Millimeter-wave-based continuous measurement method and system for cross-sectional water depth

Abstract

The present invention discloses a method and system for continuous measurement of cross-sectional water depth based on millimeter waves, which relates to the technical field of hydrological measurement. The present invention transmits frequency-modulated continuous wave signals and receives mixed echo signals through a millimeter wave radar array and a multi-input multi-output antenna. Wavelet transform and short-time Fourier transform are used to extract time-frequency features, and the reflection peaks are classified in combination with a Gaussian mixture model. The ice thickness, air interlayer thickness and initial water level are calculated based on the time delay difference, and the error is dynamically corrected through an adaptive Kalman filter. A dynamic focusing algorithm is used to synthesize the water surface and riverbed reflection signals, and a sediment content compensation and flow velocity correction model are introduced to optimize the inversion of water depth and riverbed elevation. An ant population algorithm is used to optimize the three-dimensional interpolation weights to generate a high-precision cross-sectional continuous water depth distribution map. The present invention solves the problems of poor penetration, weak anti-interference ability and low continuous water depth accuracy, and realizes full-section dynamic monitoring of ice thickness, water level, water depth and riverbed morphology.

Description

Technical Field

[0001] The present invention belongs to the technical field of hydrological measurement, and in particular, relates to a method and system for continuous measurement of cross-sectional water depth based on millimeter waves. Background Art

[0002] In cold-region water management, monitoring hydrological parameters (such as ice thickness, subglacial water level, water depth, and riverbed morphology) influenced by ice cover is a core challenge for flood control, water resource management, and ecological protection. Traditional methods rely on manual icebreaking measurements or fixed sensor networks, which present significant safety risks, delayed data updates, and insufficient spatial coverage. The advancement of millimeter-wave radar technology, with its high penetration and anti-interference capabilities, has provided a new approach for subglacial hydrological monitoring. However, existing millimeter-wave radar systems still face technical bottlenecks in multi-media penetration signal separation, dynamic environmental error compensation, and terrain modeling accuracy.

[0003] The existing technology has problems such as poor adaptability to ice environment, large dynamic environment error, large limitations of single-point measurement and low continuous water depth accuracy. Summary of the Invention

[0004] In response to the problems in the related art, the present invention provides a millimeter wave-based continuous measurement method for cross-sectional water depth to overcome the above-mentioned technical problems existing in the existing related art.

[0005] To solve the above technical problems, the present invention is achieved through the following technical solutions:

[0006] S1. Receive data from the radar in the millimeter-wave radar cluster to obtain a mixed echo signal data set;

[0007] S2, using wavelet transform and short-time Fourier transform, combined with the mixed echo signal data set, to obtain the denoised time-frequency feature matrix set;

[0008] Classify the reflection peaks in the denoised time-frequency feature matrix and combine them with the reflection peak amplitude threshold to obtain the reflection peak delay matrix;

[0009] S3. Calculating an ice layer thickness set and an air layer thickness set according to the reflection peak time delay matrix;

[0010] Obtaining a water level elevation set based on the ice layer thickness set and the air layer thickness set in combination with an adaptive Kalman filter algorithm;

[0011] S4. Using the millimeter wave propagation velocity set in water corrected by the sediment content compensation algorithm and the reflection peak delay difference set between the riverbed and the water surface corrected by the water flow velocity set, combined with the water level elevation set, a water depth dataset and a riverbed elevation dataset are obtained;

[0012] S5. Using the optimized three-dimensional interpolation algorithm, combined with the water depth data set and the riverbed elevation set, to obtain a continuous water depth distribution map of the cross section;

[0013] The present invention deploys millimeter-wave radar arrays and MIMO antenna technology, combines wavelet transform-short-time Fourier transform, adaptive Kalman filter dynamic correction, dynamic focusing beam synthesis and sediment content-velocity joint compensation algorithm, and uses ant population algorithm to optimize three-dimensional interpolation; it realizes high-precision dynamic monitoring of ice thickness, air interlayer, water level elevation, water depth and riverbed morphology in the cross-section of water areas during the ice-covered period, overcoming the shortcomings of traditional technologies such as insufficient penetration, sensitivity to environmental interference and rough terrain modeling, and providing an automated and highly reliable solution for real-time perception and visualization of full-section subglacial hydrological parameters.

[0014] Preferably, the S1 comprises the following steps:

[0015] S11. Uniformly deploying a millimeter-wave radar array along a cross section of the water area to obtain a millimeter-wave radar set; the millimeter-wave radar set includes all radars deployed in the cross section of the water area; the radars in the millimeter-wave radar set must be installed at a height that covers the dynamic range of ice and water surface changes;

[0016] Configured with multiple-input multiple-output MIMO antennas to cover the target cross-sectional area through multi-angle beam scanning;

[0017] S12, each radar in the millimeter wave radar set transmits an FMCW signal to the target area, which penetrates the ice layer, air layer and water body and then returns to the radar;

[0018] Each radar in the millimeter-wave radar set receives an echo signal reflected from the target area to obtain a mixed echo signal data set; the mixed echo signal data set includes mixed echo signal data received by each millimeter-wave radar; the mixed echo signal data includes reflection components from ice layers, air interlayers, water bodies, and riverbeds;

[0019] The present invention evenly deploys millimeter-wave radar arrays across the cross-section of the water area. The arrays are installed at a height that covers the dynamic range of ice and water surface changes, and are equipped with MIMO antennas for multi-angle beam scanning. Each radar transmits an FMCW signal that penetrates the ice layer, air interlayer, and water body, and then receives a mixed echo signal, forming a data set containing reflection components from the ice layer, air, water, and riverbed, laying a data foundation for multi-media parameter extraction.

[0020] Preferably, said S2 comprises the following steps:

[0021] S21, performing denoising processing on the mixed echo signal data in the mixed echo signal data set using wavelet transform to obtain a denoised mixed echo signal data set;

[0022] The signal delay and frequency characteristics of the denoised mixed echo signal data set are extracted by short-time Fourier transform to obtain the denoised time-frequency feature matrix set;

[0023] S22. Using a Gaussian mixture model, the reflection peaks of each medium in the denoised time-frequency feature matrix are classified to obtain a reflection peak classification matrix; the reflection peak classification matrix includes the reflection peaks of all ice layers, air interlayers, water bodies, and riverbeds monitored by the millimeter-wave radar;

[0024] The reflection peak amplitude threshold is set based on the difference in dielectric constant between ice and water.

[0025] A reflection peak delay matrix is obtained according to the reflection peak classification matrix and the reflection peak amplitude threshold; the reflection peak delay matrix includes the reflection peak delays of the upper surface, lower surface, water surface and riverbed of the ice layer monitored by all millimeter wave radars;

[0026] The present invention denoises the mixed echo signal through wavelet transform and extracts the time-frequency feature matrix in combination with short-time Fourier transform; uses Gaussian mixture model to classify the reflection peaks of ice layer, air interlayer, water body and riverbed, sets amplitude threshold based on the difference in dielectric constant, screens the effective reflection peak delay, and generates a matrix containing the delay of each medium interface, providing accurate data support for the calculation of parameters such as ice thickness and water level.

[0027] Preferably, the step S3 includes the following steps:

[0028] S31. Calculate an ice thickness set based on the time delay difference of the reflection peaks between the upper and lower surfaces of the ice layer and the propagation speed of millimeter waves in the ice; the ice thickness set includes the ice thicknesses monitored by all millimeter wave radars;

[0029] Based on the reflection peak delay difference from the lower surface of the ice layer to the water surface, combined with the propagation speed of millimeter waves in the air interlayer, the air interlayer thickness set is calculated; the air interlayer thickness set includes the air interlayer thickness monitored by all millimeter wave radars;

[0030] S32. Calculate an initial water level elevation set based on the ice layer thickness set, the air layer thickness set, and the radar installation points; the water level elevation set includes all water level elevations monitored by the millimeter-wave radar;

[0031] S33, using an adaptive Kalman filter algorithm to dynamically correct measurement errors caused by ice movement or water flow fluctuations to obtain a water level elevation set;

[0032] The present invention calculates the ice thickness and air layer thickness based on the time delay difference of the reflection peaks on the upper and lower surfaces of the ice layer and the air interlayer, combined with the propagation speed of millimeter waves, and infers the initial water level based on the radar installation height. It also dynamically corrects the ice layer movement and water flow fluctuation errors through adaptive Kalman filtering, and finally outputs a high-precision water level elevation set.

[0033] Preferably, said S4 comprises the following steps:

[0034] S41. Using a dynamic focusing algorithm combined with MIMO radar synthesis, obtain a water surface reflection signal set and a riverbed reflection signal set; and obtaining a reflection peak delay difference set between the riverbed and the water surface based on the water surface reflection signal set and the riverbed reflection signal set;

[0035] S42. Correcting the millimeter wave propagation velocity in water using a sediment content compensation algorithm to obtain a corrected millimeter wave propagation velocity set in water; correcting the reflection peak delay difference between the riverbed and the water surface using a flow velocity compensation algorithm to obtain a corrected reflection peak delay difference set between the riverbed and the water surface;

[0036] S43. Obtaining a water depth data set using a water depth calculation formula based on the corrected millimeter wave propagation velocity set in water and the corrected reflection peak delay difference set between the riverbed and the water surface; the water depth data set includes water depth data monitored by all millimeter wave radars;

[0037] The riverbed elevation set is obtained by combining the water level elevation set with the water depth dataset;

[0038] The present invention uses a dynamic focusing algorithm and MIMO radar to synthesize the water surface and riverbed reflection signals to obtain the time delay difference between the two; based on the sediment content compensation, the millimeter wave propagation speed in the water is corrected, and the time delay difference is optimized in combination with flow rate compensation. Finally, the accurate water depth data set and riverbed elevation set are calculated by correcting the parameters, effectively eliminating the influence of sediment content and water flow on monitoring accuracy.

[0039] Preferably, the S41 includes the following steps:

[0040] S411, scan the cross section at multiple angles using a MIMO radar array to obtain high spatial resolution echo data;

[0041] S412. Obtain the current beam pointing angle to be tracked based on the water surface ripples; adjust the beam weight based on the beam pointing angle in combination with a dynamic focusing algorithm to obtain a real-time beam weight set;

[0042] S413, using the real-time beam weight set in combination with the MIMO virtual array synthesis to obtain a water surface reflection signal set and a riverbed reflection signal set;

[0043] Obtaining a reflection peak delay difference set between the riverbed and the water surface based on the water surface reflection signal set and the riverbed reflection signal set;

[0044] The present invention acquires high-resolution echo data through multi-angle scanning of MIMO radar array, dynamically adjusts beam pointing angle and weight based on water surface ripples, combines virtual array synthesis to separate water surface and riverbed reflection signals, accurately extracts the time delay difference of reflection peaks between the two, adapts to dynamic water surface fluctuations and improves underwater target detection accuracy.

