Multi-frequency acoustic Doppler velocity profile measurement system and method for high sediment content water bodies

Abstract

This invention provides a multi-frequency acoustic Doppler velocity profile measurement system and method for high-sediment-laden water bodies, relating to the field of hydrological monitoring technology. The system includes an IP68 flow chamber, which integrates a multi-frequency transducer array, a multi-frequency signal processing board, a main control and inversion unit, a turbidity sensor, a sediment concentration sensor, an RTK GNSS / IMU positioning unit, and an RTU data transmission unit. The method includes: time-division multiplexing of the multi-frequency transducer array for calculating the original horizontal velocity using Doppler velocity analysis; Thorne-Hanes acoustic attenuation compensation based on localized coefficient calibration; constructing a multi-objective joint optimization function to invert the SSC and sediment particle size distribution; outputting a corrected high-precision velocity profile through bottom tracking and RTK / IMU trajectory reconstruction; generating a global two-dimensional velocity field through discrete grid partitioning, calculating cross-sectional flow rate, and calculating suspended sediment flux. This application resolves the contradiction between penetration depth and resolution, eliminates the overestimation of coarse sand in the Yellow River, and achieves integrated synchronous output of the flow field, concentration field, and flux in high-sediment-laden water bodies.

Description

Technical Field

[0001] This invention relates to the field of hydrological monitoring technology, and more specifically, to a multi-frequency acoustic Doppler velocity profile measurement system and method for high sediment content water bodies. Background Technology

[0002] The Acoustic Doppler Current Profiler (ADCP) is an instrument that uses the acoustic Doppler effect to measure the velocity profile of water bodies. It emits sound pulses into the water body and receives the backscattered echoes from suspended particles. It uses the Doppler frequency shift principle to calculate the radial velocity of each layer of water along the direction of the sound beam, and then obtains the horizontal velocity vector through multi-beam geometric synthesis, thereby realizing non-contact measurement of the velocity distribution of the water body profile.

[0003] However, when ADCP is applied to high-sediment-laden water bodies (such as sections of the middle and lower reaches of the Yellow River where the sediment concentration reaches 10 to 100 kg / m³), existing technologies reveal serious systemic defects. Actual tests in the Xiaolangdi Reservoir area and downstream of the Yellow River revealed that ADCP (such as Teledyne Rio Grande 600 kHz) exhibits a sharp decrease in acoustic penetration depth when the sediment concentration is not less than 30 kg / m³. Furthermore, a significant "overestimation artifact" occurs when the echo intensity is used to invert the suspended sediment concentration (SSC). Specifically, when the measured SSC is 30 kg / m³, the acoustic inversion result is as high as 50 to 60 kg / m³. In-depth analysis revealed that the root cause lies in the Urick and Thorne-Hanes models used in international sediment literature to calculate acoustic attenuation coefficients. These models primarily calibrate their empirical attenuation coefficients based on fine sand samples (0.1 to 0.5 mm particle size), while the Yellow River sediment contains 30% to 50% coarse sand (0.5 to 2 mm particle size). The acoustic attenuation characteristics of these two types of sediment differ significantly, leading to a deviation of 35% to 45% in the inversion results. This problem is not apparent in conventional clear water or fine-grained rivers, but is concentrated in rivers with high sediment content and coarse sand.

[0004] From a technical perspective, the core challenge of single-frequency ADCP solutions in high-sediment-content water bodies lies in the fact that while higher operating frequencies provide higher velocity measurement resolution, acoustic attenuation increases exponentially with increasing sediment concentration, resulting in severely insufficient penetration depth. For example, in the 1MHz band, when the sediment concentration reaches 20 kg / m³ or higher, its effective penetration depth drops sharply from over 30 m in clear water to less than 3 m. While using lower frequencies such as 300 kHz can improve penetration depth, the spatial resolution decreases significantly, failing to meet the measurement requirements for fine near-surface velocity profiles. Traditional single-frequency solutions struggle to simultaneously address both penetration depth and resolution requirements, resulting in a trade-off in high-sediment-content water bodies.

[0005] Meanwhile, the bottom-tracking function relied upon by traditional ADCP (Advanced Diffusion Probe) completely fails under high sediment load conditions. Because the bottom echo signal is overwhelmed by backscattering from a large amount of suspended sediment, ADCP cannot effectively identify the riverbed reflection signal, leading to continuous erroneous outputs from the bottom-tracking module. This results in significantly increased errors in vessel speed positioning and cross-sectional flow integration. Furthermore, most existing multi-frequency ADCP schemes only implement a "backup frequency band switching" function, essentially remaining simple combinations of single-frequency operations. They have not yet formed a truly meaningful multi-frequency joint inversion framework and cannot simultaneously use velocity profiles and SSC (Sequential Scattered Columnar Columnar Column) profiles as constraints for multi-objective optimization. Moreover, existing technologies generally assume a fixed particle size distribution, failing to consider the dynamic characteristics of particle size changes with hydrological conditions, leading to systematic biases in attenuation estimation.

[0006] In view of the above, this application is hereby submitted. Summary of the Invention

[0007] This invention aims to provide a multi-frequency acoustic Doppler velocity profile measurement system and method for high sediment content water bodies, in order to solve the problems of insufficient acoustic wave penetration depth in existing single-frequency ADCP in high sediment content water bodies, lack of localized attenuation compensation calibration system, complete failure of bottom tracking function leading to excessive ship speed positioning deviation, and the lack of a multi-frequency joint inversion framework, which makes it difficult to achieve synchronous high-precision measurement of velocity profile and suspended sediment concentration profile.

[0008] To solve the above-mentioned technical problems, the present invention is achieved through the following technical solution:

[0009] A multi-frequency acoustic Doppler velocity profile measurement system for water bodies with high sediment content includes an IP68 flow measurement box; the IP68 flow measurement box integrates a multi-frequency transducer array, a multi-frequency signal processing board, a main control and inversion unit, a turbidity sensor, a sediment content sensor, an RTK GNSS / IMU positioning unit, and an RTU data transmission unit; The multi-frequency transducer array is located at the front end of the IP68 flow measurement box, and its output is connected to the multi-frequency signal processing board. According to the ranging requirements, the transducer is selected to emit acoustic pulses in a time-division manner to detect the water depth. The multi-frequency signal processing board integrates an FPGA chip and a DSP chip to process the received data, and its output is connected to the main control and inversion unit. The main control and inversion unit is used to control the multi-frequency transducer array to emit acoustic pulses in a time-division manner, send control commands to the multi-frequency signal processing board, and communicate with the turbidity sensor, the sand content sensor, and the RTK GNSS / IMU positioning unit to obtain real-time observation data, and output the inversion results to the RTU data transmission unit. The RTK GNSS / IMU positioning unit is used to output the time series of the ship's geographic coordinates; The RTU data transmission unit is used to remotely send the inversion results to the shore monitoring platform or cloud server, and to receive remote control commands issued by the cloud and forward them to the main control and inversion unit.

