CN114966601B - Mountain landslide prediction method based on millimeter wave radar and electronic equipment - Google Patents
Mountain landslide prediction method based on millimeter wave radar and electronic equipment Download PDFInfo
- Publication number
- CN114966601B CN114966601B CN202210916809.3A CN202210916809A CN114966601B CN 114966601 B CN114966601 B CN 114966601B CN 202210916809 A CN202210916809 A CN 202210916809A CN 114966601 B CN114966601 B CN 114966601B
- Authority
- CN
- China
- Prior art keywords
- landslide
- millimeter wave
- slope
- wave radar
- displacement
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
- G01S7/411—Identification of targets based on measurements of radar reflectivity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Business, Economics & Management (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- General Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Data Mining & Analysis (AREA)
- Game Theory and Decision Science (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The application discloses mountain landslide prediction method and electronic equipment based on millimeter wave radar, belongs to the technical field of radar detection, and the method comprises the following steps: taking a corner reflector arranged on a mountain body to be detected as a target position, and acquiring phase fluctuation parameters of the corner reflector through a millimeter wave radar; modulating and training the phase fluctuation parameters through a machine learning algorithm to obtain a strong prediction function, and obtaining an optimal classification result according to the strong prediction function; and calculating the real displacement of the landslide area according to the optimal classification result, comparing the real displacement with a set landslide displacement threshold value, and early warning the landslide condition according to the comparison result. According to the method and the device, the landslide can be predicted, the prediction precision is improved through a high-precision data processing mode, the probability of false alarm is reduced, errors are effectively reduced, and the method and the device have good adaptability to complex environments and climatic conditions.
Description
Technical Field
The application belongs to the technical field of radar detection, and particularly relates to a landslide prediction method based on a millimeter wave radar, electronic equipment implementing the method, and a computer-readable storage medium containing a computer program implementing the method.
Background
Before landslide, the change degree is very weak, usually before 3 days before landslide, the change amount is 1-10mm per day, and the current prediction scheme comprises: (1) The Beidou positioning is utilized, a Beidou receiver is put in a designated mountain, and whether the mountain slides or not is predicted according to the change of the position of the receiver; but the precision of the scheme is poor; (2) Imaging the mountain by using an imaging radar, and predicting whether the mountain slides or not through imaging comparison; however, this solution requires a large amount of equipment for support, and is expensive and harsh on installation conditions.
Disclosure of Invention
The purpose of the invention is as follows: the embodiment of the application provides a landslide prediction method based on a millimeter wave radar, and aims to solve the technical problems of poor prediction precision and high prediction cost in the prior art; another object of an embodiment of the present application is to provide an electronic device; it is a third object of embodiments of the present application to provide a storage medium.
The technical scheme is as follows: the embodiment of the application provides a landslide prediction method based on a millimeter wave radar, which comprises the following steps:
taking a corner reflector arranged on a mountain body to be detected as a target position, and acquiring phase fluctuation parameters of the corner reflector through a millimeter wave radar;
modulating and training the phase fluctuation parameters through a machine learning algorithm to obtain a strong prediction function, and obtaining an optimal classification result according to the strong prediction function;
and calculating the real displacement of the landslide area according to the optimal classification result, comparing the real displacement with a set landslide displacement threshold value, and early warning the landslide condition according to a comparison result.
In some embodiments, the phase fluctuation parameters of the corner reflector are obtained by acquiring a time sequence differential interference phase diagram; the purpose of performing modulation training on the phase fluctuation parameters of the corner reflectors is to reduce phase delay caused by weather conditions and the like.
In some embodiments, the performing modulation training on the phase fluctuation parameter through a machine learning algorithm to obtain a strong prediction function, and obtaining an optimal classification result according to the strong prediction function, further includes:
collecting a plurality of time sequence differential interference phase diagrams, extracting differential interference phases and using the differential interference phases as a data set to obtain a first training sample;
calculating a strong prediction function according to the first training sample to obtain a trained strong predictor;
and classifying all the time sequence differential interference phase images through the strong predictor to obtain an optimal classification result, and obtaining high-precision deformation phases of all time points according to the optimal classification result to finish error correction.
In some embodiments, the strong predictor comprising a plurality of weak predictor combinations of weights is introduced, and different weights are configured for the singular points to be classified again, so that errors can be effectively reduced.
