CN-122017034-A - Land engineering terrain rapid survey method based on unmanned aerial vehicle

CN122017034ACN 122017034 ACN122017034 ACN 122017034ACN-122017034-A

Abstract

S1, obtaining boundaries of a target area, dividing grids, planning cruising paths of the unmanned aerial vehicle, controlling the unmanned aerial vehicle to fly against grid points in sequence, obtaining depth feedback signals generated by excitation of first sound wave signals at each grid point, and obtaining earth surface vibration response signals generated by excitation of second sound wave signals; and S2, determining a continuous and stable underground reference interface in the target area based on the depth feedback signals acquired by all grid points, wherein the invention realizes full-automatic grid cruising, data acquisition and processing based on the unmanned aerial vehicle, realizes quick general investigation of a large area, and reduces comprehensive survey cost and period.

Inventors

Wu Qingman
Hong Kemi
WANG CHENXI

Assignees

杭州市房地产测绘有限公司

Dates

Publication Date: 20260512
Application Date: 20260112

Claims (9)

1. The land engineering terrain rapid survey method based on the unmanned aerial vehicle is characterized by comprising the following steps of: S1, acquiring a boundary of a target area, meshing and planning a cruising path of an unmanned aerial vehicle, controlling the unmanned aerial vehicle to fly to each grid point in sequence, acquiring a depth feedback signal generated by excitation of a first sound wave signal at each grid point, and acquiring a surface vibration response signal generated by excitation of a second sound wave signal S2, determining a continuous and stable underground reference interface in a target area based on the depth feedback signals acquired by all grid points; s3, aiming at each grid point, determining relative geological structure information and relative engineering material parameters of the grid point based on a depth feedback signal and a surface vibration response signal of the grid point and taking an underground reference interface as a reference; s4, generating a land engineering comprehensive terrain model based on the space coordinates of all grid points, the relative geological structure information and the relative engineering material parameters.
2. The land engineering topography rapid survey method based on the unmanned aerial vehicle according to claim 1, wherein in the step S1, boundary information of a target area is obtained, the target area is meshed based on the boundary information to form a plurality of grid points, and a cruising path of the unmanned aerial vehicle is planned, wherein the cruising path is used for controlling the unmanned aerial vehicle to fly against each grid point in turn; For each grid point, the unmanned aerial vehicle is controlled to hover above the grid point, a first frequency band sound wave with a first penetration characteristic is vertically emitted to the ground, a first vibration signal sequence generated by the first frequency band sound wave and containing reflection information from different interfaces of the ground is received to serve as a depth feedback signal of the grid point; The first penetration characteristic is used for enabling the sound wave of the first frequency band to penetrate the surface and reach the underground interface, and the second penetration characteristic is used for enabling the sound wave of the second frequency band to be coupled with the surface shallow layer substances so as to excite the surface micro-vibration.
3. A method of rapid survey of land engineering terrain based on unmanned aerial vehicle according to claim 1, wherein in S2, the first vibration signal sequences included in the depth feedback signals of each grid point are time aligned based on the depth feedback signals collected at all grid points, and in each first vibration signal sequence, the peak point of the characteristic waveform is selected as zero reference of the sequence, and then the zero reference of all the sequences is aligned on time coordinates.
4. A method of rapid survey of land engineering terrain based on unmanned aerial vehicle according to claim 3, wherein in each time aligned first vibration signal sequence the following parallel sub-steps are performed: Detecting all local amplitude extreme points exceeding a preset amplitude threshold value, and recording the time corresponding to each extreme point as the arrival time of the reflected wave; For each detected reflected wave event, calculating the waveform absolute amplitude integral value or peak amplitude in a fixed width time window with the arrival time as the center as the waveform amplitude of the reflected wave; Extracting a first feature set containing arrival time, waveform amplitude and frequency distribution of the reflected wave from each aligned first vibration signal sequence; binding the arrival time, waveform amplitude and frequency distribution of the same reflected wave event into a characteristic tuple, wherein the set of characteristic tuples of all the reflected wave events in the sequence form a first characteristic set of the grid point; Based on the first feature set, determining adjacent grid points of each grid point according to a preset spatial adjacent relation, and for each pair of adjacent grid points, selecting two groups of feature tuples with closest arrival time of the reflected wave from the respective first feature set for pairing; calculating a normalized cross-correlation value of the pairing tuple on waveform amplitude, sliding and calculating in a time window taking the arrival time difference as a center, taking the maximum value as a first correlation coefficient of the reflected wave pair, and simultaneously recording the arrival time difference corresponding to the maximum correlation coefficient as a space time difference; Tracking a reflection wave group which keeps the first correlation coefficient high among adjacent points and has gentle change of the space time difference based on the first correlation coefficient and the space time difference obtained by calculation of all adjacent point pairs, and forming a reflection wave homophase shaft which can be continuously tracked in space; Performing space continuity test and feature consistency test on each tracked reflection wave phase axis, wherein the space continuity test is performed by counting the total number of grid points penetrated by the reflection wave phase axis, calculating the percentage of the total number of the grid points accounting for the target area, and judging that the space continuity requirement is met if the percentage is larger than or equal to a first preset proportion; Checking a first correlation coefficient of each grid point penetrated by the same phase shaft and at least one adjacent grid point on the same phase shaft by checking the first correlation coefficient of the same phase shaft and the at least one adjacent grid point, wherein the first correlation coefficient of all the check points is required to be higher than a first preset wide value, meanwhile, acquiring waveform amplitude and main frequency data of the same phase shaft on all the grid points, calculating standard deviation of a maximum value and a minimum value of the same phase shaft in the whole area, and meeting the requirement of feature consistency when the standard deviation is smaller than a second preset threshold value; The reflection wave event passing through the space continuity test and the feature consistency test is judged to be a stable reflection wave group; And converting the arrival time of the reflected wave corresponding to the stable reflected wave group at each grid point into depth information according to a preset wave speed model.
5. The rapid survey method of land engineering topography based on unmanned aerial vehicle according to claim 4, wherein the specific steps of the preset wave velocity model are as follows: S211, a preset wave speed model is a structured parameter list, the list divides an underground space from the earth surface to the expected depth into a plurality of hypothesized horizontal lamellar medium units based on the sequence from the earth surface, two key parameters are distributed for each medium unit, the key parameters comprise the layer serial number of the unit and a constant wave speed value uniquely corresponding to the unit; S212, when the reflection event of the stable reflection wave group is subjected to depth conversion, inputting a double travel time T of the reflection wave of the event at a specific grid point, wherein the T represents the total time of the wave vertically descending from the ground surface to a target reflection interface and returning to the ground surface; S213, starting from the first layer, performing layer-by-layer time distribution and thickness calculation, calculating the time required by the wave to vertically propagate in the medium of the layer for one unit thickness according to the wave velocity value of the first layer, distributing a part of total time T as the time consumed by the wave to travel back and forth in the first layer, deducing the thickness of the first layer based on the distributed time and the wave velocity of the first layer, and deducting the time distributed to the first layer from the total time T after the calculation of the first layer is completed, wherein the rest time is the total consumed time of the wave to travel in a deeper layer; S214, applying the process of S213 to each layer in the wave velocity model, wherein each layer uses a constant wave velocity value specific to the layer and the time allocated to the layer in the travel time left in the last step to calculate the thickness of the layer, and the iterative process is continuously performed until the round trip propagation time allocated to the latest layer exactly corresponds to the vertical distance from the top of the layer to the target reflection interface, and the target reflection interface is positioned inside the layer; And S215, accumulating the calculated thicknesses of all traversed horizons in the wave velocity model, wherein the accumulated sum is the vertical total depth from the earth surface to the target reflection interface, and repeatedly executing the logic process from S212 to S215 on each grid point covered by the stable reflection wave group, so that a depth value is generated for each point, all the depth values jointly form a data set which is continuously distributed in space, and the three-dimensional curved surface defined by the data set is determined as the underground reference interface.
6. The unmanned aerial vehicle-based land engineering terrain rapid survey method of claim 1, wherein the at least one local reflection interface is resolved based on a depth feedback signal of a current grid point, a first vibration signal sequence of the current grid point is read first, then in the sequence, a reflected wave group known to correspond to the underground reference interface is identified and excluded, and then in the remaining signal sequences, the processed first vibration signal sequence of the current grid point after time is read, and a preset amplitude value parameter is called; Traversing from the second data point to the next-to-last data point of the sequence, marking the position of which the amplitude value is larger than that of the adjacent data points before and after the data point as a candidate local peak point, marking the position of which the amplitude value is smaller than that of the adjacent data points before and after the data point as a candidate local valley point, calculating the amplitude absolute value of each candidate point, comparing the amplitude absolute value with a preset amplitude threshold value, judging the candidate point which is larger than or equal to the preset amplitude threshold value as an effective reflected wave event, recording the time index of the effective reflected wave event, comparing the adjacent effective events with the time interval smaller than the preset minimum time window, and only keeping the event with the larger amplitude absolute value, so as to output a final effective reflected wave event list, wherein each event corresponds to one local reflected interface, and recording the arrival time of the corresponding reflected wave; Determining the spatial relation of a local reflection interface relative to an underground reference interface, converting the arrival time of a reflected wave of the local reflection interface into a preliminary depth estimation value based on a preset wave velocity model, taking the known depth value of the underground reference interface at a current grid point as a reference depth, calculating the difference value between the preliminary depth estimation value of each local reflection interface and the reference depth to obtain a preliminary relative depth relation, analyzing the frequency spectrum of a surface vibration response signal of the current grid point, checking whether abnormal formants, absorption valleys or obvious frequency spectrum distortions exist on frequency components corresponding to the preliminary depth estimation values of each local reflection interface, and defining the phenomenon as characteristic response.
7. The unmanned aerial vehicle-based land engineering terrain rapid survey method of claim 6, wherein the preliminary relative depth relation is verified and corrected according to the existence of the characteristic response and the form thereof, if the characteristic response is clear, the existence of the local reflection interface is confirmed, the occurrence of the characteristic response is further restrained by utilizing the frequency characteristic of the characteristic response, if the characteristic response is not obvious, the preliminary identification result is possibly determined to be unreliable or the interface property is weak, and finally, the verified accurate depth difference and the spatial azimuth relation of each local reflection interface relative to the underground reference interface are output as the relative geological structure information of the current grid point; Separating a reference reflected signal component from a depth feedback signal and extracting features, firstly, in a first vibration signal sequence of a current grid point, according to a known arrival time window of a reflection wave group of an underground reference interface, intercepting a signal segment in the time window, wherein the signal segment is the reference reflected signal component, and then performing feature extraction operation on the intercepted signal segment, namely, performing spectrum analysis to obtain an amplitude spectrum of the intercepted signal segment, determining a frequency point with strongest energy in the amplitude spectrum as a reference main frequency of the intercepted signal segment so as to extract reference spectrum features, and simultaneously, calculating an envelope attenuation curve of the signal segment on a time domain, measuring time required by the amplitude attenuation of the signal segment to reach a certain preset proportion of an initial value or calculating the slope of an attenuation curve of the signal segment so as to extract the reference energy attenuation features; Directly reading a complete second vibration signal of a current grid point as input, performing spectrum analysis to obtain an amplitude spectrum of the second vibration signal, determining a frequency point with the strongest energy in the amplitude spectrum as a main ground surface frequency of the second vibration signal to extract ground surface spectrum characteristics, simultaneously calculating an envelope attenuation curve of the second vibration signal on a time domain, and measuring time required by the attenuation of the amplitude of the second vibration signal to an initial value with a preset proportion or calculating the slope of the attenuation curve of the second vibration signal to extract ground surface energy attenuation characteristics; Subtracting the value of the reference main frequency from the value of the ground surface main frequency to obtain a main frequency offset, wherein the offset is the offset of the ground surface spectral feature relative to the reference spectral feature, and dividing the time value required by the ground surface signal amplitude to attenuate to a preset proportion by the attenuation time value corresponding to the reference signal to obtain a ratio, and the ratio is the ratio of the ground surface energy attenuation feature relative to the reference energy attenuation feature.
8. The unmanned aerial vehicle-based land engineering terrain rapid survey method according to claim 7, wherein if the calculated main frequency offset is positive and greater than a first set margin, the relative hardness parameter of the surface of the current grid point is determined to be higher than a reference, if the offset is negative and the absolute value of the offset is greater than the first set margin, the relative hardness parameter is determined to be lower than the reference, and if the absolute value of the offset is less than or equal to the first set margin, the relative hardness parameter is determined to be close to the reference; If the calculated attenuation ratio is smaller than 1 and the difference between the calculated attenuation ratio and 1 is larger than a second setting allowance, the relative compactness parameter of the earth surface of the current grid point is judged to be higher than the reference, if the calculated attenuation ratio is larger than 1 and the difference between the calculated attenuation ratio and 1 is larger than the second setting allowance, the relative compactness parameter is judged to be lower than the reference, and if the calculated attenuation ratio is smaller than or equal to the second setting allowance, the calculated attenuation ratio and the calculated attenuation ratio are smaller than 1 and the difference between the calculated attenuation ratio and 1 are smaller than the second setting allowance, the calculated attenuation ratio and the calculated attenuation ratio are smaller than the second setting allowance, and the relative compactness parameter is judged to be close to the reference.
9. The unmanned aerial vehicle-based land engineering terrain rapid survey method according to claim 1, wherein in S4, spatial coordinates of all grid points, relative geological structure information corresponding to each grid point and relative engineering material parameters corresponding to each grid point are obtained, wherein the relative geological structure information at least comprises depths of underground reference interfaces at each grid point and depth and spatial morphology relations of local reflection interfaces relative to the underground reference interfaces, and the relative engineering material parameters at least comprise relative hardness parameters and relative compactness parameters; Generating a first continuous curved surface representing a three-dimensional form of the underground reference interface through spatial interpolation based on spatial coordinates of all grid points and the corresponding depth of the underground reference interface, then superposing the depth of the underground reference interface at each grid point and the relative depth relation of the underground reference interface at each grid point for each local reflection interface to obtain the absolute depth of the local reflection interface at each grid point, generating a subsequent continuous curved surface representing the local reflection interface through spatial interpolation based on the absolute depth and the spatial coordinates, and finally combining and visualizing the first continuous curved surface and all the subsequent continuous curved surfaces according to the up-down spatial relation to form a three-dimensional geological structure layer; Respectively taking the relative hardness parameter and the relative compactness parameter as attribute values, taking the space coordinates of grid points as positioning basis, generating continuous attribute fields respectively corresponding to the hardness distribution and the compactness distribution through two-dimensional space interpolation, and then rendering the continuous attribute fields according to a preset color mapping rule to generate a two-dimensional engineering attribute layer expressed in a plan view form; And taking the two-dimensional engineering attribute layer as a base plane diagram, and superposing and correlating the plane projection of the three-dimensional geological structure layer and the base plane diagram in a uniform geographic coordinate system to form an integrated model supporting synchronous query and visualization of the spatial position.

Description

Land engineering terrain rapid survey method based on unmanned aerial vehicle Technical Field The invention relates to the technical field of unmanned aerial vehicle surveying, in particular to a land engineering terrain rapid surveying method based on an unmanned aerial vehicle. Background In recent years, with the development of unmanned aerial vehicle technology, a survey mode based on an unmanned aerial vehicle platform carrying an optical camera, a multispectral sensor or a laser radar has become a mainstream means for acquiring surface morphology data, however, such technology can only realize high-precision measurement of surface topography, vegetation or structure surface geometry, belongs to the surface perception category, and cannot penetrate the surface to acquire underground shallow geological structure information which is critical to engineering construction, and cannot directly detect engineering mechanical properties of a surface rock-soil body. Traditional land engineering topography surveys mainly rely on manual field surveys, drilling and sampling and large-scale physical exploration equipment, and the manual surveys are low in efficiency, limited in coverage and have subjective errors, and the drilling method is high in cost, belongs to destructive detection, can only acquire punctiform information, is difficult to reflect continuous change of an area, and is often complicated in deployment, long in operation period and poor in environmental adaptability. Disclosure of Invention Aiming at the technical problems in the prior art, the invention provides a land engineering terrain rapid survey method based on an unmanned aerial vehicle. The technical scheme for solving the technical problems is as follows, the land engineering terrain rapid survey method based on the unmanned aerial vehicle comprises the following steps: S1, acquiring a boundary of a target area, meshing and planning a cruising path of an unmanned aerial vehicle, controlling the unmanned aerial vehicle to fly to each grid point in sequence, acquiring a depth feedback signal generated by excitation of a first sound wave signal at each grid point, and acquiring a surface vibration response signal generated by excitation of a second sound wave signal; S2, determining a continuous and stable underground reference interface in a target area based on the depth feedback signals acquired by all grid points; s3, aiming at each grid point, determining relative geological structure information and relative engineering material parameters of the grid point based on a depth feedback signal and a surface vibration response signal of the grid point and taking an underground reference interface as a reference; s4, generating a land engineering comprehensive terrain model based on the space coordinates of all grid points, the relative geological structure information and the relative engineering material parameters. In a preferred embodiment, in the step S1, boundary information of a target area is obtained, the target area is meshed based on the boundary information to form a plurality of grid points, and a cruising path of the unmanned aerial vehicle is planned, wherein the cruising path is used for controlling the unmanned aerial vehicle to fly to each grid point in turn; for each grid point, controlling the unmanned aerial vehicle to hover above the grid point, vertically transmitting a first frequency band sound wave with a first penetration characteristic to the ground, and receiving a first vibration signal sequence which is generated by the first frequency band sound wave and contains reflection information from different interfaces of the ground as a depth feedback signal of the grid point; Transmitting a second-frequency-band acoustic wave with a second penetration characteristic to the ground for each grid point to excite the ground surface, and receiving a second vibration signal representing the micro-vibration characteristic of the ground surface generated by the excitation of the second-frequency-band acoustic wave as a ground surface vibration response signal of the grid point; The first penetration characteristic is used for enabling the sound wave of the first frequency band to penetrate the surface and reach the underground interface, and the second penetration characteristic is used for enabling the sound wave of the second frequency band to be coupled with the surface shallow layer substances so as to excite the surface micro-vibration. In a preferred embodiment, in the step S2, the first vibration signal sequences included in the depth feedback signals of each grid point are time-aligned based on the depth feedback signals acquired at all grid points, and in each first vibration signal sequence, the peak point of the characteristic waveform is selected as the reference zero point of the sequence, and then the reference zero points of all the sequences are aligned on the time coordinates; In each time-aligned first vib