[0010] Preferably, the multi-frequency transducer array includes at least a 300kHz transducer for long-range detection, a 600kHz transducer for medium-range detection, and a 1000kHz transducer for short-range high-resolution detection, which are arranged side by side in space. The FPGA chip is used to generate transmission excitation pulses for each frequency band, acquire analog echo signals, complete analog-to-digital conversion, and perform preliminary Doppler frequency shift extraction. The FPGA chip transmits the digitized echo data to the DSP chip via the SPI and LVDS interfaces. The DSP chip is used to receive the digital echo data output by the FPGA and perform processing operations, and transmit the flow rate data to the main control and inversion unit via the UART interface; The RTK GNSS / IMU positioning unit integrates a dual-frequency RTK GNSS receiver module and a nine-axis IMU inertial measurement unit; The main control and inversion unit integrates an STM32H7 microcontroller, an edge GPU chip, and memory. The STM32H7 microcontroller is used to control the multi-frequency transducer array, send sampling control commands to the turbidity sensor, the sand content sensor, and the RTK GNSS / IMU positioning unit, and send the sampled data to the edge GPU chip; The memory stores a computer program that can be executed by the edge GPU chip to obtain the inversion result, and then the inversion result is sent back to the STM32H7 microcontroller, which then forwards it to the RTU data transmission unit.

[0011] This invention also provides a method for measuring multi-frequency acoustic Doppler velocity profiles in water bodies with high sediment content, comprising: S1, send control commands to the FPGA chip of the multi-frequency signal processing board to control the multi-frequency transducer array to emit acoustic pulses in a time-division manner, collect the original echo signals of the multi-frequency bands, calculate the radial velocity of a single beam based on the Doppler frequency shift, and obtain the original horizontal velocity through four-beam vector synthesis; S2, based on the original multi-band echo signal, combined with the real-time data collected by the turbidity sensor and the sediment concentration sensor, the localized Thorne-Hanes attenuation model is used to correct the two-way attenuation of the sound wave caused by sediment, and the corrected real echo intensity is obtained. S3 uses the corrected true echo intensity and the original horizontal velocity as joint observation data to construct a multi-objective joint optimization function to simultaneously invert the SSC suspended sediment concentration profile and sediment particle size distribution, and output the SSC suspended sediment content, i.e., the spatial concentration field, at any lateral and vertical position within the cross section. S4, through attitude rotation matrix and ship geographical velocity, coordinate transformation and ship motion compensation are performed on the original horizontal flow velocity to eliminate ship motion interference when the sediment of high sediment-laden water body submerges the bottom echo and the acoustic bottom tracking fails, and output a high-precision flow velocity profile corrected in the geodetic coordinate system. S5 uses a high-precision velocity profile corrected under the geodetic coordinate system, combines the ship's geographic motion trajectory to complete the lateral spatial registration of multiple profiles, performs discrete grid subdivision of the water passage section, generates a global two-dimensional velocity field to characterize the water velocity at any position within the section, and performs velocity-area summation integration on each discrete grid cell to calculate the total flow rate of the section. S6, coupled spatial concentration field and global two-dimensional velocity field, performs coupled surface integration of suspended sediment concentration and water velocity on the complete cross-section to obtain the suspended sediment transport flux of the cross-section, thus obtaining inversion results that include at least high-precision velocity profile, spatial concentration field, total cross-section flow, and cross-section suspended sediment transport flux.

[0012] Preferably, it further includes: S7 uses the measured values ​​collected by the sediment concentration sensor as a benchmark, compares them with the SSC suspended sediment concentration profile, and recalibrates the parameters of the localized Thorne-Hanes attenuation model when the deviation exceeds the standard. Iteratively optimizes the real echo intensity and spatial concentration field to complete the closed-loop adaptive correction of the entire process.

[0013] In summary, compared with the prior art, the present invention has the following beneficial effects: This invention solves the contradiction between penetration depth and resolution in water bodies with high sediment content by using a multi-frequency transducer array with division of labor and cooperation. The multi-frequency division of labor and cooperation avoids the inherent defects of the single-frequency scheme.

[0014] This invention achieves localization The calibration coefficients significantly reduced the SSC inversion bias and avoided the "overestimation illusion" caused by the mismatch between the calibration coefficients in international literature and the actual attenuation characteristics of Yellow River coarse sand.

[0015] This invention provides a reliable fallback solution for bottom tracking failure under high sand content conditions by fusing RTK GNSS / IMU integrated navigation with extended Kalman filtering for water tracking velocity.

[0016] This invention achieves synchronous joint optimization of velocity profile and SSC profile through multi-frequency joint inversion with adaptive weights; and through a dynamic calibration mechanism that automatically triggers online recalibration when deviation exceeds the limit, the localized attenuation coefficient has the ability to adapt to the dynamic evolution of water body sediment properties, which significantly improves the accuracy of flow measurement. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 This is a block diagram of a multi-frequency acoustic Doppler velocity profile measurement system for high sediment content water bodies, provided in Example 1.

[0019] Figure 2 The flowchart illustrates a method for measuring multi-frequency acoustic Doppler velocity profiles in high-sediment-laden water bodies, as provided in Example 1. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0021] Example 1 Embodiment 1 of the present invention provides a method for measuring multi-frequency acoustic Doppler velocity profiles of water bodies with high sediment content. This method can be implemented by a multi-frequency acoustic Doppler velocity profile measuring device for water bodies with high sediment content (hereinafter referred to as the measuring device), and in particular, it can be executed by one or more processors within the measuring device.

[0022] In this embodiment, the measuring device may be an electronic device equipped with a processor, which carries a computer program for the multi-frequency acoustic Doppler velocity profile measurement method for the high sediment content water body, and the computer program can be executed, such as a computer, smartphone, smart tablet, workstation, etc., which are not limited here.

[0023] like Figure 1 As shown, a multi-frequency acoustic Doppler velocity profile measurement system for high sediment-laden water bodies includes: The IP68 current meter enclosure integrates a multi-frequency transducer array, a multi-frequency signal processing board, a main control and inversion unit, a turbidity sensor, a sediment concentration sensor, an RTK GNSS / IMU positioning unit, and an RTU data transmission unit. The IP68 current meter enclosure is an IP68 waterproof underwater enclosure with a meticulously designed internal space to accommodate all the aforementioned electronic components. Constructed from high-strength titanium alloy, the enclosure exhibits excellent pressure resistance and corrosion resistance, capable of withstanding the hydrostatic pressure of deep-water environments and resisting the abrasive effects of Yellow River cement and sand. The enclosure shell features dedicated bracket mounting holes for connection and fixation to the measuring vessel hull or a fixed support. The shell surface is coated with an antifouling coating to reduce the impact of biofouling on acoustic performance.

[0024] The multi-frequency transducer array is located at the front end of the IP68 flow meter box, and its output is connected to the multi-frequency signal processing board. Based on the ranging requirements, the transducers are selected to emit acoustic pulses in a time-division manner to detect water depth. The multi-frequency transducer array includes at least a 300kHz transducer for long-range detection, a 600kHz transducer for mid-range detection, and a 1000kHz transducer for short-range high-resolution detection, arranged side-by-side in space. In a preferred embodiment, the acoustic beam axes of the three transducers are parallel to each other and form an angle of 20 to 30 degrees with the vertical axis of the flow meter box.

