation, this invention directly incorporates real-time measurement data from the device into the wind field simulation, using actual wind field data as the boundary conditions for the simulation, effectively improving the accuracy of wind speed correction results.

[0097] This method reduces the computational requirements by using a simple airflow model and a relatively small computational domain (because boundary conditions can be optimized), while maintaining a sufficiently high level of accuracy by using empirically tuned parameters.

[0098] We can also consider multiple measured data values ​​within a given time series over a given time window. In this case, time is treated as the fourth dimension, and the method used is called four-dimensional variational assimilation (4D-Var), with the objective function becoming:

[0099] ,

[0100] in This refers to the number of observations. Ultimately, this invention is not strictly limited to the use of 3D-Var and 4D-Var, and any other VDA method can be used.

[0101] Minimize the objective function In this case, the fundamental physical properties of the flow field must be followed, that is, a reasonable constraint equation for the flow field physical system must be constructed. For example, the mass conservation constraint can be expressed as... ,in Let be the wind speed. However, this invention is not limited to mass conservation constraints, and any other physical equations can be used as constraints. Generally, the physical constraint equations in 3D-Var are expressed as follows:

[0102] ,

[0103] In 4D-Var, physical constraints drive the evolution of the system state over time, therefore the constraint equations are expressed as follows:

[0104] ,

[0105] In the remainder of this invention, constraint equations in 3D-Var form are used by default.

[0106] To solve the minimum objective function with constraints The problem should be solved using the Lagrange multiplier method, where the Lagrange function is defined as:

[0107] .

[0108] The objective function under constraints is transformed into selecting appropriate x and λ such that the function The smallest problem is solving the equation. .

[0109] This invention can also use pre-existing experimental data to appropriately adjust the covariance matrix B and the objective function R, making it a semi-empirical method.

[0110] S33. When simulating radial velocity along the line of sight in the wind speed flow field model, the wind field simulation model combined with the distance weighting function is used to simulate the wind field, and the results are used for estimation.

[0111] When using a wind speed flow field model to simulate radial velocity along the line of sight, adding a specific range weighting function (RWF) under the volume of the remote sensing probe further improves the accuracy of the correction method.

[0112] For a remote sensing probe model measured at a given distance, the Radial Width Factor (RWF) assigns a weighted contribution of the simulated radial velocity at each location along the line of sight to the final desired simulated radial velocity. Centered on the desired measurement distance point, the RWF is typically a Gaussian distribution along the line of sight. As shown in Figure 2, the position of the cylindrical probe volume along the line of sight is related to the selected measurement height, and the length of the cylindrical probe is typically equal to the full width at half maximum (FWHM) value of the RWF.

[0113] As shown in the following formula, the RWF for a specific RSD will give the corresponding weight at each sampling point:

[0114] ,

[0115] in: It measures the simulated radial velocity at a distance R.

[0116] It is the number of sampling points;

[0117] It is the weight at each sampling point i, and the simulated velocity at the measurement distance R is calculated by weighting according to the specific RWF of RSD;

[0118] It is the simulated velocity at sampling point i;

[0119] It is the range of line-of-sight distance relative to sampling point i;

[0120] When simulating radial velocity along the line of sight using a wind speed flow field model, a wind field simulation model combining a distance-weighted function is used to simulate the wind field and obtain simulation results.

[0121] The simulation results are as follows:

[0122] 1. Simulate radial wind speed;

[0123] 2. The simulated real wind vector.

[0124] S34. Based on the simulated radial wind speed, the reconstructed simulated three-dimensional wind vector is calculated using the classic wind speed reconstruction algorithm.

[0125] Using the radial wind speed information obtained from the simulation, the reconstructed simulated three-dimensional wind vector is calculated through a classic wind speed reconstruction algorithm.

[0126] S35. Compare the simulated real wind vector with the reconstructed simulated three-dimensional wind vector, and establish an error model;

[0127] Then, the wind vector reconstructed using the simulated radial wind speed is compared with the simulated real wind vector to establish a reasonable error model;

[0128] The error model is as follows:

[0129] ,

[0130] Where β is the error factor.

[0131] S4. Based on the error factor β calculated by the error model, the three-dimensional wind speed vector of the acquisition point obtained by inversion in S2 is corrected.

