CN117557741A - Elevation anomaly model construction method and system based on terrain model matching - Google Patents
Elevation anomaly model construction method and system based on terrain model matching Download PDFInfo
- Publication number
- CN117557741A CN117557741A CN202410037171.5A CN202410037171A CN117557741A CN 117557741 A CN117557741 A CN 117557741A CN 202410037171 A CN202410037171 A CN 202410037171A CN 117557741 A CN117557741 A CN 117557741A
- Authority
- CN
- China
- Prior art keywords
- model
- elevation
- preliminary
- terrain
- points
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000010276 construction Methods 0.000 title claims abstract description 25
- 238000000034 method Methods 0.000 claims abstract description 61
- 238000013528 artificial neural network Methods 0.000 claims abstract description 39
- 230000002159 abnormal effect Effects 0.000 claims abstract description 37
- 238000009826 distribution Methods 0.000 claims description 28
- 238000004364 calculation method Methods 0.000 claims description 14
- 125000004122 cyclic group Chemical group 0.000 claims description 4
- 238000010801 machine learning Methods 0.000 abstract description 11
- 230000005856 abnormality Effects 0.000 abstract description 10
- 238000004458 analytical method Methods 0.000 abstract description 3
- 230000006870 function Effects 0.000 description 11
- 230000008569 process Effects 0.000 description 8
- 238000005259 measurement Methods 0.000 description 6
- 230000000306 recurrent effect Effects 0.000 description 3
- 238000012876 topography Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000005484 gravity Effects 0.000 description 2
- 238000000691 measurement method Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 210000004205 output neuron Anatomy 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 239000002689 soil Substances 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000009827 uniform distribution Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
- G06T17/05—Geographic models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Geometry (AREA)
- Computer Graphics (AREA)
- Remote Sensing (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
An elevation anomaly model construction method and system based on terrain model matching belong to the field of elevation anomaly, and comprise the steps of selecting a plurality of groups of public fitting points and a plurality of check points from a terrain model; respectively constructing a preliminary elevation anomaly model by using a quadric surface model method and a plurality of neural networks; and respectively calculating the model precision of each preliminary elevation abnormal model by utilizing the position data of the plurality of check points, and selecting the preliminary elevation abnormal model with the highest model precision as a final elevation abnormal model to output. The modeling method for the elevation abnormality in the area range is established based on the quadric surface model method and a plurality of machine learning methods, fine comparison analysis is carried out on the elevation abnormality model established by various methods, and compared with the traditional elevation abnormality modeling, the modeling method has the advantages that the surface fitting method which adopts a quadratic polynomial and the like is easy to be influenced by complex terrain, the problem of poor area precision is caused, and the elevation abnormality model with highest precision can be generated according to different terrain areas.
Description
Technical Field
The application relates to the field of elevation anomaly, in particular to an elevation anomaly model construction method and system based on terrain model matching.
Background
Elevation is one of the basic geometric elements of measurement and can be obtained by leveling, triangulation and the like. The traditional leveling accuracy is high, but because some hydropower engineering projects are located in high mountain areas, the topography is complex, vegetation covers, the traffic is inconvenient, the vision condition is poor, and leveling implementation is very difficult. Due to the characteristics of high measurement efficiency, high precision, easy operation and the like of the global navigation satellite system (Global Navigation Satellite System, GNSS), the application of the global navigation satellite system in engineering is more and more widespread in recent years.
GNSS can not directly measure the height of the measuring station, but obtains the three-dimensional coordinates of the measuring station in a space rectangular coordinate system. Taking a GPS system as an example, the three-dimensional coordinate is based on WGS-84 ellipsoids, and is different from a geodetic level according to the internal data processing basis, and the elevation abnormality of a measuring area is determined by further passing through GNSS measurement results and the geodetic level, and then the elevation of a measuring station is calculated.
The GNSS elevation anomaly determination method comprises a gravity measurement method, a adjustment conversion method, a mathematical fitting method and the like. Among them, the gravity measurement method and the differential conversion method are limited by factors such as instruments, algorithms and the like, and engineering application is less. In a number of practices for determining elevation anomalies, mathematical fitting methods are employed. And (3) at the level points of the GNSS network joint measurement part, the elevation abnormality of the level points is obtained through the normal heights and the geodetic heights of the level points, then a function model of the high Cheng Yi constant value and the plane coordinates is established through a mathematical fitting method, and the high Cheng Yi constant value of other points in the measurement area is obtained.
However, the current calculation method of the elevation abnormal value cannot adapt to various terrain conditions, and the accuracy of the calculation result is not ideal.
Disclosure of Invention
The application provides an elevation abnormal model construction method and system based on terrain model matching, which can solve the technical problems that an elevation abnormal value calculation method in the prior art cannot adapt to various terrain conditions and calculation result accuracy is not ideal.
In a first aspect, an embodiment of the present application provides a method for constructing an elevation anomaly model based on terrain model matching, where the method includes:
selecting a plurality of groups of public fitting points and a plurality of check points from a terrain model, and acquiring position data of the points, wherein the position data comprises geodetic coordinates, geodetic heights and normal heights;
respectively constructing a plurality of preliminary elevation anomaly models based on the position data of the plurality of groups of public fitting points by using a quadric surface model method and a plurality of neural networks;
and respectively calculating the model precision of each preliminary elevation abnormal model by utilizing the position data of the plurality of check points, and selecting the preliminary elevation abnormal model with the highest model precision as a final elevation abnormal model to output.
With reference to the first aspect, in one embodiment, the common fitting points are normally distributed in a terrain model; the method further comprises the steps of:
before the preliminary elevation anomaly model is built, the distribution variance of the preliminary elevation anomaly model is calculated by using the geodetic coordinates of each group of public fitting points;
after calculating the model precision of the preliminary elevation abnormal model, adjusting the distribution variance according to the precision result, and updating the model according to the public fitting point corresponding to the adjusted distribution variance until the model precision reaches a preset range.
With reference to the first aspect, in an embodiment, the plurality of checkpoints are randomly distributed in the terrain model, and the plurality of checkpoints for model accuracy calculation for all preliminary elevation anomaly models are the same.
With reference to the first aspect, in one implementation manner, the model accuracy includes an inner coincidence accuracy, which is calculated by using the following formula:
wherein,
σ 1 for representing the intra-coincidence accuracy;
i1 is a reference numeral for representing a common fitting point;
u i1 the difference between the true and proposed values for the common fitting point denoted i 1;
n 1 for representing the number of common fitting points.
With reference to the first aspect, in one implementation manner, the model accuracy includes an external coincidence accuracy, which is calculated by using the following formula:
wherein,
σ 2 for representing the accuracy of the outer compliance;
i2 is a reference numeral for indicating a check point;
u i2 a difference between the true value and the expected value for the checkpoint labeled i 2;
n 2 for indicating the number of checkpoints.
With reference to the first aspect, in an implementation manner, the plurality of neural networks include a recurrent neural network RNN, a back propagation BP neural network, and a radial basis function RBF neural network.
In a second aspect, an embodiment of the present application provides an elevation anomaly model building system based on terrain model matching, the system including:
the terrain model module is used for selecting a plurality of groups of public fitting points and a plurality of check points from the terrain model, and acquiring position data of the points, wherein the position data comprises geodetic coordinates, geodetic heights and normal heights;
the preliminary model construction module is used for constructing a plurality of preliminary elevation anomaly models based on the position data of the plurality of groups of public fitting points by using a quadric surface model method and a plurality of neural networks respectively;
and the final model construction module is used for respectively calculating the model precision of each preliminary elevation abnormal model by utilizing the position data of the plurality of check points so as to select the preliminary elevation abnormal model with the highest model precision as a final elevation abnormal model to be output.
