Method and system for continuously measuring water depth of cross section based on millimeter waves

Abstract

The invention discloses a cross section water depth continuous measurement method and system based on millimeter waves, and relates to the technical field of hydrographic measurement. According to the invention, through a millimeter wave radar array and a multiple-input-multiple-output antenna, a frequency modulation continuous wave signal is transmitted and a mixed echo signal is received; extracting time-frequency characteristics by using wavelet transform and short-time Fourier transform, and classifying reflection peaks in combination with a Gaussian mixture model; calculating the thickness of an ice layer, the thickness of an air interlayer and an initial water level based on the delay inequality, and dynamically correcting an error through adaptive Kalman filtering; a dynamic focusing algorithm is adopted to synthesize water surface and riverbed reflection signals, and a sand content compensation and flow velocity correction model is introduced to optimize water depth and riverbed elevation inversion; the three-dimensional interpolation weight is optimized through an ant population algorithm, and a high-precision cross section continuous water depth distribution diagram is generated; according to the invention, the problems of poor penetrability, weak anti-interference capability and low continuous water depth precision are solved, and full-section dynamic monitoring of the ice layer thickness, the water level, the water depth and the riverbed form is realized.

Description

Technical Field

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

[0002] In the water area management of cold regions, the monitoring of hydrological parameters (such as ice thickness, sub-ice water level, water depth, and riverbed morphology) caused by ice cover is a core challenge for flood control, water resource scheduling, and ecological protection. Traditional methods rely on manual ice-breaking measurement or fixed sensor networks, which have problems such as high safety risks, lagging data updates, and insufficient spatial coverage. With the development of millimeter-wave radar technology, its high penetration and anti-interference capabilities provide a new approach for sub-ice hydrological monitoring. However, existing millimeter-wave radar systems still have technical bottlenecks in multi-media penetration signal separation, dynamic environment error compensation, and terrain modeling accuracy.

[0003] There are problems in the prior art such as poor ice layer environment adaptability, large dynamic environment errors, great limitations in single-point measurement, and low continuous water depth accuracy. Summary of the Invention

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

[0005] To solve the above technical problems, the present invention is achieved through the following technical solutions: S1. Receive data through the radar in the millimeter-wave radar concentration to obtain a mixed echo signal data set; S2. Use wavelet transform and short-time Fourier transform, combined with the mixed echo signal data set, to obtain a denoised time-frequency feature matrix set; Classify the reflection peaks in the denoised time-frequency feature matrix, and combine with the reflection peak amplitude threshold to obtain a reflection peak time delay matrix; S3. Calculate an ice thickness set and an air interlayer thickness set according to the reflection peak time delay matrix; Obtain a water level elevation set according to the ice thickness set, air interlayer thickness set, and combined with an adaptive Kalman filtering algorithm; S4. Use the water-borne millimeter wave propagation speed set corrected by the sediment concentration compensation algorithm and the reflection peak time delay difference set between the riverbed and the water surface corrected by the water flow speed set, and combine with the water level elevation set to obtain a water depth data set and a riverbed elevation data set; S5. Use a three-dimensional interpolation algorithm optimized by an optimization algorithm, combined with the water depth data set and the riverbed elevation set, to obtain a cross-sectional continuous water depth distribution map; The present invention realizes high-precision dynamic monitoring of ice layer thickness, air interlayer, water level elevation, water depth and riverbed morphology of the cross-section of water area during the ice period by deploying a millimeter-wave radar array and MIMO antenna technology, combining wavelet transform - short-time Fourier transform, adaptive Kalman filter dynamic correction, and dynamic focusing beam synthesis and sediment concentration - flow velocity joint compensation algorithm, and optimizing three-dimensional interpolation using the ant colony algorithm; overcomes the deficiencies of traditional technologies such as insufficient penetration, sensitivity to environmental interference, and rough terrain modeling, and provides an automated and highly reliable solution for the full-section real-time perception and visualization of sub-ice hydrological parameters.

[0006] Preferably, the S1 includes the following steps: S11. Uniformly deploy a millimeter-wave radar array along the cross-section of the water area to obtain a millimeter-wave radar set; the millimeter-wave radar set contains all the radars deployed on the cross-section of the water area; the installation height of the radars in the millimeter-wave radar set needs to cover the dynamic change range of the ice layer and the water surface; Configure a multiple-input multiple-output (MIMO) antenna to cover the target cross-section area through multi-angle beam scanning; S12. Each radar in the millimeter-wave radar set emits an FMCW signal to the target area, and after penetrating the ice layer, air interlayer and water body, it returns to the radar; Each radar in the millimeter-wave radar set receives the echo signal reflected from the target area to obtain a mixed echo signal data set; the mixed echo signal data set contains the mixed echo signal data received by each millimeter-wave radar; the mixed echo signal data contains the reflection components of the ice layer, air interlayer, water body and riverbed; The present invention uniformly deploys a millimeter-wave radar array on the cross-section of the water area, whose installation height covers the dynamic change range of the ice layer and the water surface, and configures a MIMO antenna for multi-angle beam scanning; each radar emits an FMCW signal to penetrate the ice layer, air interlayer and water body and then receives the mixed echo signal, forming a data set containing the reflection components of the ice layer, air, water body and riverbed, laying a data foundation for multi-medium parameter extraction.

