CN-121978648-A - High-precision intelligent inversion method for soil humidity of complex surface based on satellite-borne GNSS-S

CN121978648ACN 121978648 ACN121978648 ACN 121978648ACN-121978648-A

Abstract

A high-precision intelligent inversion method for the soil humidity of complex earth surface based on satellite-borne GNSS-S uses multiple navigation satellites as signal sources, acquires multi-angle double-station image signals of complex earth surface through GNSS-S radar carried by low-orbit satellites, and uses the multi-view processed data to realize full-link radiometric calibration from signal transmission to receiving by adopting precision measurement equipment (1), constructs a multi-parameter characteristic database containing earth surface physical characteristics and scattering characteristics, and (3) constructs a deep learning intelligent inversion model based on GNSS-S radar observation data and characteristic database to realize quantitative inversion of soil humidity.

Inventors

ZHANG CHUANG
XIA ZHENGHUAN
ZHAO ZHILONG
LIU XIN
ZHANG TAO
LIU ZONGQIANG
ZHANG YING
JIN SHICHAO
YUE FUZHAN

Assignees

航天恒星科技有限公司

Dates

Publication Date: 20260505
Application Date: 20251230

Claims (10)

1. A complex earth surface soil humidity high-precision intelligent inversion method based on a satellite-borne GNSS-S is characterized by comprising the following steps: acquiring multi-angle GNSS-S radar images of the complex earth surface of the test area through the GNSS-S radar; carrying out full-link radiation calibration on the GNSS-S radar image transmission link; Carrying out earth surface physical characteristic acquisition and scattering characteristic actual measurement on a test area, and constructing a multi-parameter characteristic database according to the obtained information; and training and deploying an intelligent inversion model for performing complex earth surface soil humidity inversion processing.
2. The high-precision intelligent inversion method for the soil humidity of the complex earth surface based on the satellite-borne GNSS-S is characterized by comprising the following steps of: the method for acquiring the multi-angle GNSS-S radar image of the complex earth surface through the GNSS-S radar comprises the following steps: in a preset height range, a GNSS-S radar receiver is deployed on a low-orbit satellite platform, and original echo signals, scattered by complex earth surfaces, of the designated wave band signals transmitted by all navigation satellites are synchronously received; Performing double-station synthetic aperture radar imaging processing on the original echo signals to generate corresponding double-station angle complex image data; And performing multi-view processing on the double-station angle complex image data, and improving the signal-to-noise ratio of the image to a preset range to obtain a multi-angle GNSS-S radar image.
3. The high-precision intelligent inversion method for the soil humidity of the complex earth surface based on the satellite-borne GNSS-S is characterized by comprising the following steps of: The method for carrying out full-link radiation calibration comprises the following steps: Dividing the full link into a receiving link and a transmitting link, calibrating the receiving link through a beacon device, calibrating the transmitting link through a receiver, and carrying out full link joint calibration through an active transponder after the calibration is finished.
4. The complex earth surface soil humidity high-precision intelligent inversion method based on the satellite-borne GNSS-S, which is characterized by comprising the following steps of: The method for constructing the multi-parameter characteristic database comprises the following steps: acquiring multispectral satellite images of complex earth surfaces of a test area, extracting normalized vegetation indexes, and recording vegetation types and growth periods according to ground investigation; Measuring the root mean square height of soil by using a laser topography scanner, and measuring the relevant length by using a surface roughness meter; adopting an unmanned aerial vehicle to carry a GNSS-S receiver, and collecting scattering data of each angle in a typical scene; recording azimuth angles and double-station scattering angles of the navigation satellite, and extracting backward scattering coefficients and image statistical characteristics; And uniformly summarizing all the obtained parameter information to construct a multi-parameter characteristic database.
5. The high-precision intelligent inversion method for the soil humidity of the complex earth surface based on the satellite-borne GNSS-S is characterized by comprising the following steps of: the method for training and deploying the intelligent inversion model comprises the following steps: data acquisition is carried out according to the multi-parameter characteristic database, and truth value sampling specification is carried out by utilizing the required data; presetting an intelligent inversion model architecture, selecting data in a multi-parameter characteristic database as a test data set, performing performance test on the inversion model architecture, and performing model training by using the test data set after the performance test is passed until the inversion task requirement is met.
6. The high-precision intelligent inversion method for the soil humidity of the complex earth surface based on the satellite-borne GNSS-S is characterized by comprising the following steps of: the truth value sampling standard is that surface vegetation is cleared in a longitude and latitude error range appointed by a SAR image pixel center point of required data; Measuring the water content of the soil volume by using a frequency domain reflectometer, sampling according to five points of east, west, south, north and center, wherein the sampling interval is smaller than the image resolution; And taking the five-point average value as a true value label of the pixel soil humidity of the current test area, namely, the true value sampling standard.
7. The high-precision intelligent inversion method for the soil humidity of the complex earth surface based on the satellite-borne GNSS-S is characterized by comprising the following steps of: The intelligent inversion model is built by adopting a support vector regression machine matched with a radial basis function, taking 16-dimensional feature vector data in a multi-parameter feature database as an input layer architecture, and taking a soil humidity value of a complex surface of a test area as model output.
8. The high-precision intelligent inversion method for the soil humidity of the complex earth surface based on the satellite-borne GNSS-S is characterized by comprising the following steps of: The intelligent inversion model takes data in a multi-parameter characteristic database of a typical earth surface target scene as an input data set, collects and measures field data consistent with scattering information time and space according to longitude and latitude information when the data are collected, obtains the earth surface target characteristic data set as an output data set, trains the intelligent inversion model by utilizing the input data set and the output data set, and the depth intelligent inversion model obtained after training is used for giving out soil humidity information.
9. The high-precision intelligent inversion method for the soil humidity of the complex earth surface based on the satellite-borne GNSS-S is characterized by comprising the following steps of: In the process of collecting the physical characteristics of the earth surface of a test area, detecting the true value of the soil humidity of the area, selecting vegetation coverage sampling points in the earth surface area, recording longitude and latitude, altitude and vegetation parameters, removing surface coverage, inserting a soil moisture measuring instrument, recording moisture values and temperature, taking the sampling points as the center, measuring the moisture values at five points in the four directions, and averaging to realize test data collection.
10. The high-precision intelligent inversion method for the soil humidity of the complex earth surface based on the satellite-borne GNSS-S is characterized by comprising the following steps of: In the radiation calibration flow, the beacon is used for calibrating the radiation characteristic of the receiving link and is used for calibrating the gain of the receiving antenna The calibration error is less than or equal to 0.3dB; Measuring the ground power flow density of the Beidou/GNSS satellites by using a receiver, constructing a Beidou/GNSS satellite radiation power library based on deep learning to realize the prediction of the Beidou/GNSS satellite radiation power under any configuration, and calibrating the navigation satellite radiation power The calibration error is less than or equal to 0.4dB; calibrating full-link radiation characteristics by using precise active transponder, and calibrating scattering cross-sectional area of detection target in radar equation The calibration error is less than or equal to 1dB and is used for realizing the full-link variable calibration of the GNSS-S radar.