With reference to the second aspect, in one embodiment, the common fitting points are normally distributed in a terrain model;
the preliminary model construction module is further used for calculating distribution variances of the preliminary elevation abnormal models by using the geodetic coordinates of each group of public fitting points before constructing the preliminary elevation abnormal models, adjusting the distribution variances according to accuracy results after calculating model accuracy of the preliminary elevation abnormal models, and updating the models according to the public fitting points corresponding to the adjusted distribution variances until the model accuracy reaches a preset range.
With reference to the second aspect, in one embodiment, the plurality of checkpoints are randomly distributed in the terrain model, and the plurality of checkpoints for model accuracy calculation for all preliminary elevation anomaly models are the same.
With reference to the second aspect, in one embodiment, the preliminary model building module and the final model building module are each configured to perform model accuracy calculations on all preliminary elevation anomaly models, and the model accuracy includes an inner compliance accuracy and an outer compliance accuracy.
The beneficial effects that technical scheme that this application embodiment provided include:
the invention establishes a modeling method of elevation abnormality in a measuring area range based on a quadric surface model method and a plurality of machine learning methods, and carries out detailed comparison analysis on the elevation abnormality model established by various methods.
Drawings
FIG. 1 is a schematic flow chart of an embodiment of a method for constructing an elevation anomaly model based on terrain model matching in the present application;
fig. 2 is a schematic diagram of functional modules of an embodiment of an elevation anomaly model building system based on terrain model matching in the present application.
Detailed Description
In order to make the present application solution better understood by those skilled in the art, the following description will clearly and completely describe the technical solution in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art based on the embodiments herein without making any inventive effort, are intended to be within the scope of the present application.
For the purpose of making the objects, technical solutions and advantages of the present application more apparent, the embodiments of the present application will be described in further detail below with reference to the accompanying drawings.
In a first aspect, an embodiment of the present application provides a method for constructing an elevation anomaly model based on terrain model matching.
In an embodiment, referring to fig. 1, fig. 1 is a flowchart of a first embodiment of an elevation anomaly model construction method based on terrain model matching in the present application. As shown in fig. 1, the method for constructing the elevation anomaly model based on terrain model matching comprises the following steps:
and S1, selecting a plurality of groups of public fitting points and a plurality of check points from a terrain model, and acquiring position data of the points, wherein the position data comprises geodetic coordinates, geodetic heights and normal heights.
And S2, constructing a plurality of preliminary elevation anomaly models based on the position data of the plurality of groups of public fitting points by using a quadric surface model method and a plurality of neural networks respectively.
And S3, calculating the model precision of each preliminary elevation abnormal model by using the position data of the plurality of check points respectively, and selecting the preliminary elevation abnormal model with the highest model precision as a final elevation abnormal model to output.
The various neural networks include a recurrent neural network (Recurrent Neural Network, RNN), a back propagation (Backpropagation algorithm, BP) neural network, a radial basis function (Radial Basis Function, RBF) neural network.
In this embodiment, with the development of computer technology, machine learning algorithms are widely used in the fields of mathematical modeling, data processing, and the like. Practice has shown that it is feasible to make GNSS elevation anomaly determination using machine learning algorithms. The matching of the terrain and the ground features is the interaction result of the terrain and the ground features, firstly, the transformation of the ground features to the terrain needs to be discussed, and the segmentation treatment of the plane contour of the ground feature model to the terrain needs to be solved. The invention utilizes the measured data (including the geodetic coordinates, geodetic heights, normal high-level data of each point in the area) of a certain area to respectively utilize a cyclic neural network RNN, a back propagation BP neural network and a radial basis function RBF neural network to carry out elevation anomaly modeling, and compares the elevation anomaly modeling with a quadric surface fitting result of a quadric surface model method, and provides reference for engineering GNSS elevation anomaly modeling based on terrain model matching.
The invention establishes a modeling method of elevation abnormality in a region range based on a quadric surface model method and a plurality of machine learning methods, and carries out detailed comparison analysis on an elevation abnormality model established by a cyclic neural network RNN, a back propagation BP neural network and a radial basis function RBF neural network.
By adopting three machine learning algorithms such as a cyclic neural network RNN, a back propagation BP neural network and a radial basis function RBF neural network, the regional elevation anomaly can be fitted and modeled, and the result can be compared with a quadric surface fitting method. The elevation anomaly model established by using the machine learning algorithm has high precision and small residual error. Based on the same machine learning algorithm, the precision discovery of random selection part fitting point elevation anomaly modeling is researched, and when the common points are distributed more uniformly, the fitting effect is better.
Further, in an embodiment, the quadric surface model method adopts a quadratic polynomial surface to perform elevation anomaly modeling, and then the coefficients of the least square fitting function are used to calculate the elevation anomaly of the unknown point. The process is briefly deduced by assuming m elevation fitting points (the elevation fitting points are common fitting points, and form a group of common fitting points), and constructing a quadric surface f (x) by taking the fitting point plane coordinates as the known GNSS measurement results (namely, the geodetic coordinates and the geodetic height) and the level achievements (normal heights), (x, y) i ，y i ) The elevation anomaly model to describe local weighted linear regression is shown in equation (1) below:
（1）
wherein,the method is used for representing the fitting coefficient, and the physical meaning is a constant value of Cheng Yi. i is used as a reference numeral for representing a common fitting point, and i is 1-m.
According to the indirect adjustment method, an error equation shown in the following formula (2) is established:
（2）
wherein V is i Model residuals for representing points with the reference number i, namely the difference between a true value and a planned value, wherein the true value is true normal height, the planned value is normal height which is pushed to place according to the fitting result of an elevation abnormal model, the model residuals are combined according to a formula (1) and a formula (2), and a coefficient array is obtained under the least square coefficientThereby establishing quadric surface heightAnd (5) a process anomaly model. In actual surface fitting, the fitting precision has a certain relation with fitting point distribution. If the fitting points are unevenly distributed, the error of the elevation abnormal fitting surface is increased. Therefore, the method is more suitable for the area with more gentle relief.
Further, in an embodiment, the common fitting points are normally distributed in the terrain model. The method further comprises the following steps:
before the preliminary elevation anomaly model is constructed, the distribution variance of the preliminary elevation anomaly model is calculated by using the geodetic coordinates of each group of public fitting points.
And after calculating the model precision of the preliminary elevation abnormal model, adjusting the distribution variance according to the precision result, and updating the model according to the public fitting point corresponding to the adjusted distribution variance until the model precision reaches a preset range.
In this embodiment, since the accuracy of the quadric modeling method is affected by the distribution of the common fitting points and the relief situation to a certain extent, the normal distribution of the common fitting points is a relatively uniform distribution manner, which is beneficial to improving the model accuracy of the elevation anomaly model. After a plurality of groups of public fitting points are selected from the terrain model, distance difference can be calculated according to distance values between adjacent fitting points in each group, normal distribution variance is calculated according to the distance difference, after the distribution variance of all groups of public fitting points is known, a preliminary elevation anomaly model is built according to all groups of public fitting points, model precision is calculated, if the model precision is lower than a preset range, the selected public fitting points of each group are adjusted in a mode of adjusting the distribution variance of each group of public fitting points, a preliminary elevation anomaly model is rebuilt or updated according to the adjusted public fitting points, and then a model with highest precision is selected from the preliminary elevation anomaly models generated according to different methods as a final elevation anomaly model and is output.
Compared with the traditional elevation anomaly modeling, the method is easy to be influenced by terrain complexity by adopting a surface fitting method such as a quadratic polynomial, and the like, so that the problem of poor area precision is solved, and a proper public fitting point can be selected according to different terrain areas so as to generate an elevation anomaly model with highest precision.
Further, in an embodiment, the plurality of checkpoints are randomly distributed in the terrain model, and the plurality of checkpoints for performing model accuracy calculation on all the preliminary elevation anomaly models are the same.
In this embodiment, checkpoints for performing model accuracy calculation on the preliminary elevation anomaly model generated by using various methods are the same, for example, after a preliminary elevation anomaly model is generated by using a quadric surface model method, model accuracy calculation is performed by using a plurality of checkpoints (specifically, model accuracy is estimated according to the difference between the true normal height of the checkpoints and the normal height obtained by using the model). After a preliminary elevation anomaly model is generated by using the neural network, model accuracy calculation is performed by using the plurality of check points. The accuracy of the elevation anomaly model generated by using different methods can be accurately estimated.
Further, in an embodiment, the model accuracy includes an inner compliance accuracy calculated by the following formula (3):
（3）
wherein sigma 1 For representing the intra-coincidence accuracy. i1 is used as a reference numeral to denote a common fitting point. u (u) i1 The difference between the true and proposed values for the common fitting point denoted i 1. n is n 1 For representing the number of common fitting points.
The model accuracy includes an external fitting accuracy calculated by the following formula (4):
（4）
wherein sigma 2 For representing the accuracy of the outer compliance. i2 is used to denote the reference numeral of the checkpoint. u (u) i2 The difference between the true and proposed values for the checkpoint labeled i 2. n is n 2 For indicating the number of checkpoints.
Because the accuracy of the quadric surface modeling method is affected by the fitting point distribution and the topography fluctuation condition to a certain extent, the invention adopts three machine learning algorithms to conduct elevation anomaly modeling. In order to study the accuracy of the elevation anomaly model, 4 modeling schemes are designed, and elevation anomaly modeling is respectively carried out. The schemes 1-4 are respectively compared by adopting quadric surfaces, RNN, BP neural network and RBF neural network algorithms, the number of common fitting points of the 4 schemes is 64, the number of check points is 16, and the positions are unchanged.
And 3 machine learning algorithms, namely an RNN, a BP neural network and an RBF neural network, are utilized to perform model training through 64 public fitting points to obtain the relation between the plane coordinates and the elevation anomalies, so that an elevation anomaly model is established, and then the elevation anomalies of the check points are calculated. When modeling is performed by using the common fitting point, parameter tuning is performed on each algorithm in order to obtain a good modeling effect, effectively reduce the residual error of the check point model.
The human transmission layer in the BP neural network and the RBF neural network has 3 neurons which are respectively the plane coordinates of the public fitting point and the geodetic height, and the number of the output neurons is 1 and is Gao Chengyi constant. The elevation anomalies were calculated using 64 common fitting points and 16 checkpoints as model inputs, respectively, and model residual results were calculated from known leveling data, see table 1. And comparing the results of the schemes 1-4 to obtain a model residual error between the common fitting point and the check point. The common fitting point residual error is within the soil 2 cm, and the average value of the checkpoint residual errors in the schemes 2-4 is better than +1 cm.
Table 1 checkpointed elevation anomaly residual under different schemes
And obtaining a model residual error through comparison of the known value of the public fitting point and the modeling fitting value, and calculating the internal fitting precision of the model. And meanwhile, comparing the known value of the check point with the modeling fitting value, and calculating a model residual error to further obtain the external coincidence precision.
And (3) counting the inner coincidence precision and the outer coincidence precision of the schemes 1-4, wherein the results are shown in the following table 2. As can be seen from table 2 below, all 4 schemes achieved an elevation anomaly fitting accuracy better than 1.20 cm. In the aspect of internal coincidence precision, the scheme 2 is equivalent to the scheme 1 in precision, and the method shows that under the conditions of complex topography and large height difference change of a research area, the machine learning algorithm can better perform elevation anomaly modeling. Comparing the schemes 2, 3 and 4, compared with the RNN and BP neural networks, the RBF neural network with the optimized parameters obtains higher modeling precision. In terms of internal coincidence, the RBF neural network achieves sub-cm accuracy. When the external coincidence precision is verified, the 4 schemes all obtain higher elevation anomaly modeling precision. Wherein, scheme 2 is the same as scheme 1 in precision, and scheme 4 is optimal. Meanwhile, as the number of check points is small, the distribution is randomly selected, and therefore the external fitting precision is slightly lower than that of a public fitting point.
Table 2 comparison of precision under different modeling
According to the range of the area, the 32 evenly distributed public fitting points under ideal conditions are designed, and the plane coordinates of the points are calculated. And calculating the fitting point distribution of each scheme based on the distance difference of the 32 evenly distributed points. And calculating the distribution variance of the public fitting points according to the square sum of the distance difference sequence elements through an error propagation law. And designing according to the order of distribution variances from small to large, adopting the same parameter setting for each elevation anomaly model construction method to obtain elevation anomaly results and model residual errors under different common fitting point distribution conditions, and adjusting the common fitting point distribution conditions according to the results so as to improve the model precision of the elevation anomaly model.
In a second aspect, embodiments of the present application further provide an AAAA apparatus.
In an embodiment, referring to fig. 2, fig. 2 is a schematic functional block diagram of an embodiment of an elevation anomaly model building system based on terrain model matching in the present application. As shown in fig. 2, the elevation anomaly model construction system based on terrain model matching includes:
a terrain model module 2 for selecting a plurality of sets of common fitting points and a plurality of checkpoints from a terrain model and obtaining position data of the plurality of points, the position data including geodetic coordinates, geodetic altitude, and normal altitude.
And the preliminary model construction module 3 is used for constructing a plurality of preliminary elevation anomaly models based on the position data of the plurality of groups of common fitting points by using a quadric surface model method and a plurality of neural networks respectively.
And a final model construction module 4, which calculates the model precision of each preliminary elevation abnormal model by using the position data of the plurality of check points, so as to select the preliminary elevation abnormal model with the highest model precision as the final elevation abnormal model to output.
The function implementation of each module in the elevation abnormal model construction system based on the terrain model matching corresponds to each step in the elevation abnormal model construction method embodiment based on the terrain model matching, and the function and the implementation process of the function implementation are not described in detail herein.
It should be noted that, the foregoing embodiment numbers are merely for describing the embodiments, and do not represent the advantages and disadvantages of the embodiments.
From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. ROM/RAM, magnetic disk, optical disk) as described above, comprising several instructions for causing a terminal device to perform the method described in the various embodiments of the present application.
The terms "comprising" and "having" and any variations thereof in the description and claims of the present application and in the foregoing drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those listed steps or elements but may include other steps or elements not listed or inherent to such process, method, article, or apparatus. The terms "first," "second," and "third," etc. are used for distinguishing between different objects and not necessarily for describing a sequential or chronological order, and are not limited to the fact that "first," "second," and "third" are not identical.
In the description of embodiments of the present application, "exemplary," "such as," or "for example," etc., are used to indicate an example, instance, or illustration. Any embodiment or design described herein as "exemplary," "such as" or "for example" is not necessarily to be construed as preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary," "such as" or "for example," etc., is intended to present related concepts in a concrete fashion.
In the description of the embodiments of the present application, "/" means or, unless otherwise indicated, for example, a/B may represent a or B. The text "and/or" is merely an association relation describing the associated object, and indicates that three relations may exist, for example, a and/or B may indicate: the three cases where a exists alone, a and B exist together, and B exists alone, and in addition, in the description of the embodiments of the present application, "plural" means two or more than two.
In some of the processes described in the embodiments of the present application, a plurality of operations or steps occurring in a particular order are included, but it should be understood that these operations or steps may be performed out of the order in which they occur in the embodiments of the present application or in parallel, the sequence numbers of the operations merely serve to distinguish between the various operations, and the sequence numbers themselves do not represent any order of execution. In addition, the processes may include more or fewer operations, and the operations or steps may be performed in sequence or in parallel, and the operations or steps may be combined.
The foregoing description is only of the preferred embodiments of the present application, and is not intended to limit the scope of the claims, and all equivalent structures or equivalent processes using the descriptions and drawings of the present application, or direct or indirect application in other related technical fields are included in the scope of the claims of the present application.
Claims (8)
1. The method for constructing the elevation anomaly model based on terrain model matching is characterized by comprising the following steps of:
selecting a plurality of groups of public fitting points and a plurality of check points from a terrain model, and acquiring position data of the points, wherein the position data comprises geodetic coordinates, geodetic heights and normal heights; the public fitting points are normally distributed in the terrain model;
respectively constructing a plurality of preliminary elevation anomaly models based on the position data of the plurality of groups of public fitting points by using a quadric surface model method and a plurality of neural networks;
calculating the model precision of each preliminary elevation abnormal model by utilizing the position data of the plurality of check points respectively, and selecting the preliminary elevation abnormal model with the highest model precision as a final elevation abnormal model to output;
the method further comprises the steps of:
before the preliminary elevation anomaly model is built, the distribution variance of the preliminary elevation anomaly model is calculated by using the geodetic coordinates of each group of public fitting points;
after calculating the model precision of the preliminary elevation abnormal model, adjusting the distribution variance according to the precision result, and updating the model according to the public fitting point corresponding to the adjusted distribution variance until the model precision reaches a preset range.
2. The terrain model matching-based elevation anomaly model construction method of claim 1, wherein the plurality of checkpoints are randomly distributed in the terrain model and the plurality of checkpoints for model accuracy calculation for all preliminary elevation anomaly models are the same.
3. The method for constructing an elevation anomaly model based on terrain model matching according to claim 1, wherein the model accuracy comprises an inner coincidence accuracy calculated by the following formula:
wherein,
σ 1 for representing the intra-coincidence accuracy;
i1 is a reference numeral for representing a common fitting point;
u i1 the difference between the true and proposed values for the common fitting point denoted i 1;
n 1 for representing the number of common fitting points.
4. The method for constructing an elevation anomaly model based on terrain model matching according to claim 1, wherein the model accuracy comprises an external coincidence accuracy calculated by the following formula:
wherein,
σ 2 for representing the accuracy of the outer compliance;
i2 is a reference numeral for indicating a check point;
u i2 a difference between the true value and the expected value for the checkpoint labeled i 2;
n 2 for indicating the number of checkpoints.
5. The terrain model matching-based elevation anomaly model construction method of claim 1, wherein the plurality of neural networks comprises a cyclic neural network RNN, a back propagation BP neural network, a radial basis function RBF neural network.
6. An elevation anomaly model construction system based on terrain model matching, the system comprising:
the terrain model module is used for selecting a plurality of groups of public fitting points and a plurality of check points from the terrain model, and acquiring position data of the points, wherein the position data comprises geodetic coordinates, geodetic heights and normal heights; the public fitting points are normally distributed in the terrain model; the preliminary model construction module is used for constructing a plurality of preliminary elevation anomaly models based on the position data of the plurality of groups of public fitting points by using a quadric surface model method and a plurality of neural networks respectively; the preliminary model construction module is further used for calculating the distribution variance of the preliminary elevation abnormal model by using the geodetic coordinates of each group of public fitting points before constructing the preliminary elevation abnormal model, adjusting the distribution variance according to the accuracy result after calculating the model accuracy of the preliminary elevation abnormal model, and updating the model according to the public fitting points corresponding to the adjusted distribution variance until the model accuracy reaches a preset range;
and the final model construction module is used for respectively calculating the model precision of each preliminary elevation abnormal model by utilizing the position data of the plurality of check points so as to select the preliminary elevation abnormal model with the highest model precision as a final elevation abnormal model to be output.
7. The terrain model matching-based elevation anomaly model construction system of claim 6, wherein the plurality of checkpoints are randomly distributed in the terrain model and the plurality of checkpoints for model accuracy calculations for all preliminary elevation anomaly models are the same.
