A wind field inversion method and system based on a shipboard millimeter wave wind sensor

Abstract

The application provides a wind field inversion method and system based on a millimeter wave wind measuring instrument on a ship, comprising: calculating prior information using atmospheric parameters, the longitude and latitude of an observation site and an observation date, wherein the prior information comprises a prior O2 profile covariance matrix and a two-dimensional wind field prior covariance matrix; performing forward modeling on the prior information, the atmospheric parameters, background radiation and a simulated transmission process to obtain simulated brightness temperature; scanning the upper atmosphere of the observation site using an instrument to obtain actual brightness temperature; generating an instrument observation covariance matrix according to the actual brightness temperature; constructing a cost function according to the instrument observation covariance matrix constructed by the simulated brightness temperature and the actual brightness temperature; and solving the cost function to obtain the actual wind speed of the target height layer. The application constructs a cost function based on a forward modeling, inversion prior information and an instrument observation covariance matrix, and solves the cost function to obtain the actual wind speed of the target height layer, which can greatly improve the inversion accuracy of the wind speed.

Description

Technical Field

[0001] This invention relates to the field of wind field inversion technology, and in particular to a wind field inversion method and system based on a shipborne millimeter-wave anemometer. Background Technology

[0002] In nature, all objects above absolute zero radiate electromagnetic energy, and the intensity and wavelength of this radiation vary depending on the object's physical and chemical properties, temperature, and surface condition. A microwave radiometer is a receiver that operates in the microwave band and does not emit electromagnetic waves itself; it passively receives electromagnetic waves emitted by targets and the environment. The 0.118THz frequency band used in this algorithm is a widely studied electromagnetic wave band, possessing a certain degree of penetration through clouds and rain, not relying on the sun as a radiation source, and capable of operating around the clock and in all weather conditions.

[0003] Atmospheric wind fields are crucial parameters for understanding the dynamics and thermodynamics of the Earth's atmospheric system, serving as essential data for weather forecasting, space environment monitoring, and climatological research. However, at altitudes of 10–70 km, methods for acquiring wind data are extremely limited, resulting in a scarcity of usable wind field data within this range. While this invention can obtain brightness temperature via edge scanning on a spaceborne platform to invert the wind field, achieving high accuracy in wind speed measurements requires significant investment in radiometer costs. Alternatively, wind field measurements can be achieved by measuring the Doppler shift of the ozone peak absorption spectrum in the 142 GHz band from a ground-based source. However, ground-based microwave radiometers exhibit systematic drift in brightness temperature observations in individual channels or overall, leading to substantial observational errors. In contrast, a shipborne millimeter-wave radiometer, operating in the stratosphere, allows for greater propagation distances and altitudes over terahertz radiation (0.118 THz) due to lower water vapor levels, as the latter is less affected by attenuation. Compared to spaceborne terahertz radiometers, airborne platforms offer advantages such as longer dwell time, longer integration time, lower cost, and higher brightness temperature measurement accuracy. Therefore, compared to ground-based and satellite platforms, shipborne millimeter-wave wind measurement has significant advantages.

[0004] Under local thermal equilibrium conditions, molecular spectral lines generated by ground-state chemical substances can be used to invert information such as molecular content and temperature. Wind can generate Doppler shifts in molecular spectral lines, and brightness temperature information and Doppler shifts can be passively acquired using radiometers to invert the wind field. However, in windy and windless conditions, the brightness temperature acquired by the radiometer deviates in frequency with wind speed, but this deviation could be caused by wind speed, line shape, frequency, etc. This makes wind field inversion an ill-conditioned and nonlinear process, potentially involving unsolvable overdetermined equations or underdetermined equations with countless solutions. High-precision wind field inversion is not possible, and continuous wind field detection methods at altitudes of 10–70 km are extremely limited. Summary of the Invention

[0005] In order to overcome the shortcomings of the prior art, the purpose of this invention is to provide a wind field inversion method and system based on a shipborne millimeter-wave anemometer.

[0006] To achieve the above objectives, the present invention provides the following solution:

[0007] A wind field inversion method based on a shipborne millimeter-wave anemometer includes:

[0008] Prior information is calculated using atmospheric parameters, the latitude and longitude of the observation location, and the observation date. The prior information includes the prior O2 profile covariance matrix and the two-dimensional wind field prior covariance matrix.

[0009] The simulated brightness temperature was obtained by performing forward modeling on the prior information, atmospheric parameters, background radiation, and simulated atmospheric transport process using a radiative transfer model.

[0010] The instrument is used to scan the height layer at the observation site to obtain the actual brightness temperature;

[0011] The covariance matrix is ​​generated based on actual brightness temperature and instrument observations.

