Space-based GNSS-S and optical cooperation soil humidity monitoring system and on-orbit intelligent real-time inversion method

Abstract

The invention provides a space-based GNSS-S and optical collaborative soil humidity monitoring system, which is characterized in that a ground measurement and control and data processing center receives actual soil humidity information measured by a foundation soil humidity monitoring module through a Beidou short message; secondly, after space-time matching is carried out on observation data of a low-orbit soil humidity monitoring satellite within a certain period of time and actually measured soil humidity information, GNSS-S observed quantity characteristics and a mapping model of vegetation parameters extracted based on optical data and the actually measured soil humidity information are trained through a deep learning method; and then the mapping model is uploaded to a low-orbit soil humidity monitoring satellite, so that in-orbit collaborative intelligent real-time inversion of the soil humidity is realized based on observation data. Therefore, the system has the beneficial effects of high soil humidity inversion precision, strong real-time performance, wide monitoring range, high reliability, low system cost and the like.

Description

Technical Field

[0001] The present invention relates to the field of meteorological monitoring technology, and in particular to a space-based GNSS-S and optical collaborative soil moisture monitoring system and an on-orbit intelligent real-time inversion method. Background Art

[0002] Soil moisture is a key factor in regulating the water cycle between land and atmosphere, the balance of ecosystems, and energy exchange. Its fluctuations will have a significant impact on global climate change. In addition, soil moisture information plays an important role in many fields such as agriculture, meteorology, and hydrology. Therefore, real-time and accurate monitoring of soil moisture is of great significance for climate change prediction, global change research, agricultural drought monitoring, and crop yield estimation. Although the traditional in-situ measurement method is the most accurate method for estimating soil moisture, it is costly, time-consuming, and only provides point data, making it difficult to achieve high-revisit and high-efficiency soil moisture monitoring on regional and global scales.

[0003] With the rapid development of satellite remote sensing technology, it is possible to monitor soil moisture with higher temporal and spatial resolution at a lower cost and low consumption. Since the method of remote sensing monitoring of soil moisture was proposed in the 1970s, optical remote sensing and microwave remote sensing have become the main means of satellite monitoring of soil moisture. Optical remote sensing inversion of soil moisture usually obtains surface information based on thermal inertia and vegetation index, and establishes a conversion relationship with soil moisture. Optical remote sensing images have high spatial resolution and rich spectral information, but poor penetration and are easily affected by bad weather, making them unsuitable for long-term monitoring. Microwave remote sensing makes up for this defect well. Due to its longer wavelength, microwave radiation has stronger penetration and is not affected by weather conditions and acquisition time. Therefore, it can be monitored in real time all day and all weather. However, microwave remote sensing has a worse spatial resolution than optical remote sensing and will be interfered by surface vegetation scattering. Therefore, relevant scholars combined the advantages of the two, used optical remote sensing to eliminate the influence of the vegetation layer, and then studied the conversion relationship between microwave radar scattering coefficient and soil moisture. It is of great scientific significance to quantitatively invert soil moisture through multi-source remote sensing data.

[0004] However, soil moisture is a dynamic variable that requires high real-time monitoring. The currently used remote sensing monitoring methods have a long lag and cannot meet the needs of disaster monitoring and management, agricultural decision support, water resources management, water science research, etc. for real-time acquisition of soil moisture. Therefore, it is urgent to establish a low-orbit soil moisture monitoring system and an on-orbit intelligent real-time inversion method. Summary of the invention

[0005] The purpose of the present invention is to provide a space-based GNSS-S and optical collaborative soil moisture monitoring system and an on-orbit intelligent real-time inversion method, which can improve the timeliness of soil moisture information acquisition and enhance the potential of remote sensing means such as GNSS-S technology in land application fields such as agricultural decision support and disaster monitoring and management.

[0006] In order to achieve the above-mentioned object, on the one hand, the present invention provides a space-based GNSS-S and optical collaborative soil moisture monitoring system, comprising:

[0007] A low-orbit soil moisture monitoring satellite is used to capture and track ground scattering signals from navigation satellites in a detection area and collect high-resolution images, and to process the ground scattering signals and the high-resolution images in real time in combination with detection system state parameters to obtain a time-delay Doppler map and a multi-spectral high-resolution image; regularly transmit the system state parameters and first observation data including the time-delay Doppler map and the multi-spectral high-resolution image to a ground measurement, control and data processing center, and receive a mapping model of observation data features and measured soil moisture information uploaded by the ground measurement, control and data processing center, and obtain global soil moisture information through real-time inversion of on-orbit observation data using the mapping model;

[0008] A ground-based soil moisture monitoring module, used to continuously measure soil moisture and its corresponding spatiotemporal information, and upload the second observation data including the soil moisture and the spatiotemporal information to the ground measurement, control and data processing center through a first Beidou short message terminal; wherein the spatiotemporal information includes a collection time and a collection location;

[0009] The ground measurement, control and data processing center is used to regularly receive the first observation data and the system status parameters transmitted by the low-orbit soil moisture monitoring satellite, and receive the second observation data measured by the ground-based soil moisture monitoring module, and perform spatiotemporal matching between the first observation data and the first observation data to form a training set, and train the mapping model of GNSS-S observation quantity characteristics, vegetation parameters extracted based on optical data and measured soil moisture information through deep learning based on the training set, and upload the mapping model to the low-orbit soil moisture monitoring satellite.

