A shipborne GNSS-R-based wind speed inversion bias correction method

Abstract

This invention relates to the field of intelligent ship navigation, specifically to a wind speed inversion deviation correction method based on shipborne GNSS-R, comprising the following steps: acquiring time-delay Doppler image data and a reference wind speed; preprocessing the time-delay Doppler image data and outputting valid data; extracting observations related to sea surface wind speed and dividing them into training and test sets; calculating the initial inversion wind speed based on the training set and constructing initial inversion residual samples, establishing a marginal probability distribution model; establishing a joint probability distribution model based on the initial inversion residual samples and the reference wind speed; constructing test inversion residual samples based on the test set, calculating the conditional estimate of wind speed deviation based on the joint probability distribution model, correcting the deviation of the initial inversion wind speed, and calculating the corrected inversion wind speed. This invention effectively reduces the impact of systematic and random errors, significantly improves the accuracy and robustness of sea surface wind speed inversion, and provides reliable environmental perception support for intelligent ship navigation.

Description

Technical Field

[0001] This invention relates to the field of intelligent navigation of ships, and specifically to a wind speed inversion deviation correction method based on shipborne GNSS-R. Background Technology

[0002] Sea surface wind speed is a crucial environmental parameter affecting ship navigation status and safety, playing a fundamental supporting role in intelligent navigation, route planning, and ship operation control. Accurate and real-time acquisition of sea surface wind speed information around a ship is a vital prerequisite for building a ship environmental perception system and achieving intelligent decision-making, contributing to improved navigation safety and operational efficiency in complex sea conditions. Currently, sea surface wind speed data is primarily obtained through buoys, shipborne meteorological instruments, and spaceborne microwave remote sensing. While in-situ observations offer high precision, their spatial distribution is sparse and coverage is limited; spaceborne observations, while possessing advantages in terms of wide coverage, are affected by orbital periods and the complex near-shore environment, making it difficult to meet the ship's demand for timely and localized wind speed information.

[0003] In recent years, passive remote sensing technology has seen rapid development in ocean observation. Global Navigation Satellite System Reflectance (GNSS-R) utilizes sea surface scattering information from navigation satellite signals to obtain environmental parameters. It boasts advantages such as abundant signal sources, low cost, and all-weather operation, and has shown promising potential in sea surface wind speed retrieval. However, spaceborne GNSS-R is susceptible to land scattering in near-shore areas and struggles to meet the real-time observation requirements for specific routes and local sea conditions.

[0004] Therefore, there is an urgent need for a technical solution that can be deployed on shipboard platforms to achieve high-efficiency, high-resolution, and continuous observation of sea surface wind speed along the route, in order to make up for the shortcomings of existing observation methods in terms of spatiotemporal continuity and local adaptability, and to meet the needs of intelligent navigation for real-time and accurate environmental perception information. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention provides a wind speed inversion deviation correction method based on shipborne GNSS-R. By deploying the GNSS-R receiving system on a shipborne platform, it can overcome the spatial limitations of traditional shipborne point measurements, realize regional wind speed perception along the route during navigation, make up for the deficiencies of spaceborne observation in terms of temporal continuity and spatial resolution, and complement shipborne meteorological observation.

[0006] The technical solution adopted in this invention specifically includes the following steps: Delay Doppler image data is acquired using a shipborne GNSS-R receiver, and reference wind speed is obtained using an anemometer. The time-delay Doppler image data is preprocessed using quality control methods, and valid data is output. Based on the effective data, GNSS-R observations related to sea surface wind speed are extracted, and the observations are divided into training set and test set according to the proportion. The initial inversion wind speed is calculated using the cumulative distribution function based on the training set. An initial inversion residual sample is constructed based on the reference wind speed and the initial inversion wind speed. Marginal probability distribution models of the initial inversion residual sample and the reference wind speed are established respectively. Based on the statistical dependency structure between the initial inversion residual sample and the reference wind speed, a joint probability distribution model is established using the Copula function; Based on the test set, test inversion residual samples are constructed. Based on the joint probability distribution model, the conditional estimate of the wind speed deviation is calculated using the conditional distribution inversion method. The initial inversion wind speed is then corrected for deviation, and the corrected inversion wind speed is calculated.