In some embodiments, the acquiring a plurality of time-series differential interference phase patterns, extracting differential interference phases and using the differential interference phases as a data set to obtain a first training sample, further includes:
selecting M time sequence differential interference phase images from the time sequence differential interference phase images, respectively selecting N image elements with the same position from the M time sequence differential interference phase images, extracting the differential interference phase of each image element on a time sequence, taking each differential interference phase as a data set, and further taking the obtained N data sets as first training samples; wherein M and N are natural numbers.
In some embodiments, the time-series differential interference phase map is monitored periodically at predetermined times.
In some embodiments, computing a strong prediction function from the first training sample further comprises:
initializing m sets of training sample weights w in a training dataset i (ii) a Wherein, the first and the second end of the pipe are connected with each other,
and i is an integer between 1 and m;
configuring p weak predictors, and training and predicting the training samples by taking the differential interference phase of the training samples as the characteristic attribute of the samples;
after p times of training, p weak prediction functions are combined
Weighted synthesis of strong prediction functions
Obtaining a strong predictor; wherein the strong prediction function is:
wherein p is the number of weak predictors, a t Is the weight coefficient of the t-th weak predictor, and t is an integer between 1 and p.
In some embodiments, the a t Further comprising:
adjusting training sample weights w i And carrying out normalization, wherein the normalization specifically comprises the following steps:
in the formula, e c To predict the absolute value of the error of the training sample, e ɛ Is a predetermined error limit;
according to the weight w of the training sample i Calculating the prediction error rate e of the t weak predictor t (ii) a Wherein the content of the first and second substances,
and i is an integer between 1 and m, t is an integer between 1 and p, and p is the number of weak predictors;
according to the prediction error rate e of the t weak predictor t Calculating the weighting coefficient a of the t weak predictor t (ii) a Wherein the content of the first and second substances,
and t is an integer between 1 and p, p being the number of weak predictors.
In some embodiments, the strong predictor is used to classify all the time sequence differential interference phase maps to obtain an optimal classification result, and according to the optimal classification result, a high-precision deformation phase at each time point is obtained to complete error correction, further including:
constructing a time sequence classification set by using all the time sequence differential interference phase images as a unit by using image elements;
and inputting the time sequence classification set into the strong prediction function, outputting the coordinates after the correction of the image element points in the time sequence differential interference phase diagram through the strong prediction function, and taking the coordinates as an optimal classification result to obtain the corrected phase.
In some embodiments, calculating the real displacement of the landslide area according to the optimal classification result further comprises:
establishing a digital elevation model for the slope of the mountain to be detected, and extracting the slope direction and the slope of the target position according to digital elevation information;
measuring the sight line direction displacement of the millimeter-wave radar reaching the target position;
and projecting the sight line direction displacement to the maximum slope direction determined by the slope direction and the slope, thereby obtaining the real displacement of the landslide area.
In some embodiments, the slope direction and the slope of the target position are obtained by analyzing a three-dimensional terrain slope map and a slope direction map of the mountain to be detected, wherein the three-dimensional terrain slope map and the slope direction map are obtained by any one of GDAL, arcGIS or QGIS.
In some embodiments, projecting the line-of-sight direction displacement onto a maximum slope direction determined by the slope direction and the slope to obtain a corner reflector displacement parameter, further comprises:
acquiring a unit vector three-dimensional coordinate of the target position monitored by the millimeter wave radar;
according to the unit vector three-dimensional coordinate, determining the maximum slope direction of the slope direction and the slope as the direction pointing to the unit vector three-dimensional coordinate; wherein the unit vector three-dimensional coordinates are represented as follows:
wherein the content of the first and second substances,x t ，y t andh t coordinate points respectively representing three-dimensional coordinates of the unit vector,
the direction of the slope is shown as,
represents a grade;
and converting the sight line direction displacement through coordinates to obtain actual displacement along the maximum gradient direction, and using the actual displacement as the displacement parameter of the corner reflector.
In some embodiments, the corner reflector is arranged on the ground with the height of 1/3 to 1/2 of the mountain to be detected.
In some embodiments, the corner reflectors comprise at least two groups, and the millimeter wave radar is a 77GHz millimeter wave radar.
In some embodiments, the present application further provides an electronic device comprising a processor and a memory, the memory storing a computer program that, when executed on the processor, implements the millimeter wave radar-based landslide prediction method.
In some embodiments, the present application further provides a storage medium storing a computer program that, when executed on a processor, implements the millimeter wave radar-based landslide prediction method.