[0012] The cost function is constructed based on the instrument observation covariance matrix constructed from the simulated brightness temperature and the actual brightness temperature observed by the instrument.

[0013] The actual wind speed at the target height level is obtained by solving the cost function.

[0014] Preferably, the prior O2 profile covariance matrix is:

[0015]

[0016] Among them, S a This represents the prior O2 profile covariance matrix. The diagonal elements represent the variance of the prior state vector, which is the corridor distribution of oxygen in the atmosphere. Here, we take M to be 4, corresponding to the four azimuth angles of the scan. Since a three-dimensional atmospheric model is used, interpolation processing is required for the oxygen corridor. This represents the oxygen corridor distribution at the first azimuth angle. This represents the oxygen corridor distribution at the Mth azimuth angle. This represents the atmospheric correlation between the first azimuth and the Mth azimuth corridor. This represents the atmospheric correlation between the Mth azimuth and the first azimuth corridor.

[0017] Preferably, the step of generating the instrument observation covariance matrix based on the actual brightness temperature includes:

[0018] Formula used:

[0019]

[0020] Generate the instrument observation covariance matrix; where S y Represents the instrument observation covariance matrix. This indicates the brightness temperature measured in channel one. This indicates the brightness temperature measured in channel two. This indicates the brightness temperature measured in channel three. This indicates the brightness temperature measured in channel four.

[0021] Preferably, the step of constructing the cost function based on the radiative transfer model, the prior information, and the instrument observation covariance matrix includes:

[0022] A cost function is constructed based on Bayesian theory using the radiative transfer model, the prior information, and the instrument observation covariance matrix; wherein the cost function is:

[0023]

[0024] Where P(x|y) represents the conditional probability density function of x given y, x represents the wind field corridor to be inverted, y represents the observed brightness temperature with noise, and F(x) represents the radiative transfer model. a Let S represent the a priori wind field corridor, c represent the observation error, and S represent the observation error. y This represents the instrument observation covariance matrix.

[0025] Preferably, solving the cost function to obtain the actual wind speed at the target height level includes:

[0026] The line-of-sight wind speed at the target height level is obtained by solving the cost function using the least squares optimization method.

[0027] The wind speed at the target height is converted by the angle between the line of sight center and the horizontal plane to obtain the inversion corridor of the wind field in both the horizontal and vertical directions.

[0028] This invention also provides a wind field inversion system based on a shipborne millimeter-wave anemometer, comprising:

[0029] The prior information acquisition module is used to calculate prior information using atmospheric parameters, the latitude and longitude of the observation location, and the observation day. The prior information includes the prior O2 profile covariance matrix and the two-dimensional wind field prior covariance matrix.

[0030] The simulated brightness temperature is obtained by performing forward modeling on the prior information, atmospheric parameters, background radiation, and simulated transmission process using a radiative transfer model.

[0031] The actual brightness temperature acquisition module is used to scan the height layer of the observation site using the instrument to obtain the actual brightness temperature;

[0032] The instrument observation covariance matrix calculation module is used to generate the instrument observation covariance matrix based on the actual brightness temperature.

[0033] The cost function construction module is used to construct a cost function based on the instrument observation covariance matrix constructed from the simulated brightness temperature and the actual brightness temperature observed by the instrument.

[0034] The solution module is used to solve the cost function to obtain the actual wind speed at the target height level.

[0035] Preferably, the prior O2 profile covariance matrix is:

[0036]

[0037] Among them, S a This represents the prior O2 profile covariance matrix. This represents the oxygen corridor distribution at the first azimuth angle. This represents the oxygen corridor distribution at the Mth azimuth angle. This represents the atmospheric correlation between the first azimuth and the Mth azimuth corridor. This represents the atmospheric correlation between the Mth azimuth and the first azimuth corridor.

[0038] Preferably, the step of generating the instrument observation covariance matrix based on the actual brightness temperature includes:

[0039] Formula used:

[0040]

[0041] Generate the instrument observation covariance matrix; where S y Represents the instrument observation covariance matrix. This indicates the brightness temperature measured in channel one. This indicates the brightness temperature measured in channel two. This indicates the brightness temperature measured in channel three. This indicates the brightness temperature measured in channel four.

[0042] Preferably, the step of constructing the cost function based on the radiative transfer model, the prior information, and the instrument observation covariance matrix includes:

[0043] A cost function is constructed based on Bayesian theory using the radiative transfer model, the prior information, and the instrument observation covariance matrix; wherein the cost function is:

[0044]

[0045] Where P(x|y) represents the conditional probability density function of x given y, x represents the wind field corridor to be inverted, y represents the observed brightness temperature with noise, and F(x) represents the radiative transfer model. a Let S represent the a priori wind field corridor, c represent the observation error, and S represent the observation error. y This represents the instrument observation covariance matrix.