[0007] Preferably, the S2 includes the following steps: S21. Use wavelet transform to denoise the mixed echo signal data in the mixed echo signal data set to obtain a denoised mixed echo signal data set; Extract the signal delay and frequency characteristics of the denoised mixed echo signal data set through short-time Fourier transform to obtain a denoised time-frequency characteristic matrix set; S22. Use the Gaussian mixture model to classify the reflection peaks of each medium in the denoised time-frequency characteristic matrix set to obtain a reflection peak classification matrix; 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; Based on the difference in dielectric constants between ice layers and water bodies, a reflection peak amplitude threshold is set; According to the reflection peak classification matrix and the reflection peak amplitude threshold, a reflection peak time delay matrix is obtained; the reflection peak time delay matrix includes the reflection peak time delays of the upper surface, lower surface, water surface, and riverbed of the ice layer monitored by all millimeter-wave radars; In the present invention, wavelet transform is used to denoise the mixed echo signal, and the short-time Fourier transform is combined to extract the time-frequency feature matrix; the Gaussian mixture model is used to classify the reflection peaks of the ice layer, air interlayer, water body, and riverbed, and an amplitude threshold is set based on the dielectric constant difference to screen the effective reflection peak time delays, generating a matrix containing the time delays of each medium interface, providing accurate data support for the calculation of parameters such as ice thickness and water level.

[0008] Preferably, S3 includes the following steps: S31. According to the reflection peak time delay difference between the upper surface and the lower surface of the ice layer, combined with the propagation speed of millimeter waves in ice, an ice layer thickness set is calculated; the ice layer thickness set includes the ice layer thicknesses monitored by all millimeter-wave radars; According to the reflection peak time 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, an air interlayer thickness set is calculated; the air interlayer thickness set includes the air interlayer thicknesses monitored by all millimeter-wave radars; S32. According to the ice layer thickness set, the air interlayer thickness set, and the radar installation point, an initial water level elevation set is calculated; the water level elevation set contains the water level elevations monitored by all millimeter-wave radars; S33. An adaptive Kalman filtering algorithm is used to dynamically correct the measurement errors caused by ice layer movement or water flow fluctuations to obtain the water level elevation set; The present invention calculates the ice thickness and the air layer thickness based on the reflection peak time delay differences between the upper and lower surfaces of the ice layer and the air interlayer, combines the radar installation height to estimate the initial water level, and 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.

[0009] Preferably, S4 includes the following steps: S41. The dynamic focusing algorithm is used in combination with MIMO radar 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 time delay difference set between the riverbed and the water surface is obtained; S42. The sediment concentration compensation algorithm is used to correct the propagation speed of millimeter waves in water to obtain a corrected propagation speed set of millimeter waves in water; the flow velocity compensation algorithm is used to correct the reflection peak time delay difference between the riverbed and the water surface to obtain a corrected reflection peak time delay difference set between the riverbed and the water surface; S43. Obtain a water depth data set through a water depth calculation formula based on the corrected underwater millimeter wave propagation speed set and the corrected reflection peak time delay difference set between the riverbed and the water surface; the water depth data set contains all the water depth data monitored by the millimeter wave radar. Obtain a riverbed elevation set based on the water level elevation set in combination with the water depth data set. The present invention synthesizes the water surface and riverbed reflection signals through a dynamic focusing algorithm and a MIMO radar to obtain the time delay difference between them; corrects the underwater millimeter wave propagation speed based on sediment concentration compensation, combines flow velocity compensation to optimize the time delay difference, and finally calculates an accurate water depth data set and a riverbed elevation set through corrected parameters, effectively eliminating the influence of sediment and water flow on the monitoring accuracy.

[0010] Preferably, S41 includes the following steps: S411. Perform multi-angle scanning on the cross-section through a MIMO radar array to obtain high-spatial-resolution echo data. S412. Obtain the beam pointing angle that needs to be tracked currently according to the water surface ripples; adjust the beam weights according to the beam pointing angle in combination with the dynamic focusing algorithm to obtain a real-time beam weight set. S413. Use the real-time beam weight set in combination with MIMO virtual array synthesis to obtain a water surface reflection signal set and a riverbed reflection signal set. Obtain a reflection peak time delay difference set between the riverbed and the water surface according to the water surface reflection signal set and the riverbed reflection signal set. The present invention obtains high-resolution echo data through multi-angle scanning of a MIMO radar array, dynamically adjusts the beam pointing angle and weights based on the water surface ripples, combines virtual array synthesis to separate the water surface and riverbed reflection signals, accurately extracts the reflection peak time delay difference between them, adapts to dynamic water surface fluctuations and improves the underwater target detection accuracy. Preferably, S42 includes the following steps: S421. Collect the sediment concentration in 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 underwater millimeter wave propagation speed to obtain a corrected underwater millimeter wave propagation speed set. S422. Correct the reflection peak time delay difference between the riverbed and the water surface according to the real-time water flow velocity set to obtain a corrected reflection peak time delay difference set between the riverbed and the water surface. The present invention corrects the underwater millimeter wave propagation speed through real-time sediment concentration data and compensates the reflection peak time delay difference based on the water flow velocity, eliminates the interference of sediment concentration and flow velocity on signal propagation, significantly improves the inversion accuracy of water depth and riverbed elevation, and ensures the measurement reliability in a dynamic hydrological environment.

