A sensor-based water level flow data processing method and system

Abstract

The application provides a sensor-based water level and flow data processing method and system, relates to the technical field of hydrological monitoring data processing, integrates water level data, flow velocity profile data, acoustic echo feature data and water turbidity data, determines that the water body is in a high turbidity state when the water turbidity is continuously higher than a preset threshold and the water level and flow velocity change amplitude is low, deeply analyzes the acoustic echo feature data, and generates a reliability score of the flow velocity data of each depth layer in the flow velocity profile. The reliability score is used for weighted processing of the flow velocity data, low reliability data is corrected or replaced, and finally the cross-section average flow velocity is calculated based on the weighted processed flow velocity data, and the instantaneous flow is obtained by combining the cross-section area. The application overcomes the defects of data processing logic modularization and lack of cross-sensor data cross-validation in the prior art, and significantly improves the flow measurement accuracy and reliability of the hydrological monitoring system under complex water quality conditions.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hydrological monitoring data processing, in particular to a sensor-based water level and flow data processing method and system. BACKGROUND

[0002] In the field of hydrological monitoring, multifunctional hydrological monitoring systems integrate multiple sensors such as water level gauges and acoustic Doppler current profilers (ADCP) to conduct long-term automated monitoring of water levels and flows of rivers, lakes and other water bodies. These systems usually rely on pre-mapped river cross-section geometric information and real-time sensor data to calculate instantaneous flow. However, in actual operation, abnormal changes in the medium properties of the water body, especially the sharp increase in the concentration of suspended fine particles, can have a serious impact on the performance of acoustic measurement equipment, and thus cause systematic deviations in the calculated flow results.

[0003] For example, when a civil engineering project in the upstream area encounters heavy rainfall, causing a large amount of construction wastewater containing fine particles to overflow into the river and spread downstream with the water flow, the turbidity sensor at the monitoring station will capture a sharp increase in the turbidity of the water body. At this time, the working principle of the acoustic Doppler current profiler (ADCP) will be severely challenged. ADCP calculates water flow velocity by emitting acoustic pulses and receiving reflected signals, but high concentrations of fine clay and silt particles can strongly absorb and scatter acoustic energy, preventing the acoustic waves from effectively penetrating the entire water depth. This makes the ADCP-acquired water flow velocity profile data scattered and incomplete, especially in the deep water near the riverbed, where reliable velocity readings are often unavailable, and even missing data can occur.

[0004] In the face of such incomplete and degraded flow velocity profile data, the data processing unit is in a dilemma when calculating the cross-section average flow velocity. Due to the lack of complete water flow velocity information, the system has to rely on incomplete upper water data for extrapolation or use pre-set interpolation algorithms and historical averages to fill in missing deep data. For example, if the deep water data is missing, and the system only averages or extrapolates based on the upper data, the calculated cross-section average flow velocity will be systematically overestimated. In either case, the average flow velocity calculated based on incomplete or damaged data will have a significant deviation from the true cross-section average flow velocity. Ultimately, when the data processing unit multiplies this biased average flow velocity with the still accurate cross-section area obtained by the water level gauge, the resulting instantaneous flow value begins to systematically deviate from the true flow.

[0005] The deeper problem is that the existing data processing logic is modular, which collects, processes and transmits sensor data such as water level, flow rate and turbidity as independent parameters. Although the turbidity sensor faithfully records the sharp change of water turbidity, this key information can be used as a strong warning signal for the reliability decrease of ADCP flow rate measurement, but the system's data processing method does not establish a direct correlation between turbidity data and ADCP measurement effectiveness. In other words, the system has all the "clues" to find the problem, but lacks a mechanism to cross-verify and intelligently analyze these "clues" from different sensors to identify the deep problem that ADCP performance is impaired under certain water quality conditions, which in turn leads to unreliable flow calculation results. As a result, the monitoring system continues to output seemingly normal but actually incorrect flow data during the turbidity period of the water body, which may mislead the downstream water resource scheduling and environmental management decisions. SUMMARY

[0006] The present application provides a sensor-based water level flow data processing method and system, aiming to solve the problem of systematic deviation in flow calculation results caused by impaired performance of acoustic measurement equipment under abnormal changes in water medium characteristics, especially in high turbidity conditions, in the field of hydrological monitoring.

[0007] In one aspect, the present application provides a sensor-based water level flow data processing method, comprising:

[0008] Collecting water level data, flow rate profile data, acoustic echo feature data, and water turbidity data, wherein the acoustic echo feature data includes signal attenuation rate, signal-to-noise ratio, and echo intensity gradient;

[0009] If the water turbidity data is continuously higher than the preset turbidity threshold within a preset time, and the water level data change amplitude is lower than the preset water level change threshold, and the flow rate profile data change amplitude is lower than the preset flow rate profile change threshold, it is determined that the current water body is in a high turbidity state dominated by fine particles, and the water level data change amplitude is lower than the preset water level change threshold at this time;

[0010] When it is determined that the current water body is in a high turbidity state, analyzing the acoustic echo feature data to generate a reliability score for each depth layer flow rate data in the flow rate profile;

[0011] Using the reliability score, the weighted processing is performed on each depth layer flow rate data in the flow rate profile, and the flow rate data with a reliability score lower than a preset score threshold is corrected or replaced;

[0012] Based on the weighted processed flow rate data, the cross-sectional average flow rate is calculated, and combined with the corresponding water surface area of the water level data, the instantaneous flow rate is obtained.

[0013] Optionally, when determining that the current water body is in a high-turbidity state, the step of analyzing the acoustic echo characteristic data to generate a reliability score of the flow velocity data of each depth layer in the flow velocity profile comprises:

[0014] acquiring acoustic echo characteristic data using acoustic wave emission and reception modes of at least two different frequencies;

[0015] extracting acoustic wave attenuation characteristics, scattering intensity spectra, and Doppler frequency shift spectral widths of each frequency band according to the acoustic echo characteristic data;

[0016] comparing the acoustic wave attenuation characteristics, the scattering intensity spectra, and the Doppler frequency shift spectral widths of each frequency band to obtain a first comparison result reflecting scattering differences of the medium;

[0017] identifying water body medium characteristics according to the first comparison result, and adjusting weights of the acoustic wave attenuation characteristics, the scattering intensity spectra, and the Doppler frequency shift spectral widths in the calculation of the flow velocity data reliability score according to the water body medium characteristics;

[0018] generating the flow velocity data reliability score according to the adjusted weights.

[0019] Optionally, the step of performing weighted processing on the flow velocity data of each depth layer in the flow velocity profile using the reliability score comprises:

[0020] identifying continuous depth layers in the flow velocity profile whose reliability scores are lower than a preset score threshold;

[0021] analyzing a difference degree of reliability scores of adjacent layers within the continuous depth layers, and a variation gradient size of flow velocity values of the adjacent layers;

[0022] judging whether a flow velocity variation trend in the continuous depth layers has continuity and regularity according to the difference degree and the variation gradient, and if so, assigning a correction weight based on overall reliability distribution and spatial correlation to the continuous depth layers;

[0023] performing weighted processing on the flow velocity data of each depth layer in the flow velocity profile according to the correction weight.

[0024] Optionally, the step of correcting or replacing the flow velocity data whose reliability score is lower than the preset score threshold comprises:

[0025] when the reliability score of the flow velocity data is lower than a first preset threshold, combining other flow velocity data whose reliability score is higher than a second preset threshold and river cross-section geometric information to estimate target flow velocity data to replace the flow velocity data whose reliability score is lower than the first preset threshold through hydrodynamic principles; or,

[0026] When the reliability score of the flow velocity data is lower than the third preset threshold, the flow velocity data is replaced by flow velocity data calculated according to the water level data and a preset water level-flow velocity curve.

[0027] Optionally, before the step of replacing the flow velocity data with the reliability score lower than the first preset threshold by the target flow velocity data calculated according to the water level data and the preset water level-flow velocity curve, the method comprises:

[0028] Collecting bottom echo signal features of the acoustic measuring device and corresponding near-bed flow velocity data;

[0029] Comparing the bottom echo signal features with the near-bed flow velocity data to obtain a second comparison result reflecting the influence of the riverbed boundary;

[0030] According to the second comparison result, determining whether the riverbed roughness or the flow resistance coefficient changes;

[0031] When the riverbed roughness or the flow resistance coefficient changes, quantifying the change amplitude of the riverbed roughness or the flow resistance coefficient;

[0032] According to the quantified change amplitude, adjusting an empirical parameter used for hydrodynamic extrapolation;

[0033] When the reliability score of the flow velocity data is lower than the first preset threshold, the flow velocity data is replaced by flow velocity data calculated according to the adjusted empirical parameter, the flow velocity data with the reliability score higher than the second preset threshold, and the river channel cross-section geometric information.

[0034] Optionally, the steps of collecting the water level data, the flow velocity profile data, the acoustic echo feature data, and the water turbidity data comprise:

[0035] Collecting output data of each sensor;

[0036] Analyzing statistical characteristics of the output data of each sensor, the statistical characteristics including at least one of data fluctuation range, mean value drift, data jump frequency, and data missing rate;

[0037] Obtaining response data of each sensor, the response data being obtained by sending a calibration instruction or a physical excitation to the sensor;

[0038] Comparing the statistical characteristics and the response data with a preset sensor health benchmark to obtain a third comparison result;

[0039] According to the third comparison result, identifying an abnormal state of each sensor;

[0040] According to the abnormal state, the data collected by the sensor is marked, screened or replaced to ensure the effectiveness and consistency of the collected water level data, flow profile data, acoustic echo feature data and water turbidity data.