[0010] Optionally, the low-orbit soil moisture monitoring satellite includes:

[0011] A low-orbit GNSS-S system is used to receive, track and analyze direct signals from GNSS satellites and related scattered signals to obtain status data of the GNSS satellites, location information of scattered points, and time delay and Doppler information of scattered signals;

[0012] A low-orbit optical system is used to synchronously acquire multi-spectral high-resolution images of GNSS signals in ground scattering areas and extract vegetation parameters in the ground scattering areas;

[0013] The onboard computing module is used to complete the soil moisture monitoring in real time on orbit according to the GNSS-S observation quantity characteristics regularly uploaded by the ground measurement and control and data processing center, the vegetation parameters extracted based on the optical data and the mapping model of the measured soil moisture information;

[0014] A satellite-to-ground link, used to regularly transmit the first observation data and the system status parameters to the ground measurement, control and data processing center, and to upload the mapping model trained by the deep learning network to the ground measurement, control and data processing center;

[0015] An onboard storage module, used to store the data collected by the low-orbit GNSS-S system and the low-orbit optical system, and the soil moisture data analyzed and obtained by the onboard computing module;

[0016] The onboard power supply module is used to supply power to each module of the low-orbit soil moisture monitoring satellite.

[0017] Optionally, the low-orbit soil moisture monitoring satellite also includes:

[0018] GNSS-S antenna, used for simultaneously receiving dual-frequency navigation satellite scattered signals, the frequencies of the dual-frequency navigation satellite scattered signals are 1575.42MHz and 1227.60MHz, respectively, and the bandwidths are both 20.46MHz. The polarization mode of the GNSS-S antenna is left-hand circular polarization, and a phased array antenna system is adopted. The antenna normal gain is greater than 22dBi, and the number of beams is not less than 4;

[0019] Multispectral high-resolution camera, used to collect multispectral high-resolution images in the GNSS signal scattering area.

[0020] Furthermore, the multi-spectral high-resolution camera is installed on a turntable device, and the shooting area is dynamically adjusted based on the beam pointing of the GNSS-S antenna through the turntable device, so that the coverage area of ​​the multi-spectral high-resolution camera is consistent with the observation area of ​​the GNSS-S antenna.

[0021] Optionally, the foundation soil moisture monitoring module includes:

[0022] Soil moisture detector, used to collect soil moisture within 6 to 10 cm of the ground;

[0023] A satellite positioning module, used to obtain the position and measurement time of the soil moisture detector measuring the soil moisture;

[0024] A data storage device, used for storing the collected data of the soil moisture detector and the time and space information acquired by the satellite positioning module;

[0025] A solar power module, used to continuously supply power to the foundation soil moisture monitoring module;

[0026] The first Beidou short message terminal is used to upload the data information collected by the soil moisture detector and the satellite positioning module to the ground measurement, control and data processing center.

[0027] Furthermore, it includes a plurality of the ground-based soil moisture monitoring modules, and the plurality of the ground-based soil moisture monitoring modules are distributed in typical vegetation coverage scenes around the world.

[0028] Optionally, the ground measurement, control and data processing center includes:

[0029] A high-performance server is used to form a training set by temporal and spatial matching of the GNSS scattered signal data collected by the low-orbit GNSS-S system, the soil roughness and vegetation index inverted based on the images collected by the low-orbit optical system, and the true value of soil moisture actually collected by the ground-based soil moisture monitoring module; and based on the training set, continuously train the mapping model of GNSS-S observation quantity characteristics, vegetation parameters extracted based on optical data, and measured soil moisture information through a deep learning network;

[0030] The satellite-to-ground link is used to receive the first observation data and system status parameters transmitted by the low-orbit soil moisture monitoring satellite, and upload the mapping model trained by the deep learning network to the low-orbit soil moisture monitoring satellite;

[0031] The second Beidou short message terminal is used to receive the second observation data collected by the ground-based soil moisture monitoring module.

[0032] Furthermore, the ground measurement, control and data processing center is specifically used to construct a training set based on the first observation data and the second observation data accumulated over a preset time, and continuously train the mapping relationship between the GNSS-S observation quantity characteristics, the vegetation parameters extracted based on the optical data and the measured soil moisture information through a deep learning network based on the training set to obtain the mapping model and upload it to the low-orbit soil moisture monitoring satellite; and continuously receive the first observation data of the low-orbit soil moisture monitoring satellite and the second observation data of the ground-based soil moisture monitoring module, and regularly update the mapping model before uploading it to the low-orbit soil moisture monitoring satellite.