Has the advantages that: compared with the prior art, the mountain landslide prediction method based on the millimeter wave radar comprises the following steps: taking a corner reflector arranged on a mountain body to be detected as a target position, and acquiring phase fluctuation parameters of the corner reflector through a millimeter wave radar; modulating and training the phase fluctuation parameters through a machine learning algorithm to obtain a strong prediction function, and obtaining an optimal classification result according to the strong prediction function; and calculating the real displacement of the landslide area according to the optimal classification result, comparing the real displacement with a set landslide displacement threshold value, and early warning the landslide condition according to the comparison result. According to the method, when landslide is predicted, the prediction precision is improved through a high-precision data processing mode, the probability of false alarm is reduced, errors are effectively reduced, and the method has good adaptability to complex environments and climate conditions.
It can be understood that, compared with the prior art, the electronic device and the storage medium provided in the embodiments of the present application may have all technical features and beneficial effects of the method for predicting landslide based on millimeter wave radar described above, and are not described herein again.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the description below are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a flowchart of a landslide prediction method based on a millimeter wave radar according to an embodiment of the present application;
FIG. 2 is a flow chart of a machine learning algorithm provided by an embodiment of the present application;
fig. 3 is a flowchart for determining a corner reflector displacement parameter according to an embodiment of the present application;
FIG. 4 is a schematic diagram of an apparatus for implementing a prediction method according to an embodiment of the present application;
fig. 5 is a comparison between the corrected displacement time series provided in the embodiment of the present application and the monitoring result of the total station;
fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. It is to be understood that the embodiments described are only a few embodiments of the present application and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
In the description of the present application, it is noted that the terms "first", "second", and the like are used merely for distinguishing between descriptions and are not intended to indicate or imply relative importance.
Referring to fig. 1, a landslide prediction method based on a millimeter wave radar is provided, which includes the following steps:
s101: taking a corner reflector arranged on a mountain body to be detected as a target position, and acquiring phase fluctuation parameters of the corner reflector through a millimeter wave radar;
s102, carrying out modulation training on the phase fluctuation parameters through a machine learning algorithm to obtain a strong prediction function, and obtaining an optimal classification result according to the strong prediction function;
s103, calculating the real displacement of the landslide region according to the optimal classification result, comparing the real displacement with a set landslide displacement threshold value, and early warning the landslide condition according to the comparison result.
In some embodiments, the corner reflector is a radar reflector, and when radar electromagnetic waves scan the corner reflection, the electromagnetic waves are refracted and amplified on the metal corner, so that a strong echo signal is generated, and a strong echo target appears on a screen of the radar.
In some embodiments, referring to fig. 4, a corner reflector is installed on a mountain a in an area where thunderstorms are likely to occur, the corner reflector displacement parameter modulation training is performed according to the phase fluctuation condition of the corner reflector, a special corner reflector is installed at a proper position of the mountain, the phase fluctuation of the corner reflector is detected through a millimeter wave radar, and mountain displacement data is detected.
In some embodiments, the corner reflector needs to be installed at a proper position of a mountain, the installation effect is easier to measure at the ground position of 1/3 to 1/2 of the mountain waist in an area where thunderstorms easily occur according to a rainfall area diagram in the past year, and if the corner reflector is installed at the top of a mountain or the top of a tree, the signal is disturbed by the thunderstorms, so that the signal is unstable; the corner reflectors comprise at least two groups, preferably a first corner reflector and a second corner reflector; performing phase fluctuation parameter modulation training on the corner reflector by simultaneously detecting the phase of the first corner reflector and the phase of the second corner reflector to obtain a processed displacement parameter of the corner reflector; if the mountain is large, a third corner reflector can be added, and the corner reflectors are arranged at equal intervals.
In some embodiments, the millimeter wave radar adopts a 77GHz millimeter wave radar, the phase change is sensitive, an object (a corner reflector) moves by 2mm, and the phase changes by 1 pi/rad approximately; the object (corner reflector) is moved 3mm, the phase is changed by approximately 1.5 pi/rad; the object (corner reflector) is moved 5mm and the phase is changed by approximately 2.5 pi/rad. Sensitivity is frequency band dependent, so the relationship between phase change and angular anti-motion distance is only applicable to the frequency band, and the use of 77GHz millimeter wave radar has the advantage of being very sensitive to phase change.
In some embodiments, in step S102, the phase of a specific angle reversal is monitored over a period of time, a time-series differential interferometric phase diagram is obtained, and disaster warning can be performed by analyzing data. In the data processing process, data filtering and data mining are involved to improve accuracy. In order to reduce abnormal data of radar echo signals, such as phase errors caused by meteorological conditions, occlusion and other factors, the data processing precision is improved by introducing a machine learning algorithm.