[0046] Preferably, the solution module includes:

[0047] The least squares optimization unit is used to solve the cost function using the least squares optimization method to obtain the line-of-sight wind speed at the target height layer;

[0048] The line-of-sight wind speed conversion unit is used to convert the line-of-sight wind speed according to the angle between the line of sight center and the horizontal to obtain the inversion corridor of the wind field at the target height layer in both horizontal and vertical directions.

[0049] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0050] This invention provides a wind field inversion method and system based on a submarine-borne millimeter-wave anemometer, comprising: calculating prior information using atmospheric parameters, the latitude and longitude of the observation location, and the observation date; the prior information including a prior O2 profile covariance matrix and a two-dimensional wind field prior covariance matrix; performing forward modeling on the prior information, atmospheric parameters, background radiation, and simulated transmission process using a radiative transfer model to obtain simulated brightness temperature; scanning the height layer of the observation location with the instrument to obtain actual brightness temperature; generating an instrument observation covariance matrix based on the actual brightness temperature; constructing a cost function based on the instrument observation covariance matrix constructed from the simulated brightness temperature and the actual brightness temperature observed by the instrument; and solving the cost function to obtain the actual wind speed at the target height layer. This invention, by constructing a cost function based on the instrument observation covariance matrix constructed from the simulated brightness temperature and the actual brightness temperature observed by the instrument, and solving the cost function to obtain the actual wind speed at the target height layer, can significantly improve the accuracy of wind speed inversion. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0052] Figure 1 A flowchart of a wind field inversion method based on a shipborne millimeter-wave anemometer provided in an embodiment of the present invention;

[0053] Figure 2The following is a flowchart of the forward simulation brightness temperature in the embodiments provided by the present invention;

[0054] Figure 3 The brightness temperature results of the forward modeling simulation provided in the embodiments of the present invention are shown in the figure.

[0055] Figure 4 This is a schematic diagram of the scanning height of the millimeter-wave radiometer in an embodiment provided by the present invention;

[0056] Figure 5 A schematic diagram of measuring a 70km wind field using a shipborne millimeter-wave passive radiometer in an embodiment provided by the present invention;

[0057] Figure 6 The wind field corridor diagram provided in the embodiments of the present invention. Detailed Implementation

[0058] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0059] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0060] The terms "first," "second," "third," and "fourth," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, including a series of steps, processes, methods, etc., is not limited to the steps listed, but may optionally include steps not listed, or may optionally include other steps inherent to these processes, methods, products, or devices.

[0061] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0062] Please see Figure 1 A wind field inversion method based on a shipborne millimeter-wave anemometer, comprising:

[0063] Step 1: Calculate prior information using atmospheric parameters;

[0064] Forward modeling refers to simulating the output brightness temperature signal using a radiative transfer model under conditions such as a frequency of 118.75 GHz and a sensor altitude of 20 km (completely simulating the process of a millimeter-wave radiometer receiving brightness temperature signals on an airship). Prior information refers to information known before inversion, and the accuracy of the prior information and the accuracy of the radiative transfer model directly affect the accuracy of wind field inversion. Brightness temperature forward modeling calculates the radiative transfer process after inputting atmospheric distribution parameters, thereby analyzing the impact of changes in atmospheric physical parameters on the brightness temperature magnitude. See [link to previous section]. Figure 2 .

[0065] Since wind field inversion is a nonlinear, ill-conditioned inversion, it requires very good and comprehensive prior information to optimize the iterative process, reduce computation time, and increase inversion accuracy. Theoretically, the prior information for the inverted wind field parameters should be an unbiased estimate of the true values ​​of the inversion parameters in a statistical sense. Prior information includes the background covariance matrix, background profile, and initial profile.

[0066] The forward radiative transfer model and related prior information are calculated as follows:

[0067] 1) Calculation of absorption coefficient

[0068] Under local thermal equilibrium, the absorption coefficient k is described a A function of line strength S and line type F:

[0069] k a (v)=nS(T)F(v)

[0070] Where n represents the gas number density, T represents the temperature, and v has a frequency of 118.75 GHz. The linear intensity at different temperatures is calculated based on measurement data from the spectral database and the following formula:

[0071]

[0072] Regarding the two energy levels of gas transitions such as O2, and Q(T) as the partition function, this information was obtained from the High-Resolution Transmission Molecular Absorption Database (HITRAN) developed by Harvard University, where k is the Boltzmann constant, T is the atmospheric temperature, and E... f and E i T represents two energy levels that undergo a transition, and T0 and T represent the temperatures of the two energy levels.

