CN-121999576-A - Slope safety monitoring and early warning method based on foundation synthetic aperture radar

CN121999576ACN 121999576 ACN121999576 ACN 121999576ACN-121999576-A

Abstract

The invention provides a slope safety monitoring and early warning method based on a ground-based synthetic aperture radar, which relates to the technical field of slope safety detection and early warning, and the invention acquires a plurality of waveform reflection amplitude images at equal intervals by setting a reference time window, dynamically distributes importance weights to construct a current exclusive dynamic reference amplitude map, calculates a local amplitude correlation coefficient field by pixel local neighborhood window, normalizes falling amplitude and combines nonlinear amplification to form a falling index and forms a falling field, the method is free from dependence on phase interference, interference is effectively filtered, ultra-early capture of destabilization precursors such as micro-crack expansion, local looseness and the like is realized, a single dangerous core strength is formed by carrying out communication area identification on a sudden drop field in a space-time domain and weighting through Gaussian attenuation, and five-level early warning grades are directly output according to the strength and a preset threshold value.

Inventors

TAN ZEFU
WANG ZHI
LEI GUOPING
CHEN LICHUAN
ZOU JIANHUA
XU HONG
KANG YANFEI

Assignees

重庆三峡学院
重庆地质矿产研究院

Dates

Publication Date: 20260508
Application Date: 20251219

Claims (8)

