Shipborne GNSS-R significant wave height inversion method based on multiple observation quantity fields

Abstract

The invention relates to a shipborne GNSS-R significant wave height inversion method based on multiple observation quantity fields. The method comprises the following steps: acquiring real-time four-channel time delay Doppler spectrum data and significant wave height values generated by a ship along a ship route; preprocessing the real-time four-channel time delay Doppler spectrum data to obtain preprocessed four-channel time delay Doppler spectrum data; according to the preprocessed time delay Doppler spectrum data, extracting an observed quantity dependent on the significant wave height; based on 70% of observed quantity data, constructing an effective wave height inversion model of a plurality of observed quantity fields; the method comprises the following steps: extracting an observed quantity value sequence based on 30% data of observed quantity, inputting the observed quantity value sequence into a significant wave height inversion model to obtain a significant wave height inversion value, obtaining a corrected observed quantity value sequence at the same time, and establishing a multivariate parameter correction model for the corrected observed quantity value sequence to realize significant wave height inversion. According to the method, the observation quantity is corrected through the joint matrix subjected to dimension reduction reconstruction, and the inversion precision is improved.

Description

Technical Field

[0001] The invention belongs to the field of marine remote sensing and relates to a shipborne GNSS-R effective wave height inversion method based on multiple observation fields. Background Art

[0002] With the rapid development of the marine industry, marine activities are becoming more and more extensive. Accurately knowing the waves is an important prerequisite for ensuring the safety of all marine-related activities. As a representative physical quantity reflecting the waves, effective wave height is very important for maritime safety management and port and dock protection. Traditionally, the measurement of effective wave height parameters is carried out through station-based means, such as marine meteorological stations, buoys, etc., which can achieve continuous observation of relevant wave parameters, but it is limited by high costs, detection area distribution and complexity of the marine environment. With the development of satellite remote sensing technology, some single-machine radars such as satellite-borne altimeters, scatterometers and synthetic aperture radars have roughly achieved global-scale detection by actively transmitting microwave signals to the sea surface and receiving reflected signals. Due to the different degrees of problems such as the cost and accuracy of these equipment, these observation methods have limitations, and the biggest problem is that it is impossible to achieve effective wave height observations with high temporal and spatial resolution.

[0003] In recent years, remote sensing detection combined with navigation technology has become an emerging detection method. GNSS-R, as a novel passive microwave remote sensing technology, relies on more than 100 real-time navigation satellites in space to realize passive detection of the sea surface. Although satellite-borne GNSS-R can achieve global large-scale inversion, it is seriously affected by coastal areas close to the land and lacks the ability to observe in real time, making it difficult to provide more accurate information for ship navigation or coastal high-resolution observation. As a new type of maritime sensor, GNSS-R is carried on board and measures the scattering area around the mirror reflection point according to the dual-base geometric structure, which effectively complements the conventional backscattering shipborne radar and can make up for the shortcomings of other remote sensing detection technologies in time and space extension. Real-time measurement of effective wave height will help the maritime department to build a maritime monitoring network, which is of great significance to the fields of ship intelligent navigation, marine scientific research, marine engineering, etc. Summary of the invention

[0004] In order to solve the problem that for coastal areas close to the land, satellite-borne GNSS-R is seriously affected and lacks the ability of real-time observation, making it difficult to provide more accurate information for ship navigation or coastal high-resolution observation, the technical solution adopted by the present invention is: a ship-borne GNSS-R effective wave height inversion method based on multi-observation measurement fields, comprising the following steps: obtaining real-time four-channel delay Doppler spectrum data and effective wave height values ​​generated by the ship along the route;

[0005] Preprocessing the real-time four-channel delay Doppler spectrum data to obtain preprocessed four-channel delay Doppler spectrum data;

[0006] Extract the observation quantity which is dependent on the significant wave height according to the preprocessed delay Doppler spectrum data;

[0007] Based on 70% of the observation data, an effective wave height inversion model for multiple observation fields is constructed;

[0008] Based on 30% of the observed data, the observed value sequence is extracted, and the observed value sequence is input into the effective wave height inversion model to obtain the effective wave height inversion value. At the same time, the corrected observed value sequence is obtained, and a multivariate parameter correction model is established for the corrected observed value sequence to achieve effective wave height inversion.

[0009] Furthermore: the observed quantities include the leading edge slope of the normalized delay curve, the average of the trailing edge slope, the delay window length, the delay Doppler spectrum volume and the satellite altitude angle.