[0073] Since the far-field spectral lines of some gases such as O2 are difficult to express using line shape functions, their absorption coefficients need to be divided into two parts: the spectral line absorption near the center frequency (118.75 GHz) and the continuous absorption at the far-field.

[0074]

[0075] The absorption of spectral lines near the center frequency is calculated using line-by-line integration. Spectral line absorption representing the center frequency, Represents continuous absorption of the far wing.

[0076] 2) Calculation of O2 absorption mode

[0077] The oxygen absorption mode used in this invention is "O2-PWR93", but "H2O-PWR98" will be added to the gas absorption type list. Parametric calculation of absorption lines:

[0078]

[0079] Its linear strength:

[0080] S k (T)=S k (300K) / exp(b k ·Θ)

[0081] S k (300K) represents the reference line strength at T=300K, b k For parameters, at 118.75 GH Z It is 0.009. The radiation field term has been approximated using microwaves.

[0082] The pressure broadening linear function is derived using the Doppler frequency shift formula in non-relativistic theory.

[0083]

[0084] Its line width

[0085] Γ k =w k ·(P d ·Θ 0.8 +1.1·P H2O ·Θ)

[0086] The number density of oxygen is:

[0087]

[0088] ν k Y is the center frequency of the line.k Take 8.356, w k P is the width parameter. d P is the partial pressure of dry air. H2O For water vapor partial pressure, P air For atmospheric partial pressure, k B is the Boltzmann constant.

[0089] Calculation based on the normalization coefficient of non-resonant absorption in row-by-row absorption:

[0090]

[0091] Where v is the frequency, and v0 is the center frequency, which is taken as 118.75 GHz here. Its frequency and the derivative at the center of the linear body are respectively:

[0092]

[0093] It should be noted that N in NF represents the normalization parameter, and F refers to the line shape model. Here, NF is a specific normalized O2 linear absorption of VVW bond in this invention.

[0094] The oxygen continuous absorption term is proportional to the collision frequency of a single oxygen molecule with other air molecules. In forward modeling, at least oxygen and water vapor molecules must be considered simultaneously, specifically using absorption modes "O2-PWR93" and "H2O-PWR98". This is similar to simulating microwave absorption and emission under clear atmospheric conditions, as the content of other trace gas molecules is too small to be considered significantly more important than the contributions of oxygen and water vapor, and can be ignored. 3) Doppler frequency shift calculation

[0095] The average motion of the observed O2 gas and some other gases relative to the airborne platform causes a Doppler shift in the gas absorption spectrum, and this Doppler shift signal can be measured by a millimeter-wave spectrometer.

[0096] The overall wind direction is v

[0097]

[0098] The zenith angle of the wind direction is

[0099] ψv=arccos(v w / v)

[0100] And the azimuth angle is

[0101] ωv=arctan(v u / v v )

[0102] The cosine of the angle between the wind vector and the line of sight is:

[0103] cosγ=cosψv cosψu+sinψv sinψu cos(ωv-ωl)

[0104] Where ωl is the angle of the line of sight. It is the square of the meridional wind. It is the square of the zonal wind. ψv is the square of the vertical wind direction, ψu is the zenith angle of the latitudinal line of sight, ψu is the zenith angle of the meridional line of sight, and ωv is the azimuth angle.

[0105] Finally, since the wind speed did not reach the relativistic value, the Doppler shift can be calculated as follows:

[0106]

[0107] Among them, υ O Here, v is the stationary frequency, cosγ is the total wind speed, cosγ represents the angle between the wind vector and the line of sight, and c represents the speed of light. The Doppler shift of the spectral lines will cause a deviation between the brightness temperature observed by the radiometer and the brightness temperature under calm conditions. The magnitude of this deviation depends on multiple factors, including wind speed, frequency, and atmospheric profile. For the observation frequency of 118 GHz, a line-of-sight wind speed of 50 m / s will likely cause a Doppler shift of approximately Δν = 20 kHz.

