A Dam-Break Experimental Device and Method Based on Assembly Configuration Recognition and Visual Monitoring

Abstract

This invention belongs to the technical field of hydraulic experimental model facilities, and discloses a dam-break experimental device and method based on assembly configuration recognition and visual monitoring. The device includes: an aluminum profile support frame, an acrylic water tank, a gate assembly, a flexible plate or flexible gate, a lifting system, an electromagnet, and a computer vision monitoring system. This invention addresses the experimental requirements of three types: structural instability, progressive flexible failure, and impact response. It avoids the influence of equipment intrusion on experimental results and can obtain highly repeatable and high-fidelity continuous deformation data of the fluid-structure interaction, providing standardized verification data for fluid-structure interaction numerical models.

Description

Technical Field

[0001] This invention belongs to the technical field of hydraulic experimental model facilities, and particularly relates to a dam-break experimental device and method based on assembly configuration recognition and visual monitoring. Background Technology

[0002] Most existing dam-break experimental setups are geared towards studying the instantaneous failure of rigid structures. They typically only simulate the free surface evolution process after a classic rigid gate is opened, making it difficult to account for the gradual deformation and instability process of flexible gates under continuous hydrostatic pressure, and also difficult to simulate the transient impact response of dam-break flow on deformable components such as flexible plates. In addition, flexible structures under the action of water flow are usually accompanied by complex interferences such as large deformation, local occlusion, droplet splashing, and optical refraction. Contact displacement sensors can only acquire signals from a small number of discrete points, and the size of the sensors themselves can easily interfere with the local flow field and the true response of the flexible structure. Post-processing techniques that rely solely on conventional visual acquisition are prone to frequent manual adjustments of frame switching frequency, color threshold, and detection area due to changes in experimental modes, resulting in low processing efficiency and insufficient result stability.

[0003] Therefore, there is an urgent need for a fluid-structure interaction experimental device and method that can achieve switching between multiple dam-break modes on the same device platform, while taking into account both highly repeatable instantaneous opening and closing and high-precision non-contact monitoring. Summary of the Invention

[0004] To overcome the problems existing in related technologies, this invention discloses an embodiment of a dam-break experimental device and method based on assembly configuration recognition and visual monitoring, specifically involving a multi-mode dam-break experimental device and method based on assembly configuration recognition and adaptive visual monitoring. This invention solves the problems of single experimental modes, difficulty in obtaining continuous large deformations of flexible structures, and the need for multiple adjustments of visual parameters in existing technologies. It analyzes the fluid-structure interaction effect between flexible structures and water flow, ensuring the accuracy of hydrodynamic parameters and providing a new approach for hydraulic and fluid-structure interaction experimental models.

[0005] The technical solution is as follows: a dam-break experimental device based on assembly configuration recognition and visual monitoring, the device comprising: Aluminum profile support frame is used to fix and support various components to ensure liquid level stability; Acrylic water tanks, fixed on aluminum profile support frames, provide experimental sites for different dam break experiments through different channels reserved on different side walls; The gate assembly, by comprising a first rigid gate, a second rigid gate, a rigid connecting plate, and a flexible gate fixedly connected to the rigid connecting plate, is used to simulate rigid dam failure experiments and flexible gate dam failure experiments. The flexible plate is used in conjunction with the first rigid gate to simulate the impact of dam-break flow on the flexible plate. A lifting system is used to control the opening of the first rigid gate or the second rigid gate; An electromagnet, in conjunction with the lifting system, is used to attract and trigger the instantaneous release action of the lifting system before the experiment begins; The computer vision monitoring system includes a synchronous triggering and dynamic vision closed-loop control module based on the assembly configuration state vector. This module is used to identify the installation status of each component in different channels by acquiring static images, and to form an assembly configuration state vector by combining different colored dots and their initial positions. Based on this state vector, the system automatically determines the current experimental mode and calls the corresponding camera frame rate, exposure time, initial ROI, dynamic ROI search range, color enhancement threshold, marker extraction threshold, connected component area threshold, centroid displacement abrupt change threshold, and weight release height verification parameters to verify whether the current experimental mode meets the instantaneous dam failure condition.

