A hydraulic engineering quality intelligent analysis method and system

Abstract

The application discloses a kind of water conservancy project quality intelligent analysis method and system, it is related to water conservancy project quality intelligent analysis technical field, the method includes the following steps: collecting the basic information of current inland waterway under ship, the ship is divided into regular ship and occasional ship, sensor is equipped on regular ship, the motion state of ship in the process of ship route navigation and route hydrological data are collected in real time, and dam inspection database is established;Based on the route hydrological data collected by different ships in the same inland waterway under the same time period in dam inspection database, construct correlation model to analyze the flow state of route in different time periods, and integrate along time to obtain the cumulative stress of different division ranges of dam in continuous time period.The present application constructs continuous hydrological model by relying on the mobile collection and integration of discrete data of regular ship in navigation route, to intelligently analyze the cumulative stress of dam and identify the risk area to be detected.

Description

Technical Field

[0001] This invention relates to the field of intelligent analysis technology for water conservancy project quality, and more specifically, to an intelligent analysis method and system for water conservancy project quality. Background Technology

[0002] As an important component of water conservancy projects, inland river embankments undertake key functions such as flood control and drainage, waterway protection, and soil and water conservation. Their structural safety is directly related to the safety of life and property of residents along the river and the safety of navigation.

[0003] The hydrological environment of inland waterways is complex, and the cumulative stress in different areas is easily affected by hydrological data. Existing intelligent analysis of dikes mostly adopts methods such as fixed-point sensor deployment and manual inspection by boat. The monitoring points are limited, the coverage is narrow, and data processing and analysis are difficult. Only local discrete hydrological data and stress information can be obtained, and continuous monitoring of the entire waterway and dike line cannot be achieved. At the same time, manual inspection is inefficient, labor-intensive, and easily restricted by weather, water level, and navigation conditions, making it difficult to carry out high-frequency and routine inspections. The monitoring data is mostly single-point data, lacking spatiotemporal continuity, making it impossible to build a complete hydrological analysis model and identify areas with abnormal stress. As a result, preventive maintenance of dikes lacks scientific data support, and it is impossible to integrate multi-source discrete hydrological data to form a continuous spatiotemporal hydrological field. The analysis of dike stress heat zones is insufficient, leading to delayed handling of hidden dangers and increasing the risk of dike accidents. Summary of the Invention

[0004] To address the problems in related technologies, this invention proposes an intelligent analysis method and system for water conservancy engineering quality, which can effectively solve the problems of lagging data analysis of dikes and difficulty in identifying areas with abnormal stress.

[0005] Therefore, the specific technical solution adopted by the present invention is as follows: A method for intelligent analysis of the quality of water conservancy projects, comprising the following steps: S1. Collect basic information on vessels on current inland waterway routes, classify vessels into regular vessels and occasional vessels, equip regular vessels with sensors to collect real-time vessel movement status and hydrological data during navigation, and establish a dike quality inspection database. S2. Based on the hydrological data of different vessels at different sampling times in the same inland waterway within the same time period in the levee quality inspection database, a correlation model is constructed to analyze the water flow status of the waterway in different time periods. The cumulative stress of the levee in different divisions over a continuous time period is obtained by integrating along time. S3. Standardize the cumulative stress in each area of ​​the current dike, output the dike stress heat map by combining the color grading mapping function, and further mark the areas to be detected for anomalies in the stress heat map by combining the total shear stress threshold.

[0006] In a preferred embodiment, S1 includes the following steps: S11. Collect statistics on the navigation information of vessels within n consecutive cycles S of the current inland waterway route, including vessel number, number of cycles of passage within the route, number of cycles of navigation, and vessel attributes. This includes at least [number of vessels] within the n consecutive cycles S. Vessels with a periodic number of passages and a periodic number of voyages within the shipping route that both exceed a preset threshold are classified as regular vessels, while the remaining vessels are classified as occasional vessels. S12. Deploy sensors on regular vessels to collect vessel motion status and raw hydrological data as the vessels navigate the current inland waterway. S13. Establish a dike quality inspection database using MySQL, create inland waterway archives, create archives in the inland waterway archives with the frequently used vessel numbers as the index, record the movement status of the vessels during navigation and the hydrological data of the waterway, and at the same time create corresponding inland waterway dike archives and collect the coordinates of the inland waterway dikes.

[0007] In a preferred embodiment, S12 includes the following steps: S121, among which the sensors regularly deployed on ships include global navigation satellite systems, inertial measurement units and acoustic Doppler current profilers, ship motion status includes longitude, latitude, elevation, roll, pitch, heading and ship three-dimensional velocity, and raw hydrological data includes relative current vectors at different measurement depths; Where the ship's motion state vector is ,in These represent the position vector, attitude angle vector, and ship's three-dimensional velocity, respectively. Representing geographical longitude, geographical latitude, and draft elevation respectively, and attitude angle vector. These represent roll, pitch, and heading, respectively, and the ship's three-dimensional speed. ,in These represent the horizontal velocities relative to the ground in the east and north directions, and the vertical velocities of ascent and descent, respectively. Raw hydrological data ,in Representing depth z k The relative velocity vector at time t represents the sampling time. Combining this with the ship's motion, and correcting for installation errors in the acoustic Doppler current profiler, the shipboard data is converted to a geodetic coordinate system. The ship's velocity is then subtracted to obtain the true water flow velocity in the geodetic coordinate system. This process transforms the relative velocity in the raw hydrological data measured by the acoustic Doppler current profiler into a geodetic coordinate system. ,in These represent the dynamic attitude matrix and the static mounting matrix, respectively. This represents the relative velocity vector of the water flow relative to the ship, measured at sampling time t and depth r, in the coordinate system of the acoustic Doppler current profiler. This represents the vector of motion of the vessel itself.