[0007] Furthermore, the quality control includes: resampling the time-delay Doppler image data, identifying and removing outliers from the peak power extracted from the time-delay Doppler image data using the Z-score statistical analysis method, performing a relative rate of change test on the peak power between adjacent frames, and removing frames with abrupt fluctuations.

[0008] Furthermore, the observed quantity includes the mean signal-to-noise ratio, the mean signal-to-noise ratio... The calculation formula is:

[0009] in, This represents the time delay chip interval used to calculate information power. This represents the Doppler frequency shift interval used to calculate information power. Indicates the time-delay Doppler window The average information power within, This represents the average correlation power value of the noise floor in the current time-delay Doppler image.

[0010] Furthermore, the initial inversion wind speed is calculated using the cumulative distribution function. for:

[0011] in Indicates reference wind speed. The cumulative distribution function represents the mean signal-to-noise ratio. This represents the corresponding mean signal-to-noise ratio value.

[0012] Furthermore, the marginal probability distribution model includes the Weibull distribution model, the Gamma distribution model, the Lognormal distribution model, and the Normal distribution model, and the optimal distribution model is determined based on the AIC criterion.

[0013] Furthermore, the copula function includes the Gaussian copula function. Based on Sklar's theorem, any two-dimensional random variable... and joint distribution function It can be decomposed into a marginal distribution function , With a Copula function Combination forms:

[0014] Representing variables and The dependency structure, and , Each reflects its own univariate statistical characteristics. The formula for calculating the bivariate Gaussian copula, representing the Copula parameter, is as follows:

[0015] in, This represents the inverse function of the standard normal distribution. Representing variables in vector form. Represents the coefficient matrix. Observational features. The mean signal-to-noise ratio (ASNR) represents the target variable. The inversion residuals obtained by matching the cumulative distribution function (CDF) are represented by the conditional distribution of the original variables after establishing the two-dimensional joint distribution function:

[0016] Within the Copula framework, the construction of the joint distribution takes place in the probability space, based on which the target variable, given conditions... The conditional distribution under the following conditions can be expressed as:

[0017] Furthermore, the above conditional distribution describes the distribution under a given condition. The cumulative probability of the target variable given the given values. The advantage of the median lies in its robustness to outlier data and its ability to better reflect the typical level of the target variable. Based on this, the conditional median inversion is defined as:

[0018] in, This represents the value corresponding to a cumulative probability of 0.5 under given conditions. Since this value is defined in the probability space, it needs to be mapped back to the original variable space using the inverse function of the marginal distribution to obtain the corresponding inversion result.

[0019] Furthermore, the corrected inversion wind speed for:

[0020] in Indicates the initial inversion wind speed. This represents the conditional estimate of the wind speed deviation.

[0021] Compared with the prior art, the present invention has the following beneficial effects: This invention provides a wind speed inversion bias correction method based on shipborne GNSS-R. By decoupling and modeling the marginal distribution and statistical dependency structure of the initial inversion residuals and the reference wind speed, and achieving wind speed bias correction based on Copula conditional distribution inversion, it effectively improves the problem that traditional empirical methods struggle to characterize nonlinear statistical dependencies and systematic biases. This method uses the conditional median as the bias estimate, exhibiting strong robustness and generalization ability. When applied to shipborne platforms, it can achieve high-time-efficiency, high-resolution sea surface wind speed perception during navigation, providing reliable environmental information support for intelligent navigation, route planning, and ship safety, demonstrating promising engineering application prospects. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a flowchart of a wind speed inversion deviation correction method based on shipborne GNSS-R according to the present invention.

[0024] Figure 2 This is a schematic diagram of the mean signal-to-noise ratio of a wind speed inversion deviation correction method based on shipborne GNSS-R according to the present invention.