[0045] Preferably, the S42 includes the following steps:

[0046] S421. Collect the sediment concentration of each radar monitoring area to obtain a real-time sediment concentration set; based on the real-time sediment concentration set, use a sediment concentration compensation algorithm to correct the millimeter wave propagation velocity in water to obtain a corrected millimeter wave propagation velocity set in water;

[0047] S422, correcting the reflection peak delay difference between the riverbed and the water surface according to the real-time water velocity set to obtain a corrected reflection peak delay difference set between the riverbed and the water surface;

[0048] The present invention corrects the millimeter wave propagation speed in water through real-time sediment content data, and compensates for the reflection peak delay difference based on the water flow velocity, eliminating the interference of sediment content and flow velocity on signal propagation, significantly improving the inversion accuracy of water depth and riverbed elevation, and ensuring measurement reliability in dynamic hydrological environments.

[0049] Preferably, the S5 comprises the following steps:

[0050] S51. Optimizing the dynamic weight coefficient in the three-dimensional interpolation algorithm using an ant colony algorithm to obtain an optimal solution, and using the optimal solution as the dynamic weight coefficient of the three-dimensional interpolation algorithm to obtain an optimized three-dimensional interpolation algorithm;

[0051] S52, using the optimized three-dimensional interpolation algorithm, combined with the water depth data set and the riverbed elevation set, to obtain a continuous cross-sectional water depth distribution map;

[0052] The present invention optimizes the dynamic weight coefficient of three-dimensional interpolation through an ant colony algorithm to minimize the interpolation error, and combines water depth and riverbed elevation data to generate a high-precision continuous cross-sectional distribution map, thereby solving the terrain distortion problem under sparse monitoring points and improving the spatial representation capability of subglacial water bodies and riverbed morphology.

[0053] Preferably, the S51 includes the following steps:

[0054] S511. Construct an ant population and set the size of the ant population; set the maximum number of optimization iterations;

[0055] S512. Randomly set the ant population pheromone concentration according to the dynamic weight coefficient in the three-dimensional interpolation algorithm to obtain an ant population pheromone concentration set;

[0056] S513, setting an MSE that directly measures interpolation accuracy; and defining a fitness function of the ant pheromone concentration in the ant population based on the MSE;

[0057] S514. Iterate the ant population pheromone concentration set. The higher the fitness value, the higher the pheromone concentration. During each round of iteration, calculate the fitness value of each pheromone concentration in the ant population pheromone concentration set according to the fitness function. Update the pheromone concentration of each ant in the initial position set of the ant population in descending order of fitness value. In each round of iteration, obtain the optimal individual ant pheromone concentration in the ant population and the global optimal ant pheromone concentration.

[0058] S515, repeating 514, when the maximum number of optimization iterations is reached, stopping the iteration, and taking the global optimal ant pheromone concentration as the optimal solution;

[0059] The present invention constructs an ant population and randomly initializes the pheromone concentration, defines the fitness function with MSE, iteratively updates the pheromone concentration, and selects the global optimal pheromone concentration as the dynamic weight coefficient of three-dimensional interpolation, thereby achieving high-precision terrain interpolation under sparse monitoring points and effectively suppressing discrete data interpolation distortion.

[0060] A millimeter-wave-based continuous cross-sectional water depth measurement system, used to implement the above-mentioned millimeter-wave-based continuous cross-sectional water depth measurement method, includes a mixed echo signal generation module, a signal denoising and time-frequency feature classification module, an ice thickness and water level dynamic calculation module, a water depth and riverbed elevation compensation correction module, and a three-dimensional interpolation and cross-sectional terrain modeling module;

[0061] The hybrid echo signal generation module is used to deploy a millimeter-wave radar array in a water cross-section, configure a multi-input multi-output antenna to perform multi-angle beam scanning, transmit a frequency-modulated continuous wave signal to the target area, and receive the hybrid echo signal after penetrating the ice layer, air layer, and water body. The radar installation height covers the dynamic range of the ice layer and water surface, ensuring that the signal contains the reflection components of the upper and lower surfaces of the ice layer, the air layer, the water surface, and the riverbed, forming a hybrid echo signal data set, which provides the raw data basis for subsequent processing.

[0062] The signal denoising and time-frequency feature classification module is used to denoise the mixed echo signal using wavelet transform, extract the signal time-frequency features in combination with short-time Fourier transform, and generate a time-frequency feature matrix; use the Gaussian mixture model to classify the reflection peaks in the time-frequency matrix, distinguish the reflection components of ice layers, air interlayers, water bodies, and riverbeds, and filter valid signals using a preset reflection peak amplitude threshold. Finally, a reflection peak delay matrix is generated to provide key parameters for ice thickness and water level calculation;

[0063] The ice thickness and water level dynamic calculation module calculates the time delay difference between the upper and lower surfaces of the ice layer and the air interlayer based on the reflection peak delay matrix. Combined with the propagation speed of millimeter waves in different media, the ice thickness and air interlayer thickness are respectively obtained. The module uses the radar installation point height and the adaptive Kalman filter algorithm to dynamically correct errors caused by ice movement or water flow fluctuations, generate a high-precision water level elevation data set, and realize real-time monitoring of the water level under the ice.

[0064] The water depth and riverbed elevation compensation correction module uses a dynamic focusing algorithm to adjust the MIMO radar beam weights, synthesize high-resolution reflection signals from the water surface and riverbed, and obtain the time delay difference between the two reflection peaks. It introduces a sediment content compensation algorithm to correct the millimeter wave propagation speed in the water, and uses the water velocity inverted by Doppler frequency shift to perform a secondary correction on the time delay difference. Finally, combined with water level elevation data, it calculates the water depth and riverbed elevation data set under the influence of sediment content and flow velocity.

[0065] The three-dimensional interpolation and cross-sectional terrain modeling module uses an ant population algorithm to optimize the dynamic weight coefficient of the three-dimensional interpolation algorithm, and iteratively solves the optimal interpolation parameters using the mean square error as the fitness function; based on the optimized algorithm, discrete water depth and riverbed elevation data are spatially interpolated to generate a continuous underwater cross-sectional terrain distribution map, which intuitively displays the three-dimensional structural characteristics of ice layers, air interlayers, water bodies and riverbeds, providing visual support for water area ice age monitoring.