[0008] In a preferred embodiment, S2 includes the following steps: S21. Based on the data collected from different regular vessels in the same inland waterway within the same time period in the levee quality inspection database, the Gaussian process regression algorithm is used to interpolate and reconstruct the flow field in the spatial and temporal dimensions of the waterway, and a spatiotemporal Gaussian process distribution model of water flow velocity is established. Based on the navigation parameters collected by regular vessels, the continuous full-area and time-by-time water flow velocity field distribution of the inland waterway in the current time period is obtained by interpolation. A simplified boundary layer velocity model of the levee of the inland waterway is constructed. Based on the obtained waterway flow field data, the parallel velocity distribution in the near-wall area along the levee on both sides of the waterway is inverted. S22. Based on the flow velocity status of different levee ranges on both sides, calculate the wall shear stress of the water flow at each location and at each collection time along the inland waterway during the current period. The cumulative shear stress of different levee ranges within the time period is obtained by integrating along time.

[0009] In a preferred embodiment, S21 includes the following steps: S211. Extract the sampling data of different regular vessels at different sampling times under time period T for the same inland waterway from the levee quality inspection database. Reconstruct the water flow field of the waterway by spatiotemporal interpolation using Gaussian process regression. The covariance function of the Gaussian process adopts a spatiotemporally separable kernel function, in which the Matern3 / 2 kernel is used as the spatial kernel function and the squared exponential time kernel is used as the temporal kernel function. The water flow velocity at any spatial location and at any time of the waterway satisfies the prior distribution of the Gaussian process, specifically: ; in Let r be the spatial location and t be the water flow velocity corresponding to the sampling time. Represents a Gaussian process. Represents the spatiotemporal covariance kernel function; A spatiotemporally separable structure is used to construct the covariance function, decomposing the spatiotemporal joint correlation characteristics into a product of spatial and temporal dimension correlations. ,in Represents the space kernel function, Let Euclidean distance be the distance between two spatial points. For time kernel function, The time interval between two sampling moments; S212. Based on the constructed spatiotemporal model of the water flow field under time period T, the mainstream flow velocity data of the entire navigation route within the analysis time period T are obtained through spatiotemporal interpolation. A simplified boundary layer velocity model of the levee is constructed to invert the near-wall parallel flow velocity distribution at various locations s along the levee within time period T at different sampling times t, in order to calculate the wall shear stress. The simplified boundary layer velocity model of the levee is as follows: ; Where s is the longitudinal arc length coordinate along the dike, and z represents the normal height from the dike wall. The near-wall parallel flow velocity at sampling time t within the representative time period T, at position s and normal height z of the levee. This represents the water flow friction velocity at the corresponding position and time, where k is the Karman constant. Let be the surface roughness coefficient at location s of the dike. Represents the transverse attenuation function of water flow. H represents the vertical distance from the dike point to the centerline of the waterway, and H is the fixed characteristic width of the waterway.

[0010] In a preferred embodiment, S22 includes the following steps: S221. Based on the flow velocity states of different sections of the dike on both sides, a calculation model for the instantaneous shear stress of the dike wall is constructed to quantify the dynamic water flow shear force of the dike wall at different times. Based on the water density, the wall friction coefficient at various locations along the dike, and the near-wall parallel flow velocity at a preset reference height, the wall shear stress at various locations along the dike at each time is calculated. ; in, Let ρ be the shear stress of the water flow on the levee wall at position s at sampling time t, and C be the density of the water. f (s) represents the wall friction coefficient at position s of the dike. Represents the preset reference height z ref Near-wall characteristic parallel flow velocity; S222. Integrate the wall shear stress over time during period T to obtain the cumulative shear stress distribution at various locations along the embankment during that period, where: ; The cumulative shear stress at location s on the dike within time period T is represented by the dike region u, which is obtained by dividing the dike into regions at fixed intervals. , where f is the total number of dike areas, the cumulative shear stress of the dike under N consecutive time periods T is calculated, and the cumulative stress of different areas of the current dike under N consecutive time periods T is obtained.

[0011] In a preferred embodiment, step S3 includes the following steps: S31. Based on the cumulative stress of different areas of the current dike under N consecutive time periods T, the stress is standardized using Min-Max standardization. S32. Output the stress heat map of the dike by combining the color grading mapping method, and divide the dike area to be detected for anomalies.

[0012] In a preferred embodiment, S31 includes the following steps: S311. The total shear stress of different areas u of the current inland waterway embankment over N consecutive time periods T. ,in Represents the sampling period t i The total cumulative shear stress in the lower levee area u The total shear stress of the levee area u over N consecutive time periods T; Based on the maximum and minimum total shear stress values ​​of each area of ​​the dike, the cumulative stress variables of each dike section are calculated using Min-Max standardization to obtain the standardized cumulative stress intensity of each area of ​​the dike. The stress intensity ranges from 0 to 1, and the algorithm formula is as follows: ; in, These represent the maximum and minimum values ​​of the total shear stress of the levee within N consecutive time periods T, respectively. This represents the cumulative stress intensity of the dike area u.

[0013] In a preferred embodiment, S32 includes the following steps: S321. Based on the cumulative stress intensity of each divided region of the embankment after standardization, the normalized stress intensity interval [0,1] is mapped to the RGB color space through a color grading mapping function, and the global stress intensity is divided into K gradient levels. Where the boundary threshold q k According to the equal division and grading method, the thresholds at each level are evenly distributed in the interval from 0 to 1: ; Based on the graded threshold range, a unique color vector is matched for each level of stress gradient, and a segmented color mapping rule is constructed. Linear interpolation is used to achieve smooth color gradation within the range, outputting a stress heatmap of the levee. The color mapping expression is: ; in, The base color representing the kth level; S322. Determine the cumulative stress state of each area of ​​the current embankment based on the total shear stress threshold, mark the areas where the total shear stress exceeds the total shear stress threshold as areas to be detected as anomalies, and make a note in the embankment stress heat map.

[0014] A water conservancy project quality intelligent analysis system includes a route data collection module, an association model construction module, and a dike stress intelligent analysis module; The route data collection module collects basic information about vessels on the current inland waterway, including vessel number, number of periodic passages within the route, number of periodic voyages, and vessel attributes. It classifies vessels into regular vessels and occasional vessels, and deploys sensors on regular vessels to collect the vessel motion status and raw hydrological data while navigating the current inland waterway. These sensors include a global navigation satellite system, an inertial measurement unit, and an acoustic Doppler current profiler. The module collects the vessel motion status and hydrological data in real time during the navigation of the route and establishes a dike quality inspection database. The associated model construction module, based on the data collected by different regular vessels in the same inland waterway within the same time period in the levee quality inspection database, uses the Gaussian process regression algorithm to interpolate and reconstruct the flow field in the spatial and temporal dimensions of the waterway, establishes a spatiotemporal Gaussian process distribution model of water flow velocity, and interpolates the continuous full-area, time-by-time water flow velocity field distribution of the inland waterway in the current time period based on the navigation parameters collected by regular vessels. It constructs a simplified boundary layer velocity model of the inland waterway levee, and inverts the parallel velocity distribution of the near-wall region along the levee on both sides of the waterway based on the obtained water flow field data. According to the velocity state of different levee ranges on both sides, it calculates the wall shear stress of the water flow at each position and at each collection time along the levee on both sides of the inland waterway in real time. It obtains the cumulative shear stress of the levee in different ranges within the time period by integrating along time. The intelligent stress analysis module for the dikes is based on the cumulative stress of different areas of the current dike over N consecutive time periods T. It performs standardization processing through Min-Max standardization, outputs a stress heat map of the dikes using a color grading mapping method, and divides the dike areas to be detected for anomalies.

[0015] The beneficial effects of this invention are as follows: 1. This invention classifies vessels operating in inland waterways to identify regularly used vessels, deploys sensors on these vessels to collect hydrological data of the waterway during operation, constructs a waterway hydrological analysis model over a continuous time period, identifies the cumulative stress state of dikes in different areas to identify areas to be detected for anomalies, performs intelligent analysis of dike quality, identifies and filters dike areas under excessive stress for preventive maintenance, thus enhancing practicality. 2. This invention utilizes sensors mounted on regularly navigating inland waterways to collect hydrological data across the entire waterway. Based on the normal operating trajectories of the vessels, it fully covers the entire target waterway, acquiring continuous hydrological observation data distributed throughout the waterway and along the entire dike. This allows for intelligent analysis of the dike stress over continuous time periods, identifying key anomaly areas with high stress, susceptibility to scour and damage, and prominent structural hazards. The final output includes a dike stress heat map and the dike area to be detected, visually and spatially displaying the stress state and risk status of various parts along the dike. 3. This invention constructs the overall hydrological state of the waterway in continuous time and space by integrating, interpolating, and simulating discrete hydrological data collected over a continuous period. Then, through a mechanical model, it calculates the cumulative hydrodynamic load borne by different dike areas in a continuous period, so as to analyze the stress state of the dike in a continuous period and intelligently identify abnormal dike areas, thereby enhancing functionality. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is an overall flowchart of a water conservancy project quality intelligent analysis method according to an embodiment of the present invention; Figure 2 This is a block diagram of an intelligent analysis system for water conservancy project quality according to an embodiment of the present invention; Figure 3 This is a detailed flowchart of step S3 of a water conservancy project quality intelligent analysis method according to an embodiment of the present invention. Detailed Implementation

[0018] To further illustrate the various embodiments, the present invention provides accompanying drawings, which are part of the disclosure of the present invention. These drawings are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these drawings, those skilled in the art should be able to understand other possible implementation methods and the advantages of the present invention. The components in the drawings are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0019] According to an embodiment of the present invention, a method and system for intelligent analysis of the quality of water conservancy projects are provided.

[0020] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments: Example 1: like Figure 1 As shown, according to an embodiment of the present invention, a method for intelligent analysis of water conservancy project quality includes the following steps: S1. Collect basic information on vessels on current inland waterway routes, classify vessels into regular vessels and occasional vessels, equip regular vessels with sensors to collect real-time vessel movement status and hydrological data during navigation, and establish a dike quality inspection database. S11. Collect statistics on the navigation information of vessels within n consecutive cycles S of the current inland waterway route, including vessel number, number of cycles of passage within the route, number of cycles of navigation, and vessel attributes. This includes at least [number of vessels] within the n consecutive cycles S. Vessels with a periodic number of passages and a periodic number of voyages within the shipping route that both exceed a preset threshold are classified as regular vessels, while the remaining vessels are classified as occasional vessels. It should be noted that vessel attributes include passenger ships and cargo ships. The number of voyages per cycle within a route reflects the time commitment of the vessel on that route, while the number of voyages per cycle reflects the frequency of operation of the vessel on that route. n is set to 6, and S is set to 45 to ensure sufficient data to identify recurring vessels and avoid interference from short-term temporary voyages and seasonal occasional vessels. This identifies cargo ships or passenger ships traveling on the current waterway. The cycle setting can also be adjusted based on the length and popularity of the route, consulting experts in the field and using empirical methods. The preset threshold needs to be preset based on the route length and vessel attributes. By collecting and statistically analyzing the number of voyages per cycle and the number of voyages per cycle for passenger and cargo ships operating in different route length ranges, the mean plus one standard deviation is used as the preset threshold. The route length includes short-haul routes less than 100 kilometers. For medium-range routes of 100 km or less and 300 km or more, and long-range routes of 300 km or more, for each ship in the group that was determined to be operating round trips within a historical period, two core regularity indicators were statistically analyzed for passenger ships and cargo ships respectively. These indicators included the number of days the ship actually operated on the route as the periodic voyage count and the total number of round trips completed by the ship on the route as the periodic voyage count. For ships with the same ship attributes within the group, the sample mean and sample standard deviation of the above two indicators were calculated respectively. Based on statistical principles, a qualified threshold for judging whether a ship has regularity was set for each group. The threshold was determined by adding a certain multiple of the standard deviation to the statistical characteristic value to ensure that the threshold can cover most regularly operating ships and exclude ships operating sporadically. Preferably, the mean plus 1 standard deviation was used as the preset threshold.

[0021] S12. Deploy sensors on regular vessels to collect vessel motion status and raw hydrological data as the vessels navigate the current inland waterway. S121, among which the sensors regularly deployed on ships include global navigation satellite systems, inertial measurement units and acoustic Doppler current profilers, ship motion status includes longitude, latitude, elevation, roll, pitch, heading and ship three-dimensional velocity, and raw hydrological data includes relative current vectors at different measurement depths; Where the ship's motion state vector is ,in These represent the position vector, attitude angle vector, and ship's three-dimensional velocity, respectively. Representing geographical longitude, geographical latitude, and draft elevation respectively, and attitude angle vector. These represent roll, pitch, and heading, respectively, and the ship's three-dimensional speed. ,in These represent the horizontal velocities relative to the ground in the east and north directions, and the vertical velocities of ascent and descent, respectively. It should be noted that the system can obtain the ship's geographical location and three-dimensional velocity relative to the ground through the Global Navigation Satellite System, obtain the ship's real-time roll, pitch, and heading attitude angles through the inertial measurement unit, and use the acoustic Doppler current profiler to perform acoustic stratification to detect the relative velocity of the water body to the ship at different water depths. The position vector can determine the ship's absolute geographical location in the Earth coordinate system at any given moment, and the attitude angle vector can be used to construct a rotation matrix to normalize the tilted ship coordinate system to the Earth coordinate system. Roll is the ship's left and right tilt angle, pitch is the ship's bow and stern pitch angle, heading represents the bow's true north yaw angle, and draft represents the vertical height from the sea level reference plane to the ship's keel.

[0022] Raw hydrological data ,in Representing depth z k The relative velocity vector at time t represents the sampling time. Combining this with the ship's motion, and correcting for installation errors in the acoustic Doppler current profiler, the shipboard data is converted to a geodetic coordinate system. The ship's velocity is then subtracted to obtain the true water flow velocity in the geodetic coordinate system. This process transforms the relative velocity in the raw hydrological data measured by the acoustic Doppler current profiler into a geodetic coordinate system. ,in These represent the dynamic attitude matrix and the static mounting matrix, respectively. This represents the relative velocity vector of the water flow relative to the ship, measured at sampling time t and depth r, in the coordinate system of the acoustic Doppler current profiler. This represents the vector of motion of the vessel itself.

[0023] It should be noted that the static installation matrix is ​​a fixed rotation matrix used to transform the vector from the acoustic Doppler current profiler coordinate system to the ship's coordinate system. The specific transformation is determined by the pitch, roll, and heading deviation angles of the acoustic Doppler current profiler on the ship under actual conditions. This is determined through calibration measurements after installation and remains a constant. In other words, by allowing the ship to drift or travel at a constant speed in a straight line / turn in still water, the ship's speed and attitude data from the global navigation satellite system and inertial measurement unit, as well as the acoustic Doppler current profile, are recorded simultaneously. The raw flow velocity data obtained from the acoustic Doppler current profiler is fitted using the least squares method to solve for the heading, pitch, and roll installation deviation angles between the acoustic Doppler current profiler and the ship's coordinate system, thus obtaining the fixed installation rotation matrix. The dynamic attitude matrix is ​​a rotation matrix that changes with time t and is used to transform the vector from the ship's coordinate system to the geodetic coordinate system. It is obtained by synthesizing and transforming the roll, pitch, and heading data obtained in real time by the ship's inertial measurement unit through the ZYX Euler angle rotation matrix. The ship's own motion vector is obtained by differential analysis using the global navigation satellite system.

[0024] S13. Establish a dike quality inspection database using MySQL, create inland waterway archives, create archives in the inland waterway archives with the frequently used vessel numbers as the index, record the movement status of the vessels during navigation and the hydrological data of the waterway, and at the same time create corresponding inland waterway dike archives and collect the coordinates of the inland waterway dikes.

[0025] It should be noted that collecting the coordinates of the levees along inland waterways facilitates subsequent statistical analysis of the stress state of the levees within different coordinate ranges, in order to analyze the quality of the levees.

[0026] S2. Based on the hydrological data of different vessels at different sampling times in the same inland waterway within the same time period in the levee quality inspection database, a correlation model is constructed to analyze the water flow status of the waterway in different time periods. The cumulative stress of the levee in different divisions over a continuous time period is obtained by integrating along time. S21. Based on the data collected from different regular vessels in the same inland waterway within the same time period in the levee quality inspection database, the Gaussian process regression algorithm is used to interpolate and reconstruct the flow field in the spatial and temporal dimensions of the waterway, and a spatiotemporal Gaussian process distribution model of water flow velocity is established. Based on the navigation parameters collected by regular vessels, the continuous full-area and time-by-time water flow velocity field distribution of the inland waterway in the current time period is obtained by interpolation. A simplified boundary layer velocity model of the levee of the inland waterway is constructed. Based on the obtained waterway flow field data, the parallel velocity distribution in the near-wall area along the levee on both sides of the waterway is inverted. S211. Extract the sampling data of different regular vessels at different sampling times under time period T for the same inland waterway from the levee quality inspection database. Reconstruct the water flow field of the waterway by spatiotemporal interpolation using Gaussian process regression. The covariance function of the Gaussian process adopts a spatiotemporally separable kernel function, in which the Matern3 / 2 kernel is used as the spatial kernel function and the squared exponential time kernel is used as the temporal kernel function. The water flow velocity at any spatial location and at any time of the waterway satisfies the prior distribution of the Gaussian process, specifically: ; in Let r be the spatial location and t be the water flow velocity corresponding to the sampling time. Represents a Gaussian process. Represents the spatiotemporal covariance kernel function; It should be noted that, Representing a Gaussian process, a mean of 0 indicates no preset offset in the flow field. The spatiotemporal covariance kernel function is used to characterize the correlation characteristics of the flow velocity at different spatial locations and times. This represents the spatial position vector of any observation within the flight path, and another set of comparative spatial position vectors. These represent the data acquisition time and another set of comparative observation times, respectively.