[0011] Preferably, S5 includes the following steps: S51. Optimize the dynamic weight coefficient in the three-dimensional interpolation algorithm using the ant colony algorithm to obtain the optimal solution, and use the optimal solution as the dynamic weight coefficient of the three-dimensional interpolation algorithm to obtain the 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 the continuous water depth distribution map of the cross-section; In the present invention, the dynamic weight coefficient of three-dimensional interpolation is optimized by the ant colony algorithm to minimize the interpolation error, and high-precision continuous cross-section distribution maps are generated by combining water depth and riverbed elevation data, solving the problem of terrain distortion under sparse monitoring points and improving the spatial representation ability of the sub-ice water body and riverbed morphology.

[0012] Preferably, the S51 includes the following steps: S511. Construct an ant colony, set the scale of the ant colony; set the maximum optimization iteration times; S512. According to the dynamic weight coefficient in the three-dimensional interpolation algorithm, randomly set the pheromone concentration of the ant colony to obtain the ant colony pheromone concentration set; S513. Set the MSE that directly measures the interpolation accuracy; according to the MSE, define the fitness function of the pheromone concentration of the ants in the ant colony; S514. Perform iterative operations on the ant colony pheromone concentration set. The higher the fitness value, the stronger the pheromone concentration; in each round of iteration, calculate the fitness value of each pheromone concentration in the ant colony pheromone concentration set according to the fitness function, and update the pheromone concentration of each ant in the ant colony initial position set from high to low according to the fitness value. And in each round of iteration, obtain the best ant individual pheromone concentration and the global best ant pheromone concentration in the ant colony; S515. Repeat 514. When the maximum optimization iteration times is reached, stop the iteration, and use the global best ant pheromone concentration as the optimal solution; In the present invention, an ant colony is constructed and the pheromone concentration is randomly initialized, the fitness function is defined by MSE, and the pheromone concentration is iteratively updated to screen the global optimal pheromone concentration as the dynamic weight coefficient of three-dimensional interpolation, realizing high-precision terrain interpolation under sparse monitoring points and effectively suppressing the distortion of discrete data interpolation.

[0013] The cross-section water depth continuous measurement system based on millimeter wave, used to implement the above-mentioned cross-section water depth continuous measurement method based on millimeter wave, includes a mixed echo signal generation module, a signal denoising and time-frequency feature classification module, an ice layer thickness and water level dynamic calculation module, a water depth and riverbed elevation compensation and correction module, and a three-dimensional interpolation and cross-section terrain modeling module; The mixed echo signal generation module is used to deploy a millimeter-wave radar array across the water cross-section, configure a multi-input multi-output antenna for multi-angle beam scanning, transmit a frequency-modulated continuous wave signal to the target area, and receive the mixed 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 the water surface to ensure 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 mixed echo signal dataset to provide the raw data basis for subsequent processing. The signal denoising and time-frequency feature classification module is used to denoise the mixed echo signal by using wavelet transform, extract the time-frequency features of the signal in combination with the short-time Fourier transform, and generate a time-frequency feature matrix. The Gaussian mixture model is used to classify the reflection peaks in the time-frequency matrix to distinguish the reflection components of the ice layer, air layer, water body, and riverbed, and the effective signals are screened by a preset reflection peak amplitude threshold. Finally, a reflection peak delay matrix is generated to provide key parameters for calculating the ice thickness and water level. 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 layer based on the reflection peak delay matrix, and combines the propagation speed of millimeter waves in different media to obtain the ice thickness and air layer thickness respectively. The height of the radar installation point and the adaptive Kalman filter algorithm are used to dynamically correct the errors caused by the movement of the ice layer or the fluctuation of the water flow, generating a high-precision water level elevation dataset to realize the real-time monitoring of the sub-ice water level. The water depth and riverbed elevation compensation and correction module adjusts the MIMO radar beam weight through the dynamic focusing algorithm to synthesize the high-resolution reflection signals of the water surface and the riverbed, and obtains the time delay difference between their reflection peaks. The sediment concentration compensation algorithm is introduced to correct the propagation speed of millimeter waves in water, and the time delay difference is corrected twice in combination with the water flow speed inversed by the Doppler frequency shift. Finally, in combination with the water level elevation data, a water depth and riverbed elevation dataset affected by sediment concentration and flow velocity is calculated. The three-dimensional interpolation and cross-section terrain modeling module uses the ant colony algorithm to optimize the dynamic weight coefficient of the three-dimensional interpolation algorithm, and iteratively solves the optimal interpolation parameters with the mean square error as the fitness function. Based on the optimized algorithm, spatial interpolation is performed on the discrete water depth and riverbed elevation data to generate a continuous underwater cross-section terrain distribution map, visually displaying the three-dimensional structural characteristics of the ice layer, air layer, water body, and riverbed, providing visual support for the monitoring of the ice period in the water area.