[0108] Step 2: After completing the relevant atmospheric absorption calculations, a full-link simulation calculation is performed on the entire measurement process of the millimeter-wave anemometer of this invention to combine with the above calculations to achieve forward modeling of brightness temperature simulation. That is, the radiative transfer model is used to perform forward modeling on the prior information, atmospheric parameters, background radiation, and simulated atmospheric transfer process to obtain the simulated brightness temperature. Latitude and longitude are automatically obtained according to the airship's global flight mode, or manually input by the user. Since the millimeter-wave anemometer acquires a two-dimensional wind field, a three-dimensional atmospheric mode is selected during the simulation scanning and radiative transfer processes to obtain the most realistic situation. The sensor height in the simulated millimeter-wave radiometer is 20 km, the scanning height range is 10–70 km, the antenna width is 3 dB beam angle, the integration time is 0.1 s, and the system thermal noise is below 1 K. Gaussian noise with a variance of 1 K is added to the simulation. Since the millimeter-wave anemometer acquires a two-dimensional wind field, in actual observation, the feed is stationary, and the servo unit rotates with the reflector to scan, ensuring that signals from at least two perpendicular orthogonal lines of sight are received by the receiver in the 118.75 GHz frequency band. Therefore, azimuth angle measurements were taken at 45°, 135°, -45°, and -135°. The zenith angle was simulated in 0.1° increments from -3.2° to 15.6°. The vertical grid for inversion was set to 2km, taking into account gases that significantly influence this frequency band, such as oxygen, ozone, water vapor, and nitrogen. Brightness temperature was simulated through steps 1 and 2. It should be noted that these parameters will be adaptively adjusted if the actual wind field accuracy retrieved by the anemometer does not meet the user's requirements.

[0109] Step 3: Obtain prior information based on the latitude and longitude of the observation location and the observation date; the prior information includes the prior O2 profile covariance matrix and the two-dimensional wind field prior covariance matrix;

[0110] In practical applications, based on the latitude and longitude of the observation location and the number of observation days (in one year), a pre-designed prior information generation submodule is input to generate the prior information required and used for inversion, including the prior atmospheric wind field profile xa, the prior O2 profile covariance matrix Sa, etc.

[0111]

[0112] Among them, S a This represents the prior O2 profile covariance matrix. This represents the oxygen corridor distribution at the first azimuth angle. This represents the oxygen corridor distribution at the Mth azimuth angle. This represents the atmospheric correlation between the first azimuth and the Mth azimuth corridor. This represents the atmospheric correlation between the Mth azimuth and the first azimuth corridor. The generated S... aThe diagonal elements represent the variance of the prior oxygen state vector, while the off-diagonal elements represent the correlation between atmospheric layers and are mainly used for smoothing curves.

[0113] Step 4: Use a millimeter-wave radiometer to scan the height layer of the observation site to obtain the actual brightness temperature;

[0114] Please see Figure 4 The millimeter-wave radiometer primarily measures brightness temperature signals within an altitude range of 10km to 70km, with the antenna beam scanning different altitude layers. The radiometer employs a shared 118GHz reflector and a separate feed structure. During observation, the feed remains stationary, while a servo unit rotates the reflector to perform the scan. The feed is positioned at the focal point of the parabolic reflector, ensuring that during the reflector scan, the reflected signal is separated into 118GHz frequency band feeds via a frequency separation device and then received by the receiver. The millimeter-wave radiometer uses four channels with the following center frequencies and bandwidths: 118.75±0.08GHz / 60MHz, 118.75±0.2GHz / 100MHz, 118.75±0.4GHz / 200MHz, and 118.75±0.8GHz / 200MHz.

[0115] Since factors other than atmospheric conditions affect the measurement of brightness temperature signals, the receiver's own errors will also affect the final wind field accuracy. Therefore, in order to minimize the errors introduced by the radiometer, a four-channel radiometer at the same frequency is used. After obtaining the highest possible accuracy brightness temperature signal through corresponding filtering algorithms, the brightness temperature signal and Doppler frequency shift are input to the inversion algorithm center.

[0116] Step 5: Generate the instrument observation covariance matrix based on the actual brightness temperature;

[0117] Since the radiometer of this invention adopts a multi-channel parallel, mixed-frequency double-sideband method, the radiation signal output by the back-end spectrum analyzer is calculated as follows:

[0118]

[0119] in, This represents the normalized channel response function. The radiance is calculated. The radiance output from the back-end spectrum analyzer is converted into a brightness temperature signal using the Planck function. Since all four channels operate at the same frequency, but each channel has different levels of noise affecting accuracy, this invention avoids insufficient accuracy of the brightness temperature output from a single channel due to system noise. The actual brightness temperature signal received by the instrument determines the accuracy of the solution during the inversion iteration process. Therefore, this invention averages the four channels and compares the four channels with the average value to remove the brightness temperature signal that is significantly affected by noise. The resulting brightness temperature signal is then input for further processing.

[0120]

[0121] Generate the instrument observation covariance matrix. Where S y Represents the instrument observation covariance matrix. This indicates the brightness temperature measured in channel one. This indicates the brightness temperature measured in channel two. This indicates the brightness temperature measured in channel three. This indicates the brightness temperature measured in channel four.

[0122] The above instrument observation covariance matrix, except for the diagonal elements, is set under the assumption that there is no correlation between the four channels. Since the four channels are all at the same frequency, the subsequent processing here is converted into a brightness temperature signal when the receiver receives the radiance.

[0123] Step 6: Construct a cost function based on the radiative transfer model, the prior information, and the instrument observation covariance matrix;

[0124] Furthermore, before constructing the cost function, the following is also required:

[0125] 1. Generate atmospheric state vector

[0126] Due to the limited data on wind fields in the middle and upper atmosphere, the prior atmospheric state vector xa is obtained by interpolation and Doppler shift of ECMWF reanalysis data.

[0127]

[0128] u represents the meridional wind field, v represents the zonal wind field, x o21 …x o2n Δf represents the corridor of oxygen under different azimuth scanning angles in a three-dimensional atmospheric model, b is the Doppler frequency shift, and b is the error constant.

[0129] Meanwhile, the initial values ​​of the two line-of-sight wind fields are set to 50 m / s, and subsequent iterations are performed based on this. In the circumnavigation mode of the airship, the initial corridor is adjusted accordingly when it travels to different locations in order to reduce the time required for subsequent iterations.

[0130] 2. Calculate the simulated brightness temperature using a radiative transfer model.

[0131] Based on the calculation process of the radiative transfer model in the prior information calculation above, the absorption coefficient, radiative transfer, and Plank brightness temperature signal conversion are calculated respectively to generate the F(X,b) radiative transfer model, where X is the atmospheric wind field corridor and b represents the parameters in the model that are independent of the atmospheric state vector, which can be abbreviated as F(x).

[0132] Since wind field inversion is a nonlinear process, the cost function based on Bayesian theory is:

[0133]

[0134] Where P(x|y) represents the conditional probability density function of x given y, x represents the wind field corridor to be inverted, y represents the observed brightness temperature with noise, and F(x) represents the radiative transfer model. a denoted by , where c represents the a priori wind field corridor and c represents the observation error.

[0135] Step 7: Solve the cost function to obtain the actual wind speed at the target height level.

[0136] In this embodiment of the invention, the cost function can be solved using the LM iterative algorithm. The Levenberg-Marquardt method combines the advantages of gradient descent and Gauss-Newton iterative methods, and is currently the most commonly used least squares optimization method, which is also the method adopted in this algorithm.

[0137]

[0138] x i+1 This is the result of the nth wind field inversion iteration; x a and x i These are the a priori wind field profile and the preliminary estimated co-wind field profile, respectively; S a Let S be the covariance matrix of the prior wind field profile, representing the covariance associated with generating the prior wind field profile; y Let K be the instrument observation covariance matrix, representing the error caused by the instrument observation accuracy, and γ be the calculation vector obtained in the nth calculation; xi Let I be the kernel matrix calculated in the nth iteration, and K be the identity matrix. i The weight function matrix, This is a forward model.

[0139] The LM iterative algorithm requires setting appropriate stopping conditions to ensure successful inversion. If the set convergence criterion is met, the iteration stops, i.e.:

[0140]

[0141] Where, x i+1 It is a candidate solution for i+1. It is the final solution after i iterations, S y It is the instrument observation covariance matrix.

[0142] Otherwise, iteration will stop when the maximum number of iterations is reached (initially set at 10, to be adjusted if the accuracy does not meet the requirements) or when the maximum threshold is reached. Because a large maximum number of iterations may result in excessively long inversion times, which is not suitable for practical applications, the initial number of iterations is set to 10.

[0143] After obtaining the line-of-sight (LOS) wind speed, the horizontal wind speed at the corresponding altitude is calculated in Cartesian coordinates based on the current rotation angle of the servo system. Here, we take the LOS wind speed at a horizontal altitude of 70km as an example. Please refer to [link / reference]. Figure 5 At 70km, based on the antenna's 3dB beamwidth being less than 0.86 degrees and the smooth, patchy nature of the wind field, wind direction correction was performed. The vertical resolution of the wind field at this location was 3.2km, and the average horizontal wind field covered a range of 8km. At 70km, the angle between the line-of-sight center and the horizontal was 13.2°, and the obtained line-of-sight wind V... LOS Perform the conversion:

[0144] V S =V LOS *COS13.2°

[0145] Based on the above conversion, the line-of-sight winds from 10 to 70 km are converted in both horizontal and vertical directions, and the wind field corridor map can be obtained by this invention. For example... Figure 6 As shown, the two wind field corridors on the left and right are the inversion corridors in two directions. The positive and negative 20 m / s represent the actual wind speeds in the two directions, and 0 m / s represents the prior wind field corridor (in the actual inversion, the wind field corridors will be interpolated using previous data to avoid setting the same wind speed uniformly, which would affect the inversion accuracy).