[0027] A spatiotemporally separable structure is used to construct the covariance function, decomposing the spatiotemporal joint correlation characteristics into a product of spatial and temporal dimension correlations. ,in Represents the space kernel function, Let Euclidean distance be the distance between two spatial points. For time kernel function, The time interval between two sampling moments; It should be noted that the spatial kernel function is used to characterize the spatial correlation characteristics of water flow velocity at different spatial locations, while the temporal kernel function is used to characterize the temporal correlation characteristics of water flow velocity at different times. The Matern3 / 2 kernel is used as the spatial kernel function, leveraging its non-stationary spatial fitting capability to adapt to scenarios with complex waterway topography and uneven gradients in water flow spatial changes, capturing local abrupt changes in water flow characteristics. , in the formula Represents the spatial distance between points. The spatial variance of the water flow field represents the spatial fluctuation range of the flow velocity. The spatial characteristic length scale represents the spatial correlation attenuation rate of water flow. The Matern3 / 2 kernel function can fit the spatial continuous distribution law of water flow across the entire waterway, completing the global interpolation reconstruction of discrete monitoring data. For the time dimension, a quadratic exponential time-series kernel function is selected, using the time interval of ship navigation data collection as the input variable to represent the temporal correlation of inland waterway flow velocity over time. This achieves continuous temporal completion and evolution fitting of discrete time-series hydrological data. The time period T is set to 72 hours, meaning that based on the sampling data of inland waterway vessels at sampling times within 72 hours, a water flow field is constructed for the time period to analyze the state of the dikes on both sides. The spatial variance and spatial characteristic length scale of the water flow field are hyperparameters of the Gaussian process kernel function. The solution is obtained through joint iterative optimization based on the global discrete flow velocity samples collected by sensors on regularly navigable vessels along the inland waterway. Specifically, effective flow velocity observation samples are selected from the historical navigation sensor data of regularly navigable vessels to build the model training sample set, defining the hyperparameter combination. Construct a velocity vector from the observed velocities corresponding to all samples. Within the framework of Gaussian process regression, the observed flow velocity data follows a multivariate Gaussian distribution. The objective loss function is constructed with minimizing the negative logarithmic marginal likelihood function as the optimization objective. ; X is a sample matrix consisting of the spatiotemporal coordinates of all training samples, K θ To minimize the negative logarithmic marginal likelihood objective function based on the spatiotemporally separable kernel function and the current hyperparameter θ, the quasi-Newton method or conjugate gradient method is used. Each iteration includes calculating the covariance matrix, the logarithmic term of the determinant of the covariance matrix, and the product of the matrix inverse and vector based on the current iteration hyperparameters. The gradient of the objective function with respect to the hyperparameters is solved. The hyperparameters are iteratively updated according to the gradient optimization rules. The iteration optimization continues until the objective function converges to a stable interval or reaches the maximum number of iterations preset by the system. Finally, the optimal hyperparameter combination is output, and the spatial variance and spatial characteristic length scale of the water flow field are calibrated.

[0028] S212. Based on the constructed spatiotemporal model of the water flow field under time period T, the mainstream flow velocity data of the entire navigation route within the analysis time period T are obtained through spatiotemporal interpolation. A simplified boundary layer velocity model of the levee is constructed to invert the near-wall parallel flow velocity distribution at various locations s along the levee within time period T at different sampling times t, in order to calculate the wall shear stress. The simplified boundary layer velocity model of the levee is as follows: ; Where s is the longitudinal arc length coordinate along the dike, and z represents the normal height from the dike wall. The near-wall parallel flow velocity at sampling time t within the representative time period T, at position s and normal height z of the levee. This represents the water flow friction velocity at the corresponding position and time, where k is the Karman constant. Let be the surface roughness coefficient at location s of the dike. Represents the transverse attenuation function of water flow. H represents the vertical distance from the dike point to the centerline of the waterway, and H is the fixed characteristic width of the waterway.

