Settlement monitoring method based on ground observation data and D-InSAR integration

Abstract

The invention relates to the field of ground subsidence monitoring, and particularly discloses a subsidence monitoring method based on ground observation data and D-I nSAR integration. The settlement monitoring technology integrating D-I nSAR and conventional leveling is provided, the precision is improved, meanwhile, the coverage range of ground settlement monitoring is considered, large-range and full-coverage settlement fusion monitoring on the ground is achieved, multi-data-source, multi-mode and point-face-combined large-range and full-coverage settlement analysis can be conducted, and the settlement monitoring precision is improved. The D-I nSAR technology is introduced into conventional land subsidence monitoring, the field workload of leveling can be greatly saved, and the use cost is low.

Description

Technical Field

[0001] The invention belongs to the technical field of ground subsidence monitoring, and in particular relates to a subsidence monitoring method based on the integration of ground observation data and D-InSAR. Background Art

[0002] Land subsidence has become one of the major environmental geological disasters in some cities and mining areas in my country. It will not only cause serious damage to urban housing, roads, railways, bridges and other infrastructure, affecting the personal safety of the city, but also endanger the ecological environment and the safety of residents in the mining area and its surrounding areas. Therefore, it is necessary to adopt a feasible and efficient way to conduct full coverage of land subsidence monitoring in cities and mining areas, and on this basis, conduct big data analysis of subsidence observation, explore the laws of land subsidence, and realize early warning analysis of land subsidence in cities and mining areas.

[0003] Ground subsidence has become a difficult problem that urgently needs to be solved in my country's urban construction, mining safety, environmental management and other aspects. It is necessary for every city and mining area to conduct effective monitoring of ground subsidence. Subsidence monitoring is the key to improving the understanding of the subsidence process, mapping ground movement and adapting management strategies to the current situation. If reasonable monitoring methods are integrated, scientific subsidence analysis is conducted based on historical monitoring data, thereby forming a reasonable regional subsidence prediction model, a scientific ground subsidence monitoring and early warning system can be established for cities and mining areas in our province, which will be beneficial for natural resources, urban planning, geological environment, mining production and other departments to have a comprehensive understanding of the status of ground subsidence. At the same time, reasonable early warning technical services can also be established based on the prediction model to provide a scientific basis for urban planning and construction and mining safety production.

[0004] At present, the commonly used means of subsidence monitoring include precision leveling, GPS measurement, D-InSAR, time-series InSAR, etc. Leveling uses the actual measurement of the elevation values ​​of the subsidence sites to establish the subsidence results and prediction models. It is the earliest monitoring method used at home and abroad, and it is also the most common monitoring method in China. The initial investment of leveling subsidence monitoring is small, the construction process is simple, and the accuracy can meet the requirements. However, it is generally less distributed, the routes are sparse, the monitoring cycle is long, and the temporal and spatial resolution is relatively low. It is difficult to timely reflect the changing trend of the increasingly expanding regional ground subsidence, as well as the need to obtain basic data on subsidence in large areas of cities and mining areas. D-InSAR (differential interferometry) deformation measurement technology obtains interference phase through image registration, and can obtain its own deformation information after removing the phase such as its own terrain. It is one of the most popular means of subsidence monitoring. The monitoring accuracy of D-InSAR in areas with good coherence can reach millimeter level, and it can achieve coverage monitoring of large areas such as cities and mining areas. At present, there are many softwares that support D-InSAR data processing, which can save the workload of field work. However, D-InSAR is easily affected by temporal and spatial incoherence and atmospheric delay, and sometimes the monitoring results obtained are not accurate enough. The monitoring effect of D-InSAR is generally evaluated by the leveling monitoring accuracy in the same area.

[0005] Time-series InSAR (time-series interferometry) deformation measurement technology uses multiple time-phase SAR data in the same area, and separates and calculates surface deformation information through spatiotemporal analysis and preprocessing of deformation signals, atmospheric signals, DEM error phase and other signals. Although time-series InSAR overcomes the influence of spatiotemporal decorrelation and atmospheric effects to a certain extent, it requires a large number of SAR images in the study area, has low computational efficiency and is prone to loss of details, making it unsuitable for ground subsidence monitoring in large areas.

[0006] Nowadays, from the perspective of ground subsidence monitoring methods at home and abroad, it is still mainly based on manual single-point measurement such as leveling and GPS to obtain subsidence data of discrete subsidence sites. However, with the increasing application of D-InSAR, the combination of leveling and D-InSAR monitoring is conducive to the complementary advantages of the two, saving manpower and financial resources while ensuring accuracy, and has considerable research and application prospects. Summary of the invention

[0007] In order to overcome the shortcomings of the prior art, the purpose of the present invention is to provide a subsidence monitoring method based on the integration of ground observation data and D-InSAR, so as to solve the problem that traditional subsidence monitoring methods have their own usage defects and cannot achieve large-scale, high-precision and comprehensive ground subsidence monitoring and early warning.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] A subsidence monitoring method based on integration of ground observation data and D-InSAR, comprising:

[0010] The historical level observation data of the target area is mined and analyzed, and the settlement values ​​of the settlement stations in the target area are predicted using different prediction methods. The predicted values ​​are then compared with the measured results, and the accuracy is evaluated using the relative error method. The prediction method with the smallest error is determined as the prediction model for the station.

[0011] Obtain the original SAR image data of the target area, perform differential interferometry on the image data, and obtain the D-InSAR subsidence calculation results of each subsidence site in the target area;

[0012] The settlement values ​​observed at the settlement stations in the target area are compared with the D-InSAR settlement calculation results at the same point and with the same settlement observation time span. The coincidence between the D-InSAR settlement calculation results and the leveling settlement results of each settlement station is checked, and the consistency of the settlement results of the two is evaluated by significance test. The accuracy of the D-InSAR settlement results is evaluated with the leveling results as the standard. Based on the comparison of results and accuracy evaluation, the D-InSAR results are corrected to obtain the corrected D-InSAR settlement results.

[0013] Based on the corrected D-InSAR subsidence results and combined with the leveling settlement values ​​of each subsidence station, a point-to-surface settlement result is formed through a reasonable fusion method, and settlement contour lines are generated based on the settlement results. They are tracked, smoothed and annotated to visualize the settlement results and produce thematic maps of settlement results.

[0014] Preferably, the prediction method for predicting the settlement value of the settlement site in the target area includes a grey prediction GM (1,1) model, a curve fitting method and a PSO-BP algorithm.

[0015] Preferably, the step of performing differential interferometry on the image data includes: image registration of original SAR image data, differential interferometry, filtering processing and baseline estimation, combining the baseline file obtained by baseline estimation with an external DEM to simulate the SAR phase, performing differential interferometry again on the simulated SAR phase in combination with the baseline estimation, and then performing filtering processing, phase unwrapping, phase conversion and geocoding on the result of the second differential interferometry to obtain the surface deformation, that is, the D-InSAR settlement calculation result.

[0016] Preferably, the method for evaluating the accuracy of the D-InSAR sedimentation results includes a standard deviation evaluation factor.

[0017] Preferably, the method of evaluating the accuracy of D-InSAR settlement results also includes performing reasonable spatial interpolation using the results of leveling measurements at settlement sites, and comparing whether the planar settlement results generated by interpolation of the leveling results are consistent with the results directly calculated by D-InSAR.

[0018] Compared with the prior art, the present invention has the following beneficial effects:

[0019] The present invention proposes to adopt the settlement monitoring technology integrating D-InSAR and conventional leveling, which improves the accuracy while taking into account the coverage of ground settlement monitoring, realizes large-scale and full-coverage fusion settlement monitoring of the ground, and can carry out large-scale and full-coverage settlement analysis combining multiple data sources, multiple methods, and points and surfaces. The introduction of D-InSAR technology into conventional ground settlement monitoring can greatly save the field workload of leveling measurement, and has a low cost of use. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 A block diagram of a settlement monitoring method based on ground observation data and D-InSAR integration disclosed in the present invention;

[0021] Figure 2 The present invention discloses a technical roadmap for a subsidence monitoring method based on the integration of ground observation data and D-InSAR. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] Example:

[0024] See also Figure 1 - Figure 2 As shown, a subsidence monitoring method based on the integration of ground observation data and DI nSAR includes:

[0025] The historical level observation data of the target area is mined and analyzed, and the settlement values ​​of the settlement stations in the target area are predicted using different prediction methods. The predicted values ​​are then compared with the measured results, and the accuracy is evaluated using the relative error method. The prediction method with the smallest error is determined as the prediction model for the station.

[0026] The original SAR image data of the target area is obtained, and differential interferometry is performed on the image data to obtain the D-InSAR subsidence calculation results of each subsidence site in the target area. The data obtained by the ALOS-2L spaceborne SAR synthetic aperture radar launched by the Japan Space Agency and the Sentinel-1A / B earth observation satellite launched by the European Space Agency's Copernicus program can be used, and then the differential interferometry process of the image data pair is completed through the existing radar image processing software ENVISARscape;

[0027] The settlement values ​​observed at the settlement stations in the target area are compared with the D-InSAR settlement calculation results at the same point and with the same settlement observation time span. The coincidence between the D-InSAR settlement calculation results and the leveling settlement results of each settlement station is checked, and the consistency of the settlement results of the two is evaluated by significance test. The accuracy of the D-InSAR settlement results is evaluated with the leveling results as the standard. Based on the comparison of results and accuracy evaluation, the D-InSAR results are corrected to obtain the corrected D-InSAR settlement results.

[0028] Based on the corrected D-InSAR subsidence results and combined with the leveling settlement values ​​of each subsidence station, a point-to-surface settlement result is formed through a reasonable fusion method, and settlement contour lines are generated based on the settlement results. They are tracked, smoothed and annotated to visualize the settlement results and produce thematic maps of settlement results.

[0029] From the above, it can be seen that by adopting the settlement monitoring technology that integrates D-InSAR and conventional leveling, the accuracy is improved while taking into account the coverage of ground settlement monitoring, realizing large-scale and full-coverage fusion settlement monitoring of the ground, and can conduct large-scale and full-coverage settlement analysis combining multiple data sources, multiple methods, and points and surfaces. Introducing D-InSAR technology into conventional ground subsidence monitoring can greatly save the field workload of leveling measurement, and the cost of use is relatively low.

[0030] The prediction method for predicting the settlement value of the settlement site in the target area includes a grey prediction GM (1,1) model, a curve fitting method and a PSO-BP algorithm.

[0031] The GM (1,1) model is widely used in surface deformation prediction, especially in the study of settlement prediction along subway lines and mines, and has achieved relatively satisfactory prediction results. Its basic principles are as follows:

[0032] Let x (0) The non-negative equally spaced original sequence is:

[0033] x (0) ={x (0) (1),x (0) (2)…,x (0)(n)}

[0034] For the original sequence x (0) Do an accumulation to generate a new sequence:

[0035] x (1) ={x (1) (1),x (1) (2)…,x (1) (n)}

[0036] in:

[0037]

[0038] For the sequence x (1) (t) Establish the first-order differential equation

[0039]

[0040] Among them, a and u are gray parameters. According to the least squares principle, the differential equation of this formula is solved, and the parameter vector of the model is:

[0041]

[0042] Where:

[0043]

[0044] The time function of the model is:

[0045]

[0046] right Generate cumulatively and restore data:

[0047]

[0048] or

[0049]

[0050] The curve fitting method has the advantages of being simple, efficient and easy to master, and has been widely used in practice. There are many existing curve fitting methods. This study intends to select 4 commonly used curve fitting methods for settlement prediction. The calculation expressions are shown in the following table:

[0051]

[0052] The PSO-BP algorithm neural network combines the particle swarm optimization (PSO) and the back-propagation (BP) neural network, making up for the shortcomings of the BP neural network, such as slow convergence speed and easy to fall into local extreme values. This method can be used as an effective means of long-term surface settlement prediction. Its principle is to combine the PSO algorithm with the error back propagation training method of the BP neural network, take the error as the fitness function, and use the PSO algorithm to iteratively optimize the initial weights and thresholds of the BP neural network to obtain the initial weights and thresholds with high fitness. These parameters are then used in the BP neural network, and finally the values ​​that meet the accuracy requirements are output. The learning steps are as follows:

[0053] Step 1: Use the global optimal solution after PSO optimization as the initial weight w of the BP neural network ij 、w ik and threshold b j , b k Among them, w ij is the weight from the input layer to the middle layer, w ik is the weight from the hidden layer to the output layer, b j is the threshold of each neuron in the hidden layer, b k is the threshold of each neuron in the output layer, i is the number of nodes in the input layer, j is the number of nodes in the hidden layer, and k is the number of nodes in the input layer.

[0054] Step 2: After normalization, randomly select training samples x(k)=x1(k),…x k (k) and the corresponding expectation d(k) = d1(k),…d k (k).

[0055] Step 3: Calculate the input h of the hidden layer j (k), then use h j (k) and the activation function to calculate the output h of each neuron in the hidden layer j (k).

[0056] Step 4: Calculate the partial derivative δ of the error function for each neuron in the output layer based on the network prediction y(x) and the expected output d(k) k .

[0057] Step 5: Use the connection weight w from the hidden function layer to the output layer ik (k), δ of the output layer k (k) and the output h of the hidden layer j (k) Calculate the partial derivative δ of the error function with respect to each neuron in the hidden layer k (k).

[0058] Step 6: According to δ k(k) and h j (k) to correct the weight w ik (k) and threshold b k :

[0059] W jk =W jk +ηδ k (k)h j (k),

[0060] b k =b k +ηδ k (k),

[0061] According to δ j (k) and x i (k) to modify the right w ij and threshold b j :

[0062] W ij =W ij +ηδ j (k)x i (k),

[0063] b j =b j +ηδ j (k),

[0064] Step 7: Calculate the global error and determine whether the network error meets the requirements.

[0065] The steps of performing differential interferometry on the image data include: image registration of original SAR image data, differential interferometry, filtering processing and baseline estimation, combining the baseline file obtained by baseline estimation with an external DEM to simulate the SAR phase, performing differential interferometry again on the simulated SAR phase combined with the baseline estimation, and then performing filtering processing, phase unwrapping, phase conversion and geocoding on the result of the second differential interferometry to obtain the surface deformation, that is, the D-InSAR settlement calculation result.

[0066] Differential Synthetic Aperture Radar Interferometry (D-InSAR) technology refers to the use of two SAR data of the same area acquired at different time points in the same spatial position to obtain the interference phase by conjugating and multiplying the phase information in the SAR data. The interference phase can be expressed as:

[0067]

[0068] In the above formula, is the phase information of the two SAR images before and after, The interference phase caused by the earth's ellipsoid is called the flat earth effect and should be removed in the interference processing; The interference phase caused by the change of surface elevation is called terrain phase; is the deformation phase caused by the target along the radar line of sight; is the atmospheric delay phase, is the noise phase.

[0069] Through filtering, phase unwrapping, atmospheric delay phase removal and other processing, the right side of the above formula is Keep some of them and eliminate the rest, thus obtaining the surface deformation information.

[0070] The data in this paper are processed through image registration and multi-view processing, original interference pattern generation and terrain phase removal, interference pattern filtering and coherence coefficient map calculation, phase unwrapping, deformation acquisition and geocoding. The data processing flow is as follows: Figure 2 As shown in the raw SAR image data processing section.

[0071] The method for evaluating the accuracy of D-InSAR subsidence results includes a standard deviation evaluation factor, and also includes reasonable spatial interpolation using the results of leveling measurements at subsidence sites, and comparing whether the planar subsidence results generated by interpolation of the leveling results are consistent with the results directly calculated by D-InSAR.

[0072] In some embodiments, there are many settlement prediction methods based on ground measured leveling data. There may be many problems when randomly selecting a method and applying it to the study area, such as not being able to truly reflect the characteristics of ground settlement changes in the test area; there is a large difference between the estimated settlement value and the measured value, etc. To this end, based on the existing prediction methods, the present invention combines the ground observation data of the test area, summarizes the characteristics of the existing settlement status, combines the possible causes of settlement in the area, conducts a comprehensive analysis of the settlement laws of the study area, analyzes the applicability of different prediction models under different characteristic indicators, and determines a reasonable model to predict the development trend of settlement, providing an effective technical means for accurate monitoring and prediction of settlement.

[0073] Conventional D-InSAR technology is susceptible to temporal and spatial decoherence and atmospheric effects. The shooting time of the original SAR data should be as close as possible to the leveling time. The interference of the data itself will reduce the coherence of the data pair, thus affecting the effect and accuracy of differential interferometry. Therefore, data pairs with high coherence should be selected for differential interferometry. If the coherence is still very low, consider using interference superposition techniques such as permanent scatterer technology (PS) or small baseline set (SBAS) to reduce temporal and spatial decoherence factors. The settlement results of D-InSAR are checked and verified with the measured leveling data as the standard to determine whether D-InSAR can meet the accuracy requirements of grade leveling measurement, and the feasibility of fusion monitoring is analyzed from the difference in settlement results between the two.

[0074] The data of leveling measurement is the time series elevation of each settlement point, and the settlement results obtained are also single point discrete; the settlement results generated by D-InSAR are continuous in surface form. Therefore, on the premise of verifying the feasibility of fusion monitoring, a D-InSAR settlement result correction model based on ground level settlement observation data is established in an appropriate way to form accurate settlement results. The discrete settlement points and continuous settlement surfaces form the overall settlement results and are visualized.

[0075] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example" or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0076] In the drawings of the embodiments disclosed in the present invention, only the structures related to the embodiments disclosed in the present invention are involved, and other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of the present invention can be combined with each other.

Claims

1. A settlement monitoring method based on ground observation data and D-InSAR integration, characterized in that: include: The historical level observation data of the target area is mined and analyzed, and the settlement values ​​of the settlement stations in the target area are predicted using different prediction methods. The predicted values ​​are then compared with the measured results, and the accuracy is evaluated using the relative error method. The prediction method with the smallest error is determined as the prediction model for the station. Obtain the original SAR image data of the target area, perform differential interferometry on the image data, and obtain the D-InSAR subsidence calculation results of each subsidence site in the target area; The settlement values ​​observed at the settlement stations in the target area are compared with the D-InSAR settlement calculation results at the same point and with the same settlement observation time span. The coincidence between the D-InSAR settlement calculation results and the leveling settlement results of each settlement station is checked, and the consistency of the settlement results of the two is evaluated by significance test. The accuracy of the D-InSAR settlement results is evaluated with the leveling results as the standard. Based on the comparison of results and accuracy evaluation, the D-InSAR results are corrected to obtain the corrected D-InSAR settlement results. Based on the corrected D-InSAR subsidence results and combined with the leveling settlement values ​​of each subsidence station, a point-to-surface settlement result is formed through a reasonable fusion method, and settlement contour lines are generated based on the settlement results. They are tracked, smoothed and annotated to visualize the settlement results and produce thematic maps of settlement results.

2. The method for monitoring subsidence based on integration of ground observation data and D-InSAR according to claim 1, characterized in that: The prediction method for predicting the settlement value of the settlement site in the target area includes a grey prediction GM (1,1) model, a curve fitting method and a PSO-BP algorithm.

3. The method for monitoring subsidence based on integration of ground observation data and D-InSAR according to claim 1, characterized in that: The steps of performing differential interferometry on the image data include: image registration of original SAR image data, differential interferometry, filtering processing and baseline estimation, combining the baseline file obtained by baseline estimation with an external DEM to simulate the SAR phase, performing differential interferometry again on the simulated SAR phase combined with the baseline estimation, and then performing filtering processing, phase unwrapping, phase conversion and geocoding on the result of the second differential interferometry to obtain the surface deformation, that is, the D-InSAR settlement calculation result.

4. The method for monitoring subsidence based on integration of ground observation data and D-InSAR according to claim 1, characterized in that: The method for evaluating the accuracy of the D-InSAR subsidence results includes a standard deviation evaluation factor.

5. The method for monitoring subsidence based on integration of ground observation data and D-InSAR according to claim 1, characterized in that: The method for evaluating the accuracy of D-InSAR settlement results also includes performing reasonable spatial interpolation using the results of leveling measurements at settlement sites, and comparing whether the planar settlement results generated by interpolation of the leveling results are consistent with the results directly calculated by D-InSAR.

Images