[0146] Finally, the obtained wind speed is compared with relevant data. If the accuracy does not meet the requirements, adjustments will be made to the sources of error in the inversion process. If the accuracy is insufficient due to too few iterations, the number of iterations will be increased appropriately to make the iteration results more closely match the real wind field.

[0147] This invention also provides a wind field inversion system based on a shipborne millimeter-wave anemometer, comprising:

[0148] The prior information acquisition module is used to calculate prior information using atmospheric parameters, the latitude and longitude of the observation location, and the observation day. The prior information includes the prior O2 profile covariance matrix and the two-dimensional wind field prior covariance matrix.

[0149] The simulated brightness temperature is obtained by performing forward modeling on the prior information, atmospheric parameters, background radiation, and simulated transmission process using a radiative transfer model.

[0150] The actual brightness temperature acquisition module is used to scan the height layer of the observation site using the instrument to obtain the actual brightness temperature;

[0151] The instrument observation covariance matrix calculation module is used to generate the instrument observation covariance matrix based on the actual brightness temperature.

[0152] The cost function construction module is used to construct a cost function based on the instrument observation covariance matrix constructed from the simulated brightness temperature and the actual brightness temperature observed by the instrument.

[0153] The solution module is used to solve the cost function to obtain the actual wind speed at the target height level.

[0154] Preferably, the prior O2 profile covariance matrix is:

[0155]

[0156] Among them, S a This represents the prior O2 profile covariance matrix. This represents the oxygen corridor distribution at the first azimuth angle. This represents the oxygen corridor distribution at the Mth azimuth angle. This represents the atmospheric correlation between the first azimuth and the Mth azimuth corridor. This represents the atmospheric correlation between the Mth azimuth and the first azimuth corridor.

[0157] Preferably, the step of generating the instrument observation covariance matrix based on the actual brightness temperature includes:

[0158] Formula used:

[0159]

[0160] Generate the instrument observation covariance matrix; where S y Represents the instrument observation covariance matrix. This indicates the brightness temperature measured in channel one. This indicates the brightness temperature measured in channel two. This indicates the brightness temperature measured in channel three. This indicates the brightness temperature measured in channel four.

[0161] Preferably, the step of constructing the cost function based on the radiative transfer model, the prior information, and the instrument observation covariance matrix includes:

[0162] A cost function is constructed based on Bayesian theory using the radiative transfer model, the prior information, and the instrument observation covariance matrix; wherein the cost function is:

[0163]

[0164] Where P(x|y) represents the conditional probability density function of x given y, x represents the wind field corridor to be inverted, y represents the observed brightness temperature with noise, and F(x) represents the radiative transfer model. a denoted by , where c represents the a priori wind field corridor and c represents the observation error.

[0165] Preferably, the solution module includes:

[0166] The least squares optimization unit is used to solve the cost function using the least squares optimization method to obtain the line-of-sight wind speed at the target height layer;

[0167] The line-of-sight wind speed conversion unit is used to convert the line-of-sight wind speed according to the angle between the line of sight center and the horizontal to obtain the inversion corridor of the wind field at the target height layer in both horizontal and vertical directions.

[0168] Compared with the prior art, the beneficial effects of the wind field inversion system based on a shipborne millimeter-wave anemometer provided by the present invention are the same as the beneficial effects of the wind field inversion method based on a shipborne millimeter-wave anemometer described in the above technical solution, and will not be repeated here.

[0169] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. Regarding the methods disclosed in the embodiments, since they correspond to the apparatus disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the apparatus description.

[0170] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A wind field inversion method based on a shipborne millimeter-wave anemometer, characterized in that, include: Prior information is calculated using atmospheric parameters, the latitude and longitude of the observation location, and the observation date. The prior information includes the prior O2 profile covariance matrix and the two-dimensional wind field prior covariance matrix. The simulated brightness temperature is obtained by performing forward modeling on the prior information, atmospheric parameters, background radiation, and simulated transmission process using a radiative transfer model. The instrument is used to scan the height layer at the observation site to obtain the actual brightness temperature; The covariance matrix is ​​generated based on actual brightness temperature and instrument observations. The cost function is constructed based on the instrument observation covariance matrix constructed from the simulated brightness temperature and the actual brightness temperature observed by the instrument. The actual wind speed at the target height level is obtained by solving the cost function.

2. The wind field inversion method based on a shipborne millimeter-wave anemometer according to claim 1, characterized in that, The prior O2 profile covariance matrix is: Among them, S a This represents the prior O2 profile covariance matrix. This represents the oxygen corridor distribution at the first azimuth angle. This represents the oxygen corridor distribution at the Mth azimuth angle. This represents the atmospheric correlation between the first azimuth and the Mth azimuth corridor. This represents the atmospheric correlation between the Mth azimuth and the first azimuth corridor.