[0010] Further: the process of preprocessing the real-time four-channel delay Doppler spectrum data to obtain the preprocessed four-channel delay Doppler spectrum data is as follows:

[0011] Normalize the delay curves at different Doppler frequency shifts;

[0012] The four-channel delay Doppler spectrum data of each frame are annotated with the effective wave height value obtained by the wave meter. First, all NAN values ​​are removed. In addition, a standardized delay waveform is established and a threshold is selected to remove low-quality data.

[0013] Further: the number of the normalized delay curves is five, the delay window length is also the average of the five delay curve chips, and the average of the delay curve chip lengths is calculated as follows:

[0014] TDWD=τ right -τ left

[0015] Where τ right and τ left They respectively represent the right and left delay code chip values ​​intercepted in the curve.

[0016] Furthermore: the multi-observation measurement field includes multiple single observation measurement fields, and the single observation measurement field includes each observation quantity and a corresponding altitude angle and effective wave height combination; the matrix dimensions of each single observation measurement field remain consistent.

[0017] Furthermore: the satellite elevation angle range is 40°-80°, and the resolution is 0.1°.

[0018] Further: the expression of the multivariate parameter correction model is as follows:

[0019]

[0020] Where: Obs i represents different observation values, a i represents the corresponding coefficient, and b is the constant term.

[0021] Further: Based on 70% of the observation data, the process of constructing the effective wave height inversion model of multiple observation fields is as follows:

[0022] The single observation field is represented by a matrix:

[0023]

[0024] In the matrix, each point represents the LES value at the effective wave height and satellite altitude angle, and TES mn ,DDMV mn ,TDWD mn The same matrix form is maintained, and the matrix dimension changes with the effective wave height value and the elevation angle range. The matrix dimension of each single observation field remains consistent. The effective wave height inversion model F of the combined multi-observation field is:

[0025]

[0026] LES represents the leading edge slope, TES represents the trailing edge slope, DDMV represents the delay Doppler spectrum volume, TDWD represents the mean of the delay curve chip length, and m and n represent the dimensions of the matrix.

[0027] The present invention provides a shipborne GNSS-R effective wave height inversion method based on multiple observation fields, which has the following advantages:

[0028] The corrected multi-observation takes into account the mutual influence between variables, eliminates the instability of the observation caused by the altitude angle, and reduces the inversion error. This method performs a joint matrix analysis on multiple observations and takes into account the influence of the satellite altitude angle. The joint matrix reconstructed by dimensionality reduction realizes the correction of the observation. The corrected observation is used as an independent variable. The established multivariate parameter model reduces the error caused by a single observation and improves the accuracy of the inversion. Applied to shipborne platforms, it can monitor the navigation environment in real time, achieve high-resolution detection, provide safety and technical guarantees for ship intelligent navigation, maritime safety activities, etc., and improve my country's marine monitoring network. It has broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0030] Figure 1 This is the flow chart of multivariable field shipborne GNSS-R effective wave height inversion;

[0031] Figure 2 This is the geometric diagram of the scattering area formed by the shipborne GNSS-R on the sea surface;

[0032] Figure 3 It is a scatter plot of effective wave height inversion based on multi-observation fields. DETAILED DESCRIPTION

[0033] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0034] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. The following description of at least one exemplary embodiment is actually only illustrative and is by no means intended to limit the present invention and its application or use. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] Figure 1 This is the flow chart of multivariable field shipborne GNSS-R effective wave height inversion;

[0036] A shipborne GNSS-R effective wave height inversion method based on multiple observation fields includes the following steps:

[0037] S1: Obtain real-time four-channel delay Doppler spectrum data and effective wave height values ​​generated by the ship along the route;

[0038] Delay Doppler map, referred to as DDM, is the power distribution of the navigation satellite reflection signal mapped from space to the delay Doppler domain;

[0039] S2: preprocessing the real-time four-channel delay Doppler spectrum data to obtain preprocessed four-channel delay Doppler spectrum data;

[0040] S3: extracting the observation quantity that is dependent on the effective wave height according to the pre-processed delay Doppler spectrum data;

[0041] S4: Based on 70% of the observation data, construct an effective wave height inversion model for multiple observation fields;

[0042] S5: Based on 30% of the observation data, an observation value sequence is extracted, and the observation value sequence is input into the effective wave height inversion model to obtain the effective wave height inversion value, and at the same time, a corrected observation value sequence is obtained, and a multivariate parameter correction model is established for the corrected observation value sequence to achieve effective wave height inversion.