Description

High-precision intelligent inversion method for soil humidity of complex surface based on satellite-borne GNSS-S Technical Field The invention relates to a high-precision intelligent inversion method for soil humidity of a complex surface based on a satellite-borne GNSS-S, and belongs to the field of radar technology design in the electronic information industry. Background Soil humidity is a main source of moisture absorption of crops, is a carrier for dissolving and transporting soil nutrients, and plays a vital role in crop growth. Therefore, the realization of high-precision measurement of soil humidity has important significance for agricultural drought monitoring and crop yield estimation. The most promising way of measuring soil humidity is by microwave remote sensing. Compared with optical remote sensing, the microwave remote sensing has longer wavelength, has the advantages of all-day and all-weather measurement and the like, has certain penetrability to soil, and can effectively make up the defect of the optical remote sensing in soil humidity measurement. The existing microwave remote sensing technology mainly comprises technical means such as SAR (SYNTHETIC APERTURE RADAR, active microwave remote sensing), radiometers (passive microwave remote sensing), GNSS-R (Global Navigation SATELLITE SYSTEM-Reflectometry) and the like. SAR is a high-resolution active microwave imaging radar for measuring earth surface soil humidity, but because the acquired earth surface scattering information is single, SAR is difficult to completely remove the influence of factors such as vegetation coverage, earth surface roughness and the like, so that the soil humidity measurement precision under a complex earth surface environment is lower. SAR power consumption is large, full rail operation is impossible, and earth surface soil humidity cannot be measured for a long time. The radiometer is a passive detection technology, and inversion of soil humidity is achieved by receiving microwave radiation brightness temperature of soil, and soil humidity measurement accuracy in complex surface environments is low due to single surface scattering information. Radiometers face problems such as complex systems, high costs, etc. GNSS-R is a passive detection technology, and inversion of soil humidity is realized by receiving reflection signals of mirror surface areas, and the soil humidity measurement accuracy under a complex surface environment is lower due to single surface scattering information. GNSS-R can only realize soil humidity measurement of a limited number of specular reflection areas, and cannot realize continuous measurement of large breadth. In summary, the existing remote sensing satellite detection means have the core problems that under the influence of complex surface environments (multi-layer heterogeneous medium) such as soil roughness, vegetation coverage and the like, various applicable conditions of the microwave remote sensing soil humidity inversion methods have various limitations, and the high-precision monitoring requirements of the surface soil humidity in a large area, in different areas and with different characteristics are difficult to meet. The earth surface information acquired by the existing microwave remote sensing detection means is single, and only an incomplete soil humidity inversion model (similar to a 'undercolor equation') can be constructed, so that the soil humidity information of various complex earth surfaces can not be accurately output. Disclosure of Invention Aiming at various technical defects existing in the prior art, the invention provides a high-precision intelligent inversion method for the soil humidity of the complex earth surface based on a satellite-borne GNSS-S. The invention solves the technical problems by the following technical proposal: A complex earth surface soil humidity high-precision intelligent inversion method based on a satellite-borne GNSS-S comprises the following steps: acquiring multi-angle GNSS-S radar images of the complex earth surface of the test area through the GNSS-S radar; carrying out full-link radiation calibration on the GNSS-S radar image transmission link; Carrying out earth surface physical characteristic acquisition and scattering characteristic actual measurement on a test area, and constructing a multi-parameter characteristic database according to the obtained information; and training and deploying an intelligent inversion model for performing complex earth surface soil humidity inversion processing. The method for acquiring the multi-angle GNSS-S radar image of the complex earth surface through the GNSS-S radar comprises the following steps: in a preset height range, a GNSS-S radar receiver is deployed on a low-orbit satellite platform, and original echo signals, scattered by complex earth surfaces, of the designated wave band signals transmitted by all navigation satellites are synchronously received; Performing double-station syn