[0029] It should be noted that k is the Kármán constant with a value of 0.41. The surface roughness coefficient at location s of the dike is determined empirically by periodically collecting data on the dike material, surface aging, and the state of attached organisms, and consulting experts in the field. This represents the lateral attenuation function of water flow, used to characterize the velocity decay law of water flow propagating laterally from the main channel to the dike sidewall. By relying on sensors mounted on multiple regularly navigating vessels to collect lateral velocity distribution data across the entire waterway cross-section, and statistically analyzing measured lateral velocity samples under different waterway widths and different distances from the dike to the waterway centerline, an empirical attenuation function expression conforming to the current inland waterway navigation flow diffusion characteristics is obtained by fitting the waterway characteristic width and the distance from the dike to the waterway centerline as independent variables, and the measured lateral velocity attenuation ratio as the dependent variable. Specifically, the waterway characteristic width and relative distance from the bank are selected as independent variables, and the dimensionless lateral velocity attenuation ratio is used as the dependent variable. The least squares method is used to fit the lateral velocity distribution data across the entire waterway cross-section to obtain an empirical attenuation function expression conforming to the current inland waterway navigation flow diffusion characteristics, such as exponential attenuation models, polynomial attenuation models, and logarithmic attenuation models. The friction velocity of the water flow at the corresponding location and time is represented by a dynamic coupling relationship model between the friction velocity and the mainstream flow velocity of the waterway, established through a flow direction angle correction mechanism. Where: , Represents the drag coefficient. The average mainstream flow velocity of the airway obtained by reconstructing the flow at sampling time t within the representative time period T. This is the flow direction angle correction function. This represents the angle between the water flow direction and the normal direction of the embankment wall at sampling time t. The mainstream flow velocity is reconstructed using spatiotemporal interpolation of the entire water flow field across the waterway via Gaussian process regression. At each data sampling time, the average velocity of the mainstream flow region in the cross-section of the waterway is statistically calculated, and the corresponding average mainstream flow velocity at that time is output in real time. The drag coefficient needs to be preliminarily determined based on empirical methods and consultation with experts in the field, taking into account the flatness of the concrete surface of the embankment wall and the particle size of the riprap slope. The angle between the water flow direction and the normal direction of the embankment wall needs to be calculated by mathematical projection, statistical averaging, and vector decomposition of the entire water flow field across the waterway along the embankment line and in specific directions. The flow direction angle correction function... .

[0030] S22. Based on the flow velocity state of different levee ranges on both sides, calculate the wall shear stress of the water flow at each location and at each collection time along the inland waterway route during the current period, and obtain the cumulative shear stress of different levee ranges within the time period by integrating along time. S221. Based on the flow velocity states of different sections of the dike on both sides, a calculation model for the instantaneous shear stress of the dike wall is constructed to quantify the dynamic water flow shear force of the dike wall at different times. Based on the water density, the wall friction coefficient at various locations along the dike, and the near-wall parallel flow velocity at a preset reference height, the wall shear stress at various locations along the dike at each time is calculated. ; in, Let ρ be the shear stress of the water flow on the levee wall at position s at sampling time t, and ρ be the density of the water. The coefficient of wall friction at position 's' of the dike. Represents the preset reference height z ref Near-wall characteristic parallel flow velocity; It should be noted that ρ is the density of water, and the density of freshwater is usually taken as a constant of 1000 kg / m³. 3 C f (s) This implicit equation needs to be solved using the Colebrook-White formula and Newton's iterative method, with a preset reference height z. ref The near-wall characteristic parallel flow velocity needs to be transformed into a specific height z near the levee wall using the boundary layer model in S212. ref Parallel velocity distribution on the surface.

[0031] S222. Integrate the wall shear stress over time during period T to obtain the cumulative shear stress distribution at various locations along the embankment during that period, where: ; The cumulative shear stress at location s on the dike within time period T is represented by the dike region u, which is obtained by dividing the dike into regions at fixed intervals. , where f is the total number of dike areas, the cumulative shear stress of the dike under N consecutive time periods T is calculated, and the cumulative stress of different areas of the current dike under N consecutive time periods T is obtained.

[0032] It should be noted that N is set to 20, which, combined with the 72-hour time period T, means that the stress status of the levee along the waterway is statistically analyzed every 2 months. N can also be adjusted according to the actual value of T. The specific preset value can be set by consulting experts in the field, and the total interval should be controlled between 2 and 4 months. The levee is divided into areas at fixed intervals of 10m along the arc length.

[0033] Example 2: like Figure 3 As shown in Figure S3, the cumulative stress in each area of ​​the current dike is standardized, and the stress heat map of the dike is output by combining the color grading mapping function. The area to be detected for anomalies is further marked in the stress heat map by combining the total shear stress threshold. S31. Based on the cumulative stress of different areas of the current dike under N consecutive time periods T, the stress is standardized using Min-Max standardization. S311. The total shear stress of different areas u of the current inland waterway embankment over N consecutive time periods T. ,in Represents the sampling period t i The total cumulative shear stress in the lower levee area u The total shear stress of the levee area u over N consecutive time periods T; Based on the maximum and minimum total shear stress values ​​of each area of ​​the dike, the cumulative stress variables of each dike section are calculated using Min-Max standardization to obtain the standardized cumulative stress intensity of each area of ​​the dike. The stress intensity ranges from 0 to 1, and the algorithm formula is as follows: ; in, These represent the maximum and minimum values ​​of the total shear stress of the levee within N consecutive time periods T, respectively. This represents the cumulative stress intensity of the dike area u.

[0034] It should be noted that, The standardized cumulative stress intensity of the dike location u is fixed in the range of [0,1]. The closer the value is to 1, the higher the cumulative impact load of the water flow on the corresponding dike section u in a continuous period. The closer the value is to 0, the lower the stress load on the corresponding dike section u.

[0035] S32. Output the stress heat map of the dike by combining the color grading mapping method, and divide the dike area to be detected for anomalies; S321. Based on the cumulative stress intensity of each divided region of the embankment after standardization, the normalized stress intensity interval [0,1] is mapped to the RGB color space through a color grading mapping function, and the global stress intensity is divided into K gradient levels. Where the boundary threshold q k According to the equal division and grading method, the thresholds at each level are evenly distributed in the interval from 0 to 1: ; Based on the graded threshold range, a unique color vector is matched for each level of stress gradient, and a segmented color mapping rule is constructed. Linear interpolation is used to achieve smooth color gradation within the range, outputting a stress heatmap of the levee. The color mapping expression is: ; in, The base color representing the kth level; It should be noted that, This represents the base color corresponding to the k-th level. For the RGB color vector corresponding to the k-th level, the color sequence adopts a gradient rule from cool to warm tones. Low stress intensity corresponds to cool colors, and high stress intensity corresponds to warm colors, so as to achieve an intuitive visual distinction between stress strength and low stress intensity. For example, if the cumulative stress intensity level is set to K=5, that is, the stress interval [0,1] is divided into 5 gradient levels. The threshold of each level is calculated according to the equal division method. The low stress intensity interval [0,0.2) is mapped to dark cyan, the lower intensity interval [0.2,0.4) is mapped to green, the medium intensity interval [0.4,0.6) is mapped to yellow, the higher intensity interval [0.6,0.8) is mapped to orange, and the high intensity interval [0.8,1.0] is mapped to red. Through the gradient rule from cool to warm tones, an intuitive visual distinction from low risk to high risk is achieved.