[0066] Beneficial effects

[0067] The present invention has the following beneficial effects:

[0068] The present invention deploys millimeter-wave radar arrays and MIMO antenna technology, combines wavelet transform-short-time Fourier transform, adaptive Kalman filter dynamic correction, dynamic focusing beam synthesis and sediment content-velocity joint compensation algorithm, and uses ant population algorithm to optimize three-dimensional interpolation; it realizes high-precision dynamic monitoring of ice thickness, air interlayer, water level elevation, water depth and riverbed morphology in the cross-section of water areas during the ice-covered period, overcoming the shortcomings of traditional technologies such as insufficient penetration, sensitivity to environmental interference and rough terrain modeling, and providing an automated and highly reliable solution for real-time perception and visualization of full-section subglacial hydrological parameters.

[0069] Through millimeter-wave radar array and MIMO antenna technology, the present invention realizes penetrating detection of ice layers, air interlayers and water bodies, solving the problem that traditional sensors cannot obtain sub-ice water level and riverbed data due to ice obstruction; multi-angle beam scanning and high-resolution echo signal synthesis technology significantly improve the reflection signal separation capability of complex medium interfaces such as ice-air-water, ensuring the integrity and reliability of mixed echo data.

[0070] The present invention combines the time-frequency feature extraction of wavelet transform-short-time Fourier transform with the Gaussian mixture model classification algorithm to effectively remove environmental noise interference and achieve accurate classification of reflection peaks based on dielectric constant differences; it dynamically corrects measurement errors caused by ice movement and water flow fluctuations through adaptive Kalman filtering, significantly improving the stability and accuracy of water level elevation.

[0071] The present invention addresses the time delay error caused by the offset of millimeter wave propagation velocity and water flow in sandy water bodies by introducing a sediment content compensation algorithm and Doppler frequency shift inversion technology to dynamically correct the calculation parameters of water depth and riverbed elevation. This reduces the water depth error affected by sediment content and improves the accuracy of riverbed elevation inversion after flow velocity compensation.

[0072] This paper uses an ant colony algorithm to optimize the three-dimensional interpolation weight coefficient, solving the terrain distortion problem of traditional interpolation algorithms under sparse monitoring points. By iteratively generating the optimal interpolation parameters by minimizing the mean square error, the generated cross-sectional continuous water depth distribution map can accurately reflect the spatial structural characteristics of the ice layer-air interlayer-water body-riverbed, providing intuitive visualization support for ice-age safety management and scheduling decisions in water areas.

[0073] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] In order to more clearly illustrate the technical solutions of the embodiments of the invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the invention. For ordinary technicians in this field, they can also obtain drawings based on these drawings without paying any creative work.

[0075] Figure 1 Schematic diagram of the process of the millimeter wave-based continuous measurement method of cross-sectional water depth of the present invention;

[0076] Figure 2 Schematic diagram of the process of obtaining a water depth dataset and a riverbed elevation dataset in the millimeter wave-based continuous cross-sectional water depth measurement method of the present invention;

[0077] Figure 3 Schematic diagram of the process of the optimized three-dimensional interpolation algorithm in the millimeter wave-based cross-sectional water depth continuous measurement method of the present invention;

[0078] Figure 4 This is a module schematic diagram of the millimeter wave-based cross-sectional water depth continuous measurement system of the present invention. DETAILED DESCRIPTION

[0079] The following will clearly and completely describe the technical solutions in the embodiments of the invention in conjunction with the accompanying drawings. Obviously, the embodiments described are only part of the embodiments of the invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0080] In the description of the present invention, it should be understood that the terms "opening", "upper", "lower", "top", "middle", "inside" and the like indicating orientation or positional relationship are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the components or elements referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the invention.

[0081] Example 1

[0082] See also Figure 1 、 Figure 2 、 Figure 3 The present invention discloses a method and system for continuous measurement of cross-sectional water depth based on millimeter waves, comprising the following steps:

[0083] S1. Receive data from the radar in the millimeter-wave radar cluster to obtain a mixed echo signal data set;

[0084] Said S1 comprises the following steps:

[0085] S11. Deploy the millimeter-wave radar array uniformly along the cross section of the water area, and obtain the millimeter-wave radar set h={h1,h2,…,h i ,…,h g}; where h i represents the i-th radar in the cross section, g represents the total number of radars in the water cross section; the millimeter-wave radar set includes all radars deployed in the cross section; the installation height of the radars in the millimeter-wave radar set must cover the dynamic change range of the ice layer and the water surface;

[0086] Configure multiple-input multiple-output (MIMO) antennas to cover the target cross-sectional area through multi-angle beam scanning;

[0087] S12, each radar in the millimeter wave radar set transmits an FMCW signal to the target area, which penetrates the ice layer, air layer and water body and then returns to the radar;

[0088] Each radar in the millimeter-wave radar set receives the echo signal reflected from the target area, and obtains the mixed echo signal data set p={p1,p2,…,p i ,…,p g}; where p irepresents the mixed echo signal data received by the i-th millimeter-wave radar; the mixed echo signal data set contains the mixed echo signal data received by each millimeter-wave radar; the mixed echo signal data includes the reflection components of the ice layer, air interlayer, water body, and riverbed;

[0089] S2, using wavelet transform and short-time Fourier transform, combined with the mixed echo signal data set, to obtain the denoised time-frequency feature matrix set;

[0090] Classify the reflection peaks in the denoised time-frequency feature matrix and combine them with the reflection peak amplitude threshold to obtain the reflection peak delay matrix;

[0091] The S2 comprises the following steps:

[0092] S21. Use wavelet transform to perform denoising on the mixed echo signal data in the mixed echo signal data set to obtain a denoised mixed echo signal data set; the wavelet transform formula is as follows:

[0093] ;

[0094] Among them, a represents the scale parameter, b represents the displacement parameter, and p i represents the moderate echo signal, W(a,b) represents the mixed echo signal p i The wavelet transform result at scale parameter a and location parameter b is: represents the conjugate wavelet function;

[0095] The signal delay and frequency characteristics of the denoised mixed echo signal data set are extracted by short-time Fourier transform to obtain the denoised time-frequency feature matrix set;

[0096] S22. Use a Gaussian mixture model (GMM) to classify the reflection peaks of each medium in the denoised time-frequency feature matrix set to obtain a reflection peak classification matrix C. The reflection peak classification matrix is as follows:

[0097] ;

[0098] Among them, C i1 、C i2 、C i3 and C i4 Respectively represent the reflection peaks of the ice layer, air interlayer, water body and riverbed monitored by the i-th millimeter-wave radar; the reflection peak classification matrix includes the reflection peaks of the ice layer, air interlayer, water body and riverbed monitored by all millimeter-wave radars;

[0099] The Gaussian mixture model (GMM) formula is as follows,

[0100] ;

[0101] Where y represents the eigenvector in the time-frequency feature matrix, p(y) represents the probability density of the occurrence of the eigenvector y in the time-frequency feature matrix, K represents the Gaussian component (i.e., the number of media types in the time-frequency feature matrix), k|k=1,2,3,4, π k represents the mixing coefficient of the kth Gaussian component (i.e., the prior probability of the kth type of medium in the signal), β k Represents the characteristic mean vector of the k-th type of medium; represents the probability density function of the multivariate Gaussian distribution (i.e., describing the characteristic distribution of the k-th type of medium);

[0102] The reflection peak amplitude threshold is set based on the difference in dielectric constant between ice and water.

[0103] According to the reflection peak classification matrix and the reflection peak amplitude threshold, the reflection peak delay matrix D is obtained. The reflection peak delay matrix is as follows:

[0104] ;

[0105] Among them, D i1 、D i2 、D i3 and D i4 They represent the reflection peak delays of the upper surface, lower surface, water surface and riverbed of the ice layer monitored by the i-th millimeter-wave radar respectively; the reflection peak delay matrix includes the reflection peak delays of the upper surface, lower surface, water surface and riverbed of the ice layer monitored by all millimeter-wave radars;

[0106] S3. Calculating an ice layer thickness set and an air layer thickness set according to the reflection peak time delay matrix;

[0107] Obtaining a water level elevation set based on the ice layer thickness set and the air layer thickness set in combination with an adaptive Kalman filter algorithm;

[0108] The S3 includes the following steps:

[0109] S31. Calculate the ice thickness set based on the time delay difference of the reflection peaks between the upper and lower surfaces of the ice layer and the propagation speed of millimeter waves in the ice. The ice thickness set includes the ice thickness monitored by all millimeter wave radars. The calculation formula is as follows:

[0110] ;

[0111] Among them, H ice Indicates the thickness of the ice layer, v ice represents the propagation speed of millimeter waves in ice, d1 and d2 represent the time delay difference of the reflection peaks on the upper and lower surfaces of the ice layer, respectively;

[0112] Based on the reflection peak delay difference from the lower surface of the ice layer to the water surface, combined with the propagation speed of millimeter waves in the air layer, the air layer thickness set is calculated; the air layer thickness set includes the air layer thickness monitored by all millimeter wave radars; the calculation formula is as follows,

[0113] ;

[0114] Among them, H air Indicates the thickness of the air layer, v air represents the propagation speed of millimeter waves in the air layer, d2 and d3 represent the reflection peak delay difference of the lower surface and the reflection peak delay difference of the water surface respectively;

[0115] S32. Calculate an initial water level elevation set based on the ice layer thickness set, the air layer thickness set, and the radar installation points. The water level elevation set includes all water level elevations monitored by millimeter-wave radars. The calculation formula is as follows:

[0116] ;

[0117] Among them, H water Indicates water level elevation, H r Indicates the height of the installation point relative to the reference plane (such as the horizontal plane);

[0118] S33, using an adaptive Kalman filter algorithm to dynamically correct measurement errors caused by ice movement or water flow fluctuations to obtain a water level elevation set;

[0119] S4. Using the millimeter wave propagation velocity set in water corrected by the sediment content compensation algorithm and the reflection peak delay difference set between the riverbed and the water surface corrected by the water flow velocity set, combined with the water level elevation set, a water depth dataset and a riverbed elevation dataset are obtained;

[0120] The 4 steps include:

[0121] S41. Using a dynamic focusing algorithm combined with MIMO radar synthesis, obtain a water surface reflection signal set and a riverbed reflection signal set; and obtaining a reflection peak delay difference set between the riverbed and the water surface based on the water surface reflection signal set and the riverbed reflection signal set;

[0122] The S41 includes the following steps:

[0123] S411, scan the cross section at multiple angles using a MIMO radar array to obtain high spatial resolution echo data;

[0124] S412. Obtain the current beam pointing angle to be tracked based on the water surface ripples; adjust the beam weight based on the beam pointing angle in combination with the dynamic focusing algorithm to obtain a real-time beam weight set; the dynamic focusing algorithm formula is as follows:

[0125] ;

[0126] where w represents the real-time beam weight; argmin represents the minimization argument (i.e., finding the optimal solution such that || e(θ) - w H z||The one with the minimum value w vector); θ represents the beam pointing angle; e(θ) represents the steering vector dimension (i.e., the phase delay characteristic of the beam in the direction of angle θ); w H represents the conjugate transpose of the beam weight vector, which is used to synthesize the beam, and z represents the received signal matrix;