3. The wind field inversion method based on a shipborne millimeter-wave anemometer according to claim 2, characterized in that, The process of generating the instrument observation covariance matrix based on the actual brightness temperature includes: Formula used: Generate the instrument observation covariance matrix; where S y Represents the instrument observation covariance matrix. This indicates the brightness temperature measured in channel one. This indicates the brightness temperature measured in channel two. This indicates the brightness temperature measured in channel three. This indicates the brightness temperature measured in channel four.

4. The wind field inversion method based on a shipborne millimeter-wave anemometer according to claim 3, characterized in that, The construction of the cost function based on the radiative transfer model, the prior information, and the instrument observation covariance matrix includes: A cost function is constructed based on Bayesian theory using the radiative transfer model, the prior information, and the instrument observation covariance matrix; wherein the cost function is: Where P(x|y) represents the conditional probability density function of x given y, x represents the wind field corridor to be inverted, y represents the observed brightness temperature with noise, and F(x) represents the radiative transfer model. a Let S represent the a priori wind field corridor, c represent the observation error, and S represent the observation error. y This represents the instrument observation covariance matrix.

5. The wind field inversion method based on a shipborne millimeter-wave anemometer according to claim 4, characterized in that, The step of solving the cost function to obtain the actual wind speed at the target height level includes: The line-of-sight wind speed at the target height level is obtained by solving the cost function using the least squares optimization method. The wind speed at the target height is converted by the angle between the line of sight center and the horizontal plane to obtain the inversion corridor of the wind field in both the horizontal and vertical directions.

6. A wind field inversion system based on a shipborne millimeter-wave anemometer, characterized in that, include: The prior information acquisition module is used to calculate prior information using atmospheric parameters, the latitude and longitude of the observation location, and the observation day. The prior information includes the prior O2 profile covariance matrix and the two-dimensional wind field prior covariance matrix. The simulated brightness temperature is obtained by performing forward modeling on the prior information, atmospheric parameters, background radiation, and simulated transmission process using a radiative transfer model. The actual brightness temperature acquisition module is used to scan the height layer of the observation site using the instrument to obtain the actual brightness temperature; The instrument observation covariance matrix calculation module is used to generate the instrument observation covariance matrix based on the actual brightness temperature. The cost function construction module is used to construct a cost function based on the instrument observation covariance matrix constructed from the simulated brightness temperature and the actual brightness temperature observed by the instrument. The solution module is used to solve the cost function to obtain the actual wind speed at the target height level.

7. A wind field inversion system based on a shipborne millimeter-wave anemometer according to claim 6, characterized in that, The prior O2 profile covariance matrix is: Among them, S a This represents the prior O2 profile covariance matrix. This represents the oxygen corridor distribution at the first azimuth angle. This represents the oxygen corridor distribution at the Mth azimuth angle. This represents the atmospheric correlation between the first azimuth and the Mth azimuth corridor. This represents the atmospheric correlation between the Mth azimuth and the first azimuth corridor.

8. A wind field inversion system based on a shipborne millimeter-wave anemometer according to claim 7, characterized in that, The instrument observation covariance matrix calculation module includes: The instrument observation covariance matrix calculation unit is used to calculate the following formula: Generate the instrument observation covariance matrix; where S y Represents the instrument observation covariance matrix. This indicates the brightness temperature measured in channel one. This indicates the brightness temperature measured in channel two. This indicates the brightness temperature measured in channel three. This indicates the brightness temperature measured in channel four.

9. A wind field inversion system based on a shipborne millimeter-wave anemometer according to claim 8, characterized in that, The cost function construction module includes: The cost function construction unit is used to construct a cost function based on Bayesian theory using the radiative transfer model, the prior information, and the instrument observation covariance matrix; wherein the cost function is: Where P(x|y) represents the conditional probability density function of x given y, x represents the wind field corridor to be inverted, y represents the observed brightness temperature with noise, and F(x) represents the radiative transfer model. a Let S represent the a priori wind field corridor, c represent the observation error, and S represent the observation error. y This represents the instrument observation covariance matrix.

10. A wind field inversion system based on a shipborne millimeter-wave anemometer according to claim 9, characterized in that, The solution module includes: The least squares optimization unit is used to solve the cost function using the least squares optimization method to obtain the line-of-sight wind speed at the target height layer; The line-of-sight wind speed conversion unit is used to convert the line-of-sight wind speed according to the angle between the line of sight center and the horizontal to obtain the inversion corridor of the wind field at the target height layer in both horizontal and vertical directions.

Images