[0025] The 300 kHz transducer is configured for long-range detection mode, with a maximum single-beam distance of no less than 30 meters under clear water conditions. This frequency band has a longer wavelength and strong penetration capability in high-sediment-laden water bodies, making it suitable for velocity profile measurements in deep channel sections of rivers or deep-water areas of reservoirs. The 600 kHz transducer is configured for mid-range detection mode, with a maximum distance of no less than 10 meters. This frequency band strikes a balance between penetration depth and spatial resolution, making it suitable for velocity profile measurements in shallow water sections with significant changes in riverbed morphology or in the middle reaches of rivers. The 1000 kHz transducer is configured for short-range high-resolution detection mode, with a maximum distance of no less than 3 meters. This frequency band has a shorter wavelength and the highest vertical resolution, making it suitable for fine measurements of surface flow, near-shore areas, or boundary layer flow near the riverbed. The three frequency bands are designed to work in a coordinated manner according to the water depth stratification principle: the low frequency band of 300 kHz covers deep water bodies at a distance, providing flow velocity and echo intensity data in deep water areas; the mid frequency band of 600 kHz covers the intermediate water depth area, providing flow velocity data in the medium depth range; and the high frequency band of 1000 kHz covers the near-field surface water body, providing high-resolution surface flow velocity data, thereby forming a complete detection coverage without dead zones in the vertical direction.

[0026] The multi-frequency signal processing board integrates an FPGA chip and a DSP chip to process the received data, and its output is connected to the main control and inversion unit. The FPGA chip generates transmission excitation pulses for each frequency band, acquires analog echo signals, performs analog-to-digital conversion, and performs preliminary Doppler frequency shift extraction. The analog echo signal from the multi-frequency transducer is connected to the FPGA input of the multi-frequency signal processing board via a coaxial cable. The FPGA chip transmits the digitized echo data to the DSP chip for Doppler frequency shift extraction via SPI and LVDS interfaces. The DSP chip receives the digitized echo data output from the FPGA and performs processing operations (such as further spectral analysis and related calculations to extract precise Doppler frequency shift information), and transmits the flow rate data to the main control and inversion unit via a UART interface.

[0027] Preferably, the FPGA can be a Xilinx Artix-7 series chip, and the DSP can be a TI TMS320C6748 chip.

[0028] The main control and inversion unit controls the multi-frequency transducer array to emit acoustic pulses in a time-division manner, sends control commands to the multi-frequency signal processing board, and communicates with the turbidity sensor, the sediment concentration sensor, and the RTK GNSS / IMU positioning unit to acquire real-time observation data. It then outputs the inversion results to the RTU data transmission unit. The main control and inversion unit integrates an STM32H7 microcontroller, an edge GPU chip, and a memory. The STM32H7 microcontroller controls the multi-frequency transducer array, sends sampling control commands to the turbidity sensor, the sediment concentration sensor, and the RTK GNSS / IMU positioning unit, and sends the sampled data to the edge GPU chip. The memory stores a computer program that can be executed by the edge GPU chip to obtain the inversion results, which are then sent back to the STM32H7 microcontroller, which forwards them to the RTU data transmission unit.

[0029] Preferably, the edge GPU chip can be a Jetson Orin Nano chip, which is responsible for high-performance parallel computing of the multi-frequency joint inversion algorithm. The turbidity sensor can be an OBS-3+ type sensor with a range of 0 to 4000 NTU, capable of measuring the full range of water turbidity from extremely low to extremely high, providing real-time turbidity parameters for the acoustic attenuation compensation model. The sediment concentration sensor adopts a composite design using both vibration and gamma measurement principles. The vibration method is based on the principle of vibrating string density measurement, while the gamma method is based on the principle of gamma ray attenuation. The fusion of these two methods achieves a range of 0 to 200 kg / m³, providing accurate sediment concentration reference values ​​in high-sediment environments. The main control and inversion unit reads real-time sampling data from the turbidity and sediment concentration sensors via an I2C interface.

[0030] The RTK GNSS / IMU positioning unit is used to output the ship's geographic coordinates and timing. It integrates a dual-frequency RTK GNSS receiver module and a nine-axis IMU inertial measurement unit. The dual-frequency RTK GNSS receiver module simultaneously receives GPS L1 / L5 and BeiDou B1 / B2 frequency signals, achieving centimeter- to decimeter-level positioning accuracy through carrier phase differential technology. The nine-axis IMU includes a three-axis accelerometer, a three-axis gyroscope, and a three-axis magnetometer, capable of outputting the carrier's attitude angle and angular rate information in real time. The RTK GNSS / IMU positioning unit connects to the main control and inversion unit via a UART interface at a baud rate of 115200 bps, providing continuous high-precision position and attitude reference under open water conditions.

[0031] The RTU data transmission unit is used to remotely send the inversion results to the onshore monitoring platform or cloud server, and to receive remote control commands issued by the cloud and forward them to the main control and inversion unit. The RTU data transmission unit supports 4G cellular network communication and BeiDou short message dual-link communication mode to realize the real-time uploading of measurement data to the cloud server. The dual-link redundancy design ensures the reliability of data transmission.

[0032] Existing multi-frequency ADCPs lack a clear FPGA time-division transmission timing design, leading to mutual interference between multiple frequency signals. This embodiment limits FPGA time-division single-frequency transmission and three-frequency cyclic timing to avoid frequency crosstalk at the hardware level. Traditional ADCPs rely solely on MCUs for simple flow rate calculations, lacking independent edge GPUs. This solution uses an STM32H7 for overall system scheduling and an edge GPU for parallel acceleration of multi-objective optimization iterations, solving the bottleneck of real-time inversion computing power for high-sand-content, massive profile data.

[0033] Example 2 This invention also provides a method for measuring multi-frequency acoustic Doppler velocity profiles in high-sediment-laden water bodies, applied to the main control and inversion unit. Please refer to [link to relevant documentation]. Figure 2 The method includes steps S1 to S7. In practical applications, the high-sediment-laden section of the middle and lower reaches of the Yellow River is used as a typical application scenario. A detailed explanation is given using the section from the Xiaolangdi Reservoir to the downstream section of the Yellow River as an example. The sediment content in this section typically varies between 10 kg / m³ and 100 kg / m³, reaching over 80 kg / m³ during the summer flood season. The riverbed is mainly composed of coarse sand, with a median particle size of approximately 0.5 mm to 2 mm. Under these conditions, the measurement error of traditional single-frequency ADCP increases significantly.

[0034] S1, send control commands to the FPGA chip of the multi-frequency signal processing board to control the multi-frequency transducer array to emit acoustic pulses in a time-division manner, collect the original echo signals of the multi-frequency bands, calculate the radial velocity of a single beam based on the Doppler frequency shift, and obtain the original horizontal velocity through four-beam vector synthesis.

[0035] For example, a multi-frequency transducer array transmits acoustic pulse sequences in a time-division multiplexing manner under FPGA timing control. Each frequency band transmits for 50 ms, with a three-frequency cycle period of 0.3 s. Each frequency band operates independently to avoid mutual interference. The FPGA internally uses a time-division multiplexing state machine to cyclically activate the transducers in the order of 300 kHz, 600 kHz, and 1000 kHz. Within each 50 ms transmission window, pulse transmission, echo acquisition, and preliminary processing are completed before switching to the next frequency band. The 0.3 s three-frequency cycle design ensures an update rate of approximately 10 Hz for each frequency band, meeting the time resolution requirements for river flow velocity measurement.

[0036] After receiving the echo, the radial velocity of a single beam is calculated based on the Doppler frequency shift principle.

[0037] The formula for calculating the radial velocity of the single beam is: ; in, For single-beam radial flow velocity; Speed ​​of sound in water; The Doppler frequency shift extracted from the spectrum of multi-band raw echo signals can be obtained, for example, by performing autocorrelation or phase difference spectrum analysis on the multi-band raw echo signals. Its magnitude is proportional to the water flow velocity and proportional to the transmission frequency. This is the center frequency of the transducer's acoustic wave emission.

[0038] The original horizontal flow velocities of the four beams were calculated using a geometric vector synthesis method. Each ADCP was equipped with four beams arranged in a Janus four-beam configuration. The two beam groups were installed at an angle towards the upstream and downstream directions of the river, and each beam group contained two beams arranged perpendicularly to each other to eliminate the influence of instrument tilt on measurement accuracy.

[0039] The formula for obtaining the original horizontal flow velocity through four-beam vector synthesis is: ; ; in, , These represent the original horizontal and longitudinal flow velocities of the water body, respectively. , , , The radial velocities were calculated separately for each of the four beams. The angle between the beam and the instrument's vertical axis.

[0040] The combined effect of factors such as sound velocity estimation error, beam angle deviation, and sampling noise results in a flow velocity measurement error of approximately 1% to 3%, which is acceptable within the accuracy requirements for measuring high sediment content water bodies.

[0041] S2, based on the original multi-band echo signal, combined with the real-time data collected by the turbidity sensor and the sediment concentration sensor, the localized Thorne-Hanes attenuation model is used to correct the two-way attenuation of the sound wave caused by sediment, and the corrected true echo intensity is obtained.

[0042] The main control and inversion unit uses real-time data from turbidity and sediment content sensors to compensate for echo intensity according to a locally calibrated attenuation model.

[0043] The formula for calculating the corrected true echo intensity is as follows: ; ; ; ; in, The corrected true echo intensity; To measure depth, that is, the round-trip distance of sound waves; The intensity of the original echo signal in the multi-band frequency band; The total sound attenuation coefficient of the water body; The attenuation coefficient of pure water. The frequency of the transducer's acoustic wave emission center; The concentration of suspended sediment SSC; This represents the particle size distribution of sediment. The scattering attenuation coefficient of suspended sand particles; For localized frequency band calibration constants; This is a viscosity-scattering interpolation function used to characterize the attenuation law of sediment scattering; For the dimensionless scattering scale parameter of the particles; The equivalent radius of the sediment particles; The wavelength of sound waves in water.

[0044] Pure water attenuation coefficient The relationship between the sound attenuation coefficient and frequency in pure water, calculated from temperature, salinity, and frequency, follows an empirical formula, such as: ; This formula applies to the frequency range of 1 MHz to 10 MHz, and requires correction for the low-frequency band of 300 kHz. The attenuation caused by the water body itself also includes the absorption contribution from dissolved gases and suspended organic particles in the water, which is negligible in the high-sediment-laden waters of the Yellow River.

[0045] The local calibration method for the coefficient can be as follows: collect no fewer than 10 representative water samples with different sediment concentrations from the target watershed, determine the particle size distribution and acoustic attenuation coefficient of each water sample under laboratory conditions, and determine the coefficient using a nonlinear regression method. Coefficient values. Taking the Yellow River as an example, the inventors collected water samples from typical sections of the Xiaolangdi Reservoir area and downstream of the Yellow River and found that the coarse sand content in the Yellow River sediment reached 30% to 50%, with the particle size mainly distributed in the range of 0.5 mm to 2 mm. This is significantly different from the attenuation characteristics of fine sand samples with a particle size of 0.1 mm to 0.5 mm as specified in international literature. When the Urick-Thorne-Hanes model coefficients from international literature were applied to Yellow River sediment, the SSC inversion showed an overestimation artifact, with the measured 30 kg / m³ being inverted to 50 kg / m³ to 60 kg / m³, a deviation of 35% to 45%. By establishing a localized calibration model, The coefficients were redefined based on the characteristics of coarse sand in the Yellow River, reducing the inversion error to 8% to 12%.

[0046] In water bodies with high sediment content, sound waves are simultaneously attenuated by absorption from the water itself and by scattering from suspended sediment particles, resulting in a significantly greater total attenuation than in clear water. For example, with a sediment concentration of 50 kg / m³, the total attenuation coefficient in the 1 MHz band can reach several decibels per meter. The penetration depth of 1000 kHz in clear water is approximately 10 meters, but under this sediment concentration, the penetration depth drops to less than 1 meter. The total attenuation coefficient at 300 kHz is relatively low, and the penetration depth can reach over 5 meters. By compensating for echo intensity in real time, the true backscattering intensity can be recovered, providing a reliable data foundation for subsequent SSC inversion. The localized attenuation compensation model is suitable for conditions with moderate suspended sediment concentrations, does not exhibit strong multiple scattering effects, and maintains model accuracy when the sediment concentration is below approximately 200 kg / m³.

[0047] S3 uses the corrected true echo intensity and the original horizontal velocity as joint observation data to construct a multi-objective joint optimization function to simultaneously invert the SSC suspended sediment concentration profile and sediment particle size distribution, and outputs the SSC suspended sediment content, i.e., the spatial concentration field, at any lateral and vertical position within the cross section.

[0048] The master control and inversion units jointly observe the real echo intensity after multi-frequency compensation and the original horizontal velocity of Doppler flow to construct a multi-objective joint optimization problem, and simultaneously invert the SSC suspended sediment concentration profile and sediment particle size distribution.

[0049] The expression for the multi-objective joint optimization function is: ; in, For measuring depth The SSC suspended sediment concentration profile awaiting inversion; For measuring depth The particle size distribution of sediment to be inverted; The frequency of the transducer's acoustic wave emission center; Frequency band adaptive weighting; The corrected true echo intensity; A theoretical model for the acoustic scattering of sediment at corresponding frequencies; Indicates being bound by; For measuring depth The directional velocity gradient is calculated from the original horizontal velocity. It is an L-norm; The threshold for smoothing the flow velocity profile; This indicates that the minimum value is being sought.

[0050] Theoretical model of sound scattering in sediment Based on acoustic scattering theory, this invention describes the quantitative relationship between echo intensity and suspended sediment concentration and particle size distribution. For a single spherical particle, the acoustic scattering cross-section is related to the ratio of particle size to sound wavelength. When the particle size is much smaller than the sound wavelength, Rayleigh scattering theory applies; when the particle size is comparable to or larger than the sound wavelength, geometrical optics approximation or Mie scattering theory applies. In the Yellow River sediment, the particle size distribution is wide, ranging from fine silt to coarse sand, requiring a composite scattering model for description. This invention uses Thorne's formula as the basis for the scattering model, which considers the combined effects of viscous absorption and scattering attenuation. The model parameters are calibrated using measured data.

[0051] Frequency band weights The signal-to-noise ratio (SNR) can be determined according to the adaptive SNR allocation principle. In water bodies with high sediment content, the SNR of each frequency band varies significantly with water depth: the high-frequency band of 1 MHz has a higher SNR in shallow layers because high-frequency sound waves have a shorter propagation distance and less attenuation in shallow layers, and the sediment content in shallow layers is usually lower than in deeper layers; the low-frequency band of 300 kHz has a higher SNR in deep layers because low-frequency sound waves have strong penetrating power and can reach deep water bodies, while high-frequency sound waves are severely attenuated in deep layers. The weight allocation strategy of this embodiment can be: the high-frequency band of 1 MHz has a high weight in shallow layers and a low weight in deep layers; the low-frequency band of 300 kHz has a high weight in deep layers and a low weight in shallow layers; the mid-frequency band of 600 kHz has a relatively high weight in the intermediate water layer.

[0052] The formula for calculating frequency band weight is: ,in for The signal-to-noise ratio (SNR) of the frequency band at a specified water layer is estimated in real time using the statistical characteristics of the echo signal.