[0014] Beneficial effects The present invention has the following beneficial effects: The present invention realizes high-precision dynamic monitoring of ice layer thickness, air interlayer, water level elevation, water depth, and riverbed morphology of the cross-section of water area during the ice period by deploying a millimeter-wave radar array and MIMO antenna technology, combining wavelet transform - short-time Fourier transform, adaptive Kalman filter dynamic correction, dynamic focusing beam synthesis, and sediment concentration - flow velocity joint compensation algorithm, and optimizing three-dimensional interpolation by using the ant colony algorithm; overcomes the defects of insufficient penetration, environmental interference sensitivity, and rough terrain modeling of traditional technologies, and provides an automated and highly reliable solution for the full-section real-time perception and visualization of sub-ice hydrological parameters.

[0015] Through the millimeter-wave radar array and MIMO antenna technology, the present invention realizes the penetration detection of ice layer, air interlayer, and water body, and solves the problem that traditional sensors cannot obtain sub-ice water level and riverbed data due to ice layer obstruction; the multi-angle beam scanning and high-resolution echo signal synthesis technology significantly improves the separation ability of reflection signals at the interface of the complex medium of ice - air - water body, ensuring the integrity and reliability of the mixed echo data.

[0016] The present invention combines the time-frequency feature extraction of wavelet transform - short-time Fourier transform and the Gaussian mixture model classification algorithm to effectively remove environmental noise interference, and realizes the accurate classification of reflection peaks based on the dielectric constant difference; the measurement error caused by ice layer movement and water flow fluctuation is dynamically corrected by adaptive Kalman filter, significantly improving the stability and accuracy of water level elevation.

[0017] Aiming at the millimeter-wave propagation speed offset and time delay error caused by water flow in sediment-laden water, the present invention introduces a sediment concentration compensation algorithm and Doppler frequency shift inversion technology to dynamically correct the calculation parameters of water depth and riverbed elevation; reduces the water depth error under the influence of sediment concentration, and improves the inversion accuracy of riverbed elevation after flow velocity compensation.

[0018] The present invention uses the ant colony algorithm to optimize the three-dimensional interpolation weight coefficient, solves the problem of terrain distortion of traditional interpolation algorithms under sparse monitoring points; generates the optimal interpolation parameters by minimizing the mean square error iteration, and the generated continuous water depth distribution map of the cross-section can accurately reflect the spatial structure characteristics of ice layer - air interlayer - water body - riverbed, providing intuitive visual support for the safety management and dispatching decision-making of water area during the ice period.

[0019] Of course, it is not necessary for any product implementing the present invention to achieve all the above-mentioned advantages simultaneously. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] To more clearly illustrate the technical solutions of the embodiments of the invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0021] Figure 1 It is a schematic flowchart of the method for continuously measuring the cross-sectional water depth based on millimeter waves of the present invention; Figure 2 It is a schematic flowchart of the method for obtaining the water depth data set and the riverbed elevation data set in the method for continuously measuring the cross-sectional water depth based on millimeter waves of the present invention; Figure 3 It is a schematic flowchart of the method for obtaining the optimized three-dimensional interpolation algorithm in the method for continuously measuring the cross-sectional water depth based on millimeter waves of the present invention; Figure 4 It is a schematic diagram of the modules of the system for continuously measuring the cross-sectional water depth based on millimeter waves of the present invention. Detailed implementation manners

[0022] The following will clearly and completely describe the technical solutions in the embodiments of the invention with reference to the drawings in the embodiments of the invention. Obviously, the described embodiments are only some embodiments of the invention, rather than all embodiments. Based on the embodiments in the invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the invention.

[0023] In the description of the present invention, it should be understood that the terms "opening", "upper", "lower", "top", "middle", "inner", etc. indicating the orientation or position relationship are only for the convenience of describing the invention and simplifying the description, rather than indicating or implying that the components or elements referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation to the invention.

[0024] Embodiment 1 Please refer to Figure 1 , Figure 2 , Figure 3 , the present invention discloses a method and system for continuously measuring the cross-sectional water depth based on millimeter waves, including the following steps: S1. Receive data through the radars in the millimeter wave radar set to obtain a mixed echo signal data set; The S1 includes the following steps: S11. Uniformly deploy a millimeter wave radar array along the water cross-section to obtain a millimeter wave radar set h = {h1, h2,..., h i , …, h g}; where h iThe i-th radar representing the cross-section, and g represents the total number of radars in the water cross-section; the millimeter-wave radar set includes all the radars deployed in the cross-section; the installation height of the radars in the millimeter-wave radar set needs to cover the dynamic change range of the ice layer and the water surface; Configure a multi-input multi-output (MIMO) antenna to cover the target cross-section area through multi-angle beam scanning; S12. Each radar in the millimeter-wave radar set emits an FMCW signal to the target area, and after penetrating the ice layer, air layer and water body, it returns to the radar; Each radar in the millimeter-wave radar set receives the echo signal reflected by the target area to obtain a mixed echo signal data set p = {p1, p2,..., p i ,…, p g}; where p i represents 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 contains the reflection components of the ice layer, air layer, water body and riverbed; S2. Use wavelet transform and short-time Fourier transform, combined with the mixed echo signal data set, to obtain a denoised time-frequency feature matrix set; Classify the reflection peaks in the denoised time-frequency feature matrix, and combine the reflection peak amplitude threshold to obtain a reflection peak delay matrix; The S2 includes the following steps: S21. Use wavelet transform to denoise 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, ; where a represents the scale parameter, b represents the displacement parameter, p i represents the relaxation echo signal, and W(a, b) represents the wavelet transform result of the mixed echo signal p i at the scale parameter a and the position parameter b, represents the conjugate wavelet function; Extract the signal delay and frequency characteristics of the denoised mixed echo signal data set through short-time Fourier transform to obtain a denoised time-frequency feature matrix set; S22. Use the 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, and the reflection peak classification matrix is as follows, ; where C i1 , C i2 , C i3 and C i4respectively 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; The formula of the Gaussian mixture model (GMM) is as follows, ; where y represents the eigenvector in the time-frequency feature matrix, p(y) represents the probability density of the eigenvector y in the time-frequency feature matrix, K represents the Gaussian components (i.e., the number of types of media in the time-frequency feature matrix), k|k = 1, 2, 3, 4, π k represents the mixing coefficient of the k-th Gaussian component (i.e., represents the prior probability of the k-th type of media in the signal), β k represents the eigen mean vector of the k-th type of media; represents the probability density function of the multivariate Gaussian distribution (i.e., describes the characteristic distribution of the k-th type of media); Based on the difference in dielectric constants between the ice layer and the water body, set the reflection peak amplitude threshold; According to the reflection peak classification matrix and the reflection peak amplitude threshold, obtain the reflection peak delay matrix D. The reflection peak delay matrix is as follows, ; where D i1 , D i2 , D i3 and D i4 respectively 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; 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; S3. Calculate the ice layer thickness set and the air interlayer thickness set according to the reflection peak delay matrix; According to the ice layer thickness set, the air interlayer thickness set and combined with the adaptive Kalman filtering algorithm, obtain the water level elevation set; The S3 includes the following steps: S31. According to the reflection peak delay difference between the upper surface and the lower surface of the ice layer, combined with the propagation speed of millimeter waves in ice, calculate the ice layer thickness set; the ice layer thickness set includes the ice layer thicknesses monitored by all millimeter-wave radars. The calculation formula is as follows, ; where H ice represents the ice layer thickness, v ice represents the propagation speed of millimeter waves in ice, and d1 and d2 respectively represent the reflection peak delay difference between the upper surface and the lower surface of the ice layer; According to the time delay difference of the reflection peak 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 all the air interlayer thicknesses monitored by the millimeter wave radar; the calculation formula is as follows, ; where H air represents the air interlayer thickness, v air represents the propagation speed of millimeter waves in the air interlayer, d2 and d3 respectively represent the time delay difference of the reflection peak of the lower surface and the time delay difference of the reflection peak of the water surface; S32. According to the ice layer thickness set, the air interlayer thickness set and the radar installation point, the initial water level elevation set is calculated; the water level elevation set includes all the water level elevations monitored by the millimeter wave radar; the calculation formula is as follows, ; where H water represents the water level elevation, H r represents the height of the radar installation point relative to the reference plane (such as the horizontal plane); S33. The adaptive Kalman filtering algorithm is used to dynamically correct the measurement errors caused by ice layer movement or water flow fluctuations to obtain the water level elevation set; S4. The reflection peak time delay difference set of the river bed and the water surface corrected by the set of millimeter wave propagation speeds in water and the set of water flow speeds corrected by the sediment concentration compensation algorithm are used, and combined with the water level elevation set to obtain the water depth data set and the river bed elevation data set; The above S4 includes the following steps: S41. The dynamic focusing algorithm is used in combination with MIMO radar synthesis to obtain the water surface reflection signal set and the river bed reflection signal set; according to the water surface reflection signal set and the river bed reflection signal set, the reflection peak time delay difference set of the river bed and the water surface is obtained; The above S41 includes the following steps: S411. The cross section is scanned at multiple angles through the MIMO radar array to obtain high spatial resolution echo data; S412. According to the water surface ripples, the beam pointing angle that needs to be tracked currently is obtained; according to the beam pointing angle and combined with the dynamic focusing algorithm, the beam weights are adjusted to obtain the real-time beam weight set; the formula of the dynamic focusing algorithm is as follows, ; where w represents the real-time beam weight; argmin represents the minimum argument (that is, finding the H vector that makes || e(θ) - w w z|| reach the minimum value); θ represents the beam pointing angle; e(θ) represents the steering vector dimension (that is, the phase delay characteristic of the beam in the direction of angle θ); w Hdenotes the conjugate transpose of the beam weight vector, which is used for beam synthesis, and z denotes the received signal matrix; S413. Use the real-time beam weight set to combine with MIMO virtual array synthesis to obtain the water surface reflection signal set and the riverbed reflection signal set; According to the water surface reflection signal set and the riverbed reflection signal set, obtain the reflection peak time delay difference set between the riverbed and the water surface; S42. Use the sediment concentration compensation algorithm to correct the propagation speed of millimeter waves in water to obtain the corrected propagation speed set of millimeter waves in water; use the flow velocity compensation algorithm to correct the reflection peak time delay difference between the riverbed and the water surface to obtain the corrected reflection peak time delay difference set between the riverbed and the water surface; The S42 includes the following steps: S421. Collect the sediment concentration of each radar monitoring area to obtain the real-time sediment concentration set; based on the real-time sediment concentration set, use the sediment concentration compensation algorithm to correct the propagation speed of millimeter waves in water to obtain the corrected propagation speed set of millimeter waves in water; the formula of the sediment concentration compensation algorithm is as follows, ; where, v water denotes the corrected propagation speed of millimeter waves in water, v0 denotes the propagation speed of millimeter waves in water under the standard sediment concentration, δ denotes the set sediment concentration compensation, and S denotes the real-time sediment concentration; Invert the water surface flow velocity through the Doppler frequency shift to obtain the real-time water flow velocity set; the inversion formula is as follows, ; where, v flow denotes the real-time water flow velocity, f denotes the Doppler frequency shift (i.e., the frequency shift of the echo signal caused by the water surface flow velocity), λ denotes the radar wavelength, and θ denotes the radar wave incident angle; S422. Correct the reflection peak time delay difference between the riverbed and the water surface according to the real-time water flow velocity set to obtain the corrected reflection peak time delay difference set between the riverbed and the water surface; the correction formula is as follows, ; where, j denotes the corrected reflection peak time delay difference between the riverbed and the water surface, and η denotes the set flow velocity reflection peak time delay error correction coefficient; S43. According to the corrected propagation speed set of millimeter waves in water and the corrected reflection peak time delay difference set between the riverbed and the water surface, through the water depth calculation formula, obtain the water depth data set; all the water depth data monitored by the millimeter wave radar are included in the water depth data set; the water depth calculation formula is as follows, ; where, G denotes the water depth data monitored by the millimeter wave radar; According to the water level elevation set combined with the water depth data set, obtain the riverbed elevation set; the calculation formula is as follows, ; Among them, H bed represents the riverbed elevation; S5. The three-dimensional interpolation algorithm optimized by using an optimization algorithm, combined with the water depth data set and the riverbed elevation set, to obtain a continuous cross-sectional water depth distribution map; The S5 includes the following steps: S51. Use the ant colony algorithm to optimize the dynamic weight coefficient in the three-dimensional interpolation algorithm to obtain the optimal solution, and use the optimal solution as the dynamic weight coefficient of the three-dimensional interpolation algorithm to obtain the optimized three-dimensional interpolation algorithm; The S51 includes the following steps: S511. Construct an ant colony, and set the scale of the ant colony as u, then the ant colony is represented as u = {u1, u2,..., u i ,..., u q}, where u i represents the i-th ant in the ant colony; set the maximum optimization iteration number; S512. According to the dynamic weight coefficient in the three-dimensional interpolation algorithm, randomly set the pheromone concentration of the ant colony to obtain the ant colony pheromone concentration set l = {l1, l2,..., l i ,..., l q}, where l i represents the pheromone concentration of the i-th ant in the ant colony; S513. Set MSE, and MSE is used to directly measure the interpolation accuracy; according to the MSE, define the fitness function of the ant pheromone concentration in the ant colony, and the fitness function formula is as follows, ; R represents the fitness function; S514. Perform iterative operations on the ant colony pheromone concentration set. The higher the fitness value, the stronger the pheromone concentration; in each round of iteration, calculate the fitness value of each pheromone concentration in the ant colony pheromone concentration set according to the fitness function, and update the pheromone concentration of each ant in the ant colony initial position set from high to low according to the fitness value, and obtain the best ant individual pheromone concentration and the global best ant pheromone concentration in the ant colony in each round of iteration; S515. Repeat 514. When the maximum optimization iteration number is reached, stop the iteration, and use the global best ant pheromone concentration as the optimal solution; 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.