[0132] pass

[0133] ,

[0134] Finally, the wind vector measured initially by the device was corrected based on the estimated error model.

[0135] Example 3: The correction method will be described in a ground-based vertical profile monostation pulsed wind lidar, which is used to measure wind speed in the atmospheric boundary layer.

[0136] As shown in Figure 1, the single-station pulsed wind lidar emits four laser beams along different line-of-sight directions. These beams are all at approximately a 30-degree angle to the vertical axis, and their azimuth angles are spaced 90 degrees apart, pointing towards the four main directions: east, south, west, and north. The data processing module in Figure 5 uses a wind field reconstruction algorithm and a linear regression method to find the optimal value of the wind vector, minimizing the error function between the projection of the wind vector in the line-of-sight direction and the radial velocity measured in the line-of-sight direction, as illustrated in Figures 2 and 3.

[0137] The error function can be expressed as: ,

[0138] in, This is a reconstructed wind vector, where i is the line-of-sight direction index. The unit vector in the direction of view. This represents the radial velocity in the line-of-sight direction. The measured wind vector results can be obtained and stored after wind field reconstruction.

[0139] To correct wind field measurement results under non-uniform wind conditions, an embodiment of this invention embeds an additional microcontroller into the lidar. This microcontroller is dedicated to running the error analysis model shown in Figure 5. This error analysis model exchanges data with the lidar via local TCP communication. This configuration allows for real-time correction of the wind field measurement results.

[0140] The error model first discretizes the simulated wind field region into a Cartesian grid using a flow field model. Terrain elevation data is then loaded at each horizontal grid point using the coordinates of the lidar installation location and an embedded terrain database. This information is then used to vertically discretize the area from the ground to a preset maximum elevation, thereby constructing a three-dimensional Cartesian grid coordinate system, as shown in Figure 2.

[0141] Uncorrected wind measurements were then used as initial conditions in the wind speed and flow field model. For a single-station pulsed wind lidar, this translates to 10-minute average wind speed and horizontal wind direction data. After inputting the aforementioned wind field measurement data, the three-dimensional variational assimilation model was used to estimate the wind field state. The final optimized state variables are the output simulated real wind vectors. .

[0142] Background wind vector of three-dimensional variational assimilation model This is calculated using an extrapolation method of the horizontal average wind speed at different uncorrected heights at the lidar installation location. The first step is to fit the horizontal average wind speed at different heights using the power-law law of the wind profile; the second step is to extrapolate the horizontal average wind speed at each cell in the grid using the fitted power-law law. The vertical wind speed at each location is approximately zero.

[0143] The following assumptions exist, such that the background state variable error covariance matrix B can be simplified to a diagonal matrix:

[0144] 1. The model errors of the wind vector in the x, y, and z directions are uncorrelated.

[0145] 2. Spatial correlation between errors is ignored, that is, the wind vectors of each grid point are independent of each other.

[0146] The magnitude of the error covariance in the simplified diagonal matrix B is related to the relative positions of the grid points and the lidar. The closer the grid points are to the lidar installation location, the smaller the covariance; the farther the grid points are from the lidar installation location, the larger the error covariance. As shown in the example below, the elements of the diagonal matrix can have a certain numerical relationship with the distance between the grid points and the lidar installation location:

[0147] ,

[0148] Where: i is the state variable index, representing the mesh cell and velocity component;

[0149] It is the horizontal coordinate of the cell;

[0150] These are the horizontal coordinates of the lidar;

[0151] k and α are adjustable empirical parameters.

[0152] Since B is a diagonal matrix, therefore The error term of the augmented weighted least squares model in the equation can be rewritten as follows: Furthermore, due to the background wind vector at the radar location... The fact that the simulated wind vector is equal to the measured uncorrected horizontal average wind speed means that the relative difference between the simulated real wind vector and the background wind vector at the radar installation location is equal to the measured value. The deviation between the simulated real wind vector and the measured observation data has been quantified. Therefore, the equations... In This item can be ignored.

[0153] Finally, this embodiment introduces the physical constraint of mass conservation. The Lagrange function in the equation simplifies to: .

[0154] The optimization problem of the cost function is thus transformed into solving .

[0155] The equation can be expanded as follows:

[0156] ,

[0157] The last three equations can be integrated into vector form:

[0158] ,

[0159] This is a system of four simultaneous equations involving four unknown physical quantities. However, unlike more complex CFD flow models, it does not contain any nonlinear terms, such as thermal convection terms. Therefore, its solution process is not difficult.

[0160] Using the second equation, we can... Represented as:

[0161] ,

[0162] Substitute the equation back into In this embodiment of the invention, we obtained the only equation that requires numerical solution: ,

[0163] in, λ is the anisotropic diffusion term of the Lagrange multipliers, and B is the specified diffusion coefficient. Using the standard finite volume method (FVM) technique, the value of λ can be solved for each element in the domain, and then substituted into... Calculate the final output .

[0164] In addition to using the VDA method described above for wind field simulation, this embodiment also considers the influence of the lidar range weighting function RWF when calculating the simulated radial velocity.

[0165] The lidar emits laser pulse signals, and the signal intensity follows a Gaussian envelope over time. In this embodiment, the pulse duration is taken as the full width at half maximum (FWHM) value of the Gaussian envelope curve. =200ns. These pulses travel at the speed of light c along the line of sight, and therefore can also be described as a Gaussian intensity distribution in space. For this embodiment, the pulse signal is in The propagation distance within the time frame is approximately 30m.

[0166] Furthermore, when the lidar optical sensor receives backscattered echo signals from space, a certain sampling time window is required for accumulation. This accumulation time is... The distance gate is the signal at... The distance a signal travels in space within a given time period.

[0167] Sampling accumulation time Generally, it is the optimal trade-off between the total amount of light received and the size of the detector volume, and for this embodiment... =200ns.

[0168] Based on the above information, the range weighting function RWF at the preset measurement distance can be calculated. The range weighting function RWF of pulsed lidar can be expressed as:

[0169] ,

[0170] Where: c is the speed of light

[0171] R is the preset measurement distance along the line of sight.

[0172] r is the relative distance from the preset measurement position to the location of the weighted distance calculation point in the line of sight direction.

[0173] In this embodiment, the sampling interval in the line-of-sight direction is set to 1m. According to... The simulated radial velocity is calculated using the weighting function at the sampling points and the simulated velocity values. Then, using... The described wind field reconstruction algorithm outputs a reconstructed simulated wind vector.

[0174] Finally, as shown in Figure 4, the error between the simulated real wind vector and the reconstructed simulated wind vector is estimated and quantified.

[0175] In this embodiment, the error factor is expressed as the ratio between the two: This error factor can be easily used to correct the measured wind vector of RSD: .

[0176] Example 4: The data acquisition device of the present invention can be implemented not only by pulsed lidar, but also by other types of remote sensing measurement devices, including but not limited to continuous wave lidar, acoustic radar and electromagnetic radar.

[0177] Example 5: The data acquisition device of the present invention can be installed in, but is not limited to, as a cabin-mounted device, a floating device, or a vehicle-mounted device. See Figures 5-7.

[0178] Example 6: The error model calculation process can also be carried out in other ways, including but not limited to running directly on the main equipment of the remote sensing device, running on external dedicated equipment, and running on a cloud platform. As shown in Figure 5-7.

[0179] Other types of variational data assimilation methods can be used, including but not limited to four-dimensional variational data assimilation (4D-Var) and incremental variational data assimilation (Incr-Var).

[0180] The VDA method can be optimized for other variables, including but not limited to state variables such as pressure and temperature, as well as parameters such as boundary inlet velocity, surface roughness length, atmospheric stability, and eddy viscosity.

[0181] The observed data may include variables other than wind speed, including but not limited to turbulence intensity and turbulence energy dissipation rate.

[0182] The objective function can be expressed in different ways, utilizing different estimates of the difference between simulated and observed data, not limited to the augmented weighted least squares error method in the examples, including but not limited to weighted least squares error, least squares error, absolute error, relative error, and maximum error.

[0183] The objective function may include different adjustable empirical parameters, including but not limited to atmospheric stability and ground cover, or may not use any empirical parameters.

[0184] Other physical constraints besides mass conservation can be used, including but not limited to momentum conservation, thermodynamic constraints, and boundary constraints.