1. A slope safety monitoring and early warning method based on a foundation synthetic aperture radar is characterized by comprising the following steps: the method comprises the steps of 1, setting a reference time window in a back-and-forth mode by taking the current moment as an end point, and acquiring waveform reflection amplitude images of a plurality of slopes to be monitored at equal intervals in the reference time window based on a foundation synthetic aperture radar to form a monitoring image set; Step 2, dynamically distributing importance weights according to the length of the corresponding acquisition time of each waveform reflection amplitude image from the current time, constructing a dynamic reference amplitude image according to each waveform reflection amplitude image and the importance weights thereof, and associating the dynamic reference amplitude image with each waveform reflection amplitude image; Step 3, respectively constructing the same local neighborhood window for each waveform reflection amplitude image in the monitoring image set and for each pixel point and the corresponding pixel point in the dynamic reference amplitude image, calculating local amplitude values of two pixel points in the local neighborhood window, and traversing local amplitude values of all pixel points of the waveform reflection amplitude image to obtain a local amplitude correlation coefficient field; Step 4, obtaining local amplitude correlation coefficients of each pixel point according to all local amplitude correlation coefficient fields in a reference time window to form a first time sequence, carrying out normalized amplitude-fall calculation on the first time sequence, combining nonlinear amplification processing on the amplitude-fall to obtain a sudden drop index representing the short-time abrupt drop degree of the local amplitude correlation coefficients, and traversing all the sudden drop indexes to form a sudden drop field; step 5, carrying out communication area identification on the sudden drop field in a space-time domain, extracting a communication area with an average sudden drop index exceeding a preset threshold value, taking the highest point of the sudden drop index in the communication area as a dangerous center, carrying out attenuation weighting summation on all the sudden drop indexes according to the distance from the dangerous center, and obtaining dangerous nuclear strength; And 6, comparing the dangerous nuclear strength with a preset threshold value, and directly determining and outputting the early warning grade of the side slope according to the value interval in which the dangerous nuclear strength falls.
2. The slope safety monitoring and early warning method based on the ground-based synthetic aperture radar according to claim 1, wherein the logic on which the importance weights are dynamically allocated is as follows: Taking the product of the time length of the sampling time of the waveform reflection amplitude image from the current time and the exponential decay rate as a first influence coefficient, wherein the range of the exponential decay rate is as follows ; And constructing an exponential function by taking the natural constant as a base and the opposite number of the first influence coefficient as an index, and taking the output value of the exponential function as the importance weight of the waveform reflection amplitude image.
3. The slope safety monitoring and early warning method based on the ground-based synthetic aperture radar, which is characterized by comprising the following specific steps of: generating a dynamic reference amplitude chart through exponential weighted average, and performing exponential weighted average on the historical amplitude chart stored in the queue to obtain the dynamic reference amplitude chart corresponding to the current moment, wherein the calculation formula is as follows: ; In the middle of Representing the coordinates in the dynamic reference amplitude diagram as Is used to determine the magnitude value of the image pixel of (c), And Respectively the abscissa and ordinate indexes of the image pixels, For the number of waveform reflected amplitude images, Denoted as the first The amplitude waveform reflects the importance weight of the amplitude image, Representing the coordinates in the historical amplitude plot as Amplitude values of image pixels of (a).
4. The slope safety monitoring and early warning method based on the ground-based synthetic aperture radar, which is characterized in that the step of calculating a local amplitude correlation coefficient between a current amplitude map and a dynamic reference amplitude map in each local neighborhood window is as follows: Determining a local calculation window and extracting amplitude data pairs in the window, and carrying out amplitude diagram at the current moment And the generated dynamic reference amplitude map At the center of the current pixel point Respectively extracting amplitude values of all pixel points in the square window to form amplitude data pairs to be compared; For each extracted amplitude data pair, the local similarity is calculated to obtain the local amplitude correlation coefficient of the current pixel, and the complete correlation coefficient is formed after traversing and calculating all pixels, wherein the calculation formula is as follows: ; In the middle of Is the first In the amplitude image of the amplitude waveform reflection, the coordinates are Is related to the local amplitude correlation coefficient of the dynamic reference amplitude map, And Respectively the abscissa and ordinate indexes of the image pixels, For the corresponding coordinates within the window, And Respectively expressed in the current waveform reflection amplitude image and the dynamic reference amplitude image, and the coordinates are Is used for the amplitude value of the pixel point of (c), And Respectively the abscissa and ordinate indexes of the amplitude map, Represent the first In the amplitude waveform reflection amplitude image, the amplitude is as follows In local neighborhood window of (2) in pixel point coordinates As the average amplitude value of the kernel, Representing the dynamic reference amplitude graph by pixel point coordinates As a core, at a size of Average amplitude values obtained in a local neighborhood window; And performing the calculation on all pixel points in the image, and obtaining a local amplitude correlation coefficient field representing the similarity degree of the local scattering characteristics of the whole slope surface at the current moment.
5. The slope safety monitoring and early warning method based on the ground-based synthetic aperture radar, which is characterized by comprising the following specific steps of: For each pixel point in a first time sequence consisting of local amplitude-dependent coefficient fields The local amplitude correlation coefficients of corresponding pixel points in the waveform reflection amplitude images closest and farthest from the current moment are taken, a normalized amplitude-falling term is calculated, and a calculation formula is as follows: ; In the middle of In the waveform reflection amplitude image, pixel points The radar amplitude correlation coefficient at the position really effectively normalizes the amplitude falling term, And Respectively the abscissa and ordinate indexes of the image pixels, And The coordinates are shown as the following in the waveform reflection amplitude images closest and farthest from the current moment Is a local amplitude correlation coefficient of (a); Non-linear amplification is carried out on the drop amplitude, and a sudden drop index is output, and a final sudden drop index calculation formula is as follows: ; In the middle of Representing pixels in a waveform reflected amplitude image as Is used for the control of the dip index of (c), In the waveform reflection amplitude image, pixel points The radar amplitude correlation coefficient at the position really effectively normalizes the amplitude falling term.
6. The slope safety monitoring and early warning method based on the ground-based synthetic aperture radar in claim 1 is characterized by comprising the following steps of: for a sudden drop field, a pixel point with the maximum sudden drop index in the sudden drop field is selected as a starting point, four neighborhoods of the starting point are checked, and if the sudden drop index of a certain pixel point in the neighborhoods is larger than a slope instability threshold value, the pixel point becomes a new candidate point; Adding candidate points meeting the conditions into the communication area, continuously checking the four adjacent areas of the new candidate points, and repeatedly calibrating the new candidate points to be the new candidate points if the points with the sudden drop indexes of the pixel points in the adjacent areas are larger than the slope instability threshold value, and continuously adding the new candidate points into the communication area; And repeatedly checking the four adjacent domains of the new candidate points until the sudden drop indexes of the pixel points of the four adjacent domains of all the new candidate points are not larger than the slope instability threshold value, and completing the identification of the connected region.
7. The slope safety monitoring and early warning method based on the ground-based synthetic aperture radar in claim 1 is characterized in that all the falling indexes in the communication area are subjected to attenuation weighted summation according to the distance from a dangerous center to obtain dangerous nuclear strength, and the method is as follows: extracting the mutation index by identifying the connected region Greater than or equal to the dip index threshold The dangerous area with space-time communication is used for finding out the point with the highest sudden drop index in the dangerous area and marking the point as a dangerous center; And carrying out distance weighted summation on the dip indexes of all pixels in the region in a Gaussian attenuation mode to obtain the dangerous nuclear intensity in a single numerical form, wherein the calculation formula is as follows: ; In the middle of To integrate the area of the dangerous area, the internal damage degree and the concentration degree all together to obtain the dangerous nuclear strength, In order to extract the space-time connected dangerous area after connected domain analysis, Representing pixels in a waveform reflected amplitude image as Is used for the control of the dip index of (c), For the planar distance to the center of the hazard, As a center of the risk, Represents the spatial decay constant, where And The abscissa and ordinate indexes of the image pixels, respectively.
8. The method for slope safety monitoring and early warning based on ground-based synthetic aperture radar according to claim 7, wherein the method is characterized by setting a dangerous nuclear strength threshold value Judging the relationship between the dangerous nuclear strength and the monitoring and early warning, wherein the relationship is as follows: When (when) When the slope is in a stable state, no early warning is sent out; When (when) And when the side slope is unstable, a red early warning is sent out.