In some embodiments, referring to fig. 2, a flow chart of a machine learning algorithm includes:
s201, collecting a plurality of time sequence differential interference phase diagrams, extracting differential interference phases and using the differential interference phases as a data set to obtain a first training sample;
s202, calculating a strong prediction function according to the first training sample to obtain a trained strong predictor;
s203, classifying all time sequence differential interference phase images through a strong predictor to obtain an optimal classification result, and obtaining high-precision deformation phases of all time points according to the optimal classification result to finish error correction.
In some embodiments, in step S201, M time-series differential interference phase maps are selected from all the time-series differential interference phase maps periodically monitored for a predetermined time, and N image elements with the same position are respectively selected from the M time-series differential interference phase maps. Wherein M and N are natural numbers. And extracting a differential interference phase on each image element time sequence, taking the differential interference phase of each image element as a data set, and further taking the obtained N data sets as a first training sample, wherein each data set comprises M time sequence differential interference phases. Thus, the first training sample can be represented as [ (x 1, y 1), (x 2, y 2) \ 8230; (xm, ym) ], and in practice, the first training sample is a time-series differential interference phase map, and (xm, ym) represents the coordinates of the element points of the time-series differential interference phase image
In some embodiments, the step S202 further comprises:
(1) Initializing m sets of training sample weights w in a training dataset i (ii) a Wherein the content of the first and second substances,
and i is an integer between 1 and m;
(2) Configuring p weak predictors, and training and predicting the training samples by taking the differential interference phase of the training samples as the characteristic attribute of the samples;
(3) Calculating the prediction error e of the training sample c Adjusting the training sample weight w i And carrying out normalization; wherein, the normalization specifically comprises:
，
(ii) a In the formula, e c To predict the absolute value of the error of the training sample, e ɛ Is a predetermined error limit; the higher the updated weight, the larger the error, the reclassification by the next weak predictor.
(4) According to the weight w of the training sample i Calculating the prediction error rate e of the t weak predictor t (ii) a Wherein, the first and the second end of the pipe are connected with each other,
and i is an integer between 1 and m, t is an integer between 1 and p, and p is the number of weak predictors;
(5) Prediction error rate e according to t-th weak predictor t Calculating the weight coefficient a of the t weak predictor t (ii) a Wherein the content of the first and second substances,
and t is an integer between 1 and p, and p is the number of weak predictors;
(6) Stopping training after p times of iteration, and otherwise, returning to the step (2);
(7) After p times of training, according to the weight coefficient a t P weak prediction functions
Weighted synthesis of strong prediction functions
Obtaining a strong predictor; wherein the strong prediction function is:
；
wherein p is the number of weak predictors, a t Is the weight coefficient of the t-th weak predictor, and t is an integer between 1 and p.
In some embodiments, the step S203 further comprises:
constructing a time sequence classification set by using all time sequence differential interference phase images as a unit by using image elements;
and inputting the time sequence classification set into a strong prediction function, outputting the corrected coordinates of the image element points in the time sequence differential interference phase diagram through the strong prediction function, and taking the coordinates as an optimal classification result to obtain the corrected phase.
In some embodiments, the data set classified as the phase to be corrected is obtained by performing classification extraction on a large number of sequentially observed time-series differential interference phase image data sets by using a modified machine learning algorithm. And (3) respectively subtracting the delay phases of the corresponding time points from all the time sequence differential interference phase diagrams to obtain the final high-precision time sequence deformation phase of each time point. Different weights are configured for the deviation points in the iteration process, so that the classification precision is higher. That is, since weight arrangement is performed not for randomization but for use of variables and use of data, accuracy can be improved and an excessively high generalization error can be prevented.
In some embodiments, the strong predictor comprising a plurality of weak predictor combinations of weights is introduced, and different weights are configured for the singular points to be classified again, so that errors can be effectively reduced.
In some embodiments, in step S102, the phase interference difference obtained by the machine learning algorithm can only obtain the displacement in the line of sight (LOS) direction. The displacement direction of the actual landslide is often inconsistent with the LOS direction of the radar, so that the direction measured by the radar at the moment is actually the projection of the displacement vector of the actual landslide in the LOS direction. If the real displacement is to be obtained, referring to fig. 3, the step 103 of calculating the real displacement of the landslide area according to the optimal classification result further includes:
s301, establishing a digital elevation model for a slope of a mountain to be detected, and extracting the slope direction and the slope of a target position according to digital elevation information;
s302, measuring the sight line direction displacement from the millimeter wave radar to the target position;
and S303, projecting the sight line direction displacement to the maximum gradient direction determined by the slope direction and the gradient, thereby obtaining the corner reflector displacement parameter.