[0053] Flow velocity profile smoothness constraint This term is used to introduce a priori physical constraints, assuming that the flow velocity changes smoothly in the vertical direction rather than abruptly. The introduction of this constraint can effectively suppress the influence of measurement noise on the inversion results, improving the physical plausibility of the results.

[0054] The multi-objective joint optimization function can be solved using the Augmented Lagrange multiplier method or the Alternating Direction Multiplier Method (ADMM). The ADMM algorithm decomposes the original problem into several sub-problems, optimizing them alternately and iteratively until convergence. The iterative steps of the ADMM algorithm include: first, updating the SSC profile C while fixing other variables; then, updating the particle size distribution D while fixing other variables; and finally, updating the Lagrange multipliers and penalty parameters. The convergence condition is that the change in the objective function value between two consecutive iterations is less than a preset threshold or the maximum number of iterations is reached. This method achieves synchronous joint inversion of the velocity profile and the SSC profile, rather than the frequency band switching or voting averaging of traditional schemes, resulting in physically consistent inversion results.

[0055] Taking a typical cross-section of the Yellow River as an example, this section is 200 meters wide, with a maximum water depth of 15 meters and an average sediment concentration of 35 kg / m³. A three-frequency ADCP system is used to deploy a survey vessel along the cross-section, moving from the left bank to the right bank at uniform intervals. The vessel's speed is controlled between 0.5 m / s and 1 m / s to ensure sufficient sampling time at each location. At each measurement point, the three-frequency transducers simultaneously acquire echo signals. The original velocity profile is obtained through Doppler velocity calculation in step S1, and the compensated echo intensity profile is obtained through localized attenuation compensation in step S2. Then, the process proceeds to multi-frequency joint inversion in step S3. The inversion algorithm uses the compensated echo intensities of three frequency bands (300 kHz, 600 kHz, and 1000 kHz) as joint inputs, simultaneously inverting the suspended sediment concentration and particle size distribution characteristics of each layer in the vertical direction at that measurement point. The inversion results show that the sediment concentration in the surface layer (0-2 meters) is approximately 20-25 kg / m³, in the middle layer (2-10 meters) it is approximately 30-40 kg / m³, and in the bottom layer (below 10 meters) it is approximately 45-55 kg / m³, exhibiting a clear vertical stratification. The particle size distribution gradually coarsens from the surface to the bottom, with the surface layer dominated by fine silt (median particle size approximately 0.1 mm) and the bottom layer dominated by coarse sand (median particle size approximately 1.5 mm). These inversion results agree well with the laboratory analysis results of simultaneously collected water samples, with a relative deviation of less than 10%.

[0056] S4 uses the attitude rotation matrix and the ship's geographical velocity to perform coordinate transformation and ship motion compensation on the original horizontal flow velocity, so as to eliminate the interference of ship motion when the sediment in the high sediment-laden water body submerges the bottom echo and the acoustic bottom tracking fails, and outputs a high-precision flow velocity profile corrected in the geodetic coordinate system.

[0057] like Figure 2 As shown, the system monitors the acoustic bottom echo signal-to-noise ratio and lock-on status in real time. Acoustic bottom tracking is an inherent function of ADCP. It determines the distance and relative velocity between the instrument and the riverbed by detecting strong echo signals from the riverbed. When the bottom echo signal-to-noise ratio is higher than a preset threshold and the lock-on status is stable, the system uses the relative riverbed velocity obtained through traditional acoustic bottom tracking as the platform's motion reference. In clear water or moderately sediment-laden water bodies, the bottom echo signal intensity is much higher than the backscattered signal of suspended sediment, ensuring reliable bottom tracking functionality.

[0058] When the sediment content is not less than 30 kg / m³, the bottom echo may be submerged by suspended sediment scattering, causing a sharp drop in the bottom tracking signal-to-noise ratio, frequent loss of lock-on state, and failure of traditional bottom tracking functions. In this embodiment of the invention, the system automatically switches to RTK GNSS / IMU assisted trajectory reconstruction mode. At this time, the water tracking velocity is no longer mistakenly taken as the bottom tracking velocity; instead, the ADCP relative to the water velocity is corrected using the ship's velocity in geographic coordinates, achieving accurate cross-section positioning and flow integration.

[0059] The corrected expression for the high-precision velocity profile is: ; ; in, This is the corrected high-precision velocity profile; For measuring depth; Sampling time; This is the rotation matrix from the ship system to the geodetic coordinate system; The original horizontal velocity of the ship system is the velocity of the ship relative to the water. The real-time velocity of the ship's hull relative to the ground in the geodetic coordinate system; The time series of ship geographic coordinates output by the RTKGNSS / IMU positioning unit; Indicating the time sequence of ship geographical locations Regarding time Find the first-order differential.

[0060] The specific implementation method is as follows: First, the geographic coordinates of the survey vessel are obtained through RTK GNSS. Then, the ship's attitude angles, including heading, pitch, and roll angles, are obtained through the IMU, and a rotation matrix is ​​constructed. Transform the ship's system coordinates to the geodetic coordinate system, and then use the relative water velocity measured by ADCP. By rotation matrix Transform to the geodetic coordinate system, and finally determine the ship's geodetic velocity. Calculate the ship's trajectory in the geodetic coordinate system, i.e., the ship's geographic trajectory. .

[0061] The expression for the geographical trajectory of a ship is: ; For ship geographic movement trajectories, used for profile spatial registration; The variable is a dummy variable for integration, representing the integration interval [ Transition moments within; The sampling time.

[0062] The formula for ship geographic motion trajectory, i.e., the differential formula for ship trajectory, is used to solve for the ship's ground motion speed from the time series of RTK-GNSS geographic location data. This velocity serves as a necessary input to the geodetic coordinate system velocity transformation formula, used to eliminate ship motion interference and solve for high-precision velocity profiles. The ship geographic trajectory reconstruction formula obtains the ship's entire trajectory by integrating the ship's velocity relative to the ground. On the one hand, it works with the extended Kalman filter to correct IMU attitude errors and improve coordinate transformation accuracy. On the other hand, it performs channel spatial registration on the velocity profiles at each time point, providing a precise basis for cross-sectional spatial grid division for the S5 section flow integration, ensuring the spatial accuracy of flow calculation.

[0063] Taking the Xiaolangdi Reservoir area of ​​the Yellow River as an example, during the measurement period, the sediment concentration was approximately 45 kg / m³, rendering acoustic bottom tracking completely ineffective. After adopting the RTK GNSS / IMU-assisted trajectory reconstruction mode, RTK GNSS provided centimeter-level positioning accuracy, and the IMU provided high-frequency attitude information with a 100Hz sampling rate. The survey vessel traversed from the left bank to the right bank along a pre-defined cross-section. The GNSS antenna was mounted on top of the hull, the IMU was mounted in the middle of the hull, and the ADCP was connected via an extension cable and mounted on an underwater fixed bracket on the side of the hull. The algorithm outputs parameters such as the geographical coordinates of the survey vessel in real time, combining these parameters with the relative water velocity measured by the ADCP to calculate the absolute current velocity of the vessel. In areas where GNSS signals are blocked, such as under bridges or near tall buildings, the IMU enters a pure inertial calculation mode, obtaining position changes through integral acceleration, maintaining positioning accuracy for a short period.

[0064] S5 uses a high-precision velocity profile corrected in the geodetic coordinate system. It combines the ship's geographic motion trajectory to complete the lateral spatial registration of multiple profiles, performs discrete grid subdivision of the water passage, generates a global two-dimensional velocity field to characterize the water velocity at any position within the cross-section, and performs velocity-area summation integration on each discrete grid cell to calculate the total flow rate of the cross-section.

[0065] Specifically, the formula for calculating the total flow rate of the cross-section is as follows: ; in, The total instantaneous flow rate of the cross-section; , These represent the number of horizontal partitions and the number of vertical layers in the discrete mesh. For the revised first Horizontal, First High-precision average flow velocity of vertical grid cells; For the first Horizontal, First The water flow area of ​​the vertical grid cell.

[0066] When performing discrete grid partitioning, the lateral resolution is determined based on the cross-section width and the vessel's moving speed to ensure that the spacing between adjacent measuring points meets the sampling theorem requirements and avoids missing local velocity anomalies. The vertical resolution is determined based on the vertical pulse length of ADCP and the layering algorithm. For the 300 kHz frequency band, the vertical pulse length is approximately 1 meter, and for the 1000 kHz frequency band, it is approximately 0.25 meters. Data from different frequency bands are automatically fused according to water depth after joint inversion. The integration operation uses a numerical integration method, dividing the cross-section into several micro-areas, calculating the flow contribution of each micro-element separately, and then summing them to obtain the total flow. The advantages of this method are: it directly uses the velocity profile data obtained from multi-frequency joint inversion for integration, avoiding errors caused by extrapolation assumptions in traditional schemes; it has strong adaptability to irregular cross-section shapes, and the lateral segment division and vertical layer division can be flexibly adjusted according to the actual cross-section shape; multi-frequency data fusion improves the accuracy of vertical integration, especially when deep data is missing under high sediment conditions, low-frequency data can provide supplementation.

[0067] S6, coupled spatial concentration field and global two-dimensional velocity field, performs coupled surface integration of suspended sediment concentration and water velocity on the complete cross-section to obtain the suspended sediment transport flux of the cross-section, thus obtaining inversion results that include at least high-precision velocity profile, spatial concentration field, total cross-section flow, and cross-section suspended sediment transport flux.

[0068] The formula for calculating the suspended sediment transport flux of the cross-section is as follows: ; in, This refers to the instantaneous suspended sediment transport flux at the cross section; The total cross-sectional area of ​​the complete water passage; The transverse coordinates of the cross section Vertical coordinate measurement of depth The suspended sediment concentration, i.e., the spatial concentration field; for The corrected high-precision velocity profile, i.e., the global two-dimensional velocity field; It represents the area of ​​an infinitesimal element in the cross-section of the water passage.

[0069] The physical meaning of calculating the suspended sediment transport flux at a cross-section is: the mass of suspended sediment passing through the cross-section per unit time, which is an important indicator for evaluating the sediment transport capacity of a river. This calculation method is similar to the cross-sectional flow integral method, also employing numerical integration to accumulate micro-element surfaces. The difference lies in that the contribution value of each micro-element is the product of the suspended sediment concentration and the flow velocity, rather than simply the flow velocity. During the integration process, the SSC profile obtained from the multi-frequency joint inversion in step S3 is used as the concentration input, and the velocity field obtained from the Doppler measurement in step S1 and the trajectory correction in step S4 is used as the velocity input. The spatial grids of both need to be aligned to ensure the physical consistency of the calculation.

[0070] S7 uses the measured values ​​collected by the sediment concentration sensor as a benchmark, compares them with the SSC suspended sediment concentration profile, and recalibrates the parameters of the localized Thorne-Hanes attenuation model when the deviation exceeds the standard. Iteratively optimizes the real echo intensity and spatial concentration field to complete the closed-loop adaptive correction of the entire process.

[0071] Using synchronous sampling data from the sediment concentration sensor as a reference point, the system continuously compares the retrieved SSC value with the measured baseline value from the sediment concentration sensor. The comparison method is as follows: extract the SSC value retrieved from multiple frequencies and the synchronous sampling value from the sediment concentration sensor at the same water depth location, calculate the relative deviation between the two, and if the relative deviation exceeds 15% and persists for no less than three sampling cycles, trigger the online recalibration procedure.

[0072] The specific steps of the online recalibration procedure can be as follows: First, collect representative water samples near the measuring point, covering the surface, middle, and bottom layers; then, send the water samples back to the laboratory for particle size analysis and sediment concentration determination; finally, update the calibration procedure based on the laboratory analysis results. The coefficients are used to establish new localized attenuation compensation model parameters; finally, the updated parameters are downloaded to the main control and inversion unit, and the system continues to measure.

[0073] This dynamic calibration mechanism enables the localized attenuation coefficient to adapt to the dynamic evolution of water body sediment characteristics. Considering that the characteristics of Yellow River sediment change with the conditions of incoming water and sediment, and that the particle size distribution and mineral composition may change seasonally or interannually, the dynamic calibration mechanism can ensure long-term measurement accuracy.

[0074] The main control and inversion unit ultimately outputs four core hydrological parameters: a corrected high-precision velocity profile, an SSC suspended sediment concentration profile, total cross-sectional flow, and cross-sectional suspended sediment transport flux, achieving integrated synchronous output of the flow field, concentration field, and flux. Data is transmitted in real-time to a cloud server via a 4G or BeiDou short message link through the RTU data transmission unit. The cloud server is equipped with data storage, display, and analysis modules, supporting historical data query, real-time data monitoring, and visualization of measurement results.

[0075] The embodiments of this invention have achieved significant technical effects in practical applications. Taking long-term hydrological measurements of the Yellow River from the Xiaolangdi Reservoir area to the downstream section as an example, compared with the traditional single-frequency 600 kHz ADCP scheme, the SSC inversion deviation under a sediment concentration of 50 kg / m³ decreased from 35% to 45% to 8% to 12%; under the condition of bottom tracking failure, compared with the pure GPS positioning scheme, the cross-section positioning error decreased from 0.8 meters to 1.2 meters to 0.2 meters to 0.4 meters; and the root mean square error of the flow integral relative to the hydrological station's calibrated flow decreased from 18% to 6% to 8%. These improvements in technical effects demonstrate the effectiveness and superiority of the technical solution of the embodiments of this invention in solving the problem of measuring high sediment concentration water bodies.

[0076] In alternative implementations, a 150 kHz / 600 kHz dual-frequency configuration can replace the 300 kHz / 600 kHz / 1000 kHz three-frequency configuration. The 150 kHz band offers stronger long-range penetration but reduces near-field resolution, making it suitable for special applications primarily involving deep-water measurements. The FPGA chip in the multi-frequency signal processing board can be replaced with an Intel Cyclone V or a domestically produced Unisoc Logos-2 chip, achieving domestic hardware platform substitution while maintaining functional equivalence. The sediment concentration sensor can be replaced with a combined optical turbidity sensor and acoustic backscattering configuration instead of a dual-measurement configuration using vibration and gamma methods, resulting in lower costs but slightly reduced measurement accuracy.

[0077] In summary, compared with the prior art, the present invention has the following beneficial effects: (1) Multi-frequency integrated hardware architecture solves the contradiction between penetration and resolution of single-frequency devices.

[0078] Existing conventional ADCPs only use a single frequency, resulting in severe acoustic attenuation in waters with high sediment content, making it impossible to penetrate deep layers. They can only collect a small amount of surface flow velocity, leading to incomplete cross-sectional data. This solution adopts a multi-frequency time-division collaborative approach, with low frequencies responsible for deep water penetration and high frequencies ensuring high accuracy in shallow water. In high sediment content conditions of 50 kg / m³, the effective measurement depth is increased from less than 3 m to more than 25 m, allowing for the complete acquisition of full-depth water profiles.

[0079] This solution uses FPGA time-division timing control for multi-frequency transmission to avoid mutual interference between multiple frequency signals; the whole machine is IP68 integrated with multi-frequency transducers, computing units, positioning, sediment sensors, and remote transmission modules, which is different from traditional split equipment. It is simple to deploy in the field of river channels, waterproof and resistant to sediment erosion, and reduces the risk of on-site wiring and equipment failure.

[0080] This solution is equipped with STM32H7+ edge GPU heterogeneous computing power, which can complete complex inversion iterations on-site in real time without transmitting the original data back to the shore for offline calculation, thus achieving in-situ real-time hydrological output and greatly improving timeliness.

[0081] (2) The localized Thorne-Hanes attenuation model completely solves the core problem of overestimation of coarse sediment SSC.

[0082] Current acoustic sand measurement methods uniformly adopt fine sand calibration based on foreign literature. The coefficient is not applicable to rivers with high coarse sand content, such as the Yellow River. Actual measurements show that a sediment concentration of 30 kg / m³ can be used to invert sediment levels to 50-60 kg / m³, with a relative deviation of 35%-45%. This scheme uses localized, nonlinear regression analysis based on gradient water samples from the watershed. After correction, the SSC inversion deviation was reduced to 8%~12%, and the accuracy of sediment concentration calculation greatly met the hydrological specifications.

[0083] The system fully separates the intrinsic attenuation of water bodies and the attenuation of sediment particles, and compensates for echo loss through a two-way attenuation formula. This eliminates echo distortion caused by sediment scattering in high-sediment-content water bodies, providing a reliable acoustic observation basis for concentration inversion.

[0084] (3) Multi-frequency and multi-objective joint inversion has higher accuracy than the traditional single-frequency / simple averaging scheme.

[0085] This scheme abandons the simple voting and averaging method of multi-frequency data, and designs an adaptive frequency band weight to match the acoustic observation quality at different water depths; it simultaneously combines echo and Doppler current velocity observations to build an optimization model, and simultaneously solves the SSC profile and sediment particle size, which significantly reduces the cumulative error caused by step-by-step calculation.

[0086] The optimization function adds smooth constraints to the velocity profile, suppresses the concentration profile jumps caused by sediment noise in high-sediment-laden water bodies, reduces concentration noise in deep low signal-to-noise ratio regions by 60%, and the stratified distribution of sediment in the cross-section closely matches the actual water sample detection results.

[0087] (4) RTK GNSS / IMU fusion as a fallback to solve the problem of tracking failure in high sand content.

[0088] When sediment completely submerges the riverbed bottom echo in high-sediment-laden water, conventional ADCP bottom tracking fails, and relying solely on GPS compensation cannot correct the ship's attitude deflection, resulting in a cross-section positioning error of 0.8~1.2m. This solution utilizes a rotation matrix to complete coordinate transformation, and with EKF filtering, the positioning error can be reduced to 0.2~0.4m. It can automatically switch the reference between bottom tracking effective and bottom tracking failure conditions, adapting to river channel measurements across the entire range from clear water to ultra-high sediment content, thus broadening its applicability. By integrating the ship's trajectory, multi-frame profile spatial registration is completed, generating a complete global two-dimensional velocity field, overcoming the deficiency of traditional single one-dimensional profiles in fully representing the cross-sectional flow field.

[0089] (5) The flow measurement error is significantly reduced, meeting the Class I hydrological measurement standard.

[0090] Traditional single-frequency equipment suffers from incomplete profiles and insufficient hull compensation accuracy, resulting in a relative RMSE (root mean square error) of up to 18% for flow rate. This solution relies on the complete two-dimensional velocity field grid for discrete integration, accurately distinguishing the velocity differences between deep channels, shallow shoals, and shorelines, reducing artificial extrapolation assumptions, and lowering the flow rate RMSE to 6%~8%, meeting the national high-precision hydrological measurement requirements.

[0091] (6) The flow field and concentration field are coupled to output the flow rate and suspended sediment flux in one stop, improving the testing efficiency.

[0092] Existing equipment only outputs flow velocity, requiring manual stratified water sampling and secondary interpolation coupling to calculate sediment transport. This solution directly couples the global velocity field with the spatial SSC concentration field integral, simultaneously outputting four core hydrological and sediment parameters—velocity profile, SSC profile, cross-sectional flow rate, and suspended sediment flux—in a single measurement, reducing the manual workload of field testing by 90%.

[0093] (7) Closed-loop dynamic calibration mechanism to achieve long-term unattended stable monitoring.

[0094] Existing equipment Since the coefficients are fixed at the factory, changes in sediment particle size and sediment content during the flood season and dry season will permanently degrade the model, requiring frequent manual river sampling for calibration. This scheme sets a quantitative calibration trigger condition: automatic recalibration is performed when the deviation between the inverted SSC and the sensor measurement is >15% and continues for 3 consecutive sampling cycles. It adapts to seasonal changes in watershed sediment, maintains consistent accuracy over long-term continuous monitoring, and reduces manual calibration workload by 80%.

[0095] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A multi-frequency acoustic Doppler velocity profile measurement system for high sediment-laden water bodies, characterized in that, It includes an IP68 flow measurement box; the IP68 flow measurement box integrates a multi-frequency transducer array, a multi-frequency signal processing board, a main control and inversion unit, a turbidity sensor, a sand content sensor, an RTK GNSS / IMU positioning unit, and an RTU data transmission unit; The multi-frequency transducer array is located at the front end of the IP68 flow measurement box, and its output is connected to the multi-frequency signal processing board. According to the ranging requirements, the transducer is selected to emit acoustic pulses in a time-division manner to detect the water depth. The multi-frequency signal processing board integrates an FPGA chip and a DSP chip to process the received data, and its output is connected to the main control and inversion unit. The main control and inversion unit is used to control the multi-frequency transducer array to emit acoustic pulses in a time-division manner, send control commands to the multi-frequency signal processing board, and communicate with the turbidity sensor, the sand content sensor, and the RTK GNSS / IMU positioning unit to obtain real-time observation data, and output the inversion results to the RTU data transmission unit. The RTK GNSS / IMU positioning unit is used to output the time series of the ship's geographic coordinates; The RTU data transmission unit is used to remotely send the inversion results to the shore monitoring platform or cloud server, and to receive remote control commands issued by the cloud and forward them to the main control and inversion unit.

2. The multi-frequency acoustic Doppler velocity profile measurement system for high sediment content water bodies according to claim 1, characterized in that... The multi-frequency transducer array includes at least a 300kHz transducer for long-range detection, a 600kHz transducer for medium-range detection, and a 1000kHz transducer for short-range high-resolution detection, which are arranged side by side in space. The FPGA chip is used to generate transmission excitation pulses for each frequency band, acquire analog echo signals, complete analog-to-digital conversion, and perform preliminary Doppler frequency shift extraction. The FPGA chip transmits the digitized echo data to the DSP chip via the SPI and LVDS interfaces. The DSP chip is used to receive the digital echo data output by the FPGA and perform processing operations, and transmit the flow rate data to the main control and inversion unit via the UART interface; The RTK GNSS / IMU positioning unit integrates a dual-frequency RTK GNSS receiver module and a nine-axis IMU inertial measurement unit; The main control and inversion unit integrates an STM32H7 microcontroller, an edge GPU chip, and memory. The STM32H7 microcontroller is used to control the multi-frequency transducer array, send sampling control commands to the turbidity sensor, the sand content sensor, and the RTK GNSS / IMU positioning unit, and send the sampled data to the edge GPU chip; The memory stores a computer program that can be executed by the edge GPU chip to obtain the inversion result, and then the inversion result is sent back to the STM32H7 microcontroller, which then forwards it to the RTU data transmission unit.