[0025] Figure 3 This is a schematic diagram of the initial inversion wind speed of the wind speed based on the wind speed inversion deviation correction method of the present invention, which is based on shipborne GNSS-R.

[0026] Figure 4This is a schematic diagram of the corrected inversion wind speed according to the wind speed inversion deviation correction method based on shipborne GNSS-R of the present invention. Detailed Implementation

[0027] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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 should fall within the scope of protection of the present invention.

[0028] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0029] like Figure 1 As shown, this invention discloses a wind speed inversion deviation correction method based on shipborne GNSS-R, which mainly includes the following steps: Delay Doppler image data is acquired using a shipborne GNSS-R receiver, and reference wind speed is obtained using an anemometer.

[0030] The time-delay Doppler image data is preprocessed using quality control methods, and valid data is output.

[0031] Based on the effective data, GNSS-R observations related to sea surface wind speed are extracted, and the observations are divided into training and testing sets according to the proportion.

[0032] The initial inversion wind speed is calculated using the cumulative distribution function based on the training set. An initial inversion residual sample is constructed based on the reference wind speed and the initial inversion wind speed. Marginal probability distribution models of the initial inversion residual sample and the reference wind speed are established respectively.

[0033] Based on the statistical dependency structure between the initial inversion residual sample and the reference wind speed, a joint probability distribution model is established using the Copula function.

[0034] Based on the test set, test inversion residual samples are constructed. Based on the joint probability distribution model, the conditional estimate of the wind speed deviation is calculated using the conditional distribution inversion method. The initial inversion wind speed is then corrected for deviation, and the corrected inversion wind speed is calculated.

[0035] In a preferred embodiment of this application, the quality control includes: resampling the time-delay Doppler image data for 10 seconds, identifying and removing outliers from the peak power extracted from the time-delay Doppler image data using the Z-score statistical analysis method, performing a relative rate of change test on the peak power between adjacent frames, and removing frames with abrupt fluctuations.

[0036] In a preferred embodiment of this application, the observed quantity includes the mean signal-to-noise ratio, wherein the mean signal-to-noise ratio... The calculation formula is:

[0037] in, This represents the time delay chip interval used to calculate information power. This represents the Doppler frequency shift interval used to calculate information power; such as... Figure 2 As shown, Indicates the time-delay Doppler window That is, the average information power within the black box in the figure. This represents the noise floor in the current time-delay Doppler image, i.e., the average correlation power value within the two red boxes in the image.

[0038] As a preferred embodiment of this application, such as Figure 3 As shown, the initial inversion wind speed is calculated using the cumulative distribution function. for:

[0039] in Indicates reference wind speed. The cumulative distribution function represents the mean signal-to-noise ratio. This represents the corresponding mean signal-to-noise ratio value.

[0040] In a preferred embodiment of this application, the marginal probability distribution model includes the Weibull distribution model, the Gamma distribution model, the Lognormal distribution model, and the Normal distribution model, and the optimal distribution model is determined based on the AIC criterion.

[0041] In a preferred embodiment of this application, the copula function includes the Gaussian copula function. Based on Sklar's theorem, any two-dimensional random variable... and joint distribution function It can be decomposed into a marginal distribution function , With a Copula function Combination forms:

[0042] Representing variables and The dependency structure, and , Each reflects its own univariate statistical characteristics. The formula for calculating the bivariate Gaussian copula, representing the Copula parameter, is as follows:

[0043] in, This represents the inverse function of the standard normal distribution. Representing variables in vector form. Represents the coefficient matrix. Observational features. The mean signal-to-noise ratio (ASNR) represents the target variable. The inversion bias representing the cumulative distribution function (CDF) matching is given by the conditional distribution of the original variables after establishing the two-dimensional joint distribution function:

[0044] Within the Copula framework, the construction of the joint distribution takes place in the probability space, based on which the target variable, given conditions... The conditional distribution under the following conditions can be expressed as:

[0045] As a preferred embodiment of this application, the above-described conditional distribution describes the distribution under a given condition. The cumulative probability of the target variable given the given values. The advantage of the median lies in its robustness to outlier data and its ability to better reflect the typical level of the target variable. Based on this, the conditional median inversion is defined as:

[0046] in, This represents the value corresponding to a cumulative probability of 0.5 under given conditions. Since this value is defined in the probability space, it needs to be mapped back to the original variable space using the inverse function of the marginal distribution to obtain the corresponding inversion result.