[0127] S413, using the real-time beam weight set in combination with the MIMO virtual array synthesis to obtain a water surface reflection signal set and a riverbed reflection signal set;

[0128] Obtaining a reflection peak delay difference set between the riverbed and the water surface based on the water surface reflection signal set and the riverbed reflection signal set;

[0129] S42. Correcting the millimeter wave propagation velocity in water using a sediment content compensation algorithm to obtain a corrected millimeter wave propagation velocity set in water; correcting the reflection peak delay difference between the riverbed and the water surface using a flow velocity compensation algorithm to obtain a corrected reflection peak delay difference set between the riverbed and the water surface;

[0130] The S42 includes the following steps:

[0131] S421. Collect the sediment content of each radar monitoring area to obtain a real-time sediment content set; based on the real-time sediment content set, use a sediment content compensation algorithm to correct the millimeter wave propagation velocity in water to obtain a corrected millimeter wave propagation velocity set in water; the sediment content compensation algorithm formula is as follows:

[0132] ;

[0133] Among them, v water represents the corrected millimeter wave propagation velocity in water, v0 represents the millimeter wave propagation velocity in water under standard sediment concentration, δ represents the set sediment concentration compensation, and S represents the real-time sediment concentration;

[0134] The water surface velocity is inverted by Doppler frequency shift to obtain the real-time water velocity set; the inversion formula is as follows:

[0135] ;

[0136] Among them, v flow represents the real-time water velocity, f represents the Doppler shift (i.e., the frequency offset of the echo signal caused by the water surface velocity), λ represents the radar wavelength, and θ represents the radar wave incident angle;

[0137] S422. Correct the reflection peak delay difference between the riverbed and the water surface according to the real-time water velocity set to obtain a corrected reflection peak delay difference set between the riverbed and the water surface. The correction formula is as follows:

[0138] ;

[0139] Wherein, j represents the corrected reflection peak delay difference between the riverbed and the water surface, and η represents the set velocity reflection peak delay error correction coefficient;

[0140] S43. Based on the corrected millimeter wave propagation velocity set in water and the corrected reflection peak delay difference set between the riverbed and the water surface, a water depth data set is obtained using a water depth calculation formula; the water depth data set includes all water depth data monitored by millimeter wave radars; the water depth calculation formula is as follows:

[0141] ;

[0142] Where G represents the water depth data monitored by millimeter-wave radar;

[0143] According to the water level elevation set combined with the water depth data set, the riverbed elevation set is obtained; the calculation formula is as follows:

[0144] ;

[0145] Among them, H bed Indicates the riverbed elevation;

[0146] S5. Using the optimized three-dimensional interpolation algorithm, combined with the water depth data set and the riverbed elevation set, to obtain a continuous water depth distribution map of the cross section;

[0147] The S5 comprises the following steps:

[0148] S51. Optimizing the dynamic weight coefficient in the three-dimensional interpolation algorithm using an ant colony algorithm to obtain an optimal solution, and using the optimal solution as the dynamic weight coefficient of the three-dimensional interpolation algorithm to obtain an optimized three-dimensional interpolation algorithm;

[0149] The S51 includes the following steps:

[0150] S511, construct an ant population, set the size of the ant population to be u, and then the ant population is expressed as u={u1,u2,…,u i ,…,u q}, where u i Represents the i-th ant in the ant population; sets the maximum number of optimization iterations;

[0151] S512: Randomly set the ant population pheromone concentration according to the dynamic weight coefficient in the three-dimensional interpolation algorithm, and obtain the ant population pheromone concentration set l={l1,l2,…,li ,…,l q}, where l i represents the pheromone concentration of the i-th ant in the ant population;

[0152] S513. Set the MSE, which is used to directly measure the interpolation accuracy. Based on the MSE, define the fitness function of the ant pheromone concentration in the ant population. The fitness function formula is as follows:

[0153] ;

[0154] R represents the fitness function;

[0155] S514. Iterate the ant population pheromone concentration set. The higher the fitness value, the higher the pheromone concentration. During each round of iteration, calculate the fitness value of each pheromone concentration in the ant population pheromone concentration set according to the fitness function. Update the pheromone concentration of each ant in the initial position set of the ant population in descending order of fitness value. In each round of iteration, obtain the optimal individual ant pheromone concentration in the ant population and the global optimal ant pheromone concentration.

[0156] S515, repeating 514, when the maximum number of optimization iterations is reached, stopping the iteration, and taking the global optimal ant pheromone concentration as the optimal solution;

[0157] S52. Use the optimized three-dimensional interpolation algorithm, combined with the water depth data set and the riverbed elevation set, to obtain a continuous cross-sectional water depth distribution map.

[0158] Example 2

[0159] See also Figure 4 The millimeter-wave-based cross-sectional water depth continuous measurement system is used to implement the above-mentioned millimeter-wave-based cross-sectional water depth continuous measurement method, including a mixed echo signal generation module, a signal denoising and time-frequency feature classification module, an ice thickness and water level dynamic calculation module, a water depth and riverbed elevation compensation correction module, and a three-dimensional interpolation and cross-sectional terrain modeling module;

[0160] The hybrid echo signal generation module is used to deploy a millimeter-wave radar array in a water cross-section, configure a multi-input multi-output antenna to perform multi-angle beam scanning, transmit a frequency-modulated continuous wave signal to the target area, and receive the hybrid echo signal after penetrating the ice layer, air layer, and water body. The radar installation height covers the dynamic range of the ice layer and water surface, ensuring that the signal contains the reflection components of the upper and lower surfaces of the ice layer, the air layer, the water surface, and the riverbed, forming a hybrid echo signal data set, which provides the raw data basis for subsequent processing.

[0161] The signal denoising and time-frequency feature classification module is used to denoise the mixed echo signal using wavelet transform, extract the signal time-frequency features in combination with short-time Fourier transform, and generate a time-frequency feature matrix; use the Gaussian mixture model to classify the reflection peaks in the time-frequency matrix, distinguish the reflection components of ice layers, air interlayers, water bodies, and riverbeds, and filter valid signals using a preset reflection peak amplitude threshold. Finally, a reflection peak delay matrix is generated to provide key parameters for ice thickness and water level calculation;

[0162] The ice thickness and water level dynamic calculation module calculates the time delay difference between the upper and lower surfaces of the ice layer and the air interlayer based on the reflection peak delay matrix. Combined with the propagation speed of millimeter waves in different media, the ice thickness and air interlayer thickness are respectively obtained. The module uses the radar installation point height and the adaptive Kalman filter algorithm to dynamically correct errors caused by ice movement or water flow fluctuations, generate a high-precision water level elevation data set, and realize real-time monitoring of the water level under the ice.

[0163] The water depth and riverbed elevation compensation correction module uses a dynamic focusing algorithm to adjust the MIMO radar beam weights, synthesize high-resolution reflection signals from the water surface and riverbed, and obtain the time delay difference between the two reflection peaks. It introduces a sediment content compensation algorithm to correct the millimeter wave propagation speed in the water, and uses the water velocity inverted by Doppler frequency shift to perform a secondary correction on the time delay difference. Finally, combined with water level elevation data, it calculates the water depth and riverbed elevation data set under the influence of sediment content and flow velocity.

[0164] The three-dimensional interpolation and cross-sectional terrain modeling module uses an ant population algorithm to optimize the dynamic weight coefficient of the three-dimensional interpolation algorithm, and iteratively solves the optimal interpolation parameters using the mean square error as the fitness function; based on the optimized algorithm, discrete water depth and riverbed elevation data are spatially interpolated to generate a continuous underwater cross-sectional terrain distribution map, which intuitively displays the three-dimensional structural characteristics of ice layers, air interlayers, water bodies and riverbeds, providing visual support for water area ice age monitoring.

[0165] Throughout this specification, references to terms such as "one embodiment," "example," or "specific example" indicate that the specific features, structures, materials, or characteristics described in conjunction with that embodiment or example are included in at least one embodiment or example of the invention. In this specification, schematic representations of these terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0166] The preferred embodiments of the invention disclosed above are intended only to help illustrate the invention. These preferred embodiments do not exhaust all details, nor do they limit the invention to the specific embodiments described. Obviously, many modifications and variations are possible based on the content of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention.

Claims

1. A millimeter-wave-based continuous measurement method for cross-sectional water depth, characterized in that: The following steps are involved: S1. Receive data from the radar in the millimeter-wave radar cluster to obtain a mixed echo signal data set; S2, using wavelet transform and short-time Fourier transform, combined with the mixed echo signal data set, to obtain the denoised time-frequency feature matrix set; Classify the reflection peaks in the denoised time-frequency feature matrix and combine them with the reflection peak amplitude threshold to obtain the reflection peak delay matrix; S3. Calculating an ice layer thickness set and an air layer thickness set according to the reflection peak time delay matrix; Obtaining a water level elevation set based on the ice layer thickness set and the air layer thickness set in combination with an adaptive Kalman filter algorithm; S4. Using the millimeter wave propagation velocity set in water corrected by the sediment content compensation algorithm and the reflection peak delay difference set between the riverbed and the water surface corrected by the water flow velocity set, combined with the water level elevation set, a water depth dataset and a riverbed elevation dataset are obtained; S5. Use the optimized three-dimensional interpolation algorithm, combined with the water depth dataset and riverbed elevation dataset, to obtain a continuous water depth distribution map of the cross section.

2. The method for continuous measurement of cross-sectional water depth based on millimeter waves according to claim 1, characterized in that: Said S1 comprises the following steps: S11. Uniformly deploying a millimeter-wave radar array along a cross section of the water area to obtain a millimeter-wave radar set; the millimeter-wave radar set includes all radars deployed in the cross section of the water area; the radars in the millimeter-wave radar set must be installed at a height that covers the dynamic range of ice and water surface changes; Configured with multiple-input multiple-output MIMO antennas to cover the target cross-sectional area through multi-angle beam scanning; S12, each radar in the millimeter wave radar set transmits an FMCW signal to the target area, which penetrates the ice layer, air layer and water body and then returns to the radar; Each radar in the millimeter-wave radar set receives an echo signal reflected from the target area, generating a mixed echo signal dataset. The mixed echo signal dataset contains mixed echo signal data received by each millimeter-wave radar. The mixed echo signal data includes reflection components from ice layers, air interlayers, water bodies, and riverbeds.

3. The method for continuous measurement of cross-sectional water depth based on millimeter waves according to claim 1, characterized in that: The S2 comprises the following steps: S21, performing denoising processing on the mixed echo signal data in the mixed echo signal data set using wavelet transform to obtain a denoised mixed echo signal data set; The signal delay and frequency characteristics of the denoised mixed echo signal data set are extracted by short-time Fourier transform to obtain the denoised time-frequency feature matrix set; S22. Using a Gaussian mixture model, the reflection peaks of each medium in the denoised time-frequency feature matrix are classified to obtain a reflection peak classification matrix; the reflection peak classification matrix includes the reflection peaks of all ice layers, air interlayers, water bodies, and riverbeds monitored by the millimeter-wave radar; The reflection peak amplitude threshold is set based on the difference in dielectric constant between ice and water. A reflection peak delay matrix is obtained according to the reflection peak classification matrix and the reflection peak amplitude threshold; the reflection peak delay matrix includes the reflection peak delays of the upper surface, lower surface, water surface and riverbed of the ice layer monitored by all millimeter wave radars.