Description

Slope safety monitoring and early warning method based on foundation synthetic aperture radar Technical Field The invention relates to the technical field of slope safety detection and early warning, in particular to a slope safety monitoring and early warning method based on a foundation synthetic aperture radar. Background In mountain areas and engineering construction areas frequently suffering from geological disasters, slope instability is one of main risks threatening the life and property safety of people and the stability of major engineering, and particularly in high-risk scenes such as high steep slopes of surface mines, water conservancy and hydropower banks and the like, slope collapse and landslide are often strong in burst and high in destructive power. In order to realize real-time monitoring and early warning of slope instability, a foundation synthetic aperture radar is generally adopted as main monitoring equipment, and high-resolution amplitude and phase information is obtained by continuously scanning the slope surface and is used for large-area non-contact deformation monitoring. The invention is designed aiming at the early recognition and early warning requirements of slope instability under specific scenes, and utilizes the amplitude time sequence information of the foundation synthetic aperture radar to construct a complete early warning chain without relying on the traditional phase interference deformation inversion. In the prior art, slope monitoring based on a foundation synthetic aperture radar mainly relies on a phase interferometry technology, inversion of slope deformation quantity and deformation rate is achieved through multi-time interference processing, and early warning is conducted by setting a threshold value accordingly. However, the method is very easy to be influenced by various factors in practical application, and is difficult to stably acquire reliable deformation information in a complex natural environment, so that early warning is delayed or invalid. Meanwhile, the existing amplitude information is limited to coarse coherence screening, early-stage instability precursor features contained in amplitude time sequences cannot be effectively mined, such as local scattering characteristic mutation and short-time correlation abrupt drop, so that the recognition capability of precursors of microcrack expansion, local looseness and the like is insufficient, and early-stage early warning cannot be realized. In the prior art, publication number CN104515988A discloses a slope safety monitoring and early warning method based on a ground-based synthetic aperture radar, which comprises the steps of taking phase interference as a core, performing multi-phase interference treatment by depending on a permanent scatterer or a distributed scatterer, inverting slope deformation quantity and deformation rate, and setting a threshold value according to the deformation quantity and deformation rate to perform early warning. However, the method is extremely easy to interfere in a practical complex environment, so that deformation measurement accuracy is greatly reduced, early warning advance is often short, and the utilization of amplitude information is limited to coarse coherence screening, so that early instability precursors cannot be mined. According to the invention, the sudden drop field is constructed, airspace communication identification and Gaussian weighting are performed to obtain dangerous nuclear strength, the threshold value is finally and directly compared, the early warning level is output, the phase dependence is eliminated, the support of a permanent scatterer is not needed, the safe evacuation window is remarkably widened, the scattered precursors are converted into an intuitive and reliable integral dangerous scalar, and the misjudgment risk is eliminated. The above information disclosed in the background section is only for enhancement of understanding of the background of the disclosure and therefore it may include information that does not form the prior art that is already known to a person of ordinary skill in the art. Disclosure of Invention The invention aims to provide a slope safety monitoring and early warning method based on a ground-based synthetic aperture radar, which aims to solve the problems in the background technology. In order to achieve the above purpose, the present invention provides the following technical solutions: A slope safety monitoring and early warning method based on a foundation synthetic aperture radar specifically comprises the following steps: the method comprises the steps of 1, setting a reference time window in a back-and-forth mode by taking the current moment as an end point, and acquiring waveform reflection amplitude images of a plurality of slopes to be monitored at equal intervals in the reference time window based on a foundation synthetic aperture radar to form a monitoring image set; Step 2, dynamically distribut