3. A method for measuring the multi-frequency acoustic Doppler velocity profile of high-sediment-laden water bodies, based on the multi-frequency acoustic Doppler velocity profile measurement system for high-sediment-laden water bodies as described in any one of claims 1-2, characterized in that, include: S1, send control commands to the FPGA chip of the multi-frequency signal processing board to control the multi-frequency transducer array to emit acoustic pulses in a time-division manner, collect the original echo signals of the multi-frequency bands, calculate the radial velocity of a single beam based on the Doppler frequency shift, and obtain the original horizontal velocity through four-beam vector synthesis; S2, based on the original multi-band echo signal, combined with the real-time data collected by the turbidity sensor and the sediment concentration sensor, the localized Thorne-Hanes attenuation model is used to correct the two-way attenuation of the sound wave caused by sediment, and the corrected real echo intensity is obtained. S3 uses the corrected true echo intensity and the original horizontal velocity as joint observation data to construct a multi-objective joint optimization function to simultaneously invert the SSC suspended sediment concentration profile and sediment particle size distribution, and output the SSC suspended sediment content, i.e., the spatial concentration field, at any lateral and vertical position within the cross section. S4, through attitude rotation matrix and ship geographical velocity, coordinate transformation and ship motion compensation are performed on the original horizontal flow velocity to eliminate ship motion interference when the sediment of high sediment-laden water body submerges the bottom echo and the acoustic bottom tracking fails, and output a high-precision flow velocity profile corrected in the geodetic coordinate system. S5 uses a high-precision velocity profile corrected under the geodetic coordinate system, combines the ship's geographic motion trajectory to complete the lateral spatial registration of multiple profiles, performs discrete grid subdivision of the water passage section, generates a global two-dimensional velocity field to characterize the water velocity at any position within the section, and performs velocity-area summation integration on each discrete grid cell to calculate the total flow rate of the section. S6, coupled spatial concentration field and global two-dimensional velocity field, performs coupled surface integration of suspended sediment concentration and water velocity on the complete cross-section to obtain the suspended sediment transport flux of the cross-section, thus obtaining inversion results that include at least high-precision velocity profile, spatial concentration field, total cross-section flow, and cross-section suspended sediment transport flux.

4. The method for measuring multi-frequency acoustic Doppler velocity profiles in high-sediment-laden water bodies according to claim 3, characterized in that... It also includes: S7 uses the measured values ​​collected by the sediment concentration sensor as a benchmark, compares them with the SSC suspended sediment concentration profile, and recalibrates the parameters of the localized Thorne-Hanes attenuation model when the deviation exceeds the standard. Iteratively optimizes the real echo intensity and spatial concentration field to complete the closed-loop adaptive correction of the entire process.

5. The method for measuring multi-frequency acoustic Doppler velocity profiles in high-sediment-laden water bodies according to claim 3, characterized in that... The formula for calculating the radial velocity of the single beam is as follows: ； in, For single-beam radial flow velocity; Speed ​​of sound in water; The Doppler frequency shift is extracted from the spectrum of the original multi-band echo signal; The frequency of the transducer's acoustic wave emission center; The formula for obtaining the original horizontal flow velocity through four-beam vector synthesis is: ； ； in, , These represent the original horizontal and longitudinal flow velocities of the water body, respectively. , , , The radial flow velocities calculated for each of the four beams; The angle between the beam and the instrument's vertical axis.

6. The method for measuring multi-frequency acoustic Doppler velocity profiles in high-sediment-laden water bodies according to claim 3, characterized in that... The formula for calculating the corrected true echo intensity is as follows: ； ； ； ； in, The corrected true echo intensity; To measure depth, that is, the round-trip distance of sound waves; The intensity of the original echo signal in the multi-band frequency band; The total sound attenuation coefficient of the water body; The attenuation coefficient of pure water. The frequency of the transducer's acoustic wave emission center; The concentration of suspended sediment SSC; This represents the particle size distribution of sediment. The scattering attenuation coefficient of suspended sand particles; For localized frequency band calibration constants; This is a viscosity-scattering interpolation function used to characterize the attenuation law of sediment scattering; For the dimensionless scattering scale parameter of the particles; The equivalent radius of the sediment particles; The wavelength of sound waves in water.

7. The method for measuring multi-frequency acoustic Doppler velocity profiles in high-sediment-laden water bodies according to claim 6, characterized in that... The expression for the multi-objective joint optimization function is: ； in, For measuring depth The SSC suspended sediment concentration profile awaiting inversion; For measuring depth The particle size distribution of sediment to be inverted; The frequency of the transducer's acoustic wave emission center; Frequency band adaptive weighting; The corrected true echo intensity; A theoretical model for the acoustic scattering of sediment at corresponding frequencies; Indicates being bound by; For measuring depth The directional velocity gradient is calculated from the original horizontal velocity. It is an L-norm; The threshold for smoothing the flow velocity profile; This indicates that the minimum value is being sought.

8. The method for measuring multi-frequency acoustic Doppler velocity profiles in high-sediment-laden water bodies according to claim 3, characterized in that... The corrected expression for the high-precision velocity profile is: ； ； in, This is the corrected high-precision velocity profile; For measuring depth; Sampling time; This is the rotation matrix from the ship system to the geodetic coordinate system; The original horizontal velocity of the ship system is the velocity of the ship relative to the water. The real-time velocity of the ship's hull relative to the ground in the geodetic coordinate system; The time series of ship geographic coordinates output by the RTKGNSS / IMU positioning unit; Indicating the time sequence of ship geographical locations Regarding time Find the first-order differential.

9. The method for measuring multi-frequency acoustic Doppler velocity profiles in high-sediment-laden water bodies according to claim 8, characterized in that... The expression for the ship's geographical movement trajectory is: ； For ship geographic movement trajectories, used for profile spatial registration; The variable is a dummy variable for integration, representing the integration interval [ Transition moments within; Sampling time; The formula for calculating the total flow rate of the cross-section is: ； in, The total instantaneous flow rate of the cross-section; , These represent the number of horizontal partitions and the number of vertical layers in the discrete mesh. For the revised first Horizontal, First High-precision average flow velocity of vertical grid cells; For the first Horizontal, First The water flow area of ​​the vertical grid cell.

10. The method for measuring multi-frequency acoustic Doppler velocity profiles in high-sediment-laden water bodies according to claim 9, characterized in that... The formula for calculating the suspended sediment transport flux of the cross section is as follows: ； in, This refers to the instantaneous suspended sediment transport flux at the cross section; The total cross-sectional area of ​​the complete water passage; The transverse coordinates of the cross section Vertical coordinate measurement of depth The suspended sediment concentration, i.e., the spatial concentration field; for The corrected high-precision velocity profile, i.e., the global two-dimensional velocity field; It represents the area of ​​an infinitesimal element in the cross-section of the water passage.

Images