[0025] Example 2 Please refer toFigure 4 , a millimeter-wave based cross-sectional water depth continuous measurement system for implementing 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 layer thickness and water level dynamic calculation module, a water depth and riverbed elevation compensation and correction module, and a three-dimensional interpolation and cross-sectional terrain modeling module; The mixed echo signal generation module is used to deploy a millimeter-wave radar array across the water area, configure multiple-input multiple-output antennas for multi-angle beam scanning, transmit a frequency-modulated continuous wave signal to the target area, and receive the mixed echo signal after penetrating the ice layer, air layer, and water body; the radar installation height covers the dynamic change range of the ice layer and the water surface to ensure 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 mixed echo signal dataset to provide the original data basis for subsequent processing; The signal denoising and time-frequency feature classification module is used to perform denoising processing on the mixed echo signal using wavelet transform, extract the time-frequency features of the signal in combination with the 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 to distinguish the reflection components of the ice layer, air layer, water body, and riverbed, and screen out the effective signals through a preset reflection peak amplitude threshold, and finally generate a reflection peak time delay matrix to provide key parameters for calculating the ice layer thickness and water level; The ice layer 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 layer based on the reflection peak time delay matrix, and combines the propagation speeds of millimeter waves in different media to obtain the ice layer thickness and air layer thickness respectively; uses the radar installation point height and the adaptive Kalman filtering algorithm to dynamically correct the errors caused by the movement of the ice layer or the fluctuation of the water flow, generates a high-precision water level elevation dataset, and realizes the real-time monitoring of the sub-ice water level; The water depth and riverbed elevation compensation and correction module adjusts the MIMO radar beam weights through the dynamic focusing algorithm, synthesizes the high-resolution reflection signals of the water surface and the riverbed, and obtains the reflection peak time delay difference between the two; introduces a sediment concentration compensation algorithm to correct the propagation speed of millimeter waves in water, and combines the water flow speed inversed by the Doppler frequency shift to perform secondary correction on the time delay difference; finally, combines the water level elevation data to calculate the water depth and riverbed elevation dataset affected by sediment concentration and flow velocity; The three-dimensional interpolation and cross-sectional terrain modeling module uses the ant colony algorithm to optimize the dynamic weight coefficient of the three-dimensional interpolation algorithm, and uses the mean square error as the fitness function to iteratively solve the optimal interpolation parameters; based on the optimized algorithm, performs spatial interpolation on the discrete water depth and riverbed elevation data to generate a continuous underwater cross-sectional terrain distribution map, intuitively showing the three-dimensional structural characteristics of the ice layer, air layer, water body, and riverbed, and providing visual support for ice period monitoring of the water area.

[0026] In the description of this specification, the descriptions referring to the terms "one embodiment", "example", "specific example", etc. mean that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0027] The preferred embodiments of the invention disclosed above are only used to assist in the explanation of the invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific embodiments described. Obviously, many modifications and variations can be made according to the content of this specification. These embodiments are selected and specifically described in this specification in order to better explain the principle and practical application of the invention, so that those skilled in the art can understand and utilize the invention well.

Claims

1. A millimeter-wave-based cross-sectional water depth continuous measurement method, characterized in that, It includes the following steps: S1. Receive data through the radars concentrated in the millimeter-wave radar to obtain a mixed echo signal dataset; S2. Use wavelet transform and short-time Fourier transform, combined with the mixed echo signal dataset, to obtain a denoised time-frequency feature matrix set; Classify the reflection peaks in the denoised time-frequency feature matrix, and combine with the reflection peak amplitude threshold to obtain a reflection peak time delay matrix; S3. Calculate the ice layer thickness set and the air interlayer thickness set according to the reflection peak time delay matrix; According to the ice layer thickness set, the air interlayer thickness set and combined with the adaptive Kalman filtering algorithm, obtain the water level elevation set; S4. Use the set of millimeter-wave propagation speeds in water corrected by the sediment concentration compensation algorithm and the set of reflection peak time delay differences between the riverbed and the water surface corrected by the water flow speed set, and combine with the water level elevation set to obtain the water depth dataset and the riverbed elevation dataset; S5. Use the three-dimensional interpolation algorithm optimized by the optimization algorithm, combined with the water depth dataset and the riverbed elevation set, to obtain a cross-sectional continuous water depth distribution map.

2. The cross-sectional water depth continuous measurement method based on millimeter wave according to claim 1, wherein The S1 includes the following steps: S11. Uniformly deploy a millimeter-wave radar array along the water cross-section to obtain a millimeter-wave radar set; the millimeter-wave radar set contains all the radars deployed on the water cross-section; the installation height of the radars in the millimeter-wave radar set needs to cover the dynamic change range of the ice layer and the water surface; Configure a multiple-input multiple-output (MIMO) antenna to cover the target cross-sectional area through multi-angle beam scanning; S12. Each radar in the millimeter-wave radar set emits an FMCW signal to the target area, penetrates the ice layer, the air interlayer and the water body, and then returns to the radar; Each radar in the millimeter-wave radar set receives the echo signal reflected from the target area to obtain a mixed echo signal dataset; the mixed echo signal dataset contains the mixed echo signal data received by each millimeter-wave radar; the mixed echo signal data contains the reflection components of the ice layer, the air interlayer, the water body and the riverbed.