4. The method for continuous measurement of cross-sectional water depth based on millimeter waves according to claim 1, characterized in that: The S3 includes the following steps: S31. Calculate an ice thickness set based on the time delay difference of the reflection peaks between the upper and lower surfaces of the ice layer and the propagation speed of millimeter waves in the ice; the ice thickness set includes the ice thicknesses monitored by all millimeter wave radars; Based on the reflection peak delay difference from the lower surface of the ice layer to the water surface, combined with the propagation speed of millimeter waves in the air interlayer, the air interlayer thickness set is calculated; the air interlayer thickness set includes the air interlayer thickness monitored by all millimeter wave radars; S32. Calculate an initial water level elevation set based on the ice layer thickness set, the air layer thickness set, and the radar installation points; the water level elevation set includes all water level elevations monitored by the millimeter wave radar; S33. Adaptive Kalman filter algorithm is used to dynamically correct the measurement errors caused by ice movement or water flow fluctuations to obtain the water level elevation set.

5. The method for continuous measurement of cross-sectional water depth based on millimeter waves according to claim 1, characterized in that: The S4 comprises the following steps: S41. Using a dynamic focusing algorithm combined with MIMO radar synthesis, obtain a water surface reflection signal set and a riverbed reflection signal set; and obtaining a reflection peak delay difference set between the riverbed and the water surface based on the water surface reflection signal set and the riverbed reflection signal set; S42. Correcting the millimeter wave propagation velocity in water using a sediment content compensation algorithm to obtain a corrected millimeter wave propagation velocity set in water; correcting the reflection peak delay difference between the riverbed and the water surface using a flow velocity compensation algorithm to obtain a corrected reflection peak delay difference set between the riverbed and the water surface; S43. Obtaining a water depth data set using a water depth calculation formula based on the corrected millimeter wave propagation velocity set in water and the corrected reflection peak delay difference set between the riverbed and the water surface; the water depth data set includes water depth data monitored by all millimeter wave radars; The riverbed elevation set is obtained by combining the water level elevation set with the water depth dataset.

6. The method for continuous measurement of cross-sectional water depth based on millimeter waves according to claim 5, characterized in that: The S41 includes the following steps: S411, scan the cross section at multiple angles using a MIMO radar array to obtain high spatial resolution echo data; S412. Obtain the current beam pointing angle to be tracked based on the water surface ripples; adjust the beam weight based on the beam pointing angle in combination with a dynamic focusing algorithm to obtain a real-time beam weight set; S413, using the real-time beam weight set in combination with the MIMO virtual array synthesis to obtain a water surface reflection signal set and a riverbed reflection signal set; According to the water surface reflection signal set and the riverbed reflection signal set, a reflection peak delay difference set between the riverbed and the water surface is obtained.

7. The method for continuous measurement of cross-sectional water depth based on millimeter waves according to claim 5, characterized in that: The S42 includes the following steps: S421. Collect the sediment concentration of each radar monitoring area to obtain a real-time sediment concentration set; based on the real-time sediment concentration set, use a sediment concentration compensation algorithm to correct the millimeter wave propagation velocity in water to obtain a corrected millimeter wave propagation velocity set in water; S422. Correct the reflection peak delay difference between the riverbed and the water surface according to the real-time water velocity set to obtain a corrected reflection peak delay difference set between the riverbed and the water surface.

8. The method for continuous measurement of cross-sectional water depth based on millimeter waves according to claim 1, characterized in that: The S5 comprises the following steps: S51. Optimizing the dynamic weight coefficient in the three-dimensional interpolation algorithm using an ant colony algorithm to obtain an optimal solution, and using the optimal solution as the dynamic weight coefficient of the three-dimensional interpolation algorithm to obtain an optimized three-dimensional interpolation algorithm; S52. Use the optimized three-dimensional interpolation algorithm, combined with the water depth data set and the riverbed elevation set, to obtain a continuous cross-sectional water depth distribution map.

9. The method for continuous measurement of cross-sectional water depth based on millimeter waves according to claim 8, characterized in that: The S51 includes the following steps: S511, build an ant population and set the size of the ant population; maximum number of optimization iterations; S512. Randomly set the ant population pheromone concentration according to the dynamic weight coefficient in the three-dimensional interpolation algorithm to obtain an ant population pheromone concentration set; S513, setting an MSE that directly measures interpolation accuracy; and defining a fitness function of the ant pheromone concentration in the ant population based on the MSE; S514, performing an iterative operation on the ant population pheromone concentration set; during each round of iteration, calculating the fitness value of each pheromone concentration in the ant population pheromone concentration set according to the fitness function, updating the pheromone concentration of each ant in the ant population initial position set, and obtaining the optimal individual ant pheromone concentration in the ant population and the global optimal ant pheromone concentration during each round of iteration; S515, repeating 514, when the maximum number of optimization iterations is reached, stopping the iteration, and taking the global optimal ant pheromone concentration as the optimal solution.

10. The millimeter wave-based cross-sectional water depth continuous measurement system is characterized by: A method for continuous measurement of cross-sectional water depth based on millimeter waves as described in any one of claims 1 to 9 is implemented, wherein the system includes a mixed echo signal generation module, a signal denoising and time-frequency feature classification module, an ice thickness and water level dynamic calculation module, a water depth and riverbed elevation compensation correction module, and a three-dimensional interpolation and cross-sectional terrain modeling module.

Images