[0036] S322. Determine the cumulative stress state of each area of ​​the current embankment based on the total shear stress threshold, mark the areas where the total shear stress exceeds the total shear stress threshold as areas to be detected as anomalies, and make a note in the embankment stress heat map.

[0037] It should be noted that the total shear stress threshold needs to be determined by collecting O historical total shear stress values ​​for each area of ​​the current route and integrating them into a historical dataset. The collected historical data is then used to determine the total shear stress threshold using the 3 sigma principle. When the total shear stress in a region exceeds the total shear stress threshold for N consecutive time periods T, it indicates an abnormal risk, and further on-site testing of the current levee area is required. Here, O is set to 15.

[0038] Example 3: like Figure 2 As shown, a water conservancy project quality intelligent analysis system includes a route data collection module, an association model construction module, and a dike stress intelligent analysis module. The route data collection module collects basic information about vessels on the current inland waterway, including vessel number, number of periodic passages within the route, number of periodic voyages, and vessel attributes. Vessels are classified into regular vessels and occasional vessels. Sensors are deployed on regular vessels to collect the vessel motion status and raw hydrological data while navigating the current inland waterway. These sensors include the Global Navigation Satellite System, inertial measurement units, and acoustic Doppler current profilers. The module collects the vessel motion status and hydrological data in real time during the route navigation process and establishes a dike quality inspection database. The correlation model construction module, based on the data collected by different regular vessels in the same inland waterway within the same time period in the levee quality inspection database, uses the Gaussian process regression algorithm to interpolate and reconstruct the flow field in the spatial and temporal dimensions of the waterway, and establishes a spatiotemporal Gaussian process distribution model of water flow velocity. Based on the navigation parameters collected by regular vessels, it interpolates to obtain the continuous full-domain, time-by-time water flow velocity field distribution of the inland waterway in the current time period, and constructs a simplified boundary layer velocity model of the inland waterway levee. Based on the obtained waterway flow field data, it inverts the parallel velocity distribution in the near-wall region along the levee on both sides of the waterway. Based on the velocity state of different levee ranges on both sides, it calculates the wall shear stress of the water flow at each position and at each collection time along the levee on both sides of the inland waterway in real time. By integrating along time, it obtains the cumulative shear stress of the levee in different ranges within the time period. The intelligent stress analysis module for dikes is based on the cumulative stress of different areas of the current dike over N consecutive time periods T. It performs standardization processing through Min-Max standardization, and outputs a dike stress heat map by combining color grading mapping, and divides the dike areas to be detected for anomalies.

[0039] In summary, this invention classifies vessels operating in inland waterways to identify regularly used vessels, deploys sensors on these vessels to collect hydrological data during their operation, constructs a hydrological analysis model for the waterway over a continuous time period, identifies the cumulative stress state of levees in different areas to identify areas to be detected for anomalies, performs intelligent analysis of levee quality, and identifies and filters levee areas under excessive stress for preventative maintenance, thus enhancing its practicality. By relying on regularly navigating inland waterways equipped with sensors to collect hydrological data across the entire waterway, and based on the normal operating trajectories of the vessels to fully cover the entire target waterway, it is possible to obtain hydrological observation data continuously distributed throughout the waterway and along the entire dike. This allows for intelligent analysis of the dike stress over continuous time periods, identifying key anomaly areas with high stress, prone to scour and damage, and prominent structural hazards. The final output of the dike stress heat map and the dike areas to be detected are presented in a visual and spatial manner, intuitively displaying the stress status and risk situation of each part along the dike.

[0040] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent analysis of the quality of water conservancy projects, characterized in that, The method includes the following steps: S1. Collect basic information on vessels on current inland waterway routes, classify vessels into regular vessels and occasional vessels, equip regular vessels with sensors to collect real-time vessel movement status and waterway hydrological data during navigation, and establish a dike quality inspection database. S2. Based on the hydrological data of different vessels at different sampling times in the same inland waterway within the same time period in the levee quality inspection database, a correlation model is constructed to analyze the water flow status of the waterway in different time periods. The cumulative stress of the levee in different divisions over a continuous time period is obtained by integrating along time. S3. Standardize the cumulative stress in each area of ​​the current dike, output the dike stress heat map by combining the color grading mapping function, and further mark the areas to be detected for anomalies in the stress heat map by combining the total shear stress threshold.

2. The intelligent analysis method for water conservancy project quality according to claim 1, characterized in that, S1 includes the following steps: S11. Collect statistics on the navigation information of vessels within n consecutive cycles S of the current inland waterway route, including vessel number, number of cycles of passage within the route, number of cycles of navigation, and vessel attributes. This includes at least [number of vessels] within the n consecutive cycles S. Vessels with a periodic number of passages and a periodic number of voyages within the shipping route that both exceed a preset threshold are classified as regular vessels, while the remaining vessels are classified as occasional vessels. S12. Deploy sensors on regular vessels to collect vessel motion status and raw hydrological data as the vessels navigate the current inland waterway. S13. Establish a dike quality inspection database using MySQL, create inland waterway archives, create archives in the inland waterway archives with the regular vessel number as the index, record the movement status of the vessels during navigation and the hydrological data of the waterway, and at the same time create corresponding inland waterway dike archives and collect the coordinates of the inland waterway dikes.

3. The intelligent analysis method for water conservancy project quality according to claim 2, characterized in that, S12 includes the following steps: S121, among which the sensors regularly deployed on ships include global navigation satellite systems, inertial measurement units, and acoustic Doppler current profilers. Ship motion status includes longitude, latitude, elevation, roll, pitch, heading, and three-dimensional velocity of the ship. Raw hydrological data includes relative current vectors at different measurement depths.

4. The intelligent analysis method for water conservancy project quality according to claim 3, characterized in that, S2 includes the following steps: S21. Based on the data collected from different regular vessels in the same inland waterway within the same time period in the levee quality inspection database, the Gaussian process regression algorithm is used to interpolate and reconstruct the flow field in the spatial and temporal dimensions of the waterway, and a spatiotemporal Gaussian process distribution model of water flow velocity is established. Based on the navigation parameters collected by regular vessels, the continuous full-area and time-by-time water flow velocity field distribution of the inland waterway in the current time period is obtained by interpolation. A simplified boundary layer velocity model of the levee of the inland waterway is constructed. Based on the obtained waterway flow field data, the parallel velocity distribution in the near-wall area along the levee on both sides of the waterway is inverted. S22. Based on the flow velocity status of different levee ranges on both sides, calculate the wall shear stress of the water flow at each location and at each collection time along the inland waterway during the current period. The cumulative shear stress of different levee ranges within the time period is obtained by integrating along time.

5. The intelligent analysis method for water conservancy project quality according to claim 4, characterized in that, S21 includes the following steps: S211. Extract the sampling data of different regular vessels at different sampling times under time period T for the same inland waterway from the levee quality inspection database. Reconstruct the water flow field of the waterway by spatiotemporal interpolation using Gaussian process regression. The covariance function of the Gaussian process adopts a spatiotemporally separable kernel function, in which the Matern3 / 2 kernel is used as the spatial kernel function and the squared exponential time kernel is used as the temporal kernel function. The water flow velocity at any spatial location and at any time of the waterway satisfies the prior distribution of the Gaussian process, specifically: ; in Let r be the spatial location and t be the water flow velocity corresponding to the sampling time. Represents a Gaussian process. Represents the spatiotemporal covariance kernel function; S212. Based on the constructed spatiotemporal model of the water flow field under time period T, the mainstream flow velocity data of the entire navigation route and the entire time series within the analysis time period T are obtained through spatiotemporal interpolation. A simplified boundary layer flow velocity model of the embankment is constructed to invert the near-wall parallel flow velocity distribution at each location s along the embankment at different sampling times t within time period T, so as to calculate the wall shear stress.

6. The intelligent analysis method for water conservancy project quality according to claim 5, characterized in that, S22 includes the following steps: S221. Based on the flow velocity state of different sections of the dike on both sides, construct a calculation model for the instantaneous shear stress of the dike wall, quantify the dynamic water flow shear force of the dike wall at different times, and calculate the wall shear stress at each position along the dike at each time based on the density of water, the wall friction coefficient at each position along the dike, and the near-wall parallel flow velocity at a preset reference height. S222. Integrate the wall shear stress over time within time period T to obtain the cumulative shear stress distribution at various locations along the dike during this time period. Divide the dike into regions at fixed intervals to obtain the dike region u. Where f is the total number of dike areas, the cumulative shear stress of the dike is calculated over N consecutive time periods T, and the cumulative stress of different areas of the current dike under N consecutive time periods T is obtained.

7. The intelligent analysis method for water conservancy project quality according to claim 6, characterized in that, S3 includes the following steps: S31. Based on the cumulative stress of different areas of the current dike under N consecutive time periods T, the stress is standardized using Min-Max standardization. S32. Output the stress heat map of the dike by combining the color grading mapping method, and divide the dike area to be detected for anomalies.

8. The intelligent analysis method for water conservancy project quality according to claim 7, characterized in that, S31 includes the following steps: S311. The total shear stress of different areas u of the current inland waterway embankment over N consecutive time periods T. ,in Represents the sampling period t i The total cumulative shear stress in the lower levee area u The total shear stress of the levee area u over N consecutive time periods T; Based on the maximum and minimum values ​​of total shear stress in each area of ​​the dike, the cumulative stress variables of each dike section are calculated using Min-Max standardization to obtain the standardized cumulative stress intensity of each area of ​​the dike, with the stress intensity ranging from 0 to 1.

9. The intelligent analysis method for water conservancy project quality according to claim 8, characterized in that, S32 includes the following steps: S321. Based on the cumulative stress intensity of each divided region of the embankment after standardization, the normalized stress intensity interval [0,1] is mapped to the RGB color space through a color grading mapping function, and the global stress intensity is divided into K gradient levels. Where the boundary threshold q k According to the equal division and grading method, the thresholds of each level are evenly distributed in the interval from 0 to 1; Based on the graded threshold range, a unique color vector is matched for each level of force gradient, a segmented color mapping rule is constructed, and a smooth color gradient within the range is achieved through linear interpolation, outputting a heat map of the levee's force. S322. Determine the cumulative stress state of each area of ​​the current embankment based on the total shear stress threshold, mark the areas where the total shear stress exceeds the total shear stress threshold as areas to be detected as anomalies, and make a note in the embankment stress heat map.

10. A smart analysis system for the quality of water conservancy projects, characterized in that, The system adopts a water conservancy project quality intelligent analysis method as described in any one of claims 1-9, including a route data collection module, an association model construction module, and a dike stress intelligent analysis module.

Images