In some embodiments, in step S301, the slope and the gradient of the target position are obtained by analyzing a three-dimensional terrain slope map and a slope map of the mountain to be measured, where the three-dimensional terrain slope map and the slope map are obtained by any one of GDAL, arcGIS, or QGIS, and as a preferred scheme, the slope map and the slope map of the three-dimensional terrain are generated by using gdaldem slope/aspect command of GDAL; meanwhile, the obtained slope map and the slope map need to be subjected to smooth filtering, such as an average value and a gaussian filter with a proper window size, so as to suppress noise caused by the small-scale slope and the slope.
In some embodiments, the step S303 further comprises:
acquiring a unit vector three-dimensional coordinate of a millimeter wave radar monitoring target position;
determining the maximum slope direction of the slope direction and the slope as the direction pointing to the unit vector three-dimensional coordinate according to the unit vector three-dimensional coordinate; wherein, the unit vector three-dimensional coordinates are expressed as follows:
wherein the content of the first and second substances,x t ，y t andh t coordinate points respectively representing three-dimensional coordinates of the unit vector,
the direction of the slope is shown as,
represents a grade;
for line of sight (LOS) direction displacement obtained by a radar at any point in a scene, the actual displacement of the displacement along the maximum slope direction can be obtained through coordinate conversion and used as a displacement parameter of the corner reflector.
In some embodiments, the step S303 further comprises: let Δ R be the displacement in the line of sight (LOS) direction measured by the radar, the actual displacement Δ R in the maximum slope direction can be expressed as: Δ R =Δr × cos Φ, where Φ is the angle between the line of sight (LOS) direction and the maximum slope direction.
In some embodiments, the step S103 further comprises: and (4) setting the mountain displacement threshold value as T, and when the obtained real displacement of the landslide area exceeds the mountain displacement threshold value T, sending an alarm, wherein the alarm information is sent to a specific server through a network.
In some embodiments, after preprocessing the SAR radar echo data of a landslide region, a time sequence deformation field of a specific region is obtained through conventional time sequence differential interference processing, and deformation displacement before correction is obtained. Meanwhile, the original differential interference phase diagram is subjected to learning training by the time sequence classification method, the phase after the region classification is finally obtained, the position of the millimeter wave radar is used as the difference between the original differential interference phase and the delay phase obtained through the optimal classification, and the corrected deformation displacement time sequence is obtained. The delay phase includes phase errors due to meteorological causes, such as rain, fog, snow, steam, etc., among others. And monitoring the displacement of the position where the millimeter wave radar is placed by using a total station as a comparison example in the experimental process. Referring to fig. 5, a broken line represents a time displacement sequence formed by the observation results at the position where the millimeter wave radar is placed, and a trace point represents the total station observation results at the position where the millimeter wave radar is placed. The horizontal axis is a time axis, and the vertical axis is the percentage of the actual displacement in the maximum displacement obtained after coordinate conversion under the condition of corner reflector phase fluctuation. As can be seen from FIG. 5, the correction method of the application can enable the system monitoring precision error to be less than 1mm, and the method of the application effectively reduces the error and has good adaptability to complex environments and climate conditions.
In some embodiments, referring to fig. 6, the present application further provides a computer device, which may be a server or a terminal, comprising a processor connected to a system bus, a memory and a communication interface, wherein the processor is configured to provide a control computing capability of the computer device; the memory has stored thereon a computer program that, when executed by the processor, implements a millimeter wave radar-based landslide prediction method. The memory includes a computer storage medium which is a nonvolatile storage medium storing an operating system and a computer program, and an internal memory which provides an environment for the operating system and the computer program to run. The communication interface of the computer equipment is used for carrying out wired or wireless communication with an external terminal, and the wireless communication can be achieved through WIFI, a mobile cellular network and the like.
In some embodiments, all or part of the processes of the above-described embodiments of the method may be implemented by a computer program, and the computer program may be stored in a non-volatile computer-readable storage medium, and when executed, may include the processes of the above-described embodiments of the method for predicting landslide based on millimeter wave radar. Any reference to memory, databases, or other media used in the embodiments provided herein can include at least one of non-volatile and volatile memory. The nonvolatile Memory may include a Read Only Memory (ROM), a magnetic tape, a floppy disk, a flash Memory, an optical Memory, a high density embedded nonvolatile Memory, a resistance change Memory, and the like. Volatile Memory can include Random Access Memory (RAM), external cache Memory, and the like. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM), for example.
