A method for acquiring vegetation area data using bistatic InSAR navigation satellites based on multi-factor optimization

Abstract

The present invention discloses a method for acquiring vegetation area data using a navigation satellite dual-base InSAR (InSAR) based on multi-factor optimization. The present invention first selects experimental scenarios based on the type of surface vegetation coverage; then calculates radar observation indicators to determine the available experimental points in each scenario; then, by analyzing the seasonal vegetation normalized index of the experimental scenario, the duration of the experiment is determined; finally, based on a multi-objective joint optimization algorithm, the data collection time and satellite combination of each experimental point are designed to obtain the final data collection experimental design scheme. The present invention comprehensively considers factors such as vegetation coverage type, radar observation indicators, and experimental time duration to ultimately determine the data collection experimental design scheme for vegetation scenario research. Data is then collected based on this data and a data set is established for subsequent signal processing and phase analysis, providing a feasible method for accumulating raw data for navigation satellite dual-base InSAR vegetation scenario research.

Description

Technical Field

[0001] The present invention relates to the technical field of bistatic synthetic aperture radar, and in particular to a method for acquiring vegetation area data of a navigation satellite bistatic InSAR based on multi-factor optimization. Background Art

[0002] Global Navigation Satellite System-based Bistatic Interferometric SAR (GNSS-based InBSAR) utilizes in-orbit navigation satellites as transmitters and a ground-based receiver. This system employs Persistent Scatter InSAR (PS-InSAR) technology, and the receiver can be airborne, vehicle-mounted, or even fixed. Due to the abundant resources of in-orbit navigation satellites, short revisit periods, and the lack of a transmitter, the Navigation Satellite Bistatic InSAR system can measure millimeter-level deformations in the observed area at a low hardware cost. It is a radar remote sensing technology widely used for landslide early warning in areas prone to geological hazards.

[0003] However, geological disasters often occur in areas with high vegetation cover, and the received scene reflection signals are affected by multiple factors, including vegetation, topography, surface roughness, and soil composition. Due to seasonal changes in vegetation, the signals containing surface deformation information are severely contaminated by decorrelated noise. Furthermore, the wide variety, complex structure, and random distribution of vegetation hinder the establishment of vegetation reflectance models, posing challenges to the extraction and separation of received signals. This severely limits the deformation measurement accuracy of navigation satellite bistatic InSAR systems in vegetation-covered areas. Furthermore, research on vegetation impacts based on navigation satellite bistatic InSAR is in its infancy, and existing research is mostly based on spaceborne synthetic aperture radar interferometry systems, which cannot be directly applied to navigation satellite bistatic InSAR systems. Therefore, a method for acquiring vegetation scene datasets based on navigation satellite bistatic InSAR systems is urgently needed to collect field data from navigation satellite bistatic InSAR systems in vegetation scenes, and to further study the impact of vegetation attributes on the deformation measurement performance of navigation satellite bistatic InSAR systems. Summary of the Invention

[0004] In light of this, the present invention provides a method for acquiring vegetation area data using a navigation satellite bistatic InSAR system based on multi-factor optimization. This method primarily addresses the issue of evaluating the impact of vegetation on deformation measurement performance in navigation satellite bistatic InSAR systems. Based on a multi-factor joint optimization strategy, a data acquisition experimental design scheme for studying vegetation scenarios using a navigation satellite bistatic InSAR system is proposed. First, experimental scenarios are selected based on the type of surface vegetation cover. Then, radar observation indicators are calculated to determine the available experimental points for each scenario. Next, the experimental duration is determined by analyzing the seasonal normalized vegetation index of the experimental scenario. Finally, based on a multi-objective joint optimization algorithm, the acquisition time and satellite combination for each experimental point are designed to obtain the final data acquisition experimental design scheme.

[0005] The method for acquiring vegetation area data using a navigation satellite bistatic InSAR system based on multi-factor optimization of the present invention comprises:

[0006] Step 1: Select the experimental scene based on the vegetation cover type;

[0007] Step 2: Selecting experimental points in each experimental scene, where the locations of the experimental points meet the observation conditions required by the navigation satellite bistatic InSAR system;

[0008] Step 3: Determine the experimental duration T of each experimental point Dur ; The experimental duration T Dur satisfy:

[0009] Vari(NDVI(T Dur ))≥k·Vari(NDVI year ) (7)

[0010] in, P NIR and P RED Respectively represent the reflectance of the near infrared band and red band of the experimental point; NDVI (T Dur ) is the experimental duration T Dur NDVI value of the experimental point surface, NDVI year is the NDVI value of the experimental point in the past year, Vari(x)=max(x)-min(x) represents the maximum fluctuation amplitude of the input data x, and k∈(0,1) is the scale factor;

[0011] Step 4: For each experimental point, a multi-objective joint optimization model for the sampling time and satellite combination of the experimental point is constructed based on the satellite resolution, elevation angle, and theoretical deformation measurement accuracy. By solving the multi-objective joint optimization model, the sampling time and satellite combination of the experimental point are obtained.

[0012] Step 5: For the experimental points selected in step 2, determine the sampling time and satellite combination according to step 4, and conduct the experiment for a duration of T Dur The data is measured to obtain the vegetation area measurement data of the corresponding vegetation cover type.

[0013] Preferably, in step 1, four typical surface coverage areas, namely, evergreen broad-leaved forest, shrubs, herbs, and exposed rocks, are selected as experimental scenes.

[0014] Preferably, in step 2, the position of the experimental point satisfies the range resolution ρ of the navigation star bistatic InSAR system. r Index and azimuth resolution ρ a Indicators, and position precision dilution PDOP indicators; among them,

[0015]

[0016]

[0017]

[0018] Where c is the speed of light, B is the bandwidth of the transmitted signal, β is the bistatic angle, Θ is the unit vector along the bisector of β, and Γ TA is the unit vector in the direction of the line connecting the transmitter and the center of the experimental point, Z is the unit vector directly above, λ is the signal wavelength, T int is the synthetic aperture, ω TA is the angular velocity of the transmitter relative to the center of the experimental point; the superscript T indicates the matrix transpose; D 11 , D 22 , D 33 The theoretical deformation measurement accuracy in the east-west, north-south, and up-down directions respectively; D 11 , D 22 , D 33 Calculate according to the observation matrix H:

[0019]

[0020]

[0021] in, is the deformation measurement accuracy of the i-th satellite, P Si is the position of the i-th satellite at the observation time, i = 1, 2, ..., M, M is the total number of currently available satellites; P A is the center position of the experimental point, P R is the position of the receiver, and t is the center moment of the current synthetic aperture time.

[0022] Preferably, in step 4, the multi-objective joint optimization model is:

[0023]

[0024]

[0025] Among them, t is the acquisition time, SC is the satellite combination, S reso is the satellite resolution, S reso_thres is the satellite resolution threshold, θ is the satellite elevation angle, θ thres is the satellite elevation angle threshold, ACC is the satellite theoretical deformation inversion accuracy, ACC thres is the satellite theoretical deformation inversion accuracy threshold, and eff is the monitoring efficiency function.

[0026] Preferably, the eff is the detection efficiency using the PS point coverage within the experimental point range.

[0027] Beneficial effects:

[0028] This invention, based on a multi-factor joint optimization strategy, comprehensively considers factors such as vegetation cover type, radar observation indicators, and experimental duration to ultimately determine a data acquisition experimental design for vegetation scenario studies. Data is then collected based on this design and a dataset is created for subsequent signal processing and phase analysis. This provides a feasible method for accumulating raw data for bistatic InSAR vegetation scenario studies using navigation satellites. Furthermore, this invention lays the foundation for establishing a vegetation scattering model, thereby enabling high-precision surface deformation measurements in vegetated areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is the flow chart of the algorithm of the present invention;

[0030] Figure 2 Simplified scattering mechanism models for different types of vegetation;

[0031] Figure 3 This is a configuration diagram of the navigation satellite bistatic InSAR system in the present invention;

[0032] Figure 4 Results of experimental point selection using the present invention. DETAILED DESCRIPTION

[0033] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0034] The present invention provides a method for acquiring vegetation area data using a navigation satellite dual-base InSAR based on multi-factor optimization. The method comprises the following steps: first, experimental scenarios are selected based on the type of surface vegetation cover; second, radar observation indicators are calculated to determine the available experimental points in each scenario; third, the experiment duration is determined by analyzing the monthly vegetation normalized index of the experimental site; and finally, a multi-target joint optimization is performed based on factors such as satellite resolution, elevation angle, and theoretical deformation measurement accuracy to design the data acquisition time and satellite combination for each experimental point to obtain the optimal monitoring efficiency and accuracy, thereby obtaining the final data acquisition experiment design scheme.