[0047] As a preferred embodiment of this application, such as Figure 4 As shown, the corrected inversion wind speed for:

[0048] in Indicates the initial inversion wind speed. This represents the conditional estimate of the wind speed deviation.

[0049] As a preferred embodiment of this application, The technical solution of this invention proposes a GNSS-R sea surface wind speed inversion and deviation correction method suitable for shipborne platforms, which can build high-resolution, real-time local wind field perception capabilities, providing important support for intelligent navigation and marine monitoring.

[0050] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A wind speed inversion bias correction method based on shipborne GNSS-R, characterized in that, Includes the following steps: Delay Doppler image data is acquired using a shipborne GNSS-R receiver, and reference wind speed is obtained using an anemometer. The time-delay Doppler image data is preprocessed using quality control methods, and valid data is output. Based on the effective data, GNSS-R observations related to sea surface wind speed are extracted, and the observations are divided into training set and test set according to the proportion. The initial inversion wind speed is calculated using the cumulative distribution function based on the training set. An initial inversion residual sample is constructed based on the reference wind speed and the initial inversion wind speed. Marginal probability distribution models of the initial inversion residual sample and the reference wind speed are established respectively. Based on the statistical dependency structure between the initial inversion residual sample and the reference wind speed, a joint probability distribution model is established using the Copula function. Based on the test set, test inversion residual samples are constructed. Based on the joint probability distribution model, the conditional estimate of the wind speed deviation is calculated using the conditional distribution inversion method. The initial inversion wind speed is then corrected for deviation, and the corrected inversion wind speed is calculated.

2. The wind speed inversion deviation correction method based on shipborne GNSS-R according to claim 1, characterized in that, The quality control includes: resampling the time-delay Doppler image data; identifying and removing outliers from the peak power extracted from the time-delay Doppler image data using the Z-score statistical analysis method; performing a relative rate of change test on the peak power between adjacent frames; and removing frames with abrupt fluctuations.

3. The wind speed inversion deviation correction method based on shipborne GNSS-R according to claim 1, characterized in that, The observed quantity includes the mean signal-to-noise ratio (SNR). The calculation formula is: in, This represents the time delay chip interval used to calculate information power. This represents the Doppler frequency shift interval used to calculate information power. Indicates the time-delay Doppler window The average information power within, This represents the average correlation power value of the noise floor in the current time-delay Doppler image.

4. The wind speed inversion deviation correction method based on shipborne GNSS-R according to claim 3, characterized in that, The initial inversion wind speed is calculated using the cumulative distribution function. for: in Indicates reference wind speed. The cumulative distribution function represents the mean signal-to-noise ratio. This represents the corresponding mean signal-to-noise ratio value.

5. The wind speed inversion deviation correction method based on shipborne GNSS-R according to claim 1, characterized in that, The marginal probability distribution models include the Weibull distribution model, the Gamma distribution model, the Lognormal distribution model, and the Normal distribution model, and the optimal distribution model is determined based on the AIC criterion.

6. The wind speed inversion deviation correction method based on shipborne GNSS-R according to claim 1, characterized in that, The copula function includes the Gaussian copula function.

7. The wind speed inversion deviation correction method based on shipborne GNSS-R according to claim 1, characterized in that, The conditional distribution inversion calculates the conditional estimate of wind speed deviation by solving for the median corresponding to a conditional cumulative probability of 0.

5.

8. The wind speed inversion deviation correction method based on shipborne GNSS-R according to claim 1, characterized in that, The corrected inversion wind speed for: in Indicates the initial inversion wind speed. This represents the conditional estimate of the wind speed deviation.

Images