Slope thin-layer water flow roll wave evolution analysis method and system based on three-dimensional reconstruction

Abstract

The application discloses a slope thin-layer flow roll wave evolution analysis method and system based on three-dimensional reconstruction, and the method comprises the following steps: acquiring a synchronous water depth time sequence of multiple sections along a flow direction; reconstructing a three-dimensional space-time model of a thin-layer flow roll wave based on a space position and a water depth sequence; performing lower envelope surface fitting on wave form data to construct a static water base surface, separating a wave water surface and calculating a wave deviation amount; calculating geometric morphological characteristic parameters such as a shape factor, shape diversity and integral absolute slope of each section based on the deviation amount; and dividing roll wave evolution stages according to the shape factor along a flow direction. The application realizes visualization of a three-dimensional shape of a roll wave and accurate separation of a wave surface, and through construction of a systematic geometric characteristic parameter system, objective and automatic quantitative identification of roll wave evolution stages is realized.

Description

Technical Field

[0001] This invention belongs to the field of soil and water conservation monitoring and hydraulic measurement technology, and specifically relates to a method and system for analyzing the evolution of rolling waves in thin-layer water flow on slopes based on three-dimensional reconstruction. Background Technology

[0002] Rolling waves are a periodic, unstable fluctuation phenomenon generated by thin-layer water flow under specific hydraulic conditions. Their occurrence and development directly affect the prediction of slope erosion intensity and soil loss. Therefore, accurately quantifying and analyzing the evolution characteristics of rolling waves is of great significance for revealing the mechanism of slope erosion and constructing high-precision soil erosion prediction models.

[0003] To achieve the observation and measurement of rolling waves, some automated detection methods have been proposed in the prior art. For example, Chinese invention patent application CN105466527A discloses a thin-layer water flow rolling wave measurement system and method based on electromagnetic sensors. This method uses electromagnetic rolling wave sensors arranged in a straight line along the water flow direction to collect water level changes, and identifies values ​​continuously higher than the average water level as rolling waves, thereby calculating the dynamic parameters of the rolling waves, such as wave height, wave velocity, wave frequency, and kinetic energy. However, such methods are essentially one-dimensional single-point measurements. By simply taking the average water level as the still water reference surface, they are not only difficult to adapt to the stripping of water surface fluctuations under complex micro-topography, but can only deduce the macroscopic dynamic laws of rolling waves, and cannot effectively capture the details of the three-dimensional physical morphological evolution of water waves throughout their occurrence and development.

[0004] For example, Chinese invention patent application CN105444987A discloses a thin-layer water flow rolling wave measurement system and method based on high-definition photography. This method uses a high-definition digital imager to photograph the water flow slope, extracts two-dimensional rolling wave feature lines by analyzing the brightness of the photos, and then calculates the displacement velocity and frequency of the rolling wave feature lines. Although this method achieves non-contact visual measurement, it relies excessively on the surface grayscale features of two-dimensional images and lacks real physical water depth data. It cannot reconstruct the three-dimensional spatial model of the rolling wave, nor can it achieve the physical separation of dynamic water bodies and static water layers. Consequently, it cannot systematically and accurately quantify the three-dimensional geometric morphology of the rolling wave, such as the fullness of the waveform and the steepness of the undulation.

[0005] In summary, under laboratory or field measurement conditions, existing technologies largely rely on the measurement of one-dimensional point dynamic parameters or the deduction of two-dimensional surface images, failing to effectively capture and reconstruct the three-dimensional morphological evolution details of shallow water laminar flows throughout their occurrence and development. Furthermore, existing technologies suffer from a significant deficiency in separating the wave surface from the still water surface in shallow water laminar flows. Although separation methods using the still water surface as a reference surface are theoretically feasible, they have not yet been introduced into the field of three-dimensional wave analysis of thin water laminar flows. The adaptability of its three-dimensional parameters and the verification of its actual separation effect remain unexplored, making it difficult for current research to objectively and quantitatively characterize the morphological evolution of rolling waves. Summary of the Invention

[0006] This invention provides a method and system for analyzing the evolution of rolling waves in thin-layer water flow on slopes based on three-dimensional reconstruction. By fusing synchronous water depth time series from multiple cross sections to construct a three-dimensional spatiotemporal physical model, a baseline simulation separation algorithm is introduced to adaptively fit and separate the dynamically fluctuating water surface from the still water baseline. Furthermore, a geometric morphological feature parameter system based on the fluctuation deviation is constructed to automatically identify the evolutionary stages. This invention aims to solve the problems of existing technologies, such as the inability to effectively capture the three-dimensional morphological evolution details of thin-layer shallow water during its entire development process, the lack of effective measurement methods for separating the dynamically fluctuating surface from the still water surface, and the difficulty in objectively and quantitatively characterizing the physical evolution law of rolling waves.

[0007] To address the aforementioned technical problems, this invention proposes a method for analyzing the evolution of roll waves in thin-layer water flow on slopes based on three-dimensional reconstruction, comprising the following steps: Obtain synchronous water depth sequences for multiple cross-sections distributed along the direction of thin-layer water flow; Based on the spatial location of each cross section and the synchronous water depth sequence, a three-dimensional spatiotemporal model of thin-layer water flow rolling wave is reconstructed and generated. The waveform data in the three-dimensional spatiotemporal model are fitted with a lower envelope surface to construct a still water datum. The undulating water surface is then separated from the still water datum, and the undulating deviation of the undulating water surface relative to the still water datum is calculated. Based on the fluctuation deviation, the geometric morphological characteristic parameters of each cross section are calculated to quantitatively characterize the morphology of the thin-layer water flow rolling wave. Based on the variation pattern of the geometric morphological characteristic parameters of each cross section along the flow path, the evolution stages of the thin-layer water flow rolling wave are identified and divided.

[0008] Preferably, the method for constructing the three-dimensional spatiotemporal model is as follows: The synchronized water depth sequences within the set time window are stacked in chronological order to form depth point cloud matrices for each cross section. The depth point cloud matrices of each section are stitched together in the order of their corresponding spatial locations to generate a point cloud dataset containing three-dimensional information on the location along the course, the lateral location, and the water depth. The point cloud dataset is meshed, and an interpolation method is used to generate a three-dimensional spatiotemporal model of continuous thin-layer water flow rolling waves.

[0009] Preferably, the method further includes a step of eliminating sidewall effect interference before performing lower envelope surface fitting: Determine the position of the lateral centerline of the three-dimensional spatiotemporal model; Using the horizontal centerline as a reference, extend the centerline to both sides by a set width and extract the center band data; The extracted center band data is used as the waveform data for subsequent lower envelope fitting and separation.

[0010] Preferably, the method for constructing the static water substrate is as follows: A virtual flexible reference surface is set up, and a reference surface simulation separation algorithm is used for iterative calculation to make the flexible reference surface fit the waveform data from top to bottom, and the lower envelope surface of the waveform data is fitted as the static water reference surface.

[0011] Preferably, the method for calculating the fluctuation deviation is as follows: Using a preset classification threshold as a standard, the waveform data is divided into a set of undulating surface points and a set of still water surface points, and the Euclidean distance from each point in the set of undulating surface points to the still water datum is calculated as the undulating deviation.

[0012] Preferably, the geometric morphological feature parameters include one or more of shape factor, shape diversity, and integral absolute slope; The shape factor is used to characterize the fullness of the waveform, the shape diversity is used to quantify the degree of fluctuation and variation of the waveform, and the integral absolute slope is used to characterize the overall steepness of the waveform.

[0013] Preferably, the formula for calculating the shape factor is:

[0014] in, Here, H is the shape factor, H is the maximum fluctuation height extracted based on the fluctuation deviation, and L is the effective calculation length of the centerline waveform. This represents the area of ​​the waveform calculated using trapezoidal integrals.

[0015] Preferably, the formulas for calculating the shape diversity and the absolute slope of the integral are as follows:

[0016]

[0017] In the formula, IAS represents the absolute slope of the integral, which is a quantification value for shape diversity, and n is the total number of sampling points. This represents the average height of the fluctuation within the cross-section. , Let be the fluctuation heights of the i-th point and the (i+1)-th point, respectively. This represents the horizontal spatial step size between two adjacent sampling points.

[0018] Preferably, the identification and division of the evolution stages of thin-layer water flow rolling waves specifically involves dividing the stages based on the shape factor values ​​of each cross-section along the water flow direction: when When the flow is in a stable phase, it is characterized by a near-rectangular cross-section and a smooth waveform. when At this time, it is identified as a transition section, characterized by undulations in the water surface and the beginning of rolling waves; when When the waveform is identified as mature, it is characterized by sharp features and fully developed rolling waves.

[0019] A second aspect of the present invention also proposes a three-dimensional reconstruction-based slope thin-layer water flow roll wave evolution analysis system, said system being used to implement the method as described in the first aspect of the present invention, comprising: The data acquisition module is used to acquire synchronous water depth sequences of multiple cross-sections distributed along the direction of thin-layer water flow; The three-dimensional reconstruction module is used to reconstruct and generate a three-dimensional spatiotemporal model of thin-layer water flow rolling waves based on the spatial location of each cross section and the synchronous water depth sequence. The base plane separation module is used to fit the lower envelope surface of the waveform data in the three-dimensional spatiotemporal model, construct a still water base plane, separate the undulating water surface from the still water base plane, and calculate the undulation deviation of the undulating water surface relative to the still water base plane. The feature calculation module is used to calculate the geometric morphological feature parameters of each cross section based on the fluctuation deviation, so as to quantitatively characterize the morphology of the thin-layer water flow rolling wave. The evolution stage identification module is used to identify and classify the evolution stages of the thin-layer water flow rolling wave based on the variation law of the geometric morphological characteristic parameters of each cross section along the flow path.

[0020] Compared with the prior art, the present invention has the following technical effects: 1. The slope thin-layer water flow rolling wave evolution analysis method proposed in this invention integrates one-dimensional time-series water depth data with two-dimensional spatial cross-sectional position, successfully constructing a three-dimensional spatiotemporal model of thin-layer water flow rolling waves, and comprehensively depicting the three-dimensional spatial physical morphology of rolling wave development along the course.

[0021] 2. The slope thin-layer water flow roll wave evolution analysis method proposed in this invention breaks through the limitations of conventional fixed benchmarks by introducing a benchmark surface simulation separation algorithm. It adaptively fits the lower envelope surface of the waveform data, realizes the accurate separation of the undulating water body and the bottom still water layer, and greatly improves the purity of subsequent water depth fluctuation deviation calculation.

[0022] 3. The slope thin-layer water flow roll wave evolution analysis method proposed in this invention designs a mathematical model of waveform geometric morphology features that includes shape factor, shape diversity and integral absolute slope, and accurately transforms the complex and ever-changing transient process of fluid mechanics into objective, quantitative and multi-dimensional morphological numerical indicators.

[0023] 4. The slope thin-layer water flow rolling wave evolution analysis method proposed in this invention establishes a stage mapping relationship based on specific geometric morphological parameter thresholds, overcomes the error of manual subjective observation, and realizes automated qualitative and quantitative identification of the rolling wave stable segment, transition segment and mature segment of the entire life cycle evolution stage. Attached Figure Description

[0024] Figure 1 This is a flowchart illustrating the method described in this invention; Figure 2 This is a schematic diagram of the monitoring node arrangement according to an embodiment of the present invention; Figure 3 This is a schematic diagram of multi-section data spliced ​​in spatial order according to an embodiment of the present invention; Figure 4 This is a schematic diagram illustrating the principle and results of the reference surface simulation separation described in an embodiment of the present invention; Figure 5 This is a schematic diagram of the wave characteristics of the undulating water surface described in an embodiment of the present invention. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with specific embodiments of the present application and with reference to the accompanying drawings.

[0026] Example 1 This embodiment presents a method for analyzing the evolution of roll waves in thin-layer water flow on slopes based on three-dimensional reconstruction. Figure 1 As shown, it includes the following steps one through five: Step 1: Obtain the synchronous water depth sequence of multiple cross sections distributed along the direction of the thin-layer water flow.

[0027] In specific physical experiments or field monitoring scenarios, in order to fully capture the evolution of rolling waves, monitoring nodes need to be arranged along the flow direction (i.e., along the flow path) of the thin layer of water flow on the flume or slope.

[0028] Specifically, multiple measurement sections are set along the long flume, depending on the different measurement objectives and site conditions. For example, Figure 2 As shown, in a typical embodiment of this application, a total of 8 measurement sections (numbered 1-8) are set, and the distance between adjacent sections is set to approximately 1m. A corresponding water depth detection unit, such as a high-frequency liquid level sensor or a high-speed camera measurement component, is configured at each measurement section.

[0029] A high-performance computer is used to collaboratively control the detection units at each cross-section, synchronously acquiring water depth data at a preset sampling frequency. In this embodiment, to ensure the capture of instantaneous morphological changes in rolling waves, a high-frequency sampling frequency of 30 frames per second is set. The computer system synchronously records the continuous water depth changes at each cross-section on the same time axis, thereby obtaining a synchronous water depth time series for each cross-section. The above sampling frequency and cross-section spacing are preferred reference values ​​under the conditions of this embodiment. Those skilled in the art can flexibly adjust them according to specific water flow conditions (such as flow rate and slope) and the required observation accuracy.

[0030] Step 2: Based on the spatial location of each cross section and the synchronous water depth sequence, reconstruct and generate a three-dimensional spatiotemporal model of the thin-layer water flow rolling wave.

[0031] The method for constructing the three-dimensional spatiotemporal model includes the following steps S21 to S23: S21: Stack the synchronous water depth sequences within the set time window in chronological order to form a depth point cloud matrix for each cross section.

[0032] After obtaining the original water depth sequence, the data dimension is first transformed for a single cross section. For each frame of the acquired image or data, the physical water depth value corresponding to each pixel (or each measuring point) in the horizontal direction of the cross section is extracted to form a single-frame water depth vector representing the horizontal water depth distribution.

[0033] Subsequently, a set time window was selected, and all single-frame water depth vectors collected at a specific frequency within that time window were stacked and arranged strictly in chronological order. This transformed the original one-dimensional time series data into a single-section depth point cloud matrix containing two-dimensional features of lateral position and time.

[0034] S22: The depth point cloud matrices of each section are stitched together in the order of their corresponding spatial locations to generate a point cloud dataset containing three-dimensional information on the location along the course, the lateral location, and the water depth.

[0035] After completing the construction of the depth point cloud matrix of all individual sections, the known coordinates of each measurement section in the real physical space are used (for example, section 1 is located at the starting point, section 2 is located at 1m, and so on). The depth point cloud matrices of the above multiple independent sections are then imported into point cloud processing software or arranged and spatially aligned according to their real physical spatial positions using a data stitching algorithm.

[0036] After being stitched together, the data is uniformly mapped to a global three-dimensional coordinate system, forming a massive point cloud dataset. Each discrete point in this dataset has clear information in three dimensions: its position along the flow path (X-axis coordinate corresponding to the direction of water flow), its lateral position (Y-axis coordinate corresponding to the width of the cross-section), and its water depth (Z-axis height coordinate corresponding to the vertical direction, i.e., the aforementioned water depth value).

[0037] S23: The point cloud dataset is meshed, and an interpolation method is used to generate a three-dimensional spatiotemporal model of continuous thin-layer water flow rolling waves.

[0038] Since the point cloud dataset obtained by stitching is spatially discrete (there are physical gaps between the cross sections, such as no direct measurement data within a 1m gap), it cannot be directly used for the wave analysis of continuous water surfaces, so surface reconstruction is required.

[0039] The specific steps are as follows: First, the above point cloud dataset is processed by meshing (rasterization), and a fine raster step size is set (for example, in this embodiment, the raster step size is set to 0.05) to divide the planar mesh.

[0040] Since discrete point clouds cannot exactly fill all grid cells, an averaging method is used to calculate the projected height of each grid cell. This involves taking the average of the depth values ​​of multiple point clouds falling within the same grid cell as the height reference for that cell. For gap grid cells not covered by point cloud data, an interpolation method is used to estimate and fill the gaps using data from surrounding known grid cells. Specific interpolation algorithms can include averaging interpolation or other spatial interpolation algorithms.

[0041] Ultimately, the discrete point cloud was transformed into a smooth, continuous surface, successfully generating a three-dimensional spatiotemporal model of thin-layer water flow roll waves that can intuitively represent the development and morphological evolution of the roll waves along the entire surface, such as... Figure 3 As shown, this provides a high-precision three-dimensional physical basis for subsequent wavefront separation and feature extraction.

[0042] Step 3: Fit the lower envelope surface to the waveform data in the three-dimensional spatiotemporal model to construct a still water base surface, separate the undulating water surface from the still water base surface, and calculate the undulating deviation of the undulating water surface relative to the still water base surface.

[0043] In dynamic experiments or natural environment monitoring of shallow water thin-layer flow, the morphology of water waves is not only affected by gravity and bed friction, but also severely interfered with by the friction of the sidewalls at both sides in laboratory tanks or specific trenches, i.e., the sidewall effect in fluid mechanics. The velocity gradient at the sidewalls differs significantly from that at the center. To obtain waveforms that truly represent the hydraulic characteristics of the mainstream region, spatial data interception and cleaning of the generated three-dimensional model are necessary. Therefore, in this embodiment, before performing lower envelope fitting, the method further includes steps S31 to S33 to eliminate the interference of the sidewall effect: S31: Determine the position of the lateral centerline of the three-dimensional spatiotemporal model. S32: Using the horizontal centerline as a reference, extend the centerline to both sides by a set width and extract the center band data.

[0044] S33: The extracted center band data is used as the waveform data for subsequent lower envelope fitting and separation.

[0045] In practice, the system first locates the geometric center line of the three-dimensional spatiotemporal model in the horizontal direction (i.e., the Y-axis perpendicular to the flow direction) as the extraction reference. Using this center line as the axis of symmetry, a pre-defined physical width is extended to the left and right. For example, in a preferred embodiment of this application, a width of ±0.5cm is set on each side, thereby accurately cutting and extracting a central data band with a total width of 1cm from the flow region of the entire cross-section.

[0046] The central data band perfectly avoids the sidewall regions where the velocity gradient changes drastically. The three-dimensional point cloud contained therein is used as the core waveform data that is highly representative for subsequent separation algorithms. This not only significantly reduces the amount of data that the computer needs to process for the huge point cloud, but also effectively improves the physical authenticity and data reliability of subsequent morphological evolution analysis.

[0047] The method for constructing the static water-bearing base is as follows: A virtual flexible reference surface is set up, and a reference surface simulation separation algorithm is used for iterative calculation to make the flexible reference surface fit the waveform data from top to bottom, and the lower envelope surface of the waveform data is fitted as the static water reference surface.

[0048] Traditional techniques often use a fixed bottom of the water tank as a measurement reference, making it difficult to accurately separate the constant still water layer inherent beneath the dynamic water surface. This embodiment introduces a reference plane simulation separation method, an example of which is the Cloth Simulation Filtering algorithm.

[0049] First, the extracted center band waveform data is flipped in the algorithm space, and a virtual flexible reference surface composed of numerous virtual mass points and springs is set horizontally above it. Its physical properties are similar to a flexible fabric affected by gravity.

[0050] Subsequently, the algorithm parameters are configured according to the specific scale characteristics of the thin-layer water flow rolling wave: the processing mode is set to "steep slope mode adapted to the characteristics of slope water flow", the cloth resolution parameter is set (e.g., set to 0.05 to control the fineness of the flexible reference surface mesh), and the maximum number of iterations is set (e.g., 500 times).

[0051] Iterative calculations are initiated under the set virtual gravity, and the virtual flexible reference surface gradually descends. When the flexible surface descends and touches the inverted waveform point cloud data, the touched area stops descending, while the untouched area continues to fall and undergoes flexible deformation, ultimately adaptively and tightly fitting and enveloping the lower boundary of the entire wave point cloud. The point cloud and reference surface after iterative fitting are then flipped back to their real spatial orientation. This fitting surface constitutes the lower envelope of the waveform data, which perfectly represents the real still water surface after the wave water body has been stripped away, i.e., the still water surface mesh.

[0052] Combination Figure 4 As shown in the schematic diagram of the reference surface simulation separation principle and results, after the iterative processing of the above reference surface simulation separation algorithm, this embodiment successfully separates the continuous thin layer of water flow into two layers in three-dimensional space. Figure 4 The flat zones represent the fitted still water surface, while the undulating zones represent the purified fluctuating water surface. Through... Figure 4 The color scale on the right can be read intuitively. Along the flow direction from top to bottom, the distance from the wave surface point set to the still water base shows a three-dimensional spatial abrupt change from small to large and from gentle to dramatic. This intuitively and conclusively verifies the accuracy and effectiveness of the separation algorithm in extracting complex morphologies of thin-layer water flow.

[0053] The method for calculating the fluctuation deviation is as follows: Using a preset classification threshold as a standard, the waveform data is divided into a set of undulating surface points and a set of still water surface points, and the Euclidean distance from each point in the set of undulating surface points to the still water datum is calculated as the undulating deviation.

[0054] After generating the 3D still water datum mesh, the discrete point cloud data needs to be classified into binary categories based on physical attributes. A classification threshold is preset within the system, for example, set to 0.50 in this embodiment. The system calculates the distance between each 3D coordinate point in the center band waveform data and the fitted still water datum. If the distance is less than or equal to the preset threshold of 0.50, the point is determined to be a set of still water surface points that have not experienced significant jumps; if it is greater than the threshold, the point is determined to belong to a set of undulating surface points that have experienced significant spatial fluctuations.

[0055] Finally, for the 3D points classified as wave surface point sets, the Euclidean distance from their spatial coordinates to the still water surface grid is precisely calculated. Using this spatial Euclidean distance effectively eliminates systematic errors from the slope angle of long flume and the constant depth of bottom water, purely and accurately representing the absolute height of the water surface particles relative to the static water surface. This distance is defined as the wave deviation, providing a clean and accurate underlying data foundation for the subsequent steps of quantifying the waveform's shape factor, diversity, and other geometric characteristics.

[0056] Step 4: Based on the fluctuation deviation, calculate the geometric morphological characteristic parameters of each cross section to quantitatively characterize the morphology of the thin-layer water flow rolling wave.

[0057] After obtaining the pure wave deviation, eliminating the interference of slope and still water layer, and representing the true wavefront spatial undulation height as the vertical distance from the discrete point cloud to the reference plane, the system further transforms the physical 3D model into a mathematical morphological model that can be used for automated computer classification. Traditional dynamics deduction often only focuses on macroscopic average flow velocity and water depth, making it difficult to capture transient changes. However, this embodiment achieves accurate capture of evolution details by performing high-dimensional geometric calculations on the separated wave data.

[0058] The geometric morphological feature parameters include one or more of shape factor, shape diversity, and integral absolute slope; wherein, the shape factor is used to characterize the fullness of the waveform, the shape diversity is used to quantify the degree of fluctuation and variation of the waveform, and the integral absolute slope is used to characterize the overall steepness of the waveform.

[0059] Of the three parameters mentioned above, the shape factor characterizes the plump / sharp profile of the rolling wave in the longitudinal section; shape diversity characterizes the degree of variation in the transition from smooth to turbulent water surface, used to quantify the degree of waveform undulation; and the integral absolute slope characterizes the steepness of the wave surface during rises and falls, used to characterize the overall steepness of the waveform, with a larger IAS value indicating a steeper waveform profile. These three parameters are independent yet complementary, constituting a digital fingerprint describing the complete lifecycle of rolling waves in dynamic shallow laminar flow.

[0060] Specifically, the formula for calculating the shape factor is as follows:

[0061] in, Here, H is the shape factor, H is the maximum fluctuation height extracted based on the fluctuation deviation, and L is the effective calculation length of the centerline waveform. This represents the area of ​​the waveform calculated using trapezoidal integrals.

[0062] In the specific calculation process executed by the computer, the system first traverses all fluctuation deviations within the set effective calculation length L of the centerline, and selects the maximum value as the maximum fluctuation height H.

[0063] Meanwhile, since point cloud data or raster data is spatially discrete, the waveform area... The specific calculation uses the trapezoidal integral formula from calculus for discretization and summation. Its specific implementation logic is as follows:

[0064] In the formula, n is the total number of sampling points. , Let be the fluctuation heights of the i-th point and the (i+1)-th point, respectively, which are the sum of the areas of the tiny trapezoids formed by all adjacent pairs of points. This represents the horizontal spatial step size between two adjacent sampling points.

[0065] Calculated shape factor It is a crucial dimensionless physical quantity that directly maps the fullness of a waveform to its specific physical contour characteristics. For those skilled in the art, this quantization parameter has extremely strong physical indicative significance: specifically, when... When the waveform area occupies only half of the circumscribed rectangle, it exhibits a typical isosceles triangular peak shape; when When α is 1, it indicates that the waveform exhibits a semi-circular characteristic; when α is 1, it indicates that When the waveform area fills the circumscribed rectangle, it indicates that the waveform has a flat top or plateau shape. This specific numerical mapping relationship provides a core quantitative criterion for the machine to automatically identify the rolling wave stage in subsequent steps.

[0066] The formulas for calculating the shape diversity and the absolute slope of the integral are as follows:

[0067]

[0068] In the formula, IAS represents the absolute slope of the integral, serving as a quantification of shape diversity. This represents the average height of the fluctuation within the cross-section.

[0069] The host computer system extracts the fluctuation height of each local point. and Then, substitute the above two sets of statistical and calculus characteristic formulas: For shape diversity This formula is essentially a sample standard deviation calculation model for elevation distribution. It calculates the fluctuation height at each discrete point and the average fluctuation height across the entire cross-section. The sum of squared deviations accurately quantifies the dispersion of water surface undulations. When rolling waves have not yet developed and the water surface is smooth, the height of each point is close to the average value, and this calculated value is extremely small; as wave breakage or local turbulence intensifies, the number of extreme high and low points deviating from the average value increases, and this value will increase significantly.

[0070] For the integral absolute slope (IAS): this formula is obtained by summing up the local small slopes between adjacent points along the direction of water flow. The absolute value is used to calculate the average level. Since real waves have an upstream and downstream side, direct integration would cause the positive and negative values ​​to cancel each other out; therefore, the formula introduces an absolute value to superimpose all tilt variations, truly reflecting the absolute degree of wave surface folding. The larger the calculated IAS value, the steeper the overall waveform profile, and the more pronounced the sawtooth or sharp features of the wave crests; conversely, the smaller the value, the gentler the water surface.

[0071] To further clarify the mapping relationship between the above geometric characteristic parameters and the physical flow state, this embodiment introduces... Figure 5 The diagram shows the waveform characteristics of the undulating water surface. Figure 5 Demonstrates the unit width flow rate of a specific thin-layer water flow Under operating conditions, the extracted center band waveform longitudinal profile is obtained.

[0072] from Figure 5 The correspondence between the horizontal axis (position along the path) and the vertical axis (fluctuation height h) clearly confirms the evolutionary stage division logic in step five: in the front-end region (corresponding to the stable segment), The waveform profile is approximately rectangular, with almost no significant vertical displacement of the water surface; as the water flow advances towards the middle section (corresponding to the transition section), ), Figure 5 The middle begins to show continuous, slight peaks and troughs; and as the water flow advances to Figure 5 The back-end area (corresponding to the mature stage), The waveform evolves into multiple extremely tall, steep, and asymmetrically sharp isolated peaks (serrated / isosceles triangles). Figure 5 This visually confirms the shape factor constructed in this application. A mathematical quantification system can fit and reflect the objective laws governing the development and evolution of physical water flow waves.

[0073] Step 5: Based on the variation law of the geometric morphological characteristic parameters of each cross section along the flow path, identify and classify the evolution stages of the thin-layer water flow rolling wave.

[0074] After acquiring the geometric morphological characteristic parameters of all measurement sections, the system sequentially arranges these parameters according to the spatial position of the corresponding sections in the direction of water flow (such as the physical coordinates along the flow path numbered 1 to 8 in Example 1), and plots the flow path variation curves or constructs a state characteristic matrix. By tracking the dynamic evolution trajectory of the characteristic parameters from upstream to downstream, the system can overcome the subjective limitations of traditional manual observation and automatically and objectively identify the entire three-dimensional physical evolution life cycle of shallow water thin-layer flow roll waves from nothing to something, from slight undulations to violent fluctuations.

[0075] The specific division method is to divide the stages according to the shape factor values ​​of each cross-section along the water flow direction: when When the flow is in a stable phase, it is characterized by a near-rectangular cross-section and a smooth waveform. when At this time, it is identified as a transition section, characterized by undulations in the water surface and the beginning of rolling waves; when When the waveform is identified as mature, it is characterized by sharp features and fully developed rolling waves.

[0076] In the classification and discrimination module of the actual system, the shape factor is used as the core absolute quantitative criterion for evolutionary stage identification because it possesses extremely excellent dimensionless properties and is most directly sensitive to the mapping of flow contour features. Its specific computer recognition mapping logic is as follows: In the upstream starting region, such as measurement section 1-2 near the water supply end, the system detected that the waveform data of the water flow longitudinal profile filled most of the area of ​​the theoretically circumscribed rectangle. At this time... A value greater than 0.8 corresponds to a physical flow state where the water surface has not yet been significantly agitated by the imbalance between the bottom boundary layer frictional resistance and the gravitational component. At this point, the water wave fluctuations are minimal, and the system automatically marks it as a stable segment.

[0077] As the thin layer of water flows downstream, for example at measurement sections 3-5, the instability of the fluid dynamics intensifies along the way. The disturbance energy of the water waves begins to accumulate, and wave crests and troughs gradually appear, causing the proportion of the actual water body's wave area to decrease. When the system monitors a certain section or a series of sections... When the value falls into the range of 0.7 to 0.8, the system immediately triggers the identification signal of the transition segment, indicating that the rolling wave in this spatial position has broken away from the stable state and entered the intermittent formation and initial development period.

[0078] In the more downstream regions, after sufficient evolution, such as at measurement sections 6-8, the water ripples violently, the waveform is severely stretched and contracted, exhibiting an extreme isosceles triangle or a steep, sharp protrusion shape, with the waveform area occupying a very small proportion of the circumscribed rectangle. Once the system determines... If the value is consistently less than or equal to 0.7, the judgment result of the mature stage will be output immediately, indicating that the morphological evolution of the rolling wave has reached a peak state of full development and extreme erosive and destructive power.

[0079] In practical applications where the system executes the above-mentioned phase division, the shape diversity calculated in the early stages... While the integral absolute slope (IAS) is not used as a direct absolute threshold criterion, it is invoked by the system as an auxiliary verification and multi-dimensional validation indicator for evolutionary trends. For example, when the system utilizes... When it is detected that the water flow has entered the mature section, the data of that section will be retrieved and compared simultaneously. With IAS values; usually accompanied by The value decreases stepwise. The numerical curves of IAS will show a significant increase in the abrupt change response along the flow path. The mutual corroboration of these three characteristics greatly improves the accuracy and anti-interference capability of the automated identification system for flow regime evolution stages.

[0080] Furthermore, it should be noted that the above classification thresholds (0.7, 0.8) are typical experimental conditions of this embodiment (specific tank slope, bed surface roughness, and thin-layer water flow rate, as shown in the embodiment). The preferred benchmark reference value obtained under (etc.). Those skilled in the art, when facing natural slope erosion environments or irregularly shaped experimental flumes under different working conditions, can fully utilize the method framework provided in this application and the preliminary calibration experimental data to determine the optimal benchmark reference value. The threshold is adaptively fine-tuned (e.g., adjusted to a dividing line between 0.75 and 0.85) to adapt to specific actual water flow and waveform characteristics. Such equivalent parameter adjustments based on the core concept of this invention also fall within the scope of protection of this invention.

[0081] Example 2 This embodiment is a slope thin-layer water flow roll wave evolution analysis system based on three-dimensional reconstruction. The system is used to implement the method described in Embodiment 1, including: The data acquisition module is used to acquire synchronous water depth sequences of multiple cross-sections distributed along the direction of thin-layer water flow; The three-dimensional reconstruction module is used to reconstruct and generate a three-dimensional spatiotemporal model of thin-layer water flow rolling waves based on the spatial location of each cross section and the synchronous water depth sequence. The base plane separation module is used to fit the lower envelope surface of the waveform data in the three-dimensional spatiotemporal model, construct a still water base plane, separate the undulating water surface from the still water base plane, and calculate the undulation deviation of the undulating water surface relative to the still water base plane. The feature calculation module is used to calculate the geometric morphological feature parameters of each cross section based on the fluctuation deviation, so as to quantitatively characterize the morphology of the thin-layer water flow rolling wave. The evolution stage identification module is used to identify and classify the evolution stages of the thin-layer water flow rolling wave based on the variation law of the geometric morphological characteristic parameters of each cross section along the flow path.

[0082] The above description is only a preferred embodiment of the present invention. It should be noted that those skilled in the art can make several modifications and improvements without departing from the inventive concept of the present invention, and these all fall within the protection scope of the present invention.

Claims

1. A method for analyzing the evolution of roll waves in thin-layer water flow on slopes based on three-dimensional reconstruction, characterized in that, Includes the following steps: Obtain synchronous water depth sequences for multiple cross-sections distributed along the direction of thin-layer water flow; Based on the spatial location of each cross section and the synchronous water depth sequence, a three-dimensional spatiotemporal model of thin-layer water flow rolling wave is reconstructed and generated. The waveform data in the three-dimensional spatiotemporal model are fitted with a lower envelope surface to construct a still water datum. The undulating water surface is then separated from the still water datum, and the undulating deviation of the undulating water surface relative to the still water datum is calculated. Based on the fluctuation deviation, the geometric morphological characteristic parameters of each cross section are calculated to quantitatively characterize the morphology of the thin-layer water flow rolling wave. Based on the variation pattern of the geometric morphological characteristic parameters of each cross section along the flow path, the evolution stages of the thin-layer water flow rolling wave are identified and divided.

2. The method according to claim 1, characterized in that, The method for constructing the three-dimensional spatiotemporal model is as follows: The synchronized water depth sequences within the set time window are stacked in chronological order to form depth point cloud matrices for each cross section. The depth point cloud matrices of each section are stitched together in the order of their corresponding spatial locations to generate a point cloud dataset containing three-dimensional information on the location along the course, the lateral location, and the water depth. The point cloud dataset is meshed, and an interpolation method is used to generate a three-dimensional spatiotemporal model of continuous thin-layer water flow rolling waves.

3. The method according to claim 1, characterized in that, Before performing lower envelope fitting, the method further includes a step of eliminating sidewall effect interference: Determine the position of the lateral centerline of the three-dimensional spatiotemporal model; Using the horizontal centerline as a reference, extend the centerline to both sides by a set width and extract the center band data; The extracted center band data is used as the waveform data for subsequent lower envelope fitting and separation.

4. The method according to claim 1 or 3, characterized in that, The method for constructing the static water-bearing base is as follows: A virtual flexible reference surface is set up, and a reference surface simulation separation algorithm is used for iterative calculation to make the flexible reference surface fit the waveform data from top to bottom, and the lower envelope surface of the waveform data is fitted as the static water reference surface.

5. The method according to claim 1 or 3, characterized in that, The method for calculating the fluctuation deviation is as follows: Using a preset classification threshold as a standard, the waveform data is divided into a set of undulating surface points and a set of still water surface points, and the Euclidean distance from each point in the set of undulating surface points to the still water datum is calculated as the undulating deviation.

6. The method according to claim 1, characterized in that, The geometric morphological feature parameters include one or more of the following: shape factor, shape diversity, and absolute slope of integral. The shape factor is used to characterize the fullness of the waveform, the shape diversity is used to quantify the degree of fluctuation and variation of the waveform, and the integral absolute slope is used to characterize the overall steepness of the waveform.

7. The method according to claim 6, characterized in that, The formula for calculating the shape factor is: in, Here, H is the shape factor, H is the maximum fluctuation height extracted based on the fluctuation deviation, and L is the effective calculation length of the centerline waveform. This represents the area of ​​the waveform calculated using trapezoidal integrals.

8. The method according to claim 6, characterized in that, The formulas for calculating the shape diversity and the absolute slope of the integral are as follows: In the formula, IAS represents the absolute slope of the integral, which is a quantification value for shape diversity, and n is the total number of sampling points. This represents the average height of the fluctuation within the cross-section. , Let be the fluctuation heights of the i-th point and the (i+1)-th point, respectively. This represents the horizontal spatial step size between two adjacent sampling points.

9. The method according to claim 1, characterized in that, The identification and classification of the evolution stages of thin-layer water flow rolling waves specifically involves dividing the stages based on the shape factor values ​​of each cross-section along the water flow direction: when When the flow is in a stable phase, it is characterized by a near-rectangular cross-section and a smooth waveform. when At this time, it is identified as a transition section, characterized by undulations in the water surface and the beginning of rolling waves; when When the waveform is identified as mature, it is characterized by sharp features and fully developed rolling waves.

10. A slope thin-layer water flow roll wave evolution analysis system based on three-dimensional reconstruction, characterized in that, The system is used to implement the method as described in any one of claims 1-9, comprising: The data acquisition module is used to acquire synchronous water depth sequences of multiple cross-sections distributed along the direction of thin-layer water flow; The three-dimensional reconstruction module is used to reconstruct and generate a three-dimensional spatiotemporal model of thin-layer water flow rolling waves based on the spatial location of each cross section and the synchronous water depth sequence. The base plane separation module is used to fit the lower envelope surface of the waveform data in the three-dimensional spatiotemporal model, construct the still water base plane, separate the undulating water surface from the still water base plane, and calculate the undulation deviation of the undulating water surface relative to the still water base plane. The feature calculation module is used to calculate the geometric morphological feature parameters of each cross section based on the fluctuation deviation, so as to quantitatively characterize the morphology of the thin-layer water flow rolling wave. The evolution stage identification module is used to identify and classify the evolution stages of the thin-layer water flow rolling wave based on the variation law of the geometric morphological characteristic parameters of each cross section along the flow path.

Images