[0043] The steps S1 / S2 / S3 / S4 / S5 are performed sequentially;

[0044] By mounting the receiver on the ship, the real-time four-channel DDM data generated by the GNSS-R receiver along the route and the effective wave height value of the radar wave meter are collected. Figure 2 In order to form a large scattering area on the sea surface for shipborne GNSS-R, since the antenna installation height is related to the ship height, the scattering area will be concentrated in a smaller delay chip interval, which contains the main information of the sea surface. Therefore, the interval of the delay chip for generating DDM is set to [-2,3].

[0045] The quality of the received signal is affected by many factors, such as satellite transmission power changes, atmospheric attenuation, multipath effects, receiver sensitivity, etc., which can cause signal distortion. The obtained DDM data needs to be quality controlled. Remove the low-quality delay Doppler image that does not form an effective scattering area, remove the frame data where the peak power of the reflected signal is less than the noise floor, and use the direct signal power as the factor to normalize the delay curve.

[0046] The observed quantities include the leading edge slope of the normalized delay curve, the average of the trailing edge slope, the delay window length, the delay Doppler spectrum volume and the satellite altitude angle.

[0047] The delay-Doppler spectrum volume is the integral of the power in the specified delay and Doppler interval centered at the mirror reflection point, and the calculation formula is as follows:

[0048]

[0049] Where: τ and f represent the delay chip and Doppler frequency shift, and DDM(τ,f) represents the power value at that point. The upper and lower limits of the delay integration area are -0.3069 chips and 0.3069 chips respectively, and the upper and lower limits of the Doppler frequency shift integration area are 100 Hz and -100 Hz respectively.

[0050] The formula for calculating the leading edge slope and trailing edge slope is as follows. Since the effective wave height is related to the slope change, a single stage slope is difficult to characterize the changing relationship between the delay curve and the effective wave height. The leading edge slope and trailing edge slope obtained from the five curves are averaged.

[0051]

[0052] Among them: LES is the leading edge slope, TES is the trailing edge slope;

[0053] Where τ i Indicates the sequential delay chip value, C W (τ i ) represents the corresponding curve waveform value; the number of selected points n=3, in order to improve the correlation with the effective wave height, in this embodiment, the leading edge slope and the trailing edge slope adopt the average of the leading edge slope and the trailing edge slope of the five time delay curves with the Doppler frequency centered at 0.

[0054] The number of normalized delay curves is five, and the delay window length is also the average of the five delay curve chips. The average of the delay curve chip lengths is calculated as follows:

[0055] TDWD=τ right -τ left

[0056] Where: τ right Indicates the right delay chip value intercepted in the curve, τ left They represent the left delay chip values ​​intercepted in the curve. The threshold set in intercepting the normalized delay curve is 0.85.

[0057] In the shipborne experiment, due to the relative motion between the satellite and the ship, the satellite elevation angle will gradually change slightly. The satellite elevation angle information can be directly obtained from the receiver, which is limited by the height and angle of the shipborne experiment antenna. The multi-observation field matrix elevation angle range is 40°-80°, with a resolution of 0.1°.

[0058] The process of preprocessing the real-time four-channel DDM data to obtain the preprocessed four-channel DDM data is as follows:

[0059] Normalize the delay curves at different Doppler frequency shifts;

[0060] The effective wave height value obtained by the wave meter is annotated with the four-channel DDM of each frame. First, all NAN values ​​are removed. In addition, a standardized delay waveform is established and a threshold is selected to remove low-quality data. The threshold range is 0-1. The larger the threshold, the more data is removed. Therefore, the threshold selection is not a fixed value and needs to be adjusted according to the actual data situation.

[0061] Furthermore, an effective wave height inversion model of a multi-observation measurement field is constructed, and the effective wave height of each frame DDM is labeled and acquired with the satellite elevation angle. The multi-observation measurement field includes a plurality of single observation measurement fields, and the single observation measurement field includes each observation quantity and the corresponding elevation angle and effective wave height combination. The multi-observation measurement field is reconstructed by reducing the dimension. In the process of constructing the effective wave height inversion model of the multi-observation measurement field, the matrix dimensions of each single observation measurement field are kept consistent.

[0062] A single observation field can be represented by a matrix:

[0063]

[0064] In the matrix, each point represents the LES value at the effective wave height and satellite altitude angle, and TES mn ,DDMV mn ,TDWD mn The same matrix form is maintained, and the matrix dimension changes with the effective wave height value and the altitude angle range. The matrix dimension of each single observation field remains consistent, and the combined multi-observation field F is:

[0065]

[0066] LES represents the leading edge slope, TES represents the trailing edge slope, DDMV represents the delay Doppler spectrum volume, TDWD represents the mean value of the delay curve chip length, and m and n represent the dimensions of the matrix;

[0067] By decomposing the combined multi-observation field, the main characteristic components of the matrix are extracted to reconstruct F, and then the corrected multi-observation field model is obtained. F can be expressed as:

[0068] F=ST

[0069] Where: S represents the coefficient matrix, T represents the eigenvector matrix

[0070] Each observation value is extracted from the given 30% DDM data. When a set of observation value sequences (LES, TES, DDMV, TDWD) is input into the reconstructed multi-observation matrix, the effective wave height inversion value of each single observation value is obtained, and the corrected observation value sequence is obtained at the same time. The multivariate parameter model of the corrected value sequence is established as follows:

[0071]

[0072] Where Obs i represents the different observed values ​​after correction, a i represents the corresponding coefficient, and b is the constant term.

[0073] Figure 3 It is a scatter plot of effective wave height inversion based on multi-observation fields.

[0074] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements 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 shipborne GNSS-R effective wave height inversion method based on multiple observation fields, characterized by: The following steps are involved: Obtain real-time four-channel delay Doppler spectrum data and significant wave height values ​​generated by the ship along the route; Preprocessing the real-time four-channel delay Doppler spectrum data to obtain preprocessed four-channel delay Doppler spectrum data; Extract the observation quantity which is dependent on the significant wave height according to the preprocessed delay Doppler spectrum data; Based on 70% of the observation data, an effective wave height inversion model for multiple observation fields is constructed; Based on 30% of the observed data, the observed value sequence is extracted, and the observed value sequence is input into the effective wave height inversion model to obtain the effective wave height inversion value. At the same time, the corrected observed value sequence is obtained, and a multivariate parameter correction model is established for the corrected observed value sequence to achieve effective wave height inversion.

2. The method for inverting effective wave height based on multi-observation fields of shipborne GNSS-R according to claim 1, characterized in that: The observed quantities include the leading edge slope of the normalized delay curve, the average of the trailing edge slope, the delay window length, the delay Doppler spectrum volume and the satellite altitude angle.

3. The method for inverting effective wave height based on multi-observation fields of shipborne GNSS-R according to claim 1, characterized in that: The process of preprocessing the real-time four-channel delay Doppler spectrum data to obtain the preprocessed four-channel delay Doppler spectrum data is as follows: Normalize the delay curves at different Doppler frequency shifts; The four-channel delay Doppler spectrum data of each frame are annotated with the effective wave height value obtained by the wave meter. First, all NAN values ​​are removed. In addition, a standardized delay waveform is established and a threshold is selected to remove low-quality data.

4. The method for inverting effective wave height based on multi-observation fields of shipborne GNSS-R according to claim 2, characterized in that: The number of normalized delay curves is five, and the delay window length is also the average of the five delay curve chips. The average of the delay curve chip lengths is calculated as follows: TDWD=τ right -t left Where τ right and τ left They respectively represent the right and left delay code chip values ​​intercepted in the curve.

5. The method for inverting effective wave height based on multi-observation fields of shipborne GNSS-R according to claim 1, characterized in that: The multi-observation measurement field includes a plurality of single observation measurement fields, wherein the single observation measurement field includes each observation quantity and a corresponding altitude angle and effective wave height combination; the matrix dimensions of each single observation measurement field remain consistent.

6. The method for inverting effective wave height based on multi-observation fields of shipborne GNSS-R according to claim 1, characterized in that: The satellite elevation angle range is 40°-80°, and the resolution is 0.1°.

7. The method for inverting effective wave height based on multi-observation fields of shipborne GNSS-R according to claim 1, characterized in that: The expression of the multivariate parameter correction model is as follows: Where: Obs i represents different observation values, a i represents the corresponding coefficient, and b is the constant term.

8. The method for inverting effective wave height based on multi-observation fields of shipborne GNSS-R according to claim 1, characterized in that: Based on 70% of the observation data, the process of constructing the effective wave height inversion model of multiple observation fields is as follows: The single observation field is represented by a matrix: In the matrix, each point represents the LES value at the effective wave height and satellite altitude angle, and TES mn ,DDMV mn ,TDWD mn The same matrix form is maintained, and the matrix dimension changes with the effective wave height value and the altitude angle range. The matrix dimension of each single observation field remains consistent. The effective wave height inversion model F of the combined multi-observation field is: LES represents the leading edge slope, TES represents the trailing edge slope, DDMV represents the delay Doppler spectrum volume, TDWD represents the mean of the delay curve chip length, and m and n represent the dimensions of the matrix.

Images