3. The cross-sectional water depth continuous measurement method based on millimeter waves according to claim 1, characterized in that The S2 includes the following steps: S21. Use wavelet transform to denoise the mixed echo signal data in the mixed echo signal dataset to obtain a denoised mixed echo signal dataset; Extract the signal time delay and frequency characteristics of the denoised mixed echo signal dataset through short-time Fourier transform to obtain a denoised time-frequency feature matrix set; S22. Use the Gaussian mixture model to classify the reflection peaks of each medium in the denoised time-frequency feature matrix set to obtain a reflection peak classification matrix; the reflection peak classification matrix includes the reflection peaks of the ice layer, the air interlayer, the water body and the riverbed monitored by all millimeter-wave radars; Based on the dielectric constant difference between the ice layer and the water body, set the reflection peak amplitude threshold; According to the reflection peak classification matrix and the reflection peak amplitude threshold, obtain a reflection peak time delay matrix; the reflection peak time delay matrix includes the reflection peak time delays of the upper surface, the lower surface, the water surface and the riverbed of the ice layer monitored by all millimeter-wave radars.

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

5. The cross-sectional water depth continuous measurement method based on millimeter wave according to claim 1, characterized in that The S4 includes the following steps: S41. Use the dynamic focusing algorithm combined with MIMO radar synthesis to obtain the water surface reflection signal set and the riverbed reflection signal set; according to the water surface reflection signal set and the riverbed reflection signal set, obtain the time delay difference set of the reflection peaks between the riverbed and the water surface; S42. Use the sediment concentration compensation algorithm to correct the propagation speed of millimeter waves in water to obtain the corrected propagation speed set of millimeter waves in water; use the flow velocity compensation algorithm to correct the time delay difference of the reflection peaks between the riverbed and the water surface to obtain the corrected time delay difference set of the reflection peaks between the riverbed and the water surface; S43. According to the corrected propagation speed set of millimeter waves in water and the corrected time delay difference set of the reflection peaks between the riverbed and the water surface, through the water depth calculation formula, obtain the water depth data set; the water depth data set contains all the water depth data monitored by the millimeter wave radar; According to the water level elevation set combined with the water depth data set, obtain the riverbed elevation set.

6. The millimeter-wave-based cross-sectional water depth continuous measurement method according to claim 5, wherein The S41 includes the following steps: S411. Perform multi-angle scanning on the cross-section through the MIMO radar array to obtain high-spatial-resolution echo data; S412. According to the water surface ripples, obtain the beam pointing angle that needs to be tracked currently; according to the beam pointing angle combined with the dynamic focusing algorithm, adjust the beam weights to obtain the real-time beam weight set; S413. Use the real-time beam weight set combined with MIMO virtual array synthesis to obtain the water surface reflection signal set and the riverbed reflection signal set; According to the water surface reflection signal set and the riverbed reflection signal set, obtain the time delay difference set of the reflection peaks between the riverbed and the water surface.

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

8. The cross-sectional water depth continuous measurement method based on millimeter wave according to claim 1, characterized in that The S5 includes the following steps: S51. Use the ant colony algorithm to optimize the dynamic weight coefficient in the three-dimensional interpolation algorithm to obtain the optimal solution, and use the optimal solution as the dynamic weight coefficient of the three-dimensional interpolation algorithm to obtain the 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 the continuous water depth distribution map of the cross-section.

9. The cross-sectional water depth continuous measurement method based on millimeter wave according to claim 8, wherein, The S51 includes the following steps: S511. Construct an ant colony, and set the scale of the ant colony; the maximum optimization iteration times; S512. Randomly set the pheromone concentration of the ant population according to the dynamic weight coefficient in the three-dimensional interpolation algorithm to obtain the ant population pheromone concentration set; S513. Set the MSE that directly measures the interpolation accuracy; define the fitness function of the pheromone concentration of the ants in the ant population according to the MSE; S514. Perform iterative operations on the ant population pheromone concentration set; in each iteration process, 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, and obtain the best ant individual pheromone concentration and the global best ant pheromone concentration in the ant population in each iteration process; S515. Repeat S514. When the maximum optimization iteration number is reached, stop the iteration and use the global best ant pheromone concentration as the optimal solution.

10. A millimeter-wave-based cross-sectional water depth continuous measurement system, characterized in that, Implement the millimeter-wave-based cross-sectional water depth continuous measurement method according to any one of claims 1-9. 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 and correction module, and a three-dimensional interpolation and cross-sectional terrain modeling module.

Images