[0033] On the other hand, the present invention also provides an on-orbit intelligent real-time inversion method based on any of the above-mentioned space-based GNSS-S and optical collaborative soil moisture monitoring systems, comprising the steps of:

[0034] Preprocessing the delay-Doppler map obtained after GNSS-S signal processing based on the signal-to-noise ratio;

[0035] Normalizing the delay-Doppler image and extracting typical features to use them as a model input main vector;

[0036] Performing image processing on the multispectral high-resolution image to obtain land cover type, leaf area index, plant optical thickness, vegetation parameters of normalized vegetation parameters and soil roughness corresponding to the scattering area, and taking the vegetation parameters, the soil roughness and designated parameters as input vectors; wherein the designated parameters include the longitude and latitude of the scattering point, the scattering angle and the scanning beam gain of the downward-looking phased array antenna;

[0037] A training set, a validation set, and a test set of a mapping model are configured according to the input main vector and the input slave vector;

[0038] Setting up a deep learning architecture, determining the dimension of the input layer of the mapping model, adding hidden layers and output layers, selecting a loss function and minimizing the loss function by configuring an optimizer;

[0039] Training the mapping model according to the training data in the training set, and periodically verifying the model through the verification set to monitor the performance of the model, and dynamically adjusting hyperparameters to optimize the model performance;

[0040] Using the test set to evaluate the performance of the mapping model;

[0041] The trained mapping model is uploaded to the low-orbit soil moisture monitoring satellite, so that the low-orbit soil moisture monitoring satellite can perform on-orbit collaborative intelligent real-time inversion of soil moisture according to the actually measured GNSS-S data and multi-spectral high-resolution data through the mapping model.

[0042] Furthermore, the mapping model is trained and evaluated by a high-performance server of a ground measurement, control and data processing center; and the onboard computing module of the low-orbit soil moisture monitoring satellite uses the mapping model to perform on-orbit collaborative intelligent real-time inversion of soil moisture based on actually measured GNSS-S data and multi-spectral high-resolution data.

[0043] Compared with the current satellite remote sensing soil moisture system, the space-based GNSS-S and optical collaborative soil moisture monitoring system of the present invention is equipped with both optical and microwave observation methods on a satellite platform, and can realize collaborative observation of soil moisture in the same area, which greatly improves the reliability of satellite remote sensing monitoring of soil moisture. The on-orbit intelligent real-time inversion method of the present invention can continuously train GNSS scattered signal data, multi-spectral high-resolution optical images and measured soil moisture data through a deep learning network at the ground measurement and control and data processing center, and regularly upload the training and updated mapping model to the onboard computer, which has higher inversion accuracy and timeliness than the existing soil moisture satellite remote sensing detection system and inversion method. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 A system composition diagram of the space-based GNSS-S and optical collaborative soil moisture monitoring system provided by an embodiment of the present invention is shown;

[0045] Figure 2 An application scenario diagram of the space-based GNSS-S and optical collaborative soil moisture monitoring system provided by an embodiment of the present invention is shown;

[0046] Figure 3 A system composition diagram of the low-orbit soil moisture monitoring satellite of the space-based GNSS-S and optical collaborative soil moisture monitoring system provided by an embodiment of the present invention is shown;

[0047] Figure 4 A system composition diagram of the ground-based soil moisture monitoring module of the space-based GNSS-S and optical collaborative soil moisture monitoring system provided by an embodiment of the present invention is shown;

[0048] Figure 5 A system composition diagram of the ground measurement and control and data processing center of the space-based GNSS-S and optical collaborative soil moisture monitoring system provided by an embodiment of the present invention is shown;

[0049] Figure 6 A flowchart showing the steps of the on-orbit intelligent real-time inversion method provided by an embodiment of the present invention is shown;

[0050] Figure 7 A schematic diagram showing the system flow of the on-orbit intelligent real-time inversion method provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0052] It should be noted that references to "one embodiment", "embodiment", "example embodiment", etc. in this specification refer to the embodiment described, which may include specific features, structures or characteristics, but not every embodiment must include these specific features, structures or characteristics. In addition, such expressions do not refer to the same embodiment. Furthermore, when describing specific features, structures or characteristics in conjunction with an embodiment, whether or not there is an explicit description, it has been shown that it is within the knowledge of those skilled in the art to combine such features, structures or characteristics into other embodiments.

[0053] In addition, certain words are used in the specification and subsequent claims to refer to specific components or parts. Those with ordinary knowledge in the relevant field should understand that manufacturers can use different nouns or terms to refer to the same component or part. This specification and subsequent claims do not use differences in names as a way to distinguish components or parts, but use differences in the functions of components or parts as the criteria for distinction. "Including" and "including" mentioned throughout the specification and subsequent claims are open-ended terms and should be interpreted as "including but not limited to". In addition, the word "connected" here includes any direct and indirect electrical connection means. Indirect electrical connection means include connection through other devices.

[0054] Embodiment 1

[0055] Figure 1 The present invention shows a space-based GNSS-S and optical cooperative soil moisture monitoring system provided by an embodiment of the present invention. The system can be used for high-precision space-based soil moisture monitoring research on orbit. Specifically, the system includes a low-orbit soil moisture monitoring satellite 10, a ground-based soil moisture monitoring module 20, and a ground measurement and control and data processing center 30, wherein:

[0056] The low-orbit soil moisture monitoring satellite 10 is used to capture and track the ground scattering signals of the navigation satellite in the detection area and collect high-resolution images, and process the ground scattering signals and high-resolution images in real time in combination with the detection system state parameters to obtain a time-delay Doppler map and a multi-spectral high-resolution image; the system state parameters and the first observation data including the time-delay Doppler map and the multi-spectral high-resolution image are regularly transmitted to the ground measurement and control and data processing center 30, and a mapping model of the observation data characteristics and the measured soil moisture information uploaded by the ground measurement and control and data processing center 30 is received, and the mapping model is used to obtain the global soil moisture information in real time based on the on-orbit observation data; the ground-based soil moisture monitoring module 20 is used to continuously measure the soil moisture and its corresponding time and space Information, and upload the second observation data containing the soil moisture and spatiotemporal information to the ground measurement and control and data processing center 30 through the first Beidou short message terminal; wherein the spatiotemporal information includes the collection time and the collection location; the ground measurement and control and data processing center 30 is used to regularly receive the first observation data and system status parameters transmitted from the low-orbit soil moisture monitoring satellite 10, and receive the second observation data measured from the ground-based soil moisture monitoring module 20, and perform spatiotemporal matching between the first observation data and the first observation data to form a training set, and train the GNSS-S observation quantity characteristics, the mapping model based on the vegetation parameters extracted based on the optical data and the measured soil moisture information through deep learning based on the training set, and upload the mapping model to the low-orbit soil moisture monitoring satellite 10.

[0057] The ground-based soil moisture monitoring module 20 of this embodiment is specifically installed on the detection area, and is used to obtain the soil moisture status information in the area and upload it to the ground measurement and control and data processing center 30. At the same time, the ground measurement and control and data processing center 30 receives the first observation data and its system status parameters transmitted from the low-orbit soil moisture monitoring satellite 10, and then performs time-space matching on the first observation data and the second observation data of the measured soil moisture status. The time-space matching here specifically refers to matching two types of observation data at the same time and the same position, thereby generating several groups of matching data to form a training set; then using the constructed training set A mapping model of the mapping relationship between GNSS-S observation characteristics, vegetation parameters extracted based on optical data and measured soil moisture information is trained through deep learning, and the trained mapping model is uploaded to the low-orbit soil moisture monitoring satellite 10. The low-orbit soil moisture monitoring satellite 10 uses the mapping model to invert the soil moisture information of the corresponding detection area using the on-orbit observation data as the model input. In this way, the low-orbit soil moisture monitoring satellite 10 can realize large-scale on-orbit collaborative intelligent real-time inversion of soil moisture based on the on-orbit observation data, which has the beneficial effects of high soil moisture inversion accuracy, strong real-time performance, wide monitoring range, high reliability and low system cost.

[0058] The observation data and system status parameters obtained by the low-orbit soil moisture monitoring satellite 10 can be specifically regularly transmitted to the ground measurement, control and data processing center 30 through the Beidou short message terminal. Conversely, the ground measurement, control and data processing center 30 can also upload the trained mapping model to the low-orbit soil moisture monitoring satellite 10 through the Beidou short message terminal; of course, in other examples, other communication methods can also be used between the low-orbit soil moisture monitoring satellite 10 and the ground measurement, control and data processing center 30 to achieve data interaction; this embodiment does not limit this.

[0059] The first observation data specifically includes a time-delay Doppler image and a full-color multispectral image obtained through analysis and processing.

[0060] The ground measurement and control and data processing center 30 of this embodiment is specifically used to build a training set based on the first observation data and the second observation data collected during the preset time, and continuously train the mapping relationship between the GNSS-S observation characteristics, the vegetation parameters extracted based on the optical data and the measured soil moisture information through the deep learning network based on the training set to obtain the mapping model and upload it to the low-orbit soil moisture monitoring satellite 10; and continuously receive the first observation data of the low-orbit soil moisture monitoring satellite 10 and the second observation data of the ground-based soil moisture monitoring module 20, and regularly update the mapping model and upload it to the low-orbit soil moisture monitoring satellite 10. See Figure 2 In the specific implementation, the low-orbit soil moisture monitoring satellite 10 first collects GNSS ground scattered signals and multi-spectral high-resolution image data for no less than one year, and regularly transmits the data to the ground measurement and control and data processing center 30. At the same time, the ground-based soil moisture monitoring module 20 with full vegetation type coverage also uploads the measured time, position, and soil moisture true value to the ground measurement and control and data processing center 30; the ground measurement and control and data processing center 30 pre-processes the received GNSS scattered signal data, the soil roughness and vegetation index inverted based on the low-orbit optical system collected image, and the soil moisture true value collected by the ground-based soil moisture monitoring module 20, and then performs time-space matching, accumulating one year. After that, the deep learning network is trained to obtain a mapping model of GNSS-S observation characteristics, vegetation parameters extracted based on optical data and measured soil moisture information, and then the model is uploaded to the low-orbit soil moisture monitoring satellite 10. At this time, the low-orbit soil moisture monitoring satellite 10 has an on-orbit intelligent real-time soil moisture monitoring function based on the mapping model; after that, the ground measurement and control and data processing center 30 continuously optimizes and trains the mapping model based on the continuously received GNSS-S data, multi-spectral high-resolution images and soil moisture true values, and regularly uploads the trained and updated mapping model to the low-orbit soil moisture monitoring satellite 10 to continuously improve the soil moisture monitoring accuracy.

[0061] See also Figure 3 In an optional embodiment, the low-orbit soil moisture monitoring satellite 10 includes a low-orbit GNSS-S system 101, a low-orbit optical system 102, a satellite-borne computing module 103, a satellite-to-ground link 104, a satellite-borne storage module 105, and a satellite-borne power supply module 106, wherein:

[0062] The low-orbit GNSS-S system 101 is used to receive, track and analyze the direct signals of the GNSS satellite and the related scattered signals to obtain the status data of the GNSS satellite, the location information of the scattering point and the delay and Doppler information of the scattered signals, and provide a signal data source for soil moisture monitoring; the low-orbit optical system 102 is used to synchronously obtain the multi-spectral high-resolution image of the GNSS signal in the ground scattering area, extract the vegetation parameters in the ground scattering area, and provide auxiliary parameters for soil moisture monitoring; the onboard computing module 103 is used to extract the GNSS-S observation characteristics regularly uploaded by the ground measurement and control and data processing center 30, based on the optical data The mapping model of vegetation parameters and measured soil moisture information is used to complete soil moisture monitoring in real time on orbit; the satellite-to-ground link 104 is used to regularly transmit the first observation data and system status parameters to the ground measurement and control and data processing center 30, and upload the mapping model trained by the deep learning network to the ground measurement and control and data processing center 30; the on-board storage module 105 is used to store the data collected by the low-orbit GNSS-S system 101 and the low-orbit optical system 102, and the soil moisture data analyzed by the on-board computing module 103; the on-board power supply module 106 is used to power each module of the low-orbit soil moisture monitoring satellite 10.

[0063] Furthermore, the low-orbit soil moisture monitoring satellite 10 also includes the following key payloads: a GNSS-S antenna, which is used to simultaneously receive dual-frequency navigation satellite scattered signals, with frequencies of 1575.42MHz and 1227.60MHz respectively, and bandwidths of 20.46MHz. The polarization mode of the GNSS-S antenna is left-hand circular polarization, and a phased array antenna system is adopted. The antenna normal gain is greater than 22dBi, and the number of beams is not less than 4; a multi-spectral high-resolution camera, which is used to collect multi-spectral high-resolution images in the GNSS signal scattering area; it has the ability to collect more than 8 bands, and a single image covers an area of ​​not less than 100km×100km, with a spatial resolution better than 1m.

[0064] In order to ensure that the coverage area of ​​the multi-spectral high-resolution camera is consistent with the observation area of ​​the GNSS-S antenna, this embodiment installs the multi-spectral high-resolution camera on a turntable device, and dynamically adjusts the shooting area based on the beam pointing of the GNSS-S antenna through the turntable device to ensure that the coverage area of ​​the multi-spectral high-resolution camera is consistent with the observation area of ​​the GNSS-S antenna.

[0065] See also Figure 4 In an optional embodiment, the ground-based soil moisture monitoring module 20 includes a soil moisture detector 201, a satellite positioning module 202, a data storage device 203, a solar power module 204 and a first Beidou short message terminal 205, wherein:

[0066] The soil moisture detector 201 is used to collect soil moisture within 6 to 10 cm of the ground;

[0067] The satellite positioning module 202 is used to obtain the position and measurement time of the soil moisture detector measuring soil moisture; specifically, the satellite positioning module 202 can simultaneously receive GPS, BDS-3 (Beidou-3), GALILEO (Galileo Satellite Navigation System) and GLONASS (GLONASS Satellite Navigation System) satellite signals to obtain the position information and data sampling time of the corresponding soil detection area; the data storage device 203 is used to store the collected data of the soil moisture detector and the time and space information obtained by the satellite positioning module; that is, it is used to store information such as soil moisture collection time, position and soil moisture true value, where the soil moisture true value specifically refers to the real data actually collected by the soil moisture detector 201; the solar power supply module 204 is used to continuously power the ground-based soil moisture monitoring module 20; the first Beidou short message terminal 205 is used to upload the data information collected by the soil moisture detector 201 and the satellite positioning module 202 to the ground measurement and control and data processing center 30.

[0068] Furthermore, the present embodiment specifically includes a plurality of the foundation soil moisture monitoring modules 20, and the plurality of foundation soil moisture monitoring modules 20 are distributed in typical vegetation coverage scenes around the world. That is, the foundation soil moisture monitoring modules 20 are configured and installed in typical vegetation coverage scenes such as cold-temperate coniferous forests, deciduous broad-leaved forests, evergreen broad-leaved forests, monsoon forests, mangroves, deciduous broad-leaved shrubs, shrubs, grasslands, meadows, swamps, aquatic vegetation and bare soil.

[0069] See also Figure 5 In an optional implementation, the ground measurement and control and data processing center 30 includes a high-performance server 301, a satellite-to-ground link 302, and a second Beidou short message terminal 303, wherein:

[0070] The high-performance server 301 is used to form a training set by time-space matching the GNSS scattered signal data collected by the low-orbit GNSS-S system 101, the soil roughness and vegetation index inverted based on the image collected by the low-orbit optical system 102, and the soil moisture true value actually collected by the ground-based soil moisture monitoring module 20; and based on the training set, continuously train the GNSS-S observation quantity characteristics, the mapping model of the vegetation parameters extracted based on the optical data and the measured soil moisture information through the deep learning network; the satellite-to-ground link 302 is used to receive the first observation data and system status parameters transmitted from the low-orbit soil moisture monitoring satellite 10, and upload the mapping model trained by the deep learning network to the low-orbit soil moisture monitoring satellite 10; the second Beidou short message terminal 303 is used to receive the second observation data collected from the ground-based soil moisture monitoring module 20, that is, the soil moisture true value information.

[0071] Compared with the current satellite remote sensing soil moisture system, the space-based GNSS-S and optical collaborative soil moisture monitoring system provided in this embodiment is equipped with both optical and microwave observation methods on a single satellite platform, and can realize collaborative observation of soil moisture in a unified area, greatly improving the reliability of satellite remote sensing monitoring of soil moisture.

[0072] Embodiment 2

[0073] Figure 6 The on-orbit intelligent real-time inversion method provided by the second embodiment of the present invention is shown. The method is implemented based on the space-based GNSS-S and optical collaborative soil moisture monitoring system of the first embodiment, and specifically includes the following steps:

[0074] S101: pre-processing the delay-Doppler map obtained after GNSS-S signal processing based on the signal-to-noise ratio.

[0075] S102: normalizing the delay-Doppler image and extracting typical features to serve as the main vector of the model input; that is, using the typical features extracted from the delay-Doppler image as the main vector of the input of the mapping model between the GNSS-S observation characteristics, the vegetation parameters extracted based on the optical data and the measured soil moisture information.

[0076] S103: performing image processing on the multispectral high-resolution image to obtain land cover type, leaf area index, plant optical thickness, vegetation parameters of normalized vegetation parameters, and soil roughness corresponding to the scattering area, and using the vegetation parameters, the soil roughness, and designated parameters as input slave vectors; wherein the designated parameters include the longitude and latitude of the scattering point, the scattering angle, and the scanning beam gain of the downward-looking phased array antenna;

[0077] S104: configuring a training set, a validation set and a test set of a mapping model according to the input main vector and the input slave vector; preferably, using 70%-80% of the data for training, 10%-15% of the data for development, and the rest for testing.

[0078] S105: Setting a deep learning architecture, determining the dimension of the input layer of the mapping model, adding hidden layers and output layers, selecting a loss function, and minimizing the loss function by configuring an optimizer; that is, selecting an appropriate deep learning architecture in step S105, determining the dimension of the input layer, adding hidden layers and output layers, selecting an appropriate loss function, and minimizing the loss function by configuring an optimizer.

[0079] S106: The mapping model is trained according to the training data in the training set, and periodically verified through the verification set to monitor the performance of the model, and dynamically adjust hyperparameters to optimize the model performance.

[0080] S107: Using the test set to perform performance evaluation on the mapping model; specifically, evaluating the performance of the model on the test set, and calculating various evaluation indicators, such as root mean square error and determination coefficient.

[0081] S108: Upload the trained mapping model to the low-orbit soil moisture monitoring satellite, so that the low-orbit soil moisture monitoring satellite can perform on-orbit collaborative intelligent real-time inversion of soil moisture according to the actually measured GNSS-S data and multi-spectral high-resolution data through the mapping model.

[0082] Furthermore, the mapping model is trained and evaluated by a high-performance server of a ground measurement, control and data processing center; and the onboard computing module of the low-orbit soil moisture monitoring satellite uses the mapping model to perform on-orbit collaborative intelligent real-time inversion of soil moisture based on actually measured GNSS-S data and multi-spectral high-resolution data.

[0083] See also Figure 7 The on-orbit intelligent real-time inversion method provided in this embodiment can continuously train GNSS scattered signal data, multi-spectral high-resolution optical images and measured soil moisture data through a deep learning network in the ground measurement and control and data processing center, and regularly upload the trained and updated mapping model to the onboard computer. Compared with the existing soil moisture satellite remote sensing detection system and inversion method, it has higher inversion accuracy and timeliness.

[0084] In summary, the space-based GNSS-S and optical collaborative soil moisture monitoring system provided by the present invention receives the actual soil moisture information measured by the ground-based soil moisture monitoring module through the Beidou short message by the ground measurement and control and data processing center, and then performs spatiotemporal matching of the observation data of the low-orbit soil moisture monitoring satellite within a certain period of time with the measured soil moisture information, and then trains the mapping model of the GNSS-S observation quantity characteristics, the vegetation parameters extracted based on the optical data and the measured soil moisture information through a deep learning method, and then uploads the mapping model to the low-orbit soil moisture monitoring satellite, thereby realizing the on-orbit collaborative intelligent real-time inversion of soil moisture based on the observation data. An on-orbit intelligent real-time inversion method based on the system is also provided. In this way, the present invention can effectively improve the timeliness of soil moisture information acquisition, and enhance the potential of remote sensing means such as GNSS-S technology in land application fields such as agricultural decision support and disaster monitoring and management; it has the beneficial effects of high soil moisture inversion accuracy, strong real-time performance, wide monitoring range, high reliability, and low system cost.

[0085] Of course, the present invention may have many other embodiments. Without departing from the spirit and essence of the present invention, those skilled in the art may make various corresponding changes and modifications based on the present invention, but these corresponding changes and modifications should all fall within the scope of protection of the claims attached to the present invention.

Claims

1. A space-based GNSS-S and optical collaborative soil moisture monitoring system, characterized in that: Included are: A low-orbit soil moisture monitoring satellite is used to capture and track ground scattering signals from navigation satellites in a detection area and collect high-resolution images, and to process the ground scattering signals and the high-resolution images in real time in combination with detection system state parameters to obtain a time-delay Doppler map and a multi-spectral high-resolution image; regularly transmit the system state parameters and first observation data including the time-delay Doppler map and the multi-spectral high-resolution image to a ground measurement, control and data processing center, and receive a mapping model of observation data features and measured soil moisture information uploaded by the ground measurement, control and data processing center, and obtain global soil moisture information through real-time inversion of on-orbit observation data using the mapping model; A ground-based soil moisture monitoring module, used to continuously measure soil moisture and its corresponding spatiotemporal information, and upload the second observation data including the soil moisture and the spatiotemporal information to the ground measurement, control and data processing center through a first Beidou short message terminal; wherein the spatiotemporal information includes a collection time and a collection location; The ground measurement, control and data processing center is used to regularly receive the first observation data and the system status parameters transmitted by the low-orbit soil moisture monitoring satellite, and receive the second observation data measured by the ground-based soil moisture monitoring module, and perform spatiotemporal matching between the first observation data and the first observation data to form a training set, and train the mapping model of GNSS-S observation quantity characteristics, vegetation parameters extracted based on optical data and measured soil moisture information through deep learning based on the training set, and upload the mapping model to the low-orbit soil moisture monitoring satellite.

2. The space-based GNSS-S and optical collaborative soil moisture monitoring system according to claim 1 is characterized in that: The low-orbit soil moisture monitoring satellite includes: A low-orbit GNSS-S system is used to receive, track and analyze direct signals from GNSS satellites and related scattered signals to obtain status data of the GNSS satellites, location information of scattered points, and time delay and Doppler information of scattered signals; A low-orbit optical system is used to synchronously acquire multi-spectral high-resolution images of GNSS signals in ground scattering areas and extract vegetation parameters in the ground scattering areas; The onboard computing module is used to complete the soil moisture monitoring in real time on orbit according to the GNSS-S observation quantity characteristics regularly uploaded by the ground measurement and control and data processing center, the vegetation parameters extracted based on the optical data and the mapping model of the measured soil moisture information; A satellite-to-ground link, used to regularly transmit the first observation data and the system status parameters to the ground measurement, control and data processing center, and to upload the mapping model trained by the deep learning network to the ground measurement, control and data processing center; An onboard storage module, used to store the data collected by the low-orbit GNSS-S system and the low-orbit optical system, and the soil moisture data analyzed and obtained by the onboard computing module; The onboard power supply module is used to supply power to each module of the low-orbit soil moisture monitoring satellite.

3. The space-based GNSS-S and optical collaborative soil moisture monitoring system according to claim 2 is characterized in that: The low-orbit soil moisture monitoring satellite also includes: GNSS-S antenna, used for simultaneously receiving dual-frequency navigation satellite scattered signals, the frequencies of the dual-frequency navigation satellite scattered signals are 1575.42MHz and 1227.60MHz, respectively, and the bandwidths are both 20.46MHz. The polarization mode of the GNSS-S antenna is left-hand circular polarization, and a phased array antenna system is adopted. The antenna normal gain is greater than 22dBi, and the number of beams is not less than 4; Multispectral high-resolution camera, used to collect multispectral high-resolution images in the GNSS signal scattering area.