[0035] The method flow chart of the present invention is as follows Figure 1 As shown, the specific steps are:

[0036] Step 1: Select the experimental scene based on the surface vegetation cover type; specifically:

[0037] First, experimental scenarios are selected based on vegetation cover type. For target observation scenarios covered by vegetation, the scene reflection signal received by the receiver is the result of the combined effects of ground surface scattering and vegetation scattering, and is influenced by both surface and vegetation parameters. Surface parameters include surface roughness and soil moisture content, while vegetation parameters include factors such as the size, spatial distribution structure, height, and density of vegetation scatterers. Based on their structural characteristics, different types of vegetation are typically modeled as different media, exhibiting a variety of complex scattering mechanisms. Therefore, it is necessary to select several typical study areas with vegetation cover for field data collection to analyze the impact of different scattering mechanisms on the echo signal.

[0038] The surface vegetation cover in deformation-prone areas can be roughly divided into herbaceous vegetation and woody vegetation. First, herbaceous plants can be divided into two categories: short herbaceous plants with blade-like leaves extending vertically upward, and perennial grasses with broad leaves and runner stems, such as peanuts, soybeans, and vegetable crops. Since landslide-prone areas are mostly wild grasses, and it is difficult to distinguish between these two types of landforms during the classification process, they are uniformly classified as herbaceous vegetation cover in this experiment. Its scattering mechanism is as follows: Figure 2 (a) As shown. Secondly, woody plants can be divided into three categories: shrubs, broad-leaved forests and coniferous forests. Shrubs are short and dense woody plants. Their scattering process is as follows: Figure 2 As shown in (b), it can be simply divided into two parts: scattering of leaf layer and branch layer and ground scattering. Broadleaf forest is a forest community of dicotyledonous plants. Its scattering process is as follows Figure 2As shown in (c), it can be decomposed into three components: ground scattering, secondary scattering from ground to vegetation to ground and vegetation to ground to vegetation, and canopy scattering. Coniferous forests are primarily coniferous conifers of the coniferous family. Because coniferous forests are relatively rare in landslide-prone areas in China, they are not considered in this experiment. Furthermore, a study area without vegetation cover is required for comparative verification. Therefore, this method selects four typical surface cover areas: evergreen broad-leaved forest, shrubland, herbaceous areas, and bare rock as experimental scenarios for data collection.

[0039] Step 2: Calculate radar observation indicators and determine the available experimental points in each scenario; specifically:

[0040] Next, we determine the experimental point locations based on the selected experimental scenario. To ensure the performance of the Navigation Satellite Bistatic InSAR system within the selected experimental points, we use the geographic parameters of the experimental points and the Navigation Satellite Bistatic InSAR system parameters to calculate radar observation indicators, including theoretical resolution and position dilution of precision (PDOP). This allows us to select experimental point locations that meet the observation conditions required for deformation measurements using the Navigation Satellite Bistatic InSAR system, ensuring the validity and availability of the subsequently acquired field data sets.

[0041] The navigation star bistatic InSAR system adopts a bistatic configuration. Figure 3 As shown, the range resolution ρ of the observation area r and azimuth resolution ρ a Can be expressed as:

[0042]

[0043]

[0044] Where B is the bandwidth of the transmitted signal, T int is the synthetic aperture, β is the bistatic angle, Θ is the unit vector along the β angle bisector, ω TA is the angular velocity of the transmitter relative to the target area, Γ TA is the unit vector in the direction of the line connecting the transmitter and the observation area, Z is the unit vector directly above, λ is the signal wavelength, and c is the speed of light.

[0045] PDOP is an important indicator for evaluating position accuracy. Assuming that the theoretical deformation measurement accuracy in the east-west, north-south, and up-down directions is D 11 , D 22 , D 33 Then, the PDOP of the navigation satellite bistatic InSAR system can be expressed as:

[0046]

[0047] D 11 , D 22 , D 33 It can be calculated based on the observation matrix H as:

[0048]

[0049]

[0050] in, is the deformation measurement accuracy of a single satellite, P Si is the position of the i-th satellite at the observation time, P A is the location of the target area, P R is the position of the receiver, M is the total number of currently available satellites, and t is the center moment of the current synthetic aperture time.

[0051] Experimental locations where the radar observation indicators meet the standards under different experimental scenarios are selected as experimental points.

[0052] Step 3: Determine the duration of the experiment based on the seasonal vegetation normalized index; specifically:

[0053] Generally speaking, the temperature rises in spring, and the temperature difference is significant, so vegetation grows a lot; this trend further increases in summer, and the density and volume of vegetation reach their peak; while in autumn and winter, affected by cold air, the climate is cold and precipitation is limited, which is not conducive to vegetation growth, resulting in a decrease in the surface vegetation coverage. Therefore, even in areas covered by the same vegetation type, the performance of the deformation measurement of the navigation satellite bistatic InSAR system will vary from season to season. In order to incorporate the impact of vegetation growth and withering with seasonal changes in the performance evaluation of the deformation measurement of the navigation satellite bistatic InSAR system in vegetated areas, the duration of the data acquisition experiment should be designed. The Normalized Difference Vegetation Index (NDVI) is a widely used remote sensing indicator that can quantify the degree of vegetation coverage. The expression is:

[0054]

[0055] Among them, P NIR and P RED Represent the reflection of near-infrared band and red band respectively.

[0056] Calculate the monthly NDVI for the selected experimental area throughout the year, and the experimental duration is T Dur Should meet the following requirements:

[0057] Vari(NDVI(T Dur ))≥k·Vari(NDVIyear ) (7)

[0058] Among them, Vari(x)=max(x)-min(x) represents the maximum fluctuation range of input data x, NDVI(T Dur ) is the surface NDVI value of the experimental point during the experimental duration, NDVI year is the NDVI value of the area in the past year, k∈(0,1) is the scaling factor, and the k is determined according to the statistical results of the data, generally taking 0.75. If it is too large, the experiment time will be too long, and if it is too short, the seasonal changes of vegetation cannot be observed.

[0059] Step 4: Perform multi-objective joint optimization based on factors such as satellite resolution, elevation angle, and theoretical deformation measurement accuracy, and design the data collection time and satellite combination for each experimental point; specifically:

[0060] The Navigation Satellite Bistatic InSAR system utilizes multiple in-orbit navigation satellites as transmitters. Compared to traditional spaceborne InSAR systems, this system has the advantage of accessing a large number of available navigation satellites. Different combinations of these satellites can significantly vary in the deformation inversion accuracy and efficiency achieved at different time points. Therefore, a joint optimization is performed based on factors such as satellite resolution, elevation angle, and theoretical deformation measurement accuracy. The number of selected PS points (Persistent Scatter, which refers to pixels in InSAR images with stable phase and amplitude, representing surface targets with stable scattering characteristics) is used as the monitoring efficiency indicator. The data collection time and satellite combination for each experimental point are designed to achieve optimal measurement accuracy and efficiency. The multi-objective optimization model can be expressed as:

[0061]

[0062]

[0063] Among them, t represents the acquisition time, SC represents the satellite combination, S reso Indicates satellite resolution, S reso_thres is the satellite resolution threshold, θ represents the satellite elevation angle, θ thres is the satellite elevation angle threshold, ACC represents the satellite theoretical deformation inversion accuracy, ACC thres is the satellite theoretical deformation inversion accuracy threshold, and eff is the monitoring efficiency function.

[0064] At this point, an experimental plan for collecting measured datasets of vegetation scenes based on the navigation satellite bistatic InSAR system has been obtained, and the original dataset can be obtained by data collection based on this plan.

[0065] The following is an explanation of the processing results of the embodiment. In this embodiment, the Beidou navigation satellite was used as the transmission source, and the system parameters were shown in Table 1. According to the method proposed in the present invention, based on the surface vegetation cover type and the radar observation index optimization, the final selected experimental point site photos are as follows Figure 4 As shown, (a)-(d) represent the results of Zhijianggou, Songjiawan, Zuojiaping and Pengan mines respectively, and the specific parameters are shown in Table 2. In this embodiment, the experimental duration is designed to be 4 months.

[0066] Table 1 Parameters of the embodiment

[0067]

[0068] Table 2 Results of selected experimental points in the embodiment

[0069]

[0070] Then, based on factors such as satellite resolution, elevation angle, and theoretical deformation measurement accuracy, a joint optimization is performed to design the data collection time and satellite combination for each experimental point. Taking the results of August 22, 2023 as an example, the data collection time and satellite combination design for each experimental point in the embodiment are shown in Table 3. The data collection time and satellite combination design for only IGSO satellites are shown in Table 4.

[0071] The proposed method uses a multi-factor joint optimization strategy, comprehensively considering factors such as vegetation cover type, radar observation indicators, experimental time, and satellite combination. The method ultimately determines the data acquisition experimental design for vegetation scenario studies. Based on this design, data is collected using a BeiDou radar receiver and the resulting dataset is compiled. The resulting dataset can be used for radar imaging (e.g., using a backprojection imaging algorithm) to obtain multiple sets of experimental point observations from different satellites. This allows for subsequent phase processing and deformation information extraction. This provides a feasible approach for bistatic InSAR vegetation scenario studies using navigation satellites, laying the foundation for establishing vegetation scattering models and achieving high-precision deformation measurement of the surface in vegetated areas.

[0072] Table 3 Example of sampling time and satellite combination design for each experimental point in the embodiment

[0073]

[0074] Table 4 Example of sampling time and satellite combination design for each experimental point in the embodiment (only including IGSO satellite case)

[0075]

[0076] In summary, the above are only preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for acquiring vegetation area data using a navigation satellite bistatic InSAR system based on multi-factor optimization, characterized in that: include: Step 1: Select the experimental scene based on the vegetation cover type; Step 2: Selecting experimental points in each experimental scene, where the locations of the experimental points meet the observation conditions required by the navigation satellite bistatic InSAR system; Step 3: Determine the experimental duration of each experimental point ; Duration of the experiment satisfy: （7） in, , and represent the reflections of the near-infrared band and the red band of the experimental point respectively; Duration of the experiment The surface NDVI values ​​of the experimental points within is the NDVI value of the experimental point in the past year, Represents input data The maximum fluctuation range, is the scale factor; Step 4: For each experimental point, a multi-objective joint optimization model for the sampling time and satellite combination of the experimental point is constructed based on the satellite resolution, elevation angle, and theoretical deformation measurement accuracy. By solving the multi-objective joint optimization model, the sampling time and satellite combination of the experimental point are obtained. Among them, the multi-objective joint optimization model is: （8） in, For the collection time, For satellite combination, is the satellite resolution, is the satellite resolution threshold, is the satellite elevation angle, is the satellite elevation angle threshold, is the satellite theoretical deformation inversion accuracy, is the satellite theoretical deformation inversion accuracy threshold, is the monitoring efficiency function; Step 5: For the experimental points selected in step 2, determine the sampling time and satellite combination according to step 4, and conduct the experiment duration. The data is measured to obtain the vegetation area measurement data of the corresponding vegetation cover type.

2. The method according to claim 1, wherein In the step 1, four typical surface cover areas, namely, evergreen broad-leaved forest, shrubs, herbs, and exposed rocks, are selected as experimental scenes.

3. The method according to claim 1, wherein In step 2, the position of the experimental point meets the range resolution of the navigation satellite bistatic InSAR system. Index and azimuth resolution Indicators, and position precision dilution PDOP indicators; among them, （1） （2） （3） in, is the speed of light, is the bandwidth of the transmitted signal, is the bistatic angle, For the The unit vector in the direction of the angle bisector, is the unit vector of the line connecting the transmitter and the center of the experimental point, is the unit vector directly upward, is the signal wavelength, is the synthetic aperture, is the angular velocity of the transmitter relative to the center of the experimental point; the superscript T indicates the matrix transpose; These are the theoretical deformation measurement accuracy in the east-west, north-south, and up-down directions respectively; According to the observation matrix calculate: （4） （5） in, For the The deformation measurement accuracy of the satellite, For the The position of the satellite at the time of observation, i =1, 2,…, M , is the total number of currently available satellites; is the center position of the experimental point, is the receiver's location, t is the center moment of the current synthetic aperture time.

4. The method according to claim 1, wherein described eff The PS point coverage within the experimental point range is taken as the detection efficiency.

Images