[0185] Methods other than the Lagrange multiplier method can also be used to minimize the objective function, including but not limited to gradient descent, stochastic methods, evolutionary methods, heuristics, and metaheuristics.

[0186] Finally, when calculating the distance weighting function for the simulated radial velocity along the line of sight, segmentation sampling is not limited to the line of sight.

[0187] Example 7: The present invention can also consider introducing a cross-sectional area factor of the signal transmitted by the remote sensing device along the line of sight, sampling the measurement results of different detection volumes along the line of sight, and... In Replace with (The detected volume corresponding to the original sampling point). Alternatively, the distance weighting function (RWF) associated with a single-station remote sensing device can be omitted.

[0188] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the above embodiments do not limit the present invention in any way, and all technical solutions obtained by equivalent substitution or equivalent transformation fall within the protection scope of the present invention.

Claims

1. A correction method for measuring wind speed by a single station remote sensing device, characterized in that, Comprise the following specific steps: S1, collecting wind field data through a data information acquisition device, obtaining radial wind speed information of the current collection point through a data processing module; S2, based on the assumption of a uniform wind field, using a classical wind speed reconstruction algorithm to obtain the three-dimensional wind speed vector of the collection point; S3, establishing an error model; The error model is established, comprising: S31, obtaining geographical data information of the collection point; S32, setting the simulation initial conditions and system equations, and constructing a wind speed flow field model; The simulation initial conditions are set to the three-dimensional wind speed vector of the collection point obtained in S2; The wind speed flow field model adopts a variational data assimilation simulation model S33, when the wind speed flow field model simulates the radial velocity along the viewing direction, the wind field is simulated by combining the distance weighted function wind field simulation model, and the simulation result is obtained; the simulation result includes: simulated radial wind speed and simulated real wind vector; S34, based on the simulated radial wind speed, the reconstructed simulation three-dimensional wind vector is calculated through the classical wind speed reconstruction algorithm; S35, compare the simulated real wind vector with the reconstructed simulation three-dimensional wind vector, and establish an error model; S4, according to the error factor β calculated by the error model, the three-dimensional wind speed vector of the collection point obtained in S2 is corrected.

2. The correction method for measuring wind speed by a single station remote sensing device according to claim 1, wherein, The variational data assimilation simulation model is set to four-dimensional variational data assimilation method, three-dimensional variational data assimilation method or incremental variational data assimilation method.

3. The method of claim 1, wherein, The error model is set to: ， Wherein, β is the error factor.

4. The method of claim 1, wherein, Through ， Get the corrected wind vector and complete the data correction.

5. A correction system for a correction method of measuring wind speed by a single station remote sensing device according to any one of claims 1 to 4, characterized in that, Comprise data information acquisition device, data information processing module, geographic information database module, error analysis model, data correction module and data output module; Among them, the data information collection module is used to collect the data for transmitting signals, and the wind speed of the collection point in the wind field is sampled, and the collected data is sent to the data information processing module; the data information acquisition device comprises a remote sensing measuring device; The data information processing module is used to inverse the radial wind speed of the collection point, and the measured wind vector is obtained through the wind field reconstruction algorithm, and the measured wind vector and the radial wind speed are sent to the error analysis model; The error analysis model retrieves the relevant geographical information of the wind field in the geographic information database module, combines the measured wind vector and the radial wind speed, analyzes the error, estimates and quantifies the error, and sends the result to the data correction module; The data correction module is used to accept the measured wind vector and the radial wind speed sent by the data information processing module, compare and analyze the error estimation and quantification obtained by the error analysis model, and correct it, and send the corrected measured wind vector to the data output module; The data output module is used to output the corrected measured wind vector.

6. The correction system of claim 5, wherein The data information acquisition device is set to pulse laser radar, continuous wave laser radar, sound wave radar or electromagnetic wave radar.

7. The correction system of claim 5, wherein The data information acquisition device and the data information processing module adopt a cabin type installation device, a floating type installation device or a vehicle-mounted installation device.

8. The correction system of claim 5, wherein, The error analysis model, the data correction module and the data output module are installed on the remote sensing device main equipment, external equipment or cloud platform.

Images