4. The space-based GNSS-S and optical collaborative soil moisture monitoring system according to claim 3 is characterized in that: The multi-spectral high-resolution camera is installed on a turntable device, and the shooting area is dynamically adjusted by the turntable device based on the beam pointing of the GNSS-S antenna so that the coverage area of ​​the multi-spectral high-resolution camera is consistent with the observation area of ​​the GNSS-S antenna.

5. The space-based GNSS-S and optical collaborative soil moisture monitoring system according to claim 1 is characterized in that: The foundation soil moisture monitoring module includes: Soil moisture detector, used to collect soil moisture within 6 to 10 cm of the ground; A satellite positioning module, used to obtain the position and measurement time of the soil moisture detector measuring the soil moisture; A data storage device, used for storing the collected data of the soil moisture detector and the time and space information acquired by the satellite positioning module; A solar power module, used to continuously supply power to the foundation soil moisture monitoring module; The first Beidou short message terminal is used to upload the data information collected by the soil moisture detector and the satellite positioning module to the ground measurement, control and data processing center.

6. The space-based GNSS-S and optical collaborative soil moisture monitoring system according to claim 5 is characterized in that: It includes a plurality of the ground-based soil moisture monitoring modules, and the plurality of the ground-based soil moisture monitoring modules are distributed in typical vegetation coverage scenes around the world.

7. The space-based GNSS-S and optical collaborative soil moisture monitoring system according to claim 2 is characterized in that: The ground measurement and control and data processing center includes: A high-performance server is used to form a training set by temporal and spatial matching of the GNSS scattered signal data collected by the low-orbit GNSS-S system, the soil roughness and vegetation index inverted based on the images collected by the low-orbit optical system, and the true value of soil moisture actually collected by the ground-based soil moisture monitoring module; and based on the training set, continuously train the mapping model of GNSS-S observation quantity characteristics, vegetation parameters extracted based on optical data, and measured soil moisture information through a deep learning network; The satellite-to-ground link is used to receive the first observation data and system status parameters transmitted by the low-orbit soil moisture monitoring satellite, and upload the mapping model trained by the deep learning network to the low-orbit soil moisture monitoring satellite; The second Beidou short message terminal is used to receive the second observation data collected by the ground-based soil moisture monitoring module.

8. The space-based GNSS-S and optical collaborative soil moisture monitoring system according to claim 1, characterized in that: The ground measurement, control and data processing center is specifically used to construct a training set based on the first observation data and the second observation data accumulated over a preset time, and continuously train the mapping relationship between the GNSS-S observation quantity characteristics, the vegetation parameters extracted based on the optical data and the measured soil moisture information through a deep learning network based on the training set to obtain the mapping model and upload it to the low-orbit soil moisture monitoring satellite; and continuously receive the first observation data of the low-orbit soil moisture monitoring satellite and the second observation data of the ground-based soil moisture monitoring module, and regularly update the mapping model before uploading it to the low-orbit soil moisture monitoring satellite.

9. An on-orbit intelligent real-time inversion method based on the space-based GNSS-S and optical collaborative soil moisture monitoring system according to any one of claims 1 to 8, characterized in that: Includes steps: Preprocessing the delay-Doppler map obtained after GNSS-S signal processing based on the signal-to-noise ratio; Normalizing the delay-Doppler image and extracting typical features to use them as a model input main vector; Performing image processing on the multispectral high-resolution image to obtain land cover type, leaf area index, plant optical thickness, vegetation parameters of normalized vegetation parameters and soil roughness corresponding to the scattering area, and taking the vegetation parameters, the soil roughness and designated parameters as input vectors; wherein the designated parameters include the longitude and latitude of the scattering point, the scattering angle and the scanning beam gain of the downward-looking phased array antenna; A training set, a validation set, and a test set of a mapping model are configured according to the input main vector and the input slave vector; Setting up a deep learning architecture, determining the dimension of the input layer of the mapping model, adding hidden layers and output layers, selecting a loss function and minimizing the loss function by configuring an optimizer; Training the mapping model according to the training data in the training set, and periodically verifying the model through the verification set to monitor the performance of the model, and dynamically adjusting hyperparameters to optimize the model performance; Using the test set to evaluate the performance of the mapping model; The trained mapping model is uploaded to the low-orbit soil moisture monitoring satellite, so that the low-orbit soil moisture monitoring satellite can perform on-orbit collaborative intelligent real-time inversion of soil moisture according to the actually measured GNSS-S data and multi-spectral high-resolution data through the mapping model.

10. The on-orbit intelligent real-time inversion method according to claim 9, characterized in that: The mapping model is trained and evaluated through a high-performance server of a ground measurement, control and data processing center; the mapping model is used by an onboard computing module of the low-orbit soil moisture monitoring satellite to perform on-orbit collaborative intelligent real-time inversion of soil moisture based on actually measured GNSS-S data and multi-spectral high-resolution data.

Images