[0041] Optionally, the step of comparing the statistical features and the response data with the preset sensor health benchmark to obtain a third comparison result comprises:

[0042] Collecting output data of each sensor, the output data comprising the statistical features and the response data;

[0043] Obtaining water environment parameters, the water environment parameters comprising water temperature, salinity and flow rate;

[0044] Obtaining sensor operation parameters, the sensor operation parameters comprising sensor cumulative working time and calibration history;

[0045] According to the water environment parameters and the sensor operation parameters, adjusting the preset sensor health benchmark;

[0046] Comparing the statistical features and the response data with the adjusted sensor health benchmark to obtain a third comparison result.

[0047] Optionally, the step of marking, screening or replacing the data collected by the sensor according to the abnormal state comprises:

[0048] Identifying the type of the abnormal state, analyzing the degree of the abnormal state, and obtaining historical data performance of the sensor;

[0049] According to the type of the abnormal state, adjusting the priority and specific operation of the marking, screening or replacing; and according to the degree of the abnormal state, adjusting the first intensity of the marking, screening or replacing; and according to the historical data performance, adjusting the strategy of the marking, screening or replacing;

[0050] According to the adjusted priority and specific operation, the adjusted first intensity and the adjusted strategy, marking, screening or replacing the data collected by the sensor.

[0051] Optionally, the step of calculating the cross-section average flow rate comprises:

[0052] The cross-section average flow rate is calculated based on the following formula :

[0053]

[0054] wherein, is the flow rate value of the i-th depth layer, a reliability score of an i-th depth layer, a cross-sectional area of water represented by the i-th depth layer.

[0055] In another aspect, the application provides a sensor-based water level and flow data processing method and system, which comprises:

[0056] a data acquisition module for acquiring water level data, flow velocity profile data, acoustic echo feature data, and water turbidity data, wherein the acoustic echo feature data includes signal attenuation rate, signal-to-noise ratio, and echo intensity gradient;

[0057] a state judgment module for determining that the current water body is in a high turbidity state dominated by fine particles if the water turbidity data continuously exceeds a preset turbidity threshold within a preset time, and the change amplitudes of the water level data and the flow velocity profile data are both lower than their respective preset change thresholds;

[0058] a reliability evaluation module for analyzing the acoustic echo feature data to generate a reliability score of each depth layer flow velocity data in the flow velocity profile when it is determined that the current water body is in a high turbidity state dominated by fine particles;

[0059] a data processing module for performing weighted processing on each depth layer flow velocity data in the flow velocity profile using the reliability score, and correcting or replacing the flow velocity data with low reliability score;

[0060] a flow calculation and output module for calculating the cross-sectional average flow velocity based on the weighted processed flow velocity data, and obtaining the instantaneous flow by combining the cross-sectional area of water corresponding to the water level data.

[0061] The application determines that the water body is in a high turbidity state when the water turbidity continuously exceeds a preset threshold and the water level and flow velocity change amplitudes are low, thereby triggering in-depth analysis of the acoustic echo feature data to generate a reliability score of each depth layer flow velocity data in the flow velocity profile. On this basis, the reliability score is used to perform weighted processing on the flow velocity data, and the low reliability data is corrected or replaced. Finally, the cross-sectional average flow velocity is calculated based on the weighted processed flow velocity data, and the instantaneous flow is obtained by combining the cross-sectional area of water. This method overcomes the deficiencies of the prior art in data processing logic modularization and lack of cross-sensor data cross-validation, avoids the flow calculation deviation caused by incomplete or damaged data, significantly improves the flow measurement accuracy and reliability of the hydrological monitoring system under complex water quality conditions, and provides more accurate data support for water resource scheduling and environmental management. BRIEF DESCRIPTION OF DRAWINGS

[0062] In order to more clearly illustrate the present application, the drawings required to be used in the embodiments will be briefly introduced as follows. Obviously, those skilled in the art can obtain other drawings from these drawings without any creative effort.

[0063] Figure 1 A flow diagram of a sensor water level flow data processing method is shown.

[0064] Figure 2 A structure diagram of a sensor water level flow data processing system is shown.

[0065] Reference signs: 100, sensor water level flow data processing system; 10, data acquisition module; 20, state judgment module; 30, reliability evaluation module; 40, data processing module; 50, flow calculation and output module. DETAILED DESCRIPTION

[0066] The technical solutions in the present application will be described in detail below with reference to the drawings in the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. The components of the present application described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without any creative effort fall within the scope of protection of the present application.

[0067] It should be noted that: similar reference signs and letters represent similar items in the following drawings, therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. Meanwhile, in the description of the present application, the terms "first", "second" and the like are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0068] The traditional existing multifunctional hydrological monitoring system in the field of hydrological monitoring integrates various sensors to automatically monitor the water level and flow of rivers, lakes and other water bodies for a long time. However, in actual operation, the abnormal change of the medium characteristics of the water body, especially the turbidity of the water body caused by the sharp increase of the concentration of suspended fine particles, will seriously affect the performance of the acoustic measurement equipment, and then cause systematic deviation of the flow calculation result. The deeper problem is that the existing data processing logic is modular, which collects, processes and transmits the sensor data such as water level, flow rate and turbidity as independent parameters, and lacks a mechanism to cross-verify and intelligently analyze these "clues" from different sensors, so as to identify the performance damage of the acoustic Doppler current profiler under certain water quality conditions, and then cause the deep problem of unreliable flow calculation result.

[0069] As shown in Figure 1 An example flowchart of a sensor-based water level flow data processing method in the embodiment is shown. The present application proposes a sensor-based water level flow data processing method, which comprises:

[0070] S10, collecting water level data, flow profile data, acoustic echo feature data, and water turbidity data, wherein the acoustic echo feature data includes signal attenuation rate, signal-to-noise ratio and echo intensity gradient;

[0071] The water level data refers to the height information of the water surface relative to a certain reference plane measured and recorded in real time by a water level sensor (such as an ultrasonic water level meter, a pressure type water level meter, etc.). This data reflects the water storage capacity and water surface fluctuation of the water body.

[0072] The flow profile data refers to the flow velocity distribution information of the water body at different depth layers measured by acoustic Doppler current profiler (ADCP) and other devices. It is usually presented in the form of a series of depth layer flow rate values, reflecting the velocity structure of the water flow in the vertical direction.

[0073] The acoustic echo feature data refers to the signal characteristics obtained during the emission of acoustic waves and the reception of echoes by acoustic measurement equipment, including signal attenuation rate, signal-to-noise ratio and echo intensity gradient. These feature data can reflect the medium influence on the propagation of acoustic waves in the water body, such as absorption, scattering, etc.

[0074] The water turbidity data refers to the quantitative index of the turbidity degree of the water body measured by the turbidity sensor, usually expressed in turbidity units (such as NTU). High turbidity usually means that there are a large number of fine particles suspended in the water body.

[0075] S20, if the water body turbidity data is continuously higher than the preset turbidity threshold value within a preset time, and the water level data change amplitude is lower than a preset water level change threshold value, and the flow velocity profile data change amplitude is lower than a preset flow velocity profile change threshold value, it is determined that the current water body is in a high turbidity state dominated by fine particles, and at this time the water level data change amplitude is lower than the preset water level change threshold value;

[0076] S30, when it is determined that the current water body is in a high turbidity state, analyzing the acoustic echo feature data to generate a reliability score of the flow velocity data at each depth layer in the flow velocity profile;

[0077] The reliability score refers to a quantitative evaluation of the quality of the flow velocity data at each depth layer in the flow velocity profile. The score comprehensively considers factors such as acoustic echo feature data and is used to indicate the degree of influence of the water body medium on the flow velocity data, thereby determining its credibility.

[0078] S40, using the reliability score, the flow velocity data at each depth layer in the flow velocity profile is weighted and processed, and the flow velocity data with a reliability score lower than a preset score threshold value is corrected or replaced;

[0079] The preset turbidity threshold value, the preset water level change threshold value, the preset flow velocity profile change threshold value and the preset score threshold value are all pre-set parameters for judging the water body state, data change trend and data reliability level. These threshold values can be adjusted according to actual application scenarios, historical data analysis and expert experience.

[0080] S50, based on the weighted flow velocity data, the cross-sectional average flow velocity is calculated, and then combined with the water surface area corresponding to the water level data, the instantaneous flow is obtained.

[0081] The cross-sectional average flow velocity refers to the average speed of water flow through a certain cross-sectional area, which is one of the key parameters for calculating the instantaneous flow.

[0082] The cross-sectional area of the water surface refers to the effective area of the water flow through a certain cross-sectional area, which is usually calculated based on the water level data and the preset river cross-sectional geometric information.

[0083] The instantaneous flow refers to the volume of water passing through a certain cross-sectional area per unit time at a certain time, which is the core output index of hydrological monitoring.

[0084] The present application comprehensively analyzes multi-source sensor data, intelligently identifies the high turbidity state of the water body, and on this basis, evaluates the reliability of the flow velocity data, weights and corrects it, thereby effectively improving the accuracy and reliability of flow monitoring under complex hydrological conditions, overcoming the problem of flow calculation deviation caused by abnormal changes in the water body medium in the prior art.

[0085] First, in practical applications, data collection can be achieved in various ways. For example, water level data can be collected in real time by ultrasonic water level meters or pressure water level meters installed at monitoring stations. These water level meters can continuously monitor the water surface height and transmit the data to the data processing unit. Flow velocity profile data is usually collected by acoustic Doppler current profilers (ADCPs). ADCPs calculate the water flow velocity at different depths by emitting acoustic pulses and receiving echoes reflected by suspended particles in the water body, using the Doppler effect. Acoustic echo characteristic data, including signal attenuation rate, signal-to-noise ratio, and echo intensity gradient, are original acoustic signal parameters obtained by ADCPs when collecting flow velocity profile data. These parameters directly reflect the characteristics of acoustic wave propagation and reflection in water, for example, signal attenuation rate can indicate the degree of loss of acoustic energy during propagation, signal-to-noise ratio reflects the proportion of effective signal to background noise, and echo intensity gradient can reveal the distribution of scatterers in the water body. Water turbidity data is collected by turbidity sensors, which are usually based on optical principles and quantify turbidity by measuring the degree of scattering or absorption of light in water.

[0086] Second, if the water turbidity data is continuously higher than the preset turbidity threshold for a preset time, and the water level data change amplitude is lower than the preset water level change threshold, and the flow velocity profile data change amplitude is lower than the preset flow velocity profile change threshold, it is determined that the current water body is in a high turbidity state dominated by fine particles. This determination mechanism is one of the core innovations of the present application, which accurately identifies the high turbidity state caused by fine particles through multi-parameter linkage analysis. For example, a preset turbidity threshold of 500 NTU can be set, and a preset time of 30 minutes can be set. If the turbidity value monitored by the turbidity sensor for 30 consecutive minutes is higher than 500 NTU, the first condition is met. At the same time, the change amplitude of water level data and flow velocity profile data is monitored. For example, the preset water level change threshold can be set to 0.1 meters per hour, and the preset flow velocity profile change threshold can be set to 0.05 meters per second per hour. If the water level change amplitude is less than 0.1 meters per hour and the average change amplitude of flow velocity data at each depth layer in the flow velocity profile is less than 0.05 meters per second per hour within the 30 minutes, all conditions are met. When all conditions are met, it is determined that the current water body is in a high turbidity state dominated by fine particles. This multi-condition combination judgment avoids the misjudgment that may be caused by a single turbidity indicator. For example, when there are a large number of bubbles or large floating objects in the water body, the turbidity may also increase, but the change trend of water level and flow velocity may be different from that of the high turbidity state dominated by fine particles.

[0087] Again, when it is determined that the current water body is in a high-turbidity state, acoustic echo feature data is analyzed to generate a reliability score for each depth layer's flow velocity data in the flow velocity profile. Once it is determined that the water body is in a high-turbidity state, in-depth analysis of acoustic echo feature data is initiated. For example, a reliability score can be calculated for each depth layer in the flow velocity profile based on parameters such as signal attenuation rate, signal-to-noise ratio, and echo intensity gradient. When the water body has high turbidity, the sound wave will be more strongly absorbed and scattered during propagation, resulting in an increase in signal attenuation rate, a decrease in signal-to-noise ratio, and the echo intensity gradient may also become irregular. These changes can all be used as indicators to assess the reliability of flow velocity data. For example, an empirical model or machine learning model can be established to take these acoustic echo feature data as input and output a reliability score between 0 and 1, where 1 represents completely reliable data and 0 represents completely unreliable data. Specifically, when the signal attenuation rate exceeds a certain threshold, the reliability score will decrease accordingly; when the signal-to-noise ratio is below a certain threshold, the reliability score will also decrease. Abnormal changes in echo intensity gradient can also cause the reliability score to decrease. In this way, the degree to which each depth layer's flow velocity data is affected by the turbid water body can be quantified.

[0088] Next, using the reliability scores, the flow velocity data in the flow velocity profile is weighted and corrected or replaced for flow velocity data with a reliability score below a preset score threshold. After obtaining the reliability scores for each depth layer's flow velocity data, these scores are used to weight the flow velocity data. For example, when calculating the cross-section average flow velocity, depth layer flow velocity data with a high reliability score will be given a higher weight, while depth layer flow velocity data with a low reliability score will be given a lower weight. This ensures that more reliable data has a greater impact on the result when calculating the average flow velocity. In addition, for flow velocity data with a reliability score below a preset score threshold, correction or replacement measures are taken. For example, a preset score threshold can be set, such as 0.5. If the flow velocity data of a certain depth layer has a reliability score below 0.5, the data can be considered unreliable. At this time, various methods can be used for correction or replacement. One correction method is to use higher-reliability data from adjacent depth layers for interpolation or smoothing to correct low-reliability data. Another replacement method is to directly replace the data with a value estimated based on hydrodynamic models or historical average values when the data quality is extremely poor. For example, the flow velocity value of the depth layer can be estimated based on the geometric information of the river section and known hydrodynamic principles, combined with other reliable flow velocity data.

[0089] Finally, the cross-section average flow rate is calculated based on the weighted processed flow rate data, and then combined with the water level data corresponding to the cross-section area of the water passing through, to obtain the instantaneous flow rate. After the flow rate data of each depth layer in the flow rate profile is weighted processed and corrected or replaced, the cross-section average flow rate will be calculated using these optimized flow rate data. For example, the weighted average method can be used to multiply the flow rate value of each depth layer by its corresponding weight, then sum up, and then divide by the total weight, so as to obtain a more accurate cross-section average flow rate. After the cross-section average flow rate is calculated, the cross-section area of the water passing through corresponding to the current water level will be calculated by combining the real-time collected water level data and the pre-mapped geometric information of the river cross-section. Finally, the cross-section average flow rate calculated is multiplied by the cross-section area of the water passing through, so as to obtain the instantaneous flow rate. This method ensures the accuracy and reliability of the flow rate calculation results under complex conditions such as high turbidity of water body.

[0090] Firstly, the present application realizes intelligent identification of the high turbidity state of the water body through the step of "if the turbidity data of the water body is continuously higher than the preset turbidity threshold value within a preset time, and the change amplitude of the water level data is lower than the preset water level change threshold value, and the change amplitude of the flow rate profile data is lower than the preset flow rate profile change threshold value, it is determined that the current water body is in a high turbidity state dominated by fine particles below the water level change threshold value". Unlike the prior art which only relies on a single turbidity index, the present application considers the change trend of turbidity, water level and flow rate profile data, and can more accurately identify the high turbidity state dominated by fine particles which has the greatest impact on acoustic measurement. This multi-parameter linkage judgment mechanism effectively avoids misjudgment caused by the increase of turbidity due to other non-fine particle factors (such as bubbles, large floating objects), and ensures the pertinence and effectiveness of subsequent processing.

[0091] Secondly, after determining that the water body is in a high turbidity state, the present application further quantitatively evaluates the quality of the flow rate data through the step of "analyzing acoustic echo characteristic data to generate reliability scores of the flow rate data of each depth layer in the flow rate profile". The prior art often directly uses damaged flow rate data for calculation, or simply performs interpolation. However, the present application deeply excavates acoustic echo characteristic data (such as signal attenuation rate, signal-to-noise ratio and echo intensity gradient), which directly reflects the propagation characteristics and interference degree of sound waves in turbid water body. By establishing a reliability scoring mechanism, the present application can quantify the degree of influence of turbid water body on the flow rate data of each depth layer, and provide a scientific basis for subsequent data optimization processing.

[0092] Moreover, the present application realizes intelligent optimization of flow velocity data through the step of "weighting each depth layer flow velocity data in the flow velocity profile using reliability scores, and correcting or replacing flow velocity data with reliability scores lower than a preset score threshold". In the prior art, when the quality of flow velocity data decreases, simple averaging or extrapolation is often used, or a preset interpolation algorithm is relied on, which cannot effectively distinguish the reliability differences of data. The present application weights flow velocity data according to reliability scores, ensuring that more reliable data plays a dominant role in calculation. At the same time, for low-reliability data, the present application provides correction or replacement strategies, such as interpolation correction by combining adjacent reliable data, or estimation and replacement using hydraulic principles. This hierarchical and intelligent processing method significantly improves the accuracy and integrity of flow velocity data.

[0093] Finally, the present application calculates the cross-section average flow velocity based on the optimized flow velocity data, and obtains the instantaneous flow by combining the corresponding water passage cross-section area of the water level data. Through the above series of innovative steps, the method of the present application can effectively overcome the influence of high turbidity of water body on the performance of acoustic measurement equipment, significantly improving the accuracy and reliability of flow monitoring under complex hydrological conditions. This not only solves the systematic deviation problem of flow calculation results in the prior art, but also provides more reliable data support for water resource scheduling and environmental management decision-making, embodying the significant progress of the present application in the field of hydrological monitoring technology.

[0094] In some embodiments, the step of analyzing the acoustic echo characteristic data to generate reliability scores of each depth layer flow velocity data in the flow velocity profile when it is determined that the current water body is in a high turbidity state comprises:

[0095] Using at least two different frequency acoustic wave emission and reception modes, acoustic echo characteristic data is collected;

[0096] According to the acoustic echo characteristic data, the acoustic wave attenuation characteristics, scattering intensity spectrum and Doppler shift spectrum width of each frequency band are extracted;

[0097] Comparing the acoustic wave attenuation characteristics, scattering intensity spectrum and Doppler shift spectrum width of each frequency band, a first comparison result reflecting the scattering difference of the medium is obtained;

[0098] According to the first comparison result, the water body medium characteristics are identified, and the weights of the acoustic wave attenuation characteristics, the scattering intensity spectrum and the Doppler shift spectrum width in the calculation of the flow velocity data reliability score are adjusted according to the water body medium characteristics;

[0099] According to the adjusted weights, the flow velocity data reliability scores are generated.

[0100] Among them, the use of at least two different frequency acoustic wave emission and receiving mode, aims to obtain the response difference of different size particles in water to acoustic wave through multi-band acoustic detection. For example, high frequency acoustic wave can be used to detect fine particles, and low frequency acoustic wave can be used to detect larger particles, so as to more comprehensively capture the acoustic characteristics of water medium.

[0101] Further, according to the acoustic echo feature data, the acoustic wave attenuation characteristics, scattering intensity spectrum and Doppler shift spectrum width of each frequency band are extracted. The acoustic wave attenuation characteristics reflect the degree of energy loss due to absorption and scattering in the propagation process, which is closely related to the concentration and nature of suspended particles in water. The scattering intensity spectrum describes the intensity distribution of different frequency acoustic waves scattered by particles in water, which can be used to infer the particle size distribution. The Doppler shift spectrum width is related to the motion state and turbulence intensity of particles in water, and can reflect the stability of the flow velocity profile data.

[0102] Specifically, by comparing the acoustic wave attenuation characteristics, scattering intensity spectrum and Doppler shift spectrum width of each frequency band, the first comparison result reflecting the scattering difference of the medium is obtained. By comparing the attenuation, scattering and frequency shift characteristics of different frequency acoustic waves in the same water body, the medium characteristics such as the type, concentration and particle size distribution of suspended particles in the water body can be more accurately identified. For example, if the high frequency acoustic wave attenuation is significant and the low frequency acoustic wave attenuation is small, it may indicate that fine particles dominate in the water body.

[0103] Therefore, according to the first comparison result, the medium characteristics of the water body are identified. The medium characteristics of the water body can include suspended sediment concentration, particle size distribution, bubble content, etc. Identifying these characteristics is crucial for accurately evaluating the reliability of flow velocity data, because different medium characteristics will have different effects on acoustic flow velocity measurement.

[0104] On this basis, according to the medium characteristics of the water body, the weights of the acoustic wave attenuation characteristics, scattering intensity spectrum and Doppler shift spectrum width in the calculation of the flow velocity data reliability score are adjusted. For example, when it is identified that the concentration of fine particles in the water body is extremely high, the acoustic wave attenuation characteristics may become the main factor affecting the reliability of flow velocity data, so its weight will be adjusted accordingly. When there are a large number of bubbles in the water body, the Doppler shift spectrum width may be more indicative, and its weight will be adjusted. This dynamic weight adjustment mechanism makes the calculation of the reliability score better adapt to the complex and variable water environment.

[0105] Finally, according to the adjusted weights, the flow velocity data reliability score is generated. The score is a quantitative index for evaluating the accuracy and credibility of the flow velocity data at each depth layer in the flow velocity profile.

[0106] The technical solution of the present application introduces a multi-frequency acoustic wave detection and dynamic weight adjustment mechanism to solve the problem of inaccurate reliability evaluation of flow rate data in complex and highly turbid water bodies using traditional single frequency or fixed weight evaluation methods. It is because of the use of at least two different frequency acoustic wave transmission and reception modes that the response of different size particles in the water body to the acoustic wave can be more comprehensively captured, thereby obtaining more abundant acoustic echo characteristic data. By comparing and analyzing these multi-frequency acoustic characteristics (acoustic attenuation characteristics, scattering intensity spectrum and Doppler shift spectrum width), the medium characteristics of the water body, such as the type and concentration of suspended particles, can be accurately identified. Based on the accurate identification of the medium characteristics of the water body, the weight of each acoustic characteristic in the calculation of the reliability score of the flow rate data can be dynamically adjusted, so that the scoring model can adapt to the differences in the influence of the medium of the water body on acoustic measurement under different turbidity conditions. This adaptive weight adjustment mechanism ensures that the reliability score of the flow rate data under the condition of high turbidity dominated by fine particles can more accurately reflect the actual situation, providing a solid foundation for subsequent weighted processing, correction or replacement of flow rate data.

[0107] Through the above technical solution, when it is determined that the current water body is in a high turbidity state, the reliability score of each depth layer flow rate data in the flow rate profile can be more accurately and robustly generated. Compared with evaluation methods that do not consider multi-frequency acoustic characteristics and water body medium differences, the present technical solution significantly improves the accuracy and adaptability of the reliability score through multi-frequency acoustic detection and dynamic weight adjustment based on the medium characteristics of the water body. This makes the subsequent weighted processing and correction of flow rate data more accurate and effective, thereby improving the accuracy of the final instantaneous flow calculation, especially in high turbidity water environments dominated by fine particles, the advantages are more obvious, effectively avoiding measurement errors caused by the complexity of the water body medium.

[0108] In some embodiments, the step of weighting the flow rate data of each depth layer in the flow rate profile using the reliability score comprises:

[0109] identifying a continuous depth layer in the flow rate profile whose reliability score is lower than a preset score threshold;

[0110] analyzing the difference degree of the reliability scores of adjacent layers within the continuous depth layer and the size of the change gradient of the flow rate values of adjacent layers;

[0111] judging whether the flow rate change trend in the continuous depth layer has continuity and regularity according to the difference degree and the change gradient; if the continuity and the regularity are present, assigning a correction weight based on the overall reliability distribution and spatial correlation to the continuous depth layer;

[0112] weighting the flow rate data of each depth layer in the flow rate profile according to the correction weight.

[0113] Specifically, in the flow velocity profile, whether automatically or manually detected, the reliability scores of the flow velocity data at multiple consecutive depth layers from the water surface to the bottom are all below a preset threshold. This preset threshold can be set according to the actual application scenario, sensor performance, and data quality requirements; for example, it can be set to 0.6 or 0.7. The purpose is to locate areas that may be affected by local disturbances or measurement uncertainties.

[0114] Analyzing the degree of difference in reliability scores between adjacent layers within a continuous depth layer, and the magnitude of the gradient of velocity changes in adjacent layers, can be understood as a detailed examination of the data characteristics within the identified continuous depth layers. The degree of difference refers to the numerical difference in reliability scores between adjacent depth layers, which can be measured, for example, by calculating absolute or relative differences. The magnitude of the gradient refers to the rate of change of velocity values ​​between adjacent depth layers, which can be obtained, for example, by calculating the ratio of the velocity difference to the depth difference. The purpose is to assess the inherent consistency and physical plausibility of the data within these low-reliability regions.

[0115] In practical applications, based on the degree of difference and the gradient of change, it is determined whether the velocity change trend in the continuous depth layer has continuity and regularity. For example, a threshold can be set; if the difference in reliability scores between adjacent layers is less than this threshold, and the velocity change gradient is within a reasonable range (e.g., conforming to typical flow profile characteristics), then it is considered to have continuity and regularity. If the continuity and regularity are present, a corrected weight based on the overall reliability distribution and spatial correlation is assigned to the continuous depth layer. This corrected weight differs from simple layer-by-layer weighting; it comprehensively considers the reliability distribution characteristics of the entire velocity profile and the physical correlation between adjacent depth layers, thereby providing a more reasonable and representative weight value for the continuous depth layer.

[0116] Therefore, based on the corrected weights, the velocity data of each depth layer in the velocity profile are weighted, ensuring that even in areas with low local reliability, the velocity data can be more accurately incorporated into the construction of the overall velocity profile, taking into account its spatial context and physical laws.

[0117] The technical solution of the present application effectively solves the problem that the traditional weighted processing may ignore the spatial continuity of the flow velocity profile by introducing the analysis of the data characteristics in the continuous depth layer in the flow velocity profile. Specifically, when a continuous depth layer with a low reliability score is identified, by analyzing the difference degree of the reliability scores of adjacent layers and the size of the flow velocity value change gradient, it can be judged whether the flow velocity change in these low reliability regions still conforms to the physical law of water flow. For example, if a continuous low reliability region has little difference in reliability scores of adjacent layers and the flow velocity change gradient is gentle, it may indicate that the low reliability of this region is not caused by severe physical changes, but may be caused by local measurement noise, but the overall flow velocity trend is still continuous and regular. In this case, simply greatly reducing its weight may cause information loss. By assigning a correction weight based on the overall reliability distribution and spatial correlation, the present application can more intelligently process these data, avoid unnecessary distortion of the overall flow velocity profile due to local low reliability, and thus enable the flow velocity data after weighted processing to more accurately reflect the actual water flow conditions.

[0118] Through the above technical solution, the present application can more finely process the flow velocity profile data, especially in the scene where the water body is in a high turbidity state and local measurement may be disturbed to cause a decrease in reliability score. Compared with the basic technical solution of weighting only according to the reliability score of a single depth layer, the present application can avoid unreasonable correction of the entire flow velocity profile due to local data quality problems by identifying continuous low reliability depth layers and analyzing their internal spatial correlation and flow velocity change law. Thus, even in the case where the reliability score of some depth layers is low, a more reasonable correction weight can be assigned by considering its physical association with adjacent layers, thereby improving the accuracy and robustness of the flow velocity profile weighting processing, and ultimately making the calculation result of the cross-section average flow velocity closer to the true value and improving the reliability of flow monitoring.

[0119] In some embodiments, the step of modifying or replacing the flow velocity data with a reliability score lower than a preset score threshold comprises:

[0120] When the reliability score of the flow velocity data is lower than a first preset threshold, the target flow velocity data is estimated by combining other flow velocity data with a reliability score higher than a second preset threshold and river cross-section geometric information according to the principles of hydraulics to replace the flow velocity data with a reliability score lower than the first preset threshold; or,

[0121] When the reliability score of the flow velocity data is lower than a third preset threshold, the flow velocity data is calculated according to the water level data and a preset water level-flow relationship curve to replace the flow velocity data with a reliability score lower than the third preset threshold.

[0122] Specifically, the first preset threshold can be understood as a critical value indicating a moderate decline in the reliability of flow rate data. For example, when the reliability score of flow rate data is lower than the first preset threshold, it indicates that the data has a certain degree of unreliability, but can still be corrected by combining other relatively reliable information. The second preset threshold is usually higher than the first preset threshold, and is used to screen out relatively reliable flow rate data as a reference basis for hydraulic principle estimation. The river cross-section geometric information refers to the physical parameters such as the shape and size of the river cross-section, which is crucial for estimating flow rate by hydraulic principle. The target flow rate data estimated by hydraulic principle, for example, can utilize the Manning formula, the Chezy formula, or an empirical flow distribution model, combined with known cross-section geometric information and partially reliable flow rate data, to calculate the flow rate value of the target depth layer.