[0006] Furthermore, the aluminum profile support frame consists of aluminum profiles for fixing the water tank, main aluminum profiles, and aluminum profiles for supporting the counterweights; An acrylic water tank is fixed to an optical platform supporting the water tank by an aluminum profile, and includes a bottom wall, a first side wall, a second side wall, a third side wall, and a fourth side wall; wherein the third side wall and the fourth side wall are provided with a first channel, a second channel, and a third channel, and the bottom wall is provided with a fourth channel.

[0007] Furthermore, the gate assembly includes a first rigid gate corresponding to the first channel, a second rigid gate corresponding to the second channel, a rigid connecting plate corresponding to the third channel, and a flexible gate fixedly connected to the rigid connecting plate. The flexible plate is fixed inside the fourth channel; The lifting system includes a strong nylon rope, a directional pulley, and a counterweight. One end of the strong nylon rope is connected to the top of the first rigid gate or the second rigid gate, and the other end passes around the directional pulley and is connected to a five-kilogram counterweight, which is used to control the opening of the first rigid gate or the second rigid gate. The computer vision monitoring system also includes colored markers, a camera, and a processing computer; the colored markers include: The dots with the first color are attached to the flexible gate or flexible plate to obtain the bottom displacement of the flexible gate or the top displacement of the flexible plate, and are also used to process the computer to identify the current experimental mode. The second-colored stickers are affixed to the top of the counterweight and the top of the first or second rigid gate for computer recognition of the current experimental mode. The dots with a third color are located on the acrylic water tank wall for converting between image size and actual size; The synchronous triggering and dynamic visual closed-loop control module based on the assembly configuration state vector is configured with a pattern recognition module and a visual parameter adaptive switching module. The pattern recognition module automatically determines the current experimental mode by acquiring static images, gate installation status and sticker placement position before the experiment starts. The visual parameter adaptive switching module automatically calls the corresponding camera frame rate, exposure time, initial ROI, dynamic ROI search range, color enhancement threshold, marker extraction threshold, connected component area threshold, centroid displacement mutation threshold, and weight release height verification parameters based on the recognition results; and verifies whether the current experimental mode meets the instantaneous dam failure condition.

[0008] Furthermore, the bottom of the acrylic water tank is equipped with an optical platform to ensure that the bottom surface of the water tank is completely horizontal and the initial free liquid surface is horizontal; The first channel is located at half the height of the third or fourth side wall and is used to install the first rigid gate for rigid dam break experiments or dam break flow impact on flexible plates experiments.

[0009] Furthermore, the second and third channels adopt a coupled stepped shape. The second channel is used to install the second rigid gate, and the third channel is used to install the rigid connecting plate and the flexible gate to conduct a dam-break test of the flexible gate. A fourth channel is provided on the bottom wall for installing a flexible plate to conduct a dam-break flow impact test of the flexible plate.

[0010] Furthermore, sealing silicone grease is applied between the first rigid gate and the first channel, between the second rigid gate and the second channel, and between the rigid connecting plates connecting the third channel and the flexible gate; a hydrophobic agent is applied to the water-facing surfaces of the first and second rigid gates. By changing the release height of the counterweight, the opening speed and lifting stroke of the first or second rigid gate can be controlled to meet the instantaneous opening requirements under different initial water depths and experimental scales. ;in, The initial water depth upstream. It is the acceleration due to gravity. For the gate opening time, To consider the dimensionless time of the Froude similarity criterion.

[0011] Another objective of this invention is to provide a dam-break experiment method based on assembly configuration recognition and visual monitoring. This method is implemented using the aforementioned dam-break experiment device based on assembly configuration recognition and visual monitoring, and includes the following steps: S1. Affix colored dots to the flexible gate or flexible plate, gate assembly, and counterweight to be tested; select the appropriate channel, gate assembly, and flexible gate or flexible plate for installation according to the type of experiment. S2, fix the counterweight with an electromagnet to keep the gate assembly in the closed state; inject water to a preset height into the acrylic water tank; S3 processes static images acquired by the computer, automatically determines the current experimental mode based on the gate installation status and the position of the colored dots, and calls the corresponding camera frame rate, exposure time, initial ROI, dynamic ROI search range, color enhancement threshold, marker extraction threshold, connected component area threshold, centroid displacement mutation threshold, and weight release height verification parameters. The processing computer makes a preliminary judgment on the type of experiment by identifying whether the static image shows a first rigid gate or a second rigid gate. If it is a first rigid gate, it identifies whether there is a point attached to the top of the flexible plate. If there is no point, it is a rigid dam failure; if there is, it is a dam failure flow impacting the flexible plate. If it is a second rigid gate, it simultaneously detects the point attached to the bottom of the flexible gate. If both are present, it is a flexible gate dam failure experiment. After the device is identified, the processing computer obtains the experimental mode information, adjusts various parameters, and before the experiment, the processing computer re-checks whether the conditions for starting the target experiment are met. Depending on the different gate installations, the monitoring of the closed state should be different. After the computer confirms that the assembly configuration, weight release height, gate closure status and camera triggering status all meet the preset conditions, it controls the electromagnet to be de-energized and simultaneously triggers the camera to collect data. S4 processes the computer sequentially through frame cutting, marker enhancement, marker extraction, continuity filtering, coordinate transformation, and displacement calculation to obtain the displacement time-history curve of the flexible gate or flexible plate, and outputs fluid-structure interaction characteristic data for fluid dynamics analysis or numerical simulation verification.

[0012] In step S3, the experimental modes include: In the classic rigid dam-break experiment, the first rigid gate is inserted into the first channel, and no flexible gate or flexible plate is installed. After injecting a preset water depth upstream, the computer identifies and disconnects the electromagnet power supply, causing the counterweight to fall and instantly lift the first rigid gate. The camera starts recording synchronously, forming a classic rigid dam-break flow and acquiring the recorded video. In the flexible gate dam failure experiment, a second rigid gate was installed in the second channel, and a rigid connecting plate and the flexible gate were installed in the third channel. A sticker with the first color was affixed to the bottom of the flexible gate. After the experiment began, the computer identified the current experimental mode and changed various parameters. Then, the power to the electromagnet was disconnected, and the camera began synchronous recording. The counterweight caused the second rigid gate to open instantaneously, while the flexible gate underwent continuous large deformation after the upstream hydrostatic pressure was released. This allowed for the observation of the progressive failure process of the flexible gate. The computer then used computer vision technology to obtain fluid-structure interaction data of the displacement response at the bottom of the flexible gate. In the experiment of impacting a flexible plate by a dam-break flow, a first rigid gate was installed in the first channel, and a flexible plate was inserted into the fourth channel. A first-colored sticker was then affixed to the top of the flexible plate. After the experiment began, the computer identified the current experimental mode and changed various parameters. Then, the power to the electromagnet was disconnected, and the camera began recording synchronously. The first rigid gate opened instantaneously, and the dam-break flow advanced downstream and impacted the flexible plate, realizing the displacement of the flexible plate under the impact load. The computer then used computer vision technology to acquire fluid-structure interaction data of the displacement response at the top of the flexible plate.

[0013] In step S4, the frame-slicing step involves extracting and saving the video file frame by frame as JPG images under the control of a computer program; the program will obtain the frame rate of the video, and the experimental video will be sliced ​​into an image sequence frame by frame; The marker enhancement step includes removing connected regions with an area of ​​less than 4 pixels, and in the HSV color space, treating pixels with a brightness V below 0.35 as dark regions, and in the LAB color space, treating pixels with an a channel greater than 6 as reddish regions, and brightening and enhancing pixels that simultaneously meet the conditions of dark regions and reddish regions. The marker extraction step involves extracting the pixel coordinates of the marker points in the image coordinate system; the marker extraction step is based on the RGB difference method. The core formula of the RGB difference method is: ; In the formula, for The value of the red channel at the point. for The value of the green channel at the point. for The value of the blue channel at the point. The threshold is set to 40 for the first time. If no red dot is detected, the threshold is set to 20 for the second time. If no red dot is detected, the frame is marked as missing. The marker extraction step defines the detection region ROI and sets the minimum area threshold of the red dot region to 5 pixels. Regions smaller than this threshold are removed as noise. The marker extraction involves automatically scanning the JPG file under the target file, locating the red dots. The red dots have regular shapes and the red-covered pixel blocks are relatively concentrated. The mean value is used as the centroid, and then the centroid coordinates are output.

[0014] Furthermore, the continuous filtering step is based on the principle of motion continuity between consecutive frames, and applies an area threshold and a centroid displacement abrupt change threshold to the candidate connected components to filter out false targets caused by bubble occlusion, splashing droplets and background reflection; The coordinate transformation step is to convert pixel coordinates into actual coordinates based on the ratio of pixel distance to actual distance at a reference point outside the water tank. The displacement calculation involves plotting a displacement time-history curve and outputting fluid-structure interaction characteristic data for fluid dynamics analysis or numerical simulation verification.

[0015] Combining all the above technical solutions, the beneficial effects of this invention are as follows: First, the device includes an aluminum profile support frame, an acrylic water tank with a first, second, third, and fourth channel, a gate assembly and a flexible plate or flexible gate, a lifting system, an electromagnet, and a computer vision monitoring device. Through selective assembly of different channels and components, a single device can quickly switch between three modes: classic rigid dam failure, flexible gate dam failure, and dam failure flow impacting the flexible plate. The lifting system uses a free-fall counterweight to drive the gate to open instantaneously. The release height of the counterweight simultaneously determines the initial lifting speed and maximum lifting stroke of the gate, thus ensuring that the gate opening time meets the critical conditions for the instantaneous dam failure experiment. The processing computer is equipped with an assembly configuration recognition module and a visual parameter adaptive switching module. Before the experiment begins, the processing computer acquires static images and can automatically determine the current experimental mode based on the gate installation status and the placement of the marking points. It then calls the corresponding camera frame rate, exposure time, initial ROI, dynamic ROI search range, color enhancement threshold, marker extraction threshold, connected component area threshold, centroid displacement abrupt change threshold, and weight release height verification parameters. At the start of the experiment, after confirming that the assembly configuration, weight release height, gate closure status, and camera triggering status all meet the preset conditions, the processing computer controls the electromagnet to be de-energized and simultaneously triggers the camera to acquire data. The moment the electromagnet is de-energized or the actual gate start-up moment is used as the zero point of time. At the same time, the actual gate opening time is calculated through the markings on the top of the gate to verify whether it meets the instantaneous dam failure conditions. During the experiment, the computer collects the centroids of the points in the previous frame and predicts the position of the points in the current frame based on the initial ROI. Candidate connected components are then comprehensively scored based on color difference, area, roundness of shape, distance from the predicted position, and consistency with the direction of motion, thus eliminating splashed droplets, bubble obstruction, and reflective false targets. This invention addresses the experimental requirements of three types: structural instability, progressive flexible failure, and impact response. It avoids the influence of equipment intrusion on the experimental results and can obtain highly repeatable and high-fidelity continuous deformation data of the fluid-structure interaction, providing standardized verification data for fluid-structure interaction numerical models.

[0016] Secondly, this invention is the first to integrate three types of experiments into one, relying on a program that can identify the experiment type based on the gate installation status, automatically adjust parameters according to the experiment type, and automatically perform post-processing after completion. