阅读说明：本技术 一种基于gnss-ir的土壤含水率实时连续监测方法 (GNSS-IR-based real-time continuous monitoring method for soil moisture content ) 是由 孙世坤 朱牧原 姚一飞 于 2021-09-14 设计创作，主要内容包括：本发明公开了一种基于GNSS-IR的土壤含水率实时连续监测方法,包括：随机选取土壤湿度采样点,获取土壤湿度采样点的单天线GNSS原始观测数据,根据单天线GNSS原始观测数据,获取低高度角的SNR观测值,单天线GNSS同时接收来自卫星的低高度角的SNR信号与附近土壤的反射SNR信号,二者叠加产生单天线GNSS的SNR干涉信号,对SNR干涉信号进行去趋势处理及频谱分析,获取多径反射信号的三个特征参量：振幅A-(m)、频率f及相位利用农田采集的试验数据,建立实测土壤湿度与多径反射信号的振幅、频率及相位间的三元线性回归模型,将振幅A-(m)、频率f及相位作为自变量输入三元线性回归模型反演出土壤湿度的值。该方法具有成本低、可连续观察、应用范围广、易于操作等优点。(The invention discloses a GNSS-IR (global navigation satellite system-infrared) based real-time continuous monitoring method for soil moisture content, which comprises the following steps: randomly selecting a soil humidity sampling point, acquiring single-antenna GNSS original observation data of the soil humidity sampling point, acquiring an SNR (signal noise ratio) observation value of a low altitude angle according to the single-antenna GNSS original observation data, simultaneously receiving an SNR signal of the low altitude angle from a satellite and a reflection SNR signal of nearby soil by the single-antenna GNSS, superposing the SNR signal and the reflection SNR signal to generate an SNR interference signal of the single-antenna GNSS, performing trend removing processing and spectrum analysis on the SNR interference signal, and acquiring three characteristic parameters of a multipath reflection signal: amplitude A m Frequency f and phase Establishing a ternary linear regression model between the actually measured soil humidity and the amplitude, frequency and phase of the multipath reflection signal by using the test data collected by the farmland, and calculating the amplitude A m Frequency f and phase The value of soil moisture was inverted by inputting the ternary linear regression model as independent variables. The method has the advantages of low cost, continuous observation, wide application range, easy operation and the like.)

1. A GNSS-IR based soil moisture content real-time continuous monitoring method is characterized by comprising the following steps:

randomly selecting soil humidity sampling points, and acquiring single-antenna GNSS original observation data of the soil humidity sampling points;

acquiring an SNR observation value of a low altitude angle according to single-antenna GNSS original observation data;

based on the SNR observation value of a low altitude angle, the single-antenna GNSS receives the SNR signal of the low altitude angle from the satellite and the reflection SNR signal of the nearby soil at the same time, the SNR interference signal of the single-antenna GNSS is generated by superposition of the SNR observation value and the reflection SNR signal, the SNR interference signal is subjected to detrending processing and spectrum analysis, and three characteristic parameters of a multipath reflection signal are obtained: amplitude A_mFrequency f and phase

Establishing a ternary linear regression model between the actually measured soil humidity and the amplitude, frequency and phase of the multipath reflected signal;

will have an amplitude A_mFrequency f and phaseThe value of soil moisture was inverted by inputting the ternary linear regression model as independent variables.

2. The method for continuously monitoring the soil moisture content in real time based on GNSS-IR as claimed in claim 1, wherein the obtaining of the SNR observation value of the low altitude angle from the single antenna GNSS raw observation data comprises:

setting the erection height of a single-antenna GNSS signal receiver to be 1.72 m;

reading satellite observation data by using teqc software and introducing the satellite observation data into matlab;

and extracting the SNR observed value of the low altitude angle corresponding to the prn number of the proper satellite by using matlab.

3. The GNSS-IR based soil moisture content real-time continuous monitoring method as claimed in claim 1, wherein the three characteristic parameters of the multi-path reflected signal, amplitude A, are obtained_mFrequency f and phaseThe method comprises the following steps:

the signal-to-noise ratio SNR of the interference signal received by the single antenna GNSS signal machine is expressed as:

wherein A is_d、A_mRespectively representing the amplitudes of the direct satellite signal and the reflected satellite signal, and psi is the phase difference of the direct satellite signal and the reflected satellite signal;

obtaining sinusoidal oscillatory multipath reflected signal SNR_mSinusoidal oscillatory multipath reflection signal SNR_mExpressed as:

wherein h is the height of the antenna relative to the reflecting surface, the satellite height angle is E,is the initial phase, λ is the wavelength;

by comparing SNR_mCarrying out spectrum analysis on the signals to obtain the oscillation frequency f and the equivalent antenna height h of the multipath signals;

SNR using least squares_mFitting to obtain the amplitude A of the multipath signal_mAnd initial phase

4. The GNSS-IR based soil moisture content real-time continuous monitoring method according to claim 3, characterized in that the acquisition of sinusoidal oscillation multipath reflection signal SNR_mThe method comprises the following steps:

the distance of the reflected signal to reach the antenna is delta than the distance of the direct signal to be propagated, and delta satisfies the following conditions:

δ＝2hsinβ＝2hsin|E-θ|

wherein, the inclination angle between the reflecting surface and the horizontal plane is theta, the satellite elevation angle is E, beta is defined as the difference between the elevation angle and the inclination angle, namely the included angle between the direct signal and the reflecting surface, and h is the height of the antenna relative to the reflecting surface;

assuming that the angle between the reflecting surface and the horizontal plane is negligible, θ is equal to 0, and the above equation is expressed as:

δ＝2hsinE

then, the phase difference between the reflected signal and the direct signal is:

by the above equation, the frequency of the multipath oscillation is calculated as follows:

since the time of SNR observed each day is within a few hours, the rate of change of the equivalent antenna height is neglected, and the above equation is further simplified as:

combining the above formula, removing the direct signal irrelevant to the reflection surface characteristic, only retaining the multipath oscillation signal, fitting the signal with the sine signal to obtain the SNR of the sine oscillation multipath reflection signal_m。

5. The GNSS-IR based soil moisture content real-time continuous monitoring method of claim 4, wherein the SNR is measured_mPerforming spectral analysis on the signal, including:

SNR is analyzed by Lomb-Scargle spectrum analysis method_mThe signal is subjected to spectral analysis.

6. The GNSS-IR based real-time continuous monitoring method for soil moisture content as claimed in claim 2, wherein the low elevation angle is 2-30 °.

Technical Field

The invention relates to the technical field of soil humidity measurement, in particular to a GNSS-IR-based real-time continuous monitoring method for soil moisture content.

Background

GNSS-R (Global Navigation Satellite System reflection-try), a new microwave remote sensing technology, has been developed rapidly with the advantages of abundant signal sources, low cost, no climate restriction, etc. The method can process direct signals and reflected signals by utilizing a GNSS receiver of a foundation, and can be used for inversion of soil humidity due to the fact that the reflected signals are sensitive to the soil humidity.

For different GNSS-R technology branches, the signal reception system can be roughly divided into a single-antenna observation mode, a dual-antenna observation mode, and a three-antenna observation mode. Wherein, the single antenna method (GNSS-IR) uses a right-hand circularly polarized antenna to receive satellite signals; the dual-antenna method is used for receiving direct signals and reflected signals of the GNSS by using a left-handed circularly polarized antenna and a right-handed circularly polarized antenna respectively for inversion; compared with the dual-antenna method, the three-antenna method adds a right-hand circularly polarized antenna for receiving the right-hand component in the reflected signal, but the basic principle is not much different from the dual-antenna mode. Compared with a double-antenna transmitter and a three-antenna method, the single-antenna method only needs to erect a right-hand circularly polarized antenna, does not need to develop a special receiver, and directly adopts a geophysical or geomapping receiver, so that the cost is low, and the operation is simple.

Disclosure of Invention

Aiming at the defects of the prior art, the invention provides a single-polarized antenna GNSS soil humidity measuring method, which uses a right-hand circularly polarized antenna, utilizes the phase difference of received signals to invert the soil humidity, has strong feasibility, is convenient to operate, has low cost and can continuously observe.