[0123] The third preset threshold is usually lower than the first preset threshold, and is used to indicate a critical value for a serious decline in the reliability of flow rate data. When the reliability score of flow rate data is lower than the third preset threshold, it means that the flow rate data may have completely lost its credibility and is not suitable for processing through local correction. In this case, using the water level data and the preset water level-flow relationship curve to estimate the flow rate data is a more robust alternative technical solution. The water level-flow relationship curve (or rating curve) is established based on historical observation data, reflecting the empirical relationship between the water level and the instantaneous flow of a specific cross-section. Through this curve, the corresponding flow can be directly calculated from the current water level data, and then the cross-section average flow rate or the flow rate data of a specific depth layer can be inversely calculated.

[0124] The technical solution of the present application effectively solves the problem of different degrees of decline in the reliability of flow rate data in high-turbidity water bodies by introducing multiple reliability thresholds and differentiated correction or replacement strategies. When the reliability score of flow rate data is only moderately declined (lower than the first preset threshold but higher than the third preset threshold), it indicates that the data still has some reference value. At this time, by combining other reliable flow rate data and river cross-section geometric information, and using hydraulic principle for estimation, the existing information can be fully utilized to finely correct the locally unreliable data, ensuring that the corrected data is consistent with the actual hydrodynamic characteristics. It is precisely because of this estimation based on physical models that the corrected flow rate data has high rationality and accuracy.

[0125] Further, when the reliability score of the flow rate data seriously decreases (below a third preset threshold), the directly measured data can have been completely distorted, and if local correction is still attempted, the accuracy of the result will be difficult to guarantee. In view of this, the application instead adopts a more macro and robust method, i.e. calculating the flow rate data according to the water level data and a preset water level-flow rate relationship curve. Although this method is based on an empirical relationship, in the case of seriously unreliable data, it can provide a relatively reliable alternative value, avoiding the huge error caused by using highly unreliable measured data. It is precisely due to this hierarchical processing strategy that the application can flexibly select the most suitable processing method according to the actual situation of data reliability, thereby ensuring the accuracy and stability of the final flow calculation.

[0126] Through the above technical solutions, the application can take more fine and adaptive correction or replacement measures according to different degrees of decrease in the reliability of the flow rate data. Compared with a single correction strategy, the application distinguishes between moderately unreliable and seriously unreliable data, and respectively uses estimation based on the principles of hydraulics and calculation based on the water level-flow rate relationship curve, significantly improving the accuracy and robustness of flow rate data processing in complex water body environments. This differentiated processing method not only maximizes the use of effective data information, avoiding the problems of excessive correction or insufficient correction, but also provides a reliable backup technical solution when the data is severely distorted, thereby effectively improving the overall precision and reliability of the instantaneous flow calculation, especially in the high turbidity state dominated by fine particles, where the advantages are more obvious.

[0127] In some embodiments, the step of estimating the target flow rate data by the principles of hydraulics to replace the flow rate data with a reliability score below the first preset threshold comprises:

[0128] Collecting bottom echo signal features of the acoustic measurement device and corresponding near-bottom layer flow rate data;

[0129] Comparing the bottom echo signal features with the near-bottom layer flow rate data to obtain a second comparison result reflecting the influence of the riverbed boundary;

[0130] According to the second comparison result, determining whether the riverbed roughness or the flow resistance coefficient has changed;

[0131] When the riverbed roughness or the flow resistance coefficient changes, quantifying the change amplitude of the riverbed roughness or the flow resistance coefficient;

[0132] According to the quantified change amplitude, adjusting the empirical parameters for hydraulic extrapolation;

[0133] When the reliability score of the flow velocity data is lower than the first preset threshold, the flow velocity data is estimated by a hydraulic principle according to the adjusted empirical parameter and the flow velocity data and the river section geometry information whose reliability score is higher than the second preset threshold, and the flow velocity data whose reliability score is lower than the first preset threshold is replaced.

[0134] Specifically, collecting the bottom echo signal features of the acoustic measurement equipment and the corresponding near-bottom layer flow velocity data refers to obtaining the characteristics of the sound waves reflected by the riverbed, such as echo intensity, spectral width, etc., and the flow velocity data near the riverbed bottom through equipment such as acoustic Doppler current profiler (ADCP). These data can directly reflect the physical characteristics of the riverbed surface and the dynamics of the near-bottom flow.

[0135] Among them, comparing the bottom echo signal features with the near-bottom layer flow velocity data to obtain a second comparison result reflecting the influence of the riverbed boundary can be understood as evaluating the influence of the riverbed boundary on the flow by analyzing the correlation or difference between the bottom echo signal features and the near-bottom layer flow velocity data. For example, the intensity and stability of the echo signal may be related to the flatness of the riverbed and the sediment deposition, while the change of the near-bottom layer flow velocity is directly affected by the riverbed roughness.

[0136] In practical application, according to the second comparison result, it is judged whether the riverbed roughness or the flow resistance coefficient changes, for example, when the bottom echo signal features show that the riverbed surface roughness increases, or the near-bottom layer flow velocity data decreases significantly compared with historical data, it can be inferred that the riverbed roughness or the flow resistance coefficient may have changed. The riverbed roughness is a parameter describing the roughness of the riverbed surface, and the flow resistance coefficient reflects the resistance of the flow through the riverbed.

[0137] Further, when the riverbed roughness or the flow resistance coefficient changes, quantifying the change amplitude of the riverbed roughness or the flow resistance coefficient means converting the above judgment result into specific numerical value through specific algorithm or model, for example, calculating the percentage change or absolute change of roughness coefficient (such as Manning coefficient n) or resistance coefficient (such as Chezy coefficient C). The purpose is to provide quantitative basis for subsequent parameter adjustment.

[0138] Therefore, according to the quantified change amplitude, the empirical parameter for hydraulic extrapolation is adjusted, which means that according to the quantified change amplitude of the riverbed roughness or the flow resistance coefficient, the empirical parameter for flow velocity profile extrapolation in the hydraulic model is corrected. For example, in the Manning formula or Chezy formula, the roughness coefficient or resistance coefficient is a key parameter, and its accuracy directly affects the accuracy of flow velocity estimation. By dynamically adjusting these parameters, the hydraulic model can better adapt to the current riverbed conditions.

[0139] Finally, when the reliability score of the flow velocity data is lower than the first preset threshold, the flow velocity data is estimated by using the adjusted empirical parameter and the flow velocity data whose reliability score is higher than the second preset threshold and the river section geometry information according to the principle of hydraulics, and the flow velocity data whose reliability score is lower than the first preset threshold is replaced. This means that when the flow velocity data needs to be corrected or replaced, the fixed empirical parameter is no longer used, but the parameter adjusted in real time is used for hydraulic estimation, so as to improve the accuracy of the estimation result.

[0140] The technical solution of the present application effectively solves the accuracy problem that may occur when the traditional hydraulic estimation is used in the dynamic change of the riverbed condition by introducing the real-time monitoring and parameter adjustment mechanism for the influence of the riverbed boundary. Specifically, by collecting the bottom echo signal characteristics and the near-bottom flow velocity data, direct information about the riverbed roughness and the near-bottom flow dynamics can be obtained. By comparing and analyzing these data, the change of the riverbed roughness or the flow resistance coefficient can be identified. Once the change is identified, the change amplitude is quantified, and the empirical parameter used for hydraulic extrapolation is dynamically adjusted based on this. Because these empirical parameters can reflect the actual condition of the current riverbed in real time, the flow velocity data estimated by the principle of hydraulics is closer to the true value when the reliability of the flow velocity data is low, thereby significantly improving the accuracy of the correction or replacement of the flow velocity data.

[0141] Through the above technical solution, the present application can dynamically and adaptively adjust the empirical parameter in the hydraulic estimation model, especially in the complex environment where the water body is in a high turbidity state and the riverbed boundary condition may change, the accuracy of the correction or replacement of the flow velocity data can be significantly improved. Compared with the traditional method of using fixed empirical parameters, the present application can perceive the riverbed change in real time and quantify its influence, so that the estimation result is more accurate and reliable, thereby effectively avoiding the flow calculation error caused by the change of the riverbed condition, and improving the robustness and accuracy of the entire water level and flow data processing method.

[0142] In some embodiments, the steps of collecting water level data, flow velocity profile data, acoustic echo feature data, and water turbidity data include:

[0143] Collecting output data of each sensor;

[0144] Analyzing statistical characteristics of the output data of each sensor, the statistical characteristics including at least one of data fluctuation range, mean value drift, data jump frequency, and data missing rate;

[0145] Obtaining response data of each sensor, the response data being obtained by sending a calibration instruction or a physical excitation to the sensor;

[0146] comparing the statistical characteristics and the response data with preset sensor health criteria to obtain a third comparison result;

[0147] identifying an abnormal state of each sensor according to the third comparison result;

[0148] According to the abnormal state, the data collected by the sensor is marked, screened or replaced to ensure the effectiveness and consistency of the collected water level data, flow profile data and acoustic echo feature data, and water turbidity data.

[0149] Specifically, collecting output data of each sensor refers to obtaining original measurement data of each sensor in real time or at a fixed time, such as a water level sensor, a flow profiler, an acoustic Doppler current profiler (ADCP), and a turbidity sensor, etc. These original data are the basis for all subsequent processing.

[0150] Among them, analyzing the statistical characteristics of the output data of each sensor, the statistical characteristics including at least one of data fluctuation range, mean shift, data jump frequency and data missing rate, refers to statistical analysis of the data output by the sensor within a period of time. The data fluctuation range can reflect the stability of the data; the mean shift can indicate whether the sensor has systematic deviation; the data jump frequency can reveal whether the sensor has instantaneous failure or interference; and the data missing rate directly reflects the integrity of the data. Through these statistical characteristics, the working state and data quality of the sensor can be preliminarily judged.

[0151] In practical application, the response data of each sensor is obtained, and the response data is obtained by sending a calibration instruction or a physical excitation to the sensor. For example, a specific calibration instruction can be sent to the sensor to observe whether it returns the expected calibration response; or a known physical excitation (such as testing the turbidity sensor in a water sample with known turbidity) is applied to the sensor to verify its measurement accuracy and response speed. The purpose is to obtain the performance of the sensor under controlled conditions as a direct basis for evaluating its health status.

[0152] Further, comparing the statistical characteristics and the response data with preset sensor health criteria to obtain a third comparison result refers to comparing the statistical characteristics and the response data obtained by the above analysis with the baseline values reflecting the normal working state of the sensor set in advance. For example, if the mean shift of a certain sensor exceeds the preset threshold, or its response time to the calibration instruction is too long, it may indicate that the sensor is abnormal. The purpose is to quantify the deviation of the current state of the sensor from the ideal state.

[0153] According to the third comparison result, the abnormal state of each sensor is identified, which refers to judging whether the sensor has faults, drifts, damages or other abnormal working conditions based on the comparison result. The abnormal state can include but is not limited to sensor failure, data distortion, and decreased measurement accuracy.

[0154] Finally, according to the abnormal state, the data collected by the sensor is marked, screened or replaced to ensure the effectiveness and consistency of the collected water level data, flow profile data, acoustic echo feature data and water turbidity data. For example, the data collected by the sensor identified as abnormal can be marked for subsequent processing; for obviously incorrect or unreliable data, it can be screened and removed from the data set; for missing or damaged data, interpolation, extrapolation or other model estimation methods can be used to replace it to ensure the continuity and integrity of the data. The purpose is to improve the quality of data from the source and provide reliable input for subsequent flow calculation.

[0155] The technical solution of the present application effectively solves the problem of sensor abnormalities or poor data quality in traditional data collection by introducing sensor health monitoring and data quality evaluation mechanism. First, by analyzing the statistical characteristics of the output data of each sensor, potential problems such as data fluctuation anomaly, mean shift, data jump or loss can be found in time. Second, by obtaining the response data of the sensor and comparing it with the preset health benchmark, the abnormal state of the sensor can be more accurately identified. It is because of the comprehensive evaluation of the sensor state that when an abnormality is found, the original data collected can be marked, screened or replaced according to the type and degree of the abnormality. This active data quality management method ensures the effectiveness and consistency of water level data, flow profile data, acoustic echo feature data and water turbidity data from the data source, thereby providing a solid and reliable data foundation for subsequent flow profile reliability evaluation and instantaneous flow calculation.

[0156] Through the above technical solution, the present application can significantly improve the quality and reliability of the collected original data, effectively avoiding the problem of inaccurate subsequent flow calculation results caused by sensor failure or data abnormality. Compared with the technical solution of simple data collection, the present application can timely discover and handle potential problems at the data source through real-time monitoring of the health state of the sensor and dynamic evaluation of the data quality, thereby ensuring the effectiveness and consistency of the input data. This not only improves the accuracy of cross-section average flow velocity and instantaneous flow calculation, but also provides more accurate and reliable data support for hydrological monitoring and management.

[0157] In some embodiments, the step of comparing the statistical characteristics and the response data with the preset sensor health benchmark to obtain a third comparison result comprises:

[0158] collecting output data of each sensor, the output data including the statistical features and the response data;

[0159] obtaining water body environment parameters, the water body environment parameters including water body temperature, salinity, and flow rate;

[0160] obtaining sensor operation parameters, the sensor operation parameters including sensor cumulative working time and calibration history;

[0161] adjusting the preset sensor health benchmark according to the water body environment parameters and the sensor operation parameters;

[0162] comparing the statistical features and the response data with the adjusted sensor health benchmark to obtain a third comparison result.

[0163] The collecting of the output data of each sensor refers to obtaining various data generated by the sensor in a normal working state. These data include statistical features and response data used to evaluate the health status of the sensor. The statistical features can be understood as characteristics obtained by statistical analysis of the output data of the sensor, such as data fluctuation range, mean value drift, data jump frequency, and data missing rate, etc. These features can reflect the stability and reliability of the sensor data output. The response data is the feedback data generated by the sensor after sending specific calibration instructions or applying physical excitation to the sensor. The purpose is to detect whether the response of the sensor to the known input meets the expectation, so as to evaluate its internal working state.

[0164] Further, the obtaining of the water body environment parameters refers to obtaining the environmental condition information of the water body where the sensor is located in real time or periodically. The water body environment parameters include water body temperature, salinity, and flow rate, etc. These parameters have a significant impact on the physical characteristics and measurement accuracy of the sensor. For example, the change of water body temperature may cause the performance drift of the internal electronic components of the sensor, the change of salinity may affect the measurement result of the conductivity sensor, and the change of flow rate may interfere with the measurement of the acoustic Doppler current profiler.

[0165] At the same time, the obtaining of the sensor operation parameters refers to collecting the running history and state information of the sensor itself. The sensor operation parameters include sensor cumulative working time and calibration history. The sensor cumulative working time can reflect the aging degree of the sensor. With the increase of working time, the performance of the sensor may gradually decrease. The calibration history records the time of the last calibration, the calibration result, and the calibration method, etc. information, which is crucial for evaluating the accuracy of the current measurement data of the sensor.

[0166] On this basis, the preset sensor health benchmark is adjusted according to the water body environment parameter and the sensor operation parameter. Specifically, a dynamic model or rule base can be established to correct the original preset sensor health benchmark according to the real-time acquired water body environment parameter and sensor operation parameter. For example, when the water body temperature rises, the threshold of certain statistical characteristics can be appropriately relaxed; when the cumulative working time of the sensor reaches a certain degree, its health benchmark can be tightened to discover potential performance degradation earlier.

[0167] Finally, the statistical characteristics and the response data are compared with the adjusted sensor health benchmark to obtain a third comparison result. Through this dynamic adjustment method, it is ensured that the comparison process can fully consider the actual working conditions and the state of the sensor, so that the third comparison result can more accurately reflect the real health status of the sensor.

[0168] The technical solution of the present application introduces water body environment parameters and sensor operation parameters, and dynamically adjusts the sensor health benchmark based on these parameters, thereby solving the problem of false judgment or missed judgment of sensor abnormal state caused by traditional fixed health benchmark in complex and variable environment. Specifically, the performance of the sensor is not constant, and the statistical characteristics of its output data and the response to the calibration instruction are comprehensively affected by external environment (such as water temperature, salinity, flow rate) and internal state (such as cumulative working time, calibration history). For example, in a low temperature environment, the response speed of certain sensors may slow down or the output noise may increase, and if the fixed health benchmark in a high temperature environment is still used for comparison, normal phenomena may be incorrectly judged as abnormal. Similarly, as the cumulative working time of the sensor increases, its internal components may age, causing gradual performance degradation, and if the health benchmark is not adjusted, this slow performance degradation may not be discovered in time.

[0169] By acquiring water body environment parameters and sensor operation parameters, the present application can build a more comprehensive context for sensor health evaluation. The water body environment parameters provide real-time information about the external working conditions of the sensor, so that the health benchmark can adapt to different environmental challenges. The sensor operation parameters reflect the "life cycle" state of the sensor itself, so that the health benchmark can consider its aging and maintenance history. Thus, the preset sensor health benchmark is no longer static, but can be adaptively adjusted according to the actual working conditions, so that the comparison process is more accurate. When the statistical characteristics and the response data are compared with the adjusted sensor health benchmark, the third comparison result obtained can more accurately identify the real sensor abnormality, avoiding false positive or false negative judgments caused by environmental or aging factors.

[0170] Through the above technical solution, this application enables a more accurate and robust assessment of sensor health status. Since the sensor health benchmark can be dynamically adjusted based on water environment parameters and sensor operating parameters, it effectively avoids misjudgments and missed judgments caused by environmental changes or sensor aging, significantly improving the accuracy of sensor anomaly identification. This not only helps to promptly detect and handle sensor faults, ensuring the validity and consistency of subsequent water level data, flow velocity profile data, acoustic echo characteristic data, and water turbidity data, but also extends the sensor's service life and reduces maintenance costs.

[0171] Through this dynamic adjustment, this application ensures that the assessment of sensor health status can adapt to changes in actual operating conditions, thereby improving the reliability of data acquisition and the accuracy of the entire data processing method.

[0172] In some embodiments, the step of marking, filtering, or replacing the data collected by the sensor based on the abnormal state includes:

[0173] Identify the type of the abnormal state, analyze the degree of the abnormal state, and obtain the historical data performance of the sensor;

[0174] Based on the type of the abnormal state, adjust the priority and specific operation of the marking, filtering, or replacement; based on the degree of the abnormal state, adjust the first intensity of the marking, filtering, or replacement; and based on the historical data performance, adjust the strategy of marking, filtering, or replacement.

[0175] Based on the adjusted priority and specific operation, the adjusted first intensity, and the adjusted strategy, the data collected by the sensor is labeled, filtered, or replaced.

[0176] Specifically, identifying the type of abnormal state refers to distinguishing the specific manifestation of sensor data anomalies, such as data drift, instantaneous spikes, continuous deviations, periodic fluctuations, or complete failure. Analyzing the degree of abnormal state refers to quantifying the severity of the anomaly, such as the magnitude and duration of data deviation from the normal range. Obtaining historical data performance of the sensor involves reviewing the sensor's past operating records, including its calibration history, fault records, and data output characteristics under similar environmental conditions.

[0177] Adjusting the priority and specific operations of marking, filtering, or replacing based on the type of abnormal state means adopting different processing strategies for different types of anomalies. For example, for instantaneous spike data, filtering or replacement may be prioritized; for slowly drifting data, marking and subsequent calibration may be prioritized. Specific operations may include data interpolation, model prediction replacement, or direct deletion.

[0178] In practical applications, adjusting the first intensity of marking, screening or replacing according to the degree of abnormal state refers to determining the intensity of processing according to the severity of the anomaly. For example, a slight anomaly can only be marked or slightly corrected, while a serious anomaly can need to be completely replaced or deleted. The first intensity can be reflected in the confidence of the replaced data, the strictness of the screening or the warning level of the marking.

[0179] In addition, adjusting the strategy of marking, screening or replacing according to the historical data performance refers to using the past running experience of the sensor to optimize the current data processing method. For example, if a certain sensor has historically often appeared a certain type of anomaly in a specific environment, a preset more effective processing strategy can be used when a similar anomaly occurs again. The adjustment of the strategy can include selecting a specific replacement algorithm, adjusting the threshold of screening or changing the detail level of marking, etc.

[0180] The technical solution of the present application overcomes the possible extensive problems in the processing of sensor abnormal data by multi-dimensional and refined analysis of the abnormal state of the sensor, including identifying the type of anomaly, analyzing the degree of anomaly and obtaining the historical data performance. It is precisely due to the in-depth understanding of the abnormal state that the subsequent data processing can be adaptively adjusted according to the specific situation of the anomaly. For example, by distinguishing the type of anomaly, it can avoid taking a one-size-fits-all approach to different types of anomalies; by quantifying the degree of anomaly, it can ensure that the processing intensity matches the severity of the anomaly, avoiding excessive or insufficient processing; and by combining the historical data performance, it can use experience knowledge to further improve the intelligence and effectiveness of the processing strategy. This comprehensive consideration makes the marking, screening or replacing operation of the sensor collected data more targeted and reasonable.

[0181] Through the above technical solution, the effectiveness and consistency of the sensor collected data can be significantly improved. Since the data processing is no longer a simple binary judgment, but a dynamic adjustment based on the type, degree and historical performance of the abnormal state, it can more accurately identify and correct errors or biases in the data, maximize the retention of valid data, and effectively eliminate or correct unreliable data. This not only improves the quality of the original data, but also provides more reliable input for subsequent water level and flow calculation, thereby improving the accuracy, robustness and adaptability of the entire water level and flow data processing method, especially in complex and variable water environments.

[0182] In some embodiments, the step of calculating the cross-section average flow velocity comprises:

[0183] The cross-section average flow velocity is calculated based on the following formula :

[0184]

[0185] wherein, is the flow velocity value of the i-th depth layer, is the reliability score of the i-th depth layer, is the cross-sectional area of the i-th depth layer.

[0186] Specifically, represents the flow velocity value measured or processed for each depth layer after dividing the water depth into multiple discrete depth layers in the flow velocity profile of the water body. represents the reliability score of the i-th depth layer flow velocity data, which reflects the reliability of the flow velocity data in the collection, processing or correction process. The higher the score, the more reliable the data. represents the effective area occupied by the i-th depth layer in the cross section, which depends on the shape of the cross section and the division method of the depth layer. The formula is obtained by multiplying the flow velocity value of each depth layer with its reliability score and the cross-sectional area it represents, then adding up the results of all depth layers, and dividing by the sum of the reliability scores and cross-sectional areas of all depth layers. The cross-sectional average flow velocity is a comprehensive consideration of the flow velocity, reliability and area contribution .

[0187] The technical solution of the present application introduces a reliability score as a weighting factor, which can quantitatively process the quality difference of flow velocity data of different depth layers when calculating the cross-sectional average flow velocity. The traditional method may simply average or area-weight the flow velocity of each depth layer when calculating the cross-sectional average flow velocity, without fully considering the actual reliability of each flow velocity data. When the flow velocity data of some depth layers is low in reliability due to high turbidity or other factors, if not distinguished, it may cause large errors in the final cross-sectional average flow velocity. By incorporating the reliability score into the weighted calculation, the flow velocity data with high reliability contributes more to the average flow velocity, while the flow velocity data with low reliability contributes less to the average flow velocity, thereby effectively reducing the negative impact of low reliability data on the overall calculation result. In addition, in combination with , it ensures that the contribution of different depth layers to the total flow is accurately reflected, making the calculated cross-sectional average flow velocity more representative.

[0188] Through the above technical solution, the reliability difference of flow velocity data of each depth layer in the flow velocity profile can be fully considered when calculating the cross-sectional average flow velocity. This makes the final cross-sectional average flow velocity Not only reflects the spatial distribution characteristics of water flow, more importantly, its accuracy and reliability have been significantly improved. Especially in the case of high turbidity of water body, part of the flow data may exist greater uncertainty, the technical scheme can effectively reduce the interference of low reliability data on the calculation of average flow velocity, thereby improving the precision and stability of instantaneous flow calculation, and providing more reliable data support for hydrological monitoring and water resources management.

[0189] The embodiment of the application also provides a sensor-based water level and flow data processing system, as shown in a sensor-based water level and flow data processing system 100, the system comprises: Figure 2

[0190] A data acquisition module 10 is used for acquiring water level data, flow velocity profile data, acoustic echo feature data and water body turbidity data, wherein the acoustic echo feature data comprises signal attenuation rate, signal-to-noise ratio and echo intensity gradient.

[0191] A state judgment module 20 is used for judging that the current water body is in a high turbidity state dominated by fine particles if the water body turbidity data is continuously higher than a preset turbidity threshold value within a preset time, and the change amplitudes of the water level data and the flow velocity profile data are both lower than their respective preset change threshold values.

[0192] A reliability evaluation module 30 is used for analyzing the acoustic echo feature data to generate a reliability score of each depth layer flow velocity data in the flow velocity profile when it is judged that the current water body is in a high turbidity state dominated by fine particles.

[0193] A data processing module 40 is used for performing weighted processing on each depth layer flow velocity data in the flow velocity profile by using the reliability score, and correcting or replacing the flow velocity data with low reliability score.

[0194] A flow calculation and output module 50 is used for calculating the cross-section average flow velocity based on the weighted processed flow velocity data, and obtaining the instantaneous flow by combining the corresponding water passage cross-section area of the water level data.

[0195] ​The sensor-based water level flow data processing system presented in this application aims to solve the problem of systematic deviation in flow calculation results caused by abnormal changes in water medium characteristics, especially the sharp increase in suspended fine particle concentration leading to water turbidity, in traditional hydrological monitoring systems. The system acquires multi-source sensor data through the data acquisition module, intelligently identifies the high turbidity state of the water body by the state judgment module, and then evaluates the reliability of the flow rate data by the reliability evaluation module. Subsequently, the data processing module uses the reliability score to weight, correct or replace the flow rate data, and finally the flow calculation and output module calculates the cross-section average flow rate based on the optimized flow rate data and obtains the instantaneous flow. Through the coordinated work of each module, this system can effectively improve the accuracy and reliability of flow monitoring under complex hydrological conditions.

[0196] First, the data acquisition module continuously acquires multi-source sensor data, including water level data, flow rate profile data, acoustic echo feature data, and water turbidity data. These data are the basis for all subsequent analysis and processing. For example, the water level meter monitors the water surface height in real time, the acoustic Doppler current profiler (ADCP) obtains flow rate information and acoustic echo characteristics at different depths, and the turbidity sensor quantifies the turbidity of the water body. All these raw data are transmitted to the central processing unit.

[0197] Second, the state judgment module intelligently analyzes the collected data to identify whether the water body is in a high turbidity state dominated by fine particles. This judgment is not solely dependent on turbidity data, but also considers the persistence of water turbidity data within a preset time, the change amplitude of water level data, and the change amplitude of flow rate profile data. When the water turbidity is continuously higher than the preset threshold, and both the water level and flow rate profile data show low change amplitudes, the system can accurately determine that the current water body is in a special high turbidity state, i.e., a state dominated by fine particles and relatively stable flow. This multi-parameter linkage judgment mechanism avoids misjudgment caused by turbidity increase due to other factors (such as bubbles and large floating objects), ensuring the relevance and effectiveness of subsequent processing.

[0198] Then, when the system determines that the water body is in a high turbidity state, the reliability evaluation module is activated to generate reliability scores for the flow rate data at each depth in the flow rate profile. In high turbidity water bodies dominated by fine particles, the propagation of sound waves is severely affected, leading to a decrease in the quality of flow rate data obtained by ADCP. At this time, the system deeply analyzes acoustic echo feature data such as signal attenuation rate, signal-to-noise ratio, and echo intensity gradient. For example, high attenuation rate and low signal-to-noise ratio usually mean that the sound energy is lost and the signal quality is poor, resulting in a decrease in the reliability score of the corresponding depth layer flow rate data. In this way, the system can quantify the degree of influence of turbid water on each depth layer flow rate data, providing a basis for subsequent data processing.

[0199] Subsequently, the data processing module utilizes these reliability scores to weight the flow velocity data at each depth layer in the flow velocity profile, and corrects or replaces the flow velocity data with a reliability score lower than a preset score threshold. In the weighting process, the flow velocity data with a high reliability score is given a higher weight to ensure that it plays a dominant role in calculating the cross-section average flow velocity; while the flow velocity data with a low reliability score is given a lower weight. For those flow velocity data with an extremely low reliability score, the system takes more aggressive measures, such as interpolating correction by combining adjacent depth layer reliable data, or estimation and replacement by using the principles of hydraulics or historical water level-flow relationship curve when the data quality is extremely poor. This stage ensures that the flow velocity data used for flow calculation is optimized and quality controlled.

[0200] Finally, the flow calculation and output module calculates the cross-section average flow velocity based on the weighted and corrected flow velocity data, and then combines the corresponding overwater cross-section area of the real-time water level data to finally obtain the instantaneous flow. Through the above series of intelligent processing, the system of the present application can effectively overcome the influence of high turbidity of water body on the performance of acoustic measurement equipment, and significantly improve the accuracy and reliability of flow monitoring under complex hydrological conditions.

[0201] The sensor-based water level and flow data processing system proposed in the present application exhibits significant technical progress and innovation in solving the flow calculation deviation problem under the condition of high turbidity of water body in the field of hydrological monitoring. The sensor-based water level and flow data processing system of the present application effectively overcomes the above problems through its unique modular design and cooperative working mechanism. The system integrates multi-source data through the data acquisition module, realizes intelligent identification of high turbidity state of water body through the multi-parameter linkage of the state judgment module, and avoids the misjudgment that may be caused by a single indicator. The reliability evaluation module deeply analyzes the acoustic echo characteristic data and quantifies the reliability of the flow velocity data, providing a scientific basis for subsequent processing. The data processing module weights, corrects or replaces the flow velocity data according to the reliability score, significantly improving the accuracy and integrity of the flow velocity data. Finally, the flow calculation and output module calculates the instantaneous flow based on the optimized flow velocity data. This systematic solution not only solves the systematic deviation problem of the flow calculation result in the prior art, but also ensures the accuracy and reliability of flow monitoring under complex hydrological conditions through the close cooperation of each functional module, providing more reliable data support for water resource management.

[0202] The above only describes the embodiments of the present application and is not used to limit the protection scope of the present application. For those skilled in the art, the present application can have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A sensor-based water level flow data processing method, characterized by, The method comprises the following steps: Collecting water level data, flow velocity profile data, acoustic echo characteristic data, and water turbidity data, wherein the acoustic echo characteristic data comprises signal attenuation rate, signal-to-noise ratio, and echo intensity gradient; If the water turbidity data is continuously higher than a preset turbidity threshold value within a preset time, and the water level data change amplitude is lower than a preset water level change threshold value, and the flow velocity profile data change amplitude is lower than a preset flow velocity profile change threshold value, it is determined that the current water body is in a high turbidity state dominated by fine particles, and the water level data change amplitude is lower than the preset water level change threshold value at this time; When it is determined that the current water body is in a high turbidity state, analyzing the acoustic echo characteristic data to generate a reliability score of the flow velocity data of each depth layer in the flow velocity profile; Using the reliability score, performing weighted processing on the flow velocity data of each depth layer in the flow velocity profile, and correcting or replacing the flow velocity data with a reliability score lower than a preset score threshold value; Based on the weighted processed flow velocity data, calculating the cross-section average flow velocity, and then combining the corresponding water passage cross-section area of the water level data to obtain the instantaneous flow rate; The step of analyzing the acoustic echo characteristic data to generate a reliability score of the flow velocity data of each depth layer in the flow velocity profile when it is determined that the current water body is in a high turbidity state comprises: Using at least two different frequency acoustic wave emission and reception modes to collect acoustic echo characteristic data; According to the acoustic echo characteristic data, extracting the acoustic wave attenuation characteristics, scattering intensity spectrum, and Doppler frequency shift spectrum width of each frequency band; Comparing the acoustic wave attenuation characteristics, scattering intensity spectrum, and Doppler frequency shift spectrum width of each frequency band to obtain a first comparison result reflecting the scattering difference of the medium; According to the first comparison result, identifying the water body medium characteristics, and adjusting the weight of the acoustic wave attenuation characteristics, scattering intensity spectrum, and Doppler frequency shift spectrum width in the calculation of the flow velocity data reliability score according to the water body medium characteristics; According to the adjusted weight, generating the flow velocity data reliability score.

2. The sensor-based water level flow data processing method of claim 1, wherein, The step of using the reliability score to perform weighted processing on the flow velocity data of each depth layer in the flow velocity profile comprises: Identifying the continuous depth layers in the flow velocity profile with a reliability score lower than a preset score threshold value; Analyzing the difference degree of the reliability scores of adjacent layers in the continuous depth layers, and the change gradient size of the flow velocity values of adjacent layers; According to the difference degree and the change gradient, judging whether the flow velocity change trend in the continuous depth layers has continuity and regularity; if it has the continuity and the regularity, giving the continuous depth layers a correction weight based on the overall reliability distribution and spatial correlation; According to the correction weight, performing weighted processing on the flow velocity data of each depth layer in the flow velocity profile.

3. The sensor-based water level flow rate data processing method of claim 1, wherein, The step of correcting or replacing the flow velocity data with a reliability score lower than a preset score threshold value comprises: When the reliability score of the flow velocity data is lower than a first preset threshold, the flow velocity data is replaced by target flow velocity data estimated by hydrodynamics principle in combination with other flow velocity data whose reliability score is higher than a second preset threshold and river section geometry information; or when the reliability score of the flow velocity data is lower than a third preset threshold, the flow velocity data is replaced by flow velocity data calculated according to the water level data and a preset water level-flow relationship curve.

4. The sensor-based water level flow rate data processing method according to claim 3, wherein, The method further comprises, before the step of replacing the flow velocity data whose reliability score is lower than the first preset threshold by the target flow velocity data estimated by hydrodynamics principle: collecting bottom echo signal features of the acoustic measuring device and corresponding near-bottom layer flow velocity data; comparing the bottom echo signal features with the near-bottom layer flow velocity data to obtain a second comparison result reflecting the influence of riverbed boundary; judging whether the riverbed roughness or the flow resistance coefficient changes according to the second comparison result; when the riverbed roughness or the flow resistance coefficient changes, quantifying the change amplitude of the riverbed roughness or the flow resistance coefficient; adjusting an empirical parameter used for hydrodynamics extrapolation according to the quantified change amplitude; when the reliability score of the flow velocity data is lower than the first preset threshold, estimating the flow velocity data by hydrodynamics principle according to the adjusted empirical parameter and the flow velocity data whose reliability score is higher than the second preset threshold and the river section geometry information, and replacing the flow velocity data whose reliability score is lower than the first preset threshold by the estimated flow velocity data.

5. The sensor-based water level flow rate data processing method of claim 1, wherein, The method further comprises, in the step of collecting the water level data, the flow velocity profile data, the acoustic echo feature data and the water body turbidity data: collecting output data of each sensor; analyzing statistical features of the output data of each sensor, the statistical features including at least one of data fluctuation range, mean value drift, data jump frequency and data missing rate; obtaining response data of each sensor, the response data being obtained by sending a calibration instruction or a physical excitation to the sensor; comparing the statistical features and the response data with a preset sensor health benchmark to obtain a third comparison result; identifying an abnormal state of each sensor according to the third comparison result; according to the abnormal state, marking, screening or replacing the data collected by the sensor to ensure the effectiveness and consistency of the collected water level data, the flow velocity profile data, the acoustic echo feature data and the water body turbidity data.

6. The sensor-based water level flow rate data processing method of claim 5, wherein, The method further comprises, in the step of comparing the statistical features and the response data with the preset sensor health benchmark to obtain a third comparison result: collecting output data of each sensor, the output data including the statistical features and the response data; obtaining water body environmental parameters, the water body environmental parameters including water temperature, salinity and flow velocity; obtaining sensor operating parameters, the sensor operating parameters including sensor cumulative working time and calibration history; adjusting the preset sensor health benchmark according to the water body environmental parameters and the sensor operating parameters; comparing the statistical features and the response data with the adjusted sensor health benchmark to obtain a third comparison result.

7. The sensor-based water level flow rate data processing method according to claim 5, wherein, The step of marking, screening or replacing the data collected by the sensor according to the abnormal state comprises: identifying the type of the abnormal state, analyzing the degree of the abnormal state, and obtaining the historical data performance of the sensor; adjusting the priority and specific operation of the marking, screening or replacing according to the type of the abnormal state, adjusting the first intensity of the marking, screening or replacing according to the degree of the abnormal state, and adjusting the strategy of the marking, screening or replacing according to the historical data performance; marking, screening or replacing the data collected by the sensor according to the adjusted priority and specific operation, the adjusted first intensity and the adjusted strategy.

8. The sensor-based water level flow rate data processing method of claim 1, wherein, The step of calculating the average flow velocity of the section comprises: The cross-sectional average flow velocity is calculated based on the following formula : ; wherein, is the flow velocity value for the i-th depth layer, is the reliability score for the i-th depth layer, is the cross-sectional area of the water passage represented by the i-th depth layer.

9. A sensor-based water level flow data processing system, characterized by, The system comprises: a data acquisition module, configured to acquire water level data, flow velocity profile data, acoustic echo characteristic data and water body turbidity data, wherein the acoustic echo characteristic data comprises signal attenuation rate, signal-to-noise ratio and echo intensity gradient; a state judgment module, configured to determine that the current water body is in a high turbidity state dominated by fine particles if the water body turbidity data continuously exceeds a preset turbidity threshold value within a preset time, and the change amplitudes of the water level data and the flow velocity profile data are both lower than their respective preset change threshold values; a reliability evaluation module, configured to analyze the acoustic echo characteristic data to generate a reliability score of the flow velocity data of each depth layer in the flow velocity profile when it is determined that the current water body is in the high turbidity state dominated by fine particles; further configured to acquire acoustic echo characteristic data using at least two different frequency acoustic wave emission and reception modes; extract acoustic wave attenuation characteristics, scattering intensity spectrum and Doppler frequency shift spectrum width of each frequency segment according to the acoustic echo characteristic data; obtain a first comparison result for reflecting medium scattering differences by comparing the acoustic wave attenuation characteristics, the scattering intensity spectrum and the Doppler frequency shift spectrum width of each frequency segment; identify water body medium characteristics according to the first comparison result, and adjust the weights of the acoustic wave attenuation characteristics, the scattering intensity spectrum and the Doppler frequency shift spectrum width in the calculation of the flow velocity data reliability score according to the water body medium characteristics; generate the flow velocity data reliability score according to the adjusted weights; a data processing module, configured to perform weighted processing on the flow velocity data of each depth layer in the flow velocity profile using the reliability score, and correct or replace the flow velocity data with a low reliability score; a flow calculation and output module, configured to calculate the average flow velocity of the section based on the weighted processed flow velocity data, and obtain the instantaneous flow by combining the water level data corresponding to the water passing section area.

Images