It should be noted that the algorithms and displays presented herein are not inherently related to any particular computer, virtual machine, or other apparatus. Various general purpose devices may be used with the teachings herein. The required structure for constructing such a device will be apparent from the description above. In addition, this application is not directed to any particular programming language. It is appreciated that a variety of programming languages may be used to implement the teachings of the present invention as described herein, and any descriptions of specific languages are provided above to disclose the best mode of the invention.
In the foregoing embodiments, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to the related descriptions of other embodiments.
The method, the electronic device and the medium for predicting landslide based on the millimeter wave radar provided by the embodiment of the application are introduced in detail, and a specific example is applied to explain the principle and the implementation mode of the application, and the description of the embodiment is only used for helping to understand the technical scheme and the core idea of the application; those of ordinary skill in the art will understand that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications or substitutions do not depart from the spirit and scope of the present disclosure as defined by the appended claims.
Claims (11)
1. A landslide prediction method based on millimeter wave radar is characterized by comprising the following steps:
taking a corner reflector arranged on a mountain body to be detected as a target position, and acquiring phase fluctuation parameters of the corner reflector through a millimeter wave radar;
modulating and training the phase fluctuation parameters through a machine learning algorithm to obtain a strong prediction function, and obtaining an optimal classification result according to the strong prediction function;
calculating the real displacement of the landslide area according to the optimal classification result, comparing the real displacement with a set landslide displacement threshold value, and early warning the landslide condition according to the comparison result;
the modulating training of the phase fluctuation parameters through a machine learning algorithm to obtain a strong prediction function, and obtaining an optimal classification result according to the strong prediction function, further comprising:
collecting a plurality of time sequence differential interference phase diagrams, extracting differential interference phases and using the differential interference phases as a data set to obtain a first training sample;
calculating a strong prediction function according to the first training sample to obtain a trained strong predictor;
classifying all the time sequence differential interference phase images through the strong predictor to obtain an optimal classification result, and obtaining high-precision deformation phases of all time points according to the optimal classification result to finish error correction;
computing a strong prediction function from the first training sample, further comprising:
initializing m sets of training sample weights w in a training dataset i (ii) a Wherein the content of the first and second substances,
and i is an integer between 1 and m;
configuring p weak predictors, and training and predicting the training samples by taking the differential interference phase of the training samples as the characteristic attribute of the samples;
after p times of training, p weak prediction functions are combined
Weighted synthesis of strong prediction functions
Obtaining a strong predictor; wherein the strong prediction function is:
wherein p is the number of weak predictors, a t Is the weight coefficient of the t weak predictor, and t is an integer between 1 and p;
a is a t Further comprising:
adjusting training sample weights w i And carrying out normalization, wherein the normalization specifically comprises the following steps:
in the formula, e c To predict the absolute value of the error of the training sample, e ɛ Is a predetermined error limit;
according to the weight w of the training sample i Calculating the prediction error rate e of the t weak predictor t (ii) a Wherein the content of the first and second substances,
and i is an integer between 1 and m, t is an integer between 1 and p, and p is the number of weak predictors;
according to the prediction error rate e of the t weak predictor t Calculating the weight coefficient a of the t weak predictor t (ii) a Wherein the content of the first and second substances,
and t is an integer between 1 and p, p being the number of weak predictors.
2. The method for mountain landslide prediction based on millimeter wave radar as claimed in claim 1 wherein the collecting a plurality of time sequence differential interference phase maps, extracting differential interference phase and using as data set to obtain the first training sample further comprises:
selecting M time sequence differential interference phase images from the time sequence differential interference phase images, respectively selecting N image elements with the same position from the M time sequence differential interference phase images, extracting the differential interference phase of each image element on a time sequence, taking each differential interference phase as a data set, and further taking the obtained N data sets as first training samples; wherein M and N are natural numbers.
3. The method according to claim 2, wherein the time-series differential interferometric phase map is monitored periodically for a predetermined time.
4. The millimeter wave radar-based landslide prediction method according to claim 1, wherein all the time sequence differential interference phase maps are classified by the strong predictor to obtain an optimal classification result, and according to the optimal classification result, a high-precision deformation phase at each time point is obtained to complete error correction, further comprising:
constructing a time sequence classification set by using all the time sequence differential interference phase images as a unit by using image elements;
and inputting the time sequence classification set into the strong prediction function, outputting the coordinates after the correction of the image element points in the time sequence differential interference phase diagram through the strong prediction function, and taking the coordinates as an optimal classification result to obtain the corrected phase.
5. The millimeter wave radar-based landslide prediction method according to claim 1, wherein calculating a true displacement of a landslide region according to the optimal classification result further comprises:
establishing a digital elevation model for the slope of the mountain to be detected, and extracting the slope direction and the slope of the target position according to digital elevation information;
measuring the sight line direction displacement of the millimeter wave radar reaching the target position;
and projecting the sight line direction displacement to the maximum slope direction determined by the slope direction and the slope, thereby obtaining the real displacement of the landslide area.
6. The landslide prediction method based on millimeter wave radar as claimed in claim 5, wherein the slope direction and the gradient of the target position are obtained by analyzing a three-dimensional terrain slope map and a slope direction map of the mountain to be detected, wherein the three-dimensional terrain slope map and the slope direction map are obtained by any one of GDAL, arcGIS or QGIS.
7. The millimeter wave radar-based landslide prediction method of claim 5, wherein projecting the line-of-sight direction displacement onto a maximum slope direction determined by the slope direction and the slope to obtain corner reflector displacement parameters further comprises:
acquiring a unit vector three-dimensional coordinate of the target position monitored by the millimeter wave radar;
determining the maximum slope direction of the slope direction and the slope as the direction pointing to the unit vector three-dimensional coordinate according to the unit vector three-dimensional coordinate; wherein the unit vector three-dimensional coordinates are represented as follows:
wherein the content of the first and second substances,x t ，y t andh t coordinate points respectively representing three-dimensional coordinates of the unit vector,
which is indicative of the direction of the slope,
represents a grade;
and converting the sight line direction displacement through coordinates to obtain actual displacement along the direction of the maximum slope, and taking the actual displacement as the real displacement of the mountain landslide area.
8. The landslide prediction method based on the millimeter wave radar as claimed in claim 1, wherein the corner reflector is arranged on the ground with the height of 1/3 to 1/2 of the mountain to be measured.
9. The landslide prediction method of claim 1 wherein said corner reflectors comprise at least two groups, and wherein said millimeter wave radar is 77GHz millimeter wave radar.
10. An electronic device comprising a processor and a memory, the memory having stored thereon a computer program, characterized in that the computer program, when executed on the processor, implements the prediction method of any one of claims 1-9.
11. A computer-readable storage medium, on which a computer program is stored which, when executed on a processor, carries out the prediction method according to any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210916809.3A CN114966601B (en) | 2022-08-01 | 2022-08-01 | Mountain landslide prediction method based on millimeter wave radar and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210916809.3A CN114966601B (en) | 2022-08-01 | 2022-08-01 | Mountain landslide prediction method based on millimeter wave radar and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114966601A CN114966601A (en) | 2022-08-30 |
| CN114966601B true CN114966601B (en) | 2022-10-21 |
Family
ID=82970068
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210916809.3A Active CN114966601B (en) | 2022-08-01 | 2022-08-01 | Mountain landslide prediction method based on millimeter wave radar and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114966601B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105182339A (en) * | 2015-09-25 | 2015-12-23 | 昆明理工大学 | Method for correcting environmental influences at slope deformation monitoring on the basis of corner reflector |
| CN107504914A (en) * | 2017-07-28 | 2017-12-22 | 安徽威德萨科技有限公司 | A kind of danger zone and the deformation monitoring method of alarm |
| CN111351424A (en) * | 2020-03-31 | 2020-06-30 | 内蒙古雷远信息科技有限公司 | Deformation measuring method and radar system |
| CN111522005A (en) * | 2020-06-05 | 2020-08-11 | 中国有色金属长沙勘察设计研究院有限公司 | Deformation monitoring and terrain reconstruction method |
| CN114063062A (en) * | 2021-11-18 | 2022-02-18 | 中国有色金属长沙勘察设计研究院有限公司 | Method and device for emergency monitoring of landslide disaster |
| CN114185048A (en) * | 2022-02-15 | 2022-03-15 | 湖南吉赫信息科技有限公司 | Method, system and storage medium for extracting landslide displacement vector by foundation InSAR |
| CN114187533A (en) * | 2022-02-15 | 2022-03-15 | 西南交通大学 | GB-InSAR (GB-InSAR) atmospheric correction method based on random forest time sequence classification |
| CN114578356A (en) * | 2022-03-02 | 2022-06-03 | 中南大学 | Distributed scatterer deformation monitoring method, system and equipment based on deep learning |
-
2022
- 2022-08-01 CN CN202210916809.3A patent/CN114966601B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105182339A (en) * | 2015-09-25 | 2015-12-23 | 昆明理工大学 | Method for correcting environmental influences at slope deformation monitoring on the basis of corner reflector |
| CN107504914A (en) * | 2017-07-28 | 2017-12-22 | 安徽威德萨科技有限公司 | A kind of danger zone and the deformation monitoring method of alarm |
| CN111351424A (en) * | 2020-03-31 | 2020-06-30 | 内蒙古雷远信息科技有限公司 | Deformation measuring method and radar system |
| CN111522005A (en) * | 2020-06-05 | 2020-08-11 | 中国有色金属长沙勘察设计研究院有限公司 | Deformation monitoring and terrain reconstruction method |
| CN114063062A (en) * | 2021-11-18 | 2022-02-18 | 中国有色金属长沙勘察设计研究院有限公司 | Method and device for emergency monitoring of landslide disaster |
| CN114185048A (en) * | 2022-02-15 | 2022-03-15 | 湖南吉赫信息科技有限公司 | Method, system and storage medium for extracting landslide displacement vector by foundation InSAR |
| CN114187533A (en) * | 2022-02-15 | 2022-03-15 | 西南交通大学 | GB-InSAR (GB-InSAR) atmospheric correction method based on random forest time sequence classification |
| CN114578356A (en) * | 2022-03-02 | 2022-06-03 | 中南大学 | Distributed scatterer deformation monitoring method, system and equipment based on deep learning |
Non-Patent Citations (1)
| Title |
|---|
| 基于AdaBoost-PSO-ELM算法的滑坡位移预测研究;张晓明 等;《计算机技术与应用》;20191231;正文第84-87页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114966601A (en) | 2022-08-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6689396B2 (en) | Weather forecasting device, weather forecasting method, and program | |
| CN109100723A (en) | Upper-level winds inversion method based on Doppler radar data | |
| CN114240940B (en) | Cloud and cloud shadow detection method and device based on remote sensing image | |
| CN113344291A (en) | Urban inland inundation range forecasting method, device, medium and equipment | |
| Yin et al. | Estimating rainfall intensity using an image-based deep learning model | |
| Nascimento et al. | Trends and rainfall tropicalization in Paraná State, south of Brazil | |
| CN114898206A (en) | Short-time heavy rainfall forecasting method, computer equipment and storage medium | |
| CN114966601B (en) | Mountain landslide prediction method based on millimeter wave radar and electronic equipment | |
| Feng et al. | A hierarchical network densification approach for reconstruction of historical ice velocity fields in East Antarctica | |
| CN113625232A (en) | Method, device, medium and equipment for suppressing multipath false target in radar detection | |
| CN115877345B (en) | Method and device for supplementing missing measurement data of wind profile radar | |
| CN117029840A (en) | Mobile vehicle positioning method and system | |
| CN116879900A (en) | Thunderstorm centroid positioning and tracking prediction method based on gray processing | |
| CN116340863A (en) | Air pollutant prediction method and device, electronic equipment and readable storage medium | |
| Wang et al. | A rapid identification and warning method for severe weather via Doppler radar based on an improved TITAN algorithm | |
| CN113724229B (en) | Method and device for determining elevation difference and electronic equipment | |
| CN116243273A (en) | Photon counting laser radar data filtering method and device | |
| CN114916059A (en) | WiFi fingerprint sparse map extension method based on interval random logarithm shadow model | |
| CN111767357B (en) | Regional mining complete evaluation method and equipment, electronic equipment and storage medium | |
| CN114089443A (en) | UHF frequency band ionosphere scintillation event forecasting method based on TEC integral quantity and seasonal change coefficient | |
| CN113988466A (en) | Correction method, device and equipment for forecast rainfall data | |
| JP6997664B2 (en) | Status judgment device | |
| CN116108338B (en) | Dynamic set identification method and device for particle phase state | |
| CN116068595B (en) | Sea surface wind speed inversion method and device, electronic equipment and medium | |
| CN115511887B (en) | Drought index prediction method and device based on artificial intelligence fusion |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230327 Address after: 610000 building a, Liangjiang international, Tianfu 1st Street, hi tech Zone, Chengdu, Sichuan Patentee after: Sichuan digital transportation technology Co.,Ltd. Address before: 211111 No. 9 Jizhou East Road, Jiangning Economic and Technological Development Zone, Nanjing City, Jiangsu Province Patentee before: HAWKEYE TECHNOLOGY CO.,LTD. Patentee before: Sichuan digital transportation technology Co.,Ltd. |
|
| TR01 | Transfer of patent right |