8. The terrain model matching-based elevation anomaly model construction system of claim 6, wherein the preliminary model construction module and the final model construction module are each configured to perform model accuracy calculations on all preliminary elevation anomaly models, and the model accuracy includes an inner coincidence accuracy and an outer coincidence accuracy.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410037171.5A CN117557741A (en) | 2024-01-10 | 2024-01-10 | Elevation anomaly model construction method and system based on terrain model matching |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410037171.5A CN117557741A (en) | 2024-01-10 | 2024-01-10 | Elevation anomaly model construction method and system based on terrain model matching |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117557741A true CN117557741A (en) | 2024-02-13 |
Family
ID=89820792
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410037171.5A Pending CN117557741A (en) | 2024-01-10 | 2024-01-10 | Elevation anomaly model construction method and system based on terrain model matching |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117557741A (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109270560A (en) * | 2018-10-12 | 2019-01-25 | 东南大学 | The multidimensional Gross postionning and valued methods of region height anomaly data |
| CN113532397A (en) * | 2021-07-07 | 2021-10-22 | 天津大学 | Regional elevation anomaly fitting method based on extension algorithm |
-
2024
- 2024-01-10 CN CN202410037171.5A patent/CN117557741A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109270560A (en) * | 2018-10-12 | 2019-01-25 | 东南大学 | The multidimensional Gross postionning and valued methods of region height anomaly data |
| CN113532397A (en) * | 2021-07-07 | 2021-10-22 | 天津大学 | Regional elevation anomaly fitting method based on extension algorithm |
Non-Patent Citations (1)
| Title |
|---|
| 潘国俊: "基于机器学习的高程异常建模算法及其水电工程的应用", 《水力发电》, 30 June 2023 (2023-06-30), pages 28 - 33 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105136054B (en) | The fine deformation monitoring method of structures and system based on Three Dimensional Ground laser scanning | |
| CN114966685B (en) | Dam deformation monitoring and predicting method based on InSAR and deep learning | |
| CN109100719B (en) | Terrain map joint mapping method based on satellite-borne SAR (synthetic aperture radar) image and optical image | |
| CN113532397B (en) | Regional elevation anomaly fitting method based on expansion algorithm | |
| CN113239567A (en) | Gravity field model assisted inverse distance weighting quasi-geoid grid interpolation method | |
| CN104613966B (en) | A kind of cadastration off-line data processing method | |
| CN107422342A (en) | GNSS satellite clock correction estimates method of quality control in real time | |
| CN113465524A (en) | Point cloud data-based earth-rock dam filling real-time dam facing deformation monitoring method | |
| CN108427741B (en) | DEM relative error evaluation method based on large number of high-precision control points | |
| CN115688607A (en) | Band-shaped cross-sea quasi-geoid refinement method based on multi-channel spectrum combination | |
| CN113295149A (en) | CP III coordinate calculation method and device based on joint observation quantity | |
| CN117557741A (en) | Elevation anomaly model construction method and system based on terrain model matching | |
| CN112241844A (en) | Method and device for determining and updating environmental risk source background list of drinking water source area | |
| CN113627074B (en) | Ground wave propagation delay prediction method based on transfer learning | |
| Xue et al. | Unbiased nonlinear least squares estimations of short-distance equations | |
| CN103149551B (en) | Convex optimization range-based localization method based on linear combination | |
| CN111624655B (en) | Method and device for determining residual static correction value of first-motion wave | |
| CN109614585B (en) | Novel ionosphere region reconstruction method | |
| CN111721272A (en) | Engineering surface measurement method based on ellipsoid calculation | |
| Avramiuc et al. | Determination of a new gravimetric quasigeoid for Romania | |
| TWI399527B (en) | Interpretation method and system of outlier position of rock slope | |
| CN111145347B (en) | Method and device for correcting digital elevation data of terrain section and correction equipment | |
| CN116258023B (en) | Wind speed prediction method of wind power plant and terminal equipment | |
| Yakubu et al. | Modelling Uncertainties in Differential Global Positioning System Dataset | |
| CN118051845B (en) | Geospatial full coverage data generation method and device based on space variable parameter machine learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |