CN105806208A - Deformation abnormality detection method based on GNSS net shape changes - Google Patents

Abstract

The invention discloses a deformation abnormality detection method based on GNSS net shape changes. The characteristic that GNSS three-dimensional unconstrained free net adjustment has a high-precision geometric net shape is utilized, a GNSS net observed for a long time is used as an integral time-space observation unit, one station and one base line of a first phase of net are used as a reference position and a reference direction, and rotation and translation conversion is carried out on GNSS nets of all phases. The change characteristics of GNSS net shapes are fully measured from four aspects: the lengths of base lines mutually associated among GNSS net measuring stations, changes of included angles of base lines, changes of base line azimuth angles, and horizontal displacements of the measuring stations. According to the characteristics of crustal deformation high space, a time-space sequence matrix is constructed from the four aspects, main component time-space response analysis is carried out, and the time-space distribution and the evolution characteristics of the deformation abnormality are rapidly detected. Tests show that the influences of uncorrelated noise can be effectively separated by the changes of the whole GNSS net shape, the deformation information related to the high space is integrally highlighted, and the extraction of micro-dynamic deformation information and the detection of the deformation abnormality are facilitated.

Description

A kind of Deformation Anomalies detection method based on the change of GNSS net form

Technical field

The invention belongs to Surveying Science and Technology field, particularly relate to region On Fluctuations deformation data and extract and the detection of Deformation Anomalies spatial and temporal distributions.

Background technology

Research currently, with respect to GNSS crustal deformation information retrieval and Deformation Anomalies context of detection is more concentrate on these three aspect.One is based on the change of the GNSS length of base；Two are based on GNSS station coordinates change in displacement；Three are based on GNSS displacement field asks strain.

First method, can only be used for tension or the compression of analyzed area crustal deformation, and base direction is different from deformation direction by the change of the GNSS length of base, and acquired results is widely different.And for walking sliding or rotating deformation, baseline change is also inconspicuous.

Second method, though the change in displacement of each station be can be observed by GNSS station coordinates change in displacement, is had intuitive, but displacement depends on benchmark, uses different benchmark will obtain different displacement fields.Meanwhile, the selection of base station reference frame is a dynamic process, is not changeless, it is therefore desirable to finds suitable Shifted Reference benchmark could reflect real crustal deformation displacement characteristic, and finds suitable Shifted Reference benchmark not easily.

The third method, namely strain is asked by GNSS displacement displacement field, the dynamic changing process of areal strain field principal strain, shearing strain, the anglec of rotation and surface expansion can be obtained, as reflecting regional motion, deformation properties and the time dependent parameter of intensity by the computational methods of strain tensor.But premise assumes that unit interior media is uniform, and displacement is continuous, small deformation etc..

Only a point, the deformation data of a line of the change in location at one station or the change reflection of a length of base, can not represent the regional deformation feature of entirety.And, if this point or linear variable performance only small, then just be difficult to find overall region On Fluctuations deformation.Meanwhile, from long-term Continuous Observation, huge GNSS network, the method detection Deformation Anomalies information by phase done site by site, blanket type screening will be quite time-consuming, and efficiency is very low.

It is proposed to this end that a kind of utilization is not changed, by framework and common mode noise effects, the higher GNSS net form of relative accuracy, the method detecting Anomalies of Crustal Deformation.

Summary of the invention

It is an object of the invention to provide a kind of Deformation Anomalies detection method based on the change of GNSS net form, it is possible to the spatial and temporal distributions information of quick obtaining Deformation Anomalies.

For reaching above-mentioned purpose, the technical solution used in the present invention is: a kind of Deformation Anomalies detection method based on the change of GNSS net form, comprises the following steps:

1), the GNSS survey station of the long-term Continuous Observation in detection region is set up GNSS space-time monitoring network；

2), each phase GNSS monitoring net is carried out Baselines and three-dimensional without constraint adjustment of Free Networks；

3), the GNSS net after adjustment is carried out gauss projection and obtain the net form of a plane and each website plane coordinates；

4), utilize the website coordinate in first phase observational network as reference position, this point and another website are constituted the direction of baseline as reference direction, undertaken translating and rotation transformation by the GNSS net that other each phase is observed.The net form making each issue is all unified on a fixing point and a fixing reference direction, and coordinate after the conversion of all survey stations in being netted；

5), the coordinate after changing is utilized to ask for the change in time and space sequence that whole GNSS space-time monitors the key element of all lengths of base of network, baseline angle, azimuth of director from the gun and all Station Displacements four, as the index weighing net form change；

6), build this four key elements space-time matrix, carry out main constituent space-time response analysis；

7) Deformation Anomalies spatial and temporal distributions and Characteristics of Evolution, are determined.

Described step 4) particularly as follows: the GNSS space-time monitoring network survey station number setting structure is individual as m, observing total issue is the n phase, and the first phase reference station coordinates of selection is (x₀,y₀), the grid azimuth on reference baseline limit is α₀, then each phase GNSS space-time monitoring all station coordinates of network new coordinate after translation rotation transformation is:

${x^{\′}}_i^j=x_0+(x_i^j-x_0^j)cos({\mathrm{\α}}_0^j-{\mathrm{\α}}_0)-(y_i^j-y_0^j)sin({\mathrm{\α}}_0^j-{\mathrm{\α}}_0)---\left(1\right)$

${y^{\′}}_i^j=y_0+(y_i-y_0^j)sin({\mathrm{\α}}_0^j-{\mathrm{\α}}_0)-(y_i-y_0^j)cos({\mathrm{\α}}_0^j-{\mathrm{\α}}_0)---\left(2\right)$

In above formula, i represents survey station number, and value is i=1 ... m；J represents observation issue, and value is j=1 ... n.

Coordinate after utilizing conversion in described step (5) is asked for whole GNSS and is netted all lengths of base, baseline angle, azimuth of director from the gun and all Station Displacements, described in detail below:

The plane basic lineal vector that survey station i and survey station i+1 is constituted is:

${{\mathrm{dx}}^{\′}}_{i,i+1}^j={x^{\′}}_{i+1}^j-{x^{\′}}_i^j---\left(3\right);$

${{\mathrm{dy}}^{\′}}_{i,i+1}^j={y^{\′}}_{i+1}^j-{y^{\′}}_i^j---\left(4\right);$

The length of base that then survey station i and survey station i+1 is constituted is:

$D_{i,i+1}^j=\sqrt{{\mathrm{dx}}^{\′}{{}_{i,i+1}^j}^2+{\mathrm{dy}}^{\′}{{}_{i,i+1}^j}^2}---\left(5\right);$

Angle between baseline i～i+1 and baseline i～i+2 is:

${\mathrm{Ang}}_{D_{i,i+1},D_{i,i+2}}^j=\arccos \left(\frac{{D_{i,i+1}^j}^2+{D_{i,i+2}^j}^2-{D_{i+1,i+2}^j}^2}{2D_{i,i+1}^j{D}_{i,i+2}^j}\right)\·\frac{180}{\mathrm{\π}}\·3600---\left(6\right);$

The azimuth of baseline i～i+1 is:

$R_{i,i+1}^j=\arctan \left(\frac{{{\mathrm{dy}}^{\′}}_{i,i+1}^j}{{{\mathrm{dx}}^{\′}}_{i,i+1}^j}\right)\·\frac{180}{\mathrm{\π}}\·3600---\left(7\right);$

Survey station i displacement is:

${{\mathrm{dx}}^{\′}}_i^j={x^{\′}}_i^{j+1}-{x^{\′}}_i^j---\left(8\right);$

${{\mathrm{dy}}^{\′}}_i^j={y^{\′}}_i^{j+1}-{y^{\′}}_i^j---\left(9\right).$

Described step (6) specifically includes:

61) Time-space serial matrix, is set as X_m×n, wherein, every a line represents a Station Displacements, a length of base, an azimuth of director from the gun or the changing value of baseline angle all phases；The changing value of phase all Station Displacements, the length of base, azimuth of director from the gun or baseline angle is shown in each list；

62), for ensureing that all changes time series is all with zero for basic point, by formula (10) to matrix X_m×nIn each element implement centralization:

$X^{\′}(i,j)=X(i,j)-\frac{\underset{k=1}{\overset{m}{\Σ}}X(i,k)}{m}---\left(10\right)$

In formula, i, the ranks number of j representing matrix, i=1,2 ... m；J=1,2 ... n；

63), to the matrix X' after centralization_m×nCarry out singular value decomposition, obtain orthogonal matrix U_m×m、V_n×nWith the matrix S being made up of singular value_m×n, it is expressed as:

X'_m×n=U_m×mS_m×nV_n×n ^T(11)；

64), singular value being arranged from big to small, front several mode component sums are called holotype component, and it contains the most information of former sequence；If taking front r for holotype, then matrix X'_m×nApproximate representation is:

X_m×n'≈X'_r=U_rS_rV_r ^T(12)

In formula, S_rFor S_m×nIn before the diagonal matrix of r bigger singular value composition, V_rFor V_n×nThe matrix of r column vector composition before middle correspondence, each of which row are called time holotype vector；U_rFor U_m×mThe matrix of r column vector composition before middle correspondence, each of which row are called space holotype vector.

Described step (7) specifically includes:

71) curve time response, analyzed area deformation dynamic change characterization in time, according to time holotype vector are obtained；Obtaining roomage response figure according to space holotype vector, analyzed area deformation is the Different Effects degree of different survey stations during GNSS is netted；

72), from whether curve time response is that homogenous linear change judges the time that Deformation Anomalies occurs, if the linear trend that curve deviation is original, or present non-linear, non-uniform change or have violent concussion, then show very likely to there occurs in this time period Deformation Anomalies；

From roomage response figure, the size of each survey station roomage response judges the spatial distribution that deformation occurs, and the survey station that roomage response is bigger is relatively big by influence of crust deformation, illustrates that the regional extent that distance deformation occurs is nearer；Otherwise, the survey station that roomage response is less is less by influence of crust deformation, distance deformation generation area is described farther out, thus judges the regional extent that deformation occurs.

The present invention has the advantage that as described below:

The present invention is directed to GNSS three-dimensional be independent of benchmark without constraint adjustment of Free Networks and have the feature of high accuracy geometry net form, utilize whole GNSS to net the index system of the whole net form change of observation data construct measurement that is mutually related between all survey stations.The crustal deformation information highlighting whole region is concentrated with the change of net form.Utilize the feature of crustal deformation high spatial dependency, achieved the rapid extraction of Regional Crust deformation space time information by the method for main constituent space-time response analysis.Test shows, adopts the change of net form to have the faint deformation data ability of certain seizure, is conducive to the detection of earth's crust On Fluctuations Deformation Anomalies information.

Accompanying drawing explanation

Fig. 1 is the method flow diagram of the present invention；

Fig. 2 is curve chart time response of GNSS net length of base change；

Fig. 3 is curve chart time response of GNSS net baseline variable angle；

Fig. 4 is the roomage response figure of GNSS net length of base change；

Fig. 5 is the roomage response figure of GNSS net baseline variable angle.

Detailed description of the invention

As shown in Figure 1, the invention discloses a kind of Deformation Anomalies detection method based on the change of GNSS net form, the full name of GNSS is GLONASS (GlobalNavigationSatelliteSystem), it is to refer to all of satellite navigation system, including the whole world, region and enhancing, such as the GPS of the U.S., Muscovite Glonass, the Galileo in Europe, Chinese Beidou satellite navigation system.This sentences GPS network is that example illustrates, and specifically includes following steps:

1), the GPS survey station detecting the long-term Continuous Observation in region is constituted region GPS space-time monitoring network.

2), GPS space-time is monitored network and carries out Baselines and three-dimensional without constraint adjustment of Free Networks；Wherein Baselines and three-dimensional nothing constraint adjustment of Free Networks are existing computational methods, are no longer described in detail at this.

3), carry out Guass plane projection without the GPS space-time monitoring network after constraint adjustment of Free Networks obtain the net form of a plane of GPS network and each website plane coordinates to three-dimensional；Guass plane projection described herein is existing method, repeats no more.

4), utilize the coordinate at arbitrary station that the first phase observes as reference position in each website plane coordinates, this station and another station are constituted the direction of baseline as reference direction, the plane coordinates of the GPS network that each issue below is observed carries out translating and rotation transformation in Gaussian plane, the net form making each issue is all unified on a fixing point and a fixing reference direction, thus the coordinate after being changed.Described in detail below:

If the GPS network survey station number built is m, observing total issue is the n phase, and the first phase reference station coordinates of selection is (x₀,y₀), the grid azimuth on reference baseline limit is α₀, then each all station coordinates of phase GPS network new coordinate after translation rotation transformation is:

${x^{\′}}_i^j=x_0+(x_i^j-x_0^j)cos({\mathrm{\α}}_0^j-{\mathrm{\α}}_0)-(y_i^j-y_0^j)sin({\mathrm{\α}}_0^j-{\mathrm{\α}}_0)---\left(1\right)$

${y^{\′}}_i^j=y_0+(y_i-y_0^j)sin({\mathrm{\α}}_0^j-{\mathrm{\α}}_0)-(y_i-y_0^j)cos({\mathrm{\α}}_0^j-{\mathrm{\α}}_0)---\left(2\right)$

In formula, i represents survey station number, and value is i=1 ... m；J represents observation issue, and value is j=1 ... n.

5), the coordinate after changing is utilized to ask for all lengths of base of whole GPS network, baseline angle, azimuth of director from the gun and all Station Displacements, as four indexs weighing GPS network deformation.Specific formula for calculation is as described below:

The plane basic lineal vector that survey station i and survey station i+1 is constituted is:

${{\mathrm{dx}}^{\′}}_{i,i+1}^j={x^{\′}}_{i+1}^j-{x^{\′}}_i^j---\left(3\right);$

${{\mathrm{dy}}^{\′}}_{i,i+1}^j={y^{\′}}_{i+1}^j-{y^{\′}}_i^j---\left(4\right);$

The length of base that then survey station i and survey station i+1 is constituted is:

$D_{i,i+1}^j=\sqrt{{\mathrm{dx}}^{\′}{{}_{i,i+1}^j}^2+{\mathrm{dy}}^{\′}{{}_{i,i+1}^j}^2}---\left(5\right);$

Angle between baseline i～i+1 and baseline i～i+2 is:

${\mathrm{Ang}}_{D_{i,i+1},D_{i,i+2}}^j=\arccos \left(\frac{{D_{i,i+1}^j}^2+{D_{i,i+2}^j}^2-{D_{i+1,i+2}^j}^2}{2D_{i,i+1}^j{D}_{i,i+2}^j}\right)\·\frac{180}{\mathrm{\π}}\·3600---\left(6\right);$

The azimuth of baseline i～i+1 is:

$R_{i,i+1}^j=\arctan \left(\frac{{{\mathrm{dy}}^{\′}}_{i,i+1}^j}{{{\mathrm{dx}}^{\′}}_{i,i+1}^j}\right)\·\frac{180}{\mathrm{\π}}\·3600---\left(7\right);$

Survey station i displacement is:

${{\mathrm{dx}}^{\′}}_i^j={x^{\′}}_i^{j+1}-{x^{\′}}_i^j---\left(8\right);$

${{\mathrm{dy}}^{\′}}_i^j={y^{\′}}_i^{j+1}-{y^{\′}}_i^j---\left(9\right).$

6), build Time-space serial matrix, carry out main constituent space-time response analysis；Specifically include following steps: 61), set Time-space serial matrix as X_m×n, wherein, every a line represents a Station Displacements, a length of base, an azimuth of director from the gun or the changing value of baseline angle all phases；The changing value of phase all Station Displacements, the length of base, azimuth of director from the gun or baseline angle is shown in each list；Even every a line represents the changing value of Station Displacements all phase, then the changing value of phase all Station Displacements is shown in each list；

62), for ensureing that all changes time series is all with zero for basic point, by formula (10) to matrix X_m×nIn each element implement centralization:

$X^{\′}(i,j)=X(i,j)-\frac{\underset{k=1}{\overset{m}{\Σ}}X(i,k)}{m}---\left(10\right)$

In formula, i, the ranks number of j representing matrix, i=1,2 ... m；J=1,2 ... n；

63), to the matrix X' after centralization_m×nCarry out singular value decomposition, obtain orthogonal matrix U_m×m、V_n×nWith the matrix S being made up of singular value_m×n, it is expressed as:

X'_m×n=U_m×mS_m×nV_n×n ^T(11)

64), the size of singular value reflects the main constituent spatiotemporal mode component size to time series contribution rate, singular value is arranged from big to small, front several mode component sums are called holotype component (generally taking first principal component), and it contains the most information of former sequence；If taking front r for holotype, then matrix X'_m×nApproximate representation is:

X_m×n'≈X'_r=U_rS_rV_r ^T(12)

In formula, S_rFor S_m×nIn before the diagonal matrix of r bigger singular value composition, V_rFor V_n×nThe matrix of r column vector composition before middle correspondence, each of which row are called time holotype vector；U_rFor U_m×mThe matrix of r column vector composition before middle correspondence, each of which row are called space holotype vector, reflect the response magnitude of each deformation of standing in whole time series.

7), Deformation Anomalies Annual distribution and Deformation Anomalies spatial distribution are extracted: after utilizing principal component decomposition to obtain space-time holotype vector, by time holotype vector reflecting regional deformation process over time；By the space holotype vector reflecting regional deformation influence degree to survey stations all in GPS network, described in detail below:

71) curve time response, analyzed area deformation dynamic change characterization in time, according to time holotype vector are obtained；Obtain roomage response figure according to space holotype vector, analyzed area deformation to survey station each in GPS network Different Effects degree；

72), from whether curve time response is that homogenous linear change judges the time that Deformation Anomalies occurs, if the linear trend that curve deviation is original, or present non-linear, non-uniform change or have violent concussion, then show very likely to there occurs in this time period Deformation Anomalies；Judging, from the size of each survey station roomage response, the spatial distribution that deformation occurs, the survey station that roomage response is bigger is relatively big by influence of crust deformation, illustrates that the regional extent that distance deformation occurs is nearer；Otherwise, the survey station that roomage response is less is less by influence of crust deformation, distance deformation generation area is described farther out, thus judges the regional extent that deformation occurs.

For the crustal deformation that Lushan Ms7.0 earthquake in 2013 occurs, the odd-numbered day solution change in time and space sequence time response that the method for the invention extracts angle between the GNSS all lengths of base of net and baseline is adopted to see Fig. 2 and Fig. 3 respectively.As shown in Figure 2, length of base variation tendency is generally in linearly increasing state, once occurs in that once blocking before shake, but not fairly obvious.Versus baseline length change, the baseline variable angle that Fig. 3 reflects preseismic deformation show more prominent, it can be clearly seen that about shake first 9 months (about in June, 2012) start to occur in that bigger fluctuation, the trend of linear change before deviate from.The PC1 roomage response feature extracting the GNSS net length of base and baseline variable angle is shown in Fig. 4 and Fig. 5.In figure, different colours represents different roomage response degree, increases on the occasion of representing, and negative value represents and reduces.As shown in Figure 4, baseline SCXJ-SCLH decrement is relatively large, accounts for 23%；Next to that baseline SCXJ-SCMB and SCDF-SCSM, account for 15% and 16%.Baseline SCXJ-SCYX and baseline SCXJ-SCSM is in expansion state, respectively accounts for 16% and 12%；Other responsiveness is less, lower than 9%.Generally speaking, relevant to survey station SCXJ length of base change is relatively large.And the survey station SCXJ station that distance earthquake centre is nearest just.The roomage response of baseline variable angle as shown in Figure 5 is seen, the roomage response degree of the variable angle relevant to survey station SCXJ, SCMB is relatively big, respectively-20% and 27%.Visible, angle deformation occurs mainly in the east region of GNSS net, and this matches with epicentral location.

It follows that the present invention can obtain the dynamic evolution process of Regional Crust deformation spatial-temporal distribution characteristic and deformation, be conducive to the extraction of On Fluctuations deformation data and the detection of Deformation Anomalies information.

The GNSS survey station of survey region is set up GNSS net by the present invention, carries out three-dimensional without retraining free adjustment, coordinate result is transformed into Gaussian plane.And using a point and a limit in net as reference position and reference direction, GNSS net has been carried out rotation and the translation transformation of net form.Though floating relatively big because of three-dimensional without constraint each issue coordinate frame of adjustment of Free Networks result, but GNSS geometry network structure being constant；And not by the impact of common-mode error and uncorrelated noise, relative positional accuracy is higher.Its determined observation station position geometry is solely dependent upon the quality of observation data；The mutation analysis utilizing whole GNSS net form can efficiently separate the impact of uncorrelated noise, concentrates and highlights the deformation data that high spatial is relevant, has the micro-deformation data ability of certain seizure.

Feature according to regional deformation information high spatial dependency, using GNSS net form as research unit, net variable angle between the change of all lengths of base, baseline, azimuth of director from the gun change and all four aspects of station horizontal displacement time series from GNSS respectively and comprehensively construct the index system weighing the change of whole net form.During by the main constituent of these four aspects, null response obtains Regional Crust deformation spatial-temporal distribution characteristic and dynamic evolution process.Result of study shows, the method is more conducive to extraction and the Deformation Anomalies detection of On Fluctuations deformation data.This explores for deformation monitoring and dynamical mechanism thereof and provides certain theoretical basis and reference frame.

Claims (5)

1. the Deformation Anomalies detection method based on the change of GNSS net form, it is characterised in that: comprise the following steps:

1), the GNSS survey station of the long-term Continuous Observation in detection region is set up GNSS space-time monitoring network；

2), each phase GNSS space-time is monitored network and carries out Baselines and three-dimensional without constraint adjustment of Free Networks；

3), the result after three-dimensional nothing constraint adjustment of Free Networks is carried out gauss projection and obtain the net form of a plane and each website plane coordinates；

4), utilize the website coordinate in first phase observational network as reference position, the base direction this website and another website constituted is as reference direction, undertaken translating and rotation transformation by the GNSS net that other each phase is observed, the net form making each issue is all unified on a fixing point and a fixing reference direction, and coordinate after the conversion of all survey stations in being netted；

5) coordinate after, utilizing conversion asks for the whole GNSS change in time and space sequence netting the key element of all lengths of base, baseline angle, azimuth of director from the gun and all Station Displacements four；

6), build this four key elements space-time matrix, carry out main constituent space-time response analysis；

7), analyze deformation characteristics, extract Deformation Anomalies spatial and temporal distributions.

2. the Deformation Anomalies detection method based on the change of GNSS net form as claimed in claim 1, it is characterized in that: described step 4) particularly as follows: the GNSS space-time monitoring network survey station number setting structure is individual as m, observing total issue is the n phase, and the first phase reference station coordinates of selection is (x₀,y₀), the grid azimuth on reference baseline limit is α₀, then each phase GNSS space-time monitoring all station coordinates of network new coordinate after translation rotation transformation is:

${x^{\′}}_i^j=x_0+(x_i^j-x_0^j)cos({\mathrm{\α}}_0^j-{\mathrm{\α}}_0)-(y_i^j-y_0^j)sin({\mathrm{\α}}_0^j-{\mathrm{\α}}_0)---\left(1\right)$

${y^{\′}}_i^j=y_0+(y_i-y_0^j)sin({\mathrm{\α}}_0^j-{\mathrm{\α}}_0)-(y_i-y_0^j)cos({\mathrm{\α}}_0^j-{\mathrm{\α}}_0)---\left(2\right)$

In above formula, i represents survey station number, and value is i=1......m；J represents observation issue, and value is j=1......n.

3. the Deformation Anomalies detection method based on the change of GNSS net form as claimed in claim 2, it is characterized in that: the coordinate after utilizing conversion in described step (5) is asked for whole GNSS and netted all lengths of base, baseline angle, azimuth of director from the gun and all Station Displacements, described in detail below:

The plane basic lineal vector that survey station i and survey station i+1 is constituted is:

${{\mathrm{dx}}^{\′}}_{i,i+1}^j={x^{\′}}_{i+1}^j-{x^{\′}}_i^j---\left(3\right);$

${{\mathrm{dy}}^{\′}}_{i,i+1}^j={y^{\′}}_{i+1}^j-{y^{\′}}_i^j---\left(4\right);$

The length of base that then survey station i and survey station i+1 is constituted is:

$D_{i,i+1}^j=\sqrt{{\mathrm{dx}}^{\′}{{}_{i,i+1}^j}^2+{\mathrm{dy}}^{\′}{{}_{i,i+1}^j}^2}---\left(5\right);$

Angle between baseline i～i+1 and baseline i～i+2 is:

${\mathrm{Ang}}_{D_{i,i+1},D_{i,i+2}}^j=\arccos \left(\frac{{D_{i,i+1}^j}^2+{D_{i,i+2}^j}^2-{D_{i+1,i+2}^j}^2}{2D_{i,i+1}^j{D}_{i,i+2}^j}\right)\·\frac{\mathrm{180}}{\mathrm{\π}}\·3600---\left(6\right);$

The azimuth of baseline i～i+1 is:

$R_{i,i+1}^j=arctan\left(\frac{{{\mathrm{dy}}^{\′}}_{i,i+1}^j}{{{\mathrm{dx}}^{\′}}_{i,i+1}^j}\right)\·\frac{180}{\mathrm{\π}}\·3600---\left(7\right);$

Survey station i displacement is:

${{\mathrm{dx}}^{\′}}_i^j={x^{\′}}_i^{j+1}-{x^{\′}}_i^j---\left(8\right);$

${{\mathrm{dy}}^{\′}}_i^j={y^{\′}}_i^{j\mathrm{+1}}-{y^{\′}}_i^j---\left(9\right).$

4. the Deformation Anomalies detection method based on the change of GNSS net form as claimed in claim 3, it is characterised in that: described step (6) specifically includes:

61) Time-space serial matrix, is set as X_m×n, wherein, every a line represents a Station Displacements, a length of base, an azimuth of director from the gun or the changing value of baseline angle all phases；The changing value of phase all Station Displacements, the length of base, azimuth of director from the gun or baseline angle is shown in each list；

62), for ensureing that all changes time series is all with zero for basic point, by formula (10) to matrix X_m×nIn each element implement centralization:

$X^{\′}(i,j)=X(i,j)-\frac{\underset{k=1}{\overset{m}{\Σ}}X(i,k)}{m}---\left(10\right)$

In formula, i, the ranks number of j representing matrix, i=1,2 ... m；J=1,2 ... n；

63), to the matrix X' after centralization_m×nCarry out singular value decomposition, obtain orthogonal matrix U_m×m、V_n×nWith the matrix S being made up of singular value_m×n, it is expressed as:

X'_m×n=U_m×mS_m×nV_n×n ^T(11)；

64), singular value being arranged from big to small, front several mode component sums are called holotype component, and it contains the most information of former sequence；If taking front r for holotype, then matrix X'_m×nApproximate representation is:

X_m×n'≈X'_r=U_rS_rV_r ^T(12)

In formula, S_rFor S_m×nIn before the diagonal matrix of r bigger singular value composition, V_rFor V_n×nThe matrix of r column vector composition before middle correspondence, each of which row are called time holotype vector；U_rFor U_m×mThe matrix of r column vector composition before middle correspondence, each of which row are called space holotype vector.

5. the Deformation Anomalies detection method based on the change of GNSS net form as claimed in claim 4, it is characterised in that: described step (7) specifically includes:

71) curve time response, analyzed area deformation dynamic change characterization in time, according to time holotype vector are obtained；Obtaining roomage response figure according to space holotype vector, analyzed area deformation is to the Different Effects degree of each survey station in GNSS net；

72), from whether curve time response is that homogenous linear change judges the time that Deformation Anomalies occurs, if the linear trend that curve deviation is original, or present non-linear, non-uniform change or have violent concussion, then show very likely to there occurs in this time period Deformation Anomalies；

From roomage response figure, the size of each survey station roomage response judges the spatial distribution that deformation occurs, and the survey station that roomage response is bigger is relatively big by influence of crust deformation, illustrates that the regional extent that this survey station distance deformation occurs is nearer；Otherwise, the survey station that roomage response is less is less by influence of crust deformation, this survey station distance deformation generation area is described farther out, thus judges the regional extent that deformation occurs.