The embodiment of the invention provides a GNSS-IR based real-time continuous monitoring method for soil moisture content, which comprises the following steps:

randomly selecting soil humidity sampling points, and acquiring single-antenna GNSS original observation data of the soil humidity sampling points;

acquiring an SNR observation value of a low altitude angle according to single-antenna GNSS original observation data;

the single-antenna GNSS receives the SNR signal of a low altitude angle from a satellite and the reflection SNR signal of the nearby soil at the same time, the SNR interference signal of the single-antenna GNSS is generated by superposition of the SNR signal and the reflection SNR signal, trend removing processing and spectrum analysis are carried out on the SNR interference signal, and three characteristic parameters of a multipath reflection signal are obtained: amplitude A_mFrequency f and phase

Establishing a ternary linear regression model between the actually measured soil humidity and the amplitude, frequency and phase of the multipath reflected signal by using the test data collected by the farmland;

will have an amplitude A_mFrequency f and phaseThe value of soil moisture was inverted by inputting the ternary linear regression model as independent variables.

Further, acquiring the SNR observation value at a low altitude angle from the raw observation data of the single-antenna GNSS includes:

setting the erection height of a single-antenna GNSS signal receiver to be 1.72 m;

reading satellite observation data by using teqc software and introducing the satellite observation data into matlab;

and extracting SNR observed values of 2-30-degree elevation angles corresponding to the prn number of the proper satellite by using matlab.

Further, three characteristic parameters of the multipath reflected signal, amplitude A, are obtained_mFrequency f and phaseThe method comprises the following steps:

the signal-to-noise ratio SNR of the interference signal received by the single antenna GNSS signal machine is expressed as:

wherein A is_d、A_mVibration representing direct and reflected signals of satellite respectivelyAmplitude psi is the phase difference between the two;

obtaining sinusoidal oscillatory multipath reflected signal SNR_mSinusoidal oscillatory multipath reflection signal SNR_mExpressed as:

wherein h is the height of the antenna relative to the reflecting surface, the satellite height angle is E,is the initial phase, λ is the wavelength;

by comparing SNR_mCarrying out spectrum analysis on the signals to obtain the oscillation frequency f of the multipath signals and the effective antenna height h;

finally, SNR is measured by using least square method_mFitting to obtain the amplitude A of the multipath signal_mAnd initial phase

Further, a sinusoidal oscillation multipath reflection signal SNR is obtained_mThe method comprises the following steps:

the distance of the reflected signal to reach the antenna is delta than the distance of the direct signal to be propagated, and delta satisfies the following conditions:

δ＝2h sinβ＝2h sin|E-θ|

wherein, the inclination angle between the reflecting surface and the horizontal plane is theta, the satellite elevation angle is E, beta is defined as the difference between the elevation angle and the inclination angle, namely the included angle between the direct signal and the reflecting surface, and h is the height of the antenna relative to the reflecting surface;

assuming that the angle between the reflecting surface and the horizontal plane is neglected to be about 0, the above equation is expressed as:

δ＝2hsinE

then, the phase difference between the reflected signal and the direct signal is:

by the above equation, the frequency of the multipath oscillation is calculated as follows:

since the time of SNR observed each day is within a few hours, the rate of change of effective antenna height is neglected, and the above equation is further simplified as:

combining the above formula, removing the direct signal irrelevant to the reflection surface characteristic, only retaining the multipath oscillation signal, fitting the signal with the sine signal to obtain the SNR of the sine oscillation multipath reflection signal_m。

Further, for SNR_mThe Lomb-Scargle spectrum analysis method, namely an LS method is adopted for the spectrum analysis of the signal.

The embodiment of the invention provides a GNSS-IR based real-time continuous monitoring method for soil moisture content, which has the following beneficial effects compared with the prior art:

the invention has the soil humidity measuring process, the calculated amount is small, and the calculation can be carried out by a computer. Compared with the traditional soil humidity monitoring, the invention has the advantages of low cost, continuous observation, wide application range, easy operation and the like. The later stage also can combine together with wisdom irrigation system and play a role, has fairly big application potential.

Drawings

FIG. 1 is a block diagram of a method for continuously monitoring soil moisture content in real time based on GNSS-IR according to an embodiment of the present invention;

FIG. 2 is a schematic view of an application scenario of a GNSS receiver of a GNSS-IR based soil moisture content real-time continuous monitoring method according to an embodiment of the present invention;

fig. 3 is a correlation analysis of a soil moisture measured value and an inversion value of a test sample of a GNSS-IR based soil moisture content real-time continuous monitoring method according to an embodiment of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Referring to fig. 1 to 3, an embodiment of the present invention provides a GNSS-IR based real-time continuous monitoring method for soil moisture content, including:

randomly selecting soil humidity sampling points, and acquiring single-antenna GNSS original observation data of the soil humidity sampling points;

acquiring an SNR observation value of a low altitude angle according to single-antenna GNSS original observation data;

the single-antenna GNSS receives the SNR signal of a low altitude angle from a satellite and the reflection SNR signal of the nearby soil at the same time, the SNR interference signal of the single-antenna GNSS is generated by superposition of the SNR signal and the reflection SNR signal, trend removing processing and spectrum analysis are carried out on the SNR interference signal, and three characteristic parameters of a multipath reflection signal are obtained: amplitude A_mFrequency f and phase

Establishing a ternary linear regression model between the actually measured soil humidity and the amplitude, frequency and phase of the multipath reflected signal by using the test data collected by the farmland;

will have an amplitude A_mFrequency f and phaseThe value of soil moisture was inverted by inputting the ternary linear regression model as independent variables.

The specific embodiment is as follows:

step 1: and receiving and preprocessing the GNSS original observation signal to obtain a signal-to-noise ratio (SNR) observation value at a low altitude angle.

The specific implementation of step 1 is as follows:

as shown in FIG. 2, a single antenna GNSS signal receiver is installed under a ground-based condition, and the erection height is 1.72 m. And reading satellite observation data by using teqc software, introducing the satellite observation data into matlab, and extracting an SNR (signal to noise ratio) observation value of a 2-30-degree elevation angle corresponding to a prn signal of a proper satellite by using the matlab.

Step 2: and extracting characteristic parameters. The GNSS signal receiver simultaneously receives direct signals from satellites and reflected signals of nearby soil, and the direct signals and the reflected signals are superposed to generate GNSS interference signals. The SNR interference signal is subjected to detrending processing and spectrum analysis, and three characteristic parameters corresponding to the SNR interference signal can be obtained: amplitude A_mFrequency f and phase

The specific implementation of step 2 is as follows:

the signal-to-noise ratio SNR of the interference signal received by the GNSS antenna can be represented by the following formula:

wherein A is_d、A_mWhich respectively represent the amplitudes of the direct and reflected satellite signals, phi is the phase difference between the two. Since the amplitude of the reflected signal is smaller than the direct signal, the amplitude of the interference signal depends mainly on the direct signal. And carrying out polynomial fitting on the interference signals, namely removing direct signals and highlighting signal fluctuation caused by multipath signals.

The frequency of the SNR waveform depends on the relative positional relationship between the satellite, the reflecting surface, and the receiving antenna. In general, the reflecting surface may have a certain inclination θ with respect to the horizontal plane, where δ satisfies the following condition, assuming that the satellite altitude is E, β is defined as the difference between the altitude and the inclination (i.e. the included angle between the direct signal and the reflecting surface), h is the height of the antenna relative to the reflecting surface, and the distance that the reflected signal needs to travel to reach the antenna than the direct signal is δ:

δ＝2h sinβ＝2h sin|E-θ| (2)

assuming that the angle between the reflecting surface and the horizontal is negligible (θ equals approximately 0), the above equation can be simplified as:

δ＝2hsinE (3)

from the above, it can be derived that the phase difference between the reflected signal and the direct signal is:

the extraction of the phase difference can be completed.

From the above equation, the frequency of the multipath oscillation can be calculated as follows:

and because the time of the observed SNR is within a few hours each day, the rate of change of the equivalent antenna height is negligible, and the above equation can be further simplified as:

the expression of the multipath oscillation frequency shows that h and f are in a linear relation, and the higher the antenna erection is, the larger the multipath oscillation frequency is. Combining the above formula, the direct signal irrelevant to the reflection surface characteristic is removed, only the multipath oscillation signal is reserved, and the signal can be fitted by a sine signal. Multipath reflected signal SNR manifested as sinusoidal oscillation_mCan be expressed as:

by comparing SNR_mThe signal is subjected to spectrum analysis, and the oscillation frequency f and the equivalent antenna height h of the multipath signal can be obtained.

Spectral analysis of time domain signals, the most common method in signal processing is fourier transformation, but fourier analysis is only applicable toTime domain continuous and evenly distributed signal. And SNR acquired by foundation test_mThe multipath signal has the characteristic of non-uniformity in a time domain, in the test process, due to factors such as surface vegetation, soil roughness, blockage of trees and buildings and the like, the data also contains a large amount of noise, a false power spectrum peak value is generated in Fourier analysis, and large errors can exist in the phase and amplitude of the periodic signal. Thus to SNR_mWhen the signal is subjected to spectrum analysis, ordinary Fourier analysis is not applicable.

For this reason, a Lomb-Scargle spectrum analysis method (i.e., LS method) for SNR, which can process a non-uniform signal, is introduced in consideration of factors such as non-uniformity and length limitation of a time-domain sequence_mSpectral analysis was performed. The method can effectively extract weak periodic signals from the time domain sequence, and can weaken false signals generated by nonuniformity of the time domain sequence to a certain extent. In addition, the method can also give the false alarm probability and significance of the existence of each frequency component signal. The Lomb-Scargle method is therefore very suitable for SNR_mSpectral analysis of (2).

Finally, SNR is measured by using least square method_mFitting is carried out to obtain the amplitude A of the multipath signal_mAnd initial phaseAnd then, by using the test data collected by the farmland, an inversion empirical model between the actually measured soil humidity and the frequency, amplitude and phase of the multipath signal can be respectively established.

And step 3: inversion of soil moisture. Utilizing the three characteristic parameters obtained in the step 2 to obtain the amplitude A_mFrequency f and phaseAnd establishing a ternary linear regression model of the measured value of the soil humidity, and performing inversion on the soil humidity by using the ternary linear regression model.

The specific implementation of step 3 is as follows:

the GNSS single-antenna ternary linear regression method is characterized in that a ternary linear regression model of actually measured soil humidity is established by utilizing interference characteristic parameters in a training set, then the interference characteristic parameters in a testing set are input to obtain an inverted soil humidity value, and the effectiveness of the model is evaluated through correlation coefficients and error analysis.

The mathematical expression for the ternary linear regression can be represented by the following sub-formula:

y＝β₀+β₁x₁+β₂x₂+β₃x₃+ε (8)

wherein y represents a dependent variable, x_iDenotes the independent variable, β_iRepresenting the regression coefficients, epsilon represents the random error term.

In soil moisture inversion model, SNR is measured_mIs defined as the independent variable x_iThe output soil humidity value is used as a dependent variable Y, and the regression method is to determine beta₀、β₁、β₂、β₃And the values of the regression coefficients are equal.

Equation (8) can be expressed in a matrix form:

Y＝Xβ+ε (9)，

wherein

After the multiple linear regression model is determined, the unknown parameters need to be estimated. The method adopted is generally a common least square method.

Is provided withRespectively is a parameter beta₀、β₁、β₂、β₃Then y obtains an observed value that can be expressed as:

e_korder toIs y_kThe estimated values of (c) are:

the result of the above formula is the observed value y_kThe regression fit value of (1).

According to the least squares method, the sum of squares of deviations Q of the observed values from the regression values should be made, with the Q requirement being as small as possible. Q is the sum of the errors.

The minimum value in the formula (13) is calculated according to the extreme value principle, so that the minimum value meeting the condition can be obtained.

The least squares estimate of the regression coefficient β that can be obtained by finally solving the matrix equation is:

in unary linear regression, by using the judgment coefficient R²To express the degree of fitting,

where SST denotes the sum of the squares of the total deviations:

SSR represents the regression sum of squares:

SSE represents the sum of the squares of the residuals:

the relationship among the three satisfies:

SST＝SSR+SSE (18)

coefficient of judgment R²Is defined as:

from this, it can be seen that the coefficient R is determined²The smaller the fit, the more general the fit is. Determination coefficient R²Closer to 1 indicates a higher degree of fit.

Substituting the calculation result of the regression coefficient into the formula (8) to obtain a ternary linear regression model, and substituting the interference characteristic parameter needing to be predicted as an independent variable into the model to invert the value of the soil humidity. The correlation analysis result of the inversion value of the test sample and the measured soil moisture value is shown in fig. 3.

TABLE 1 summary of models created by the present invention using the sps software

	R2	Significance (P value)	F statistic	Root mean square deviation of RMSE
					Ternary regression model	0.906	0.000	135.020	0.5465

Although the embodiments of the present invention have been disclosed in the foregoing for illustrative purposes, those skilled in the art will appreciate that various modifications, additions and substitutions are possible, without departing from the scope and spirit of the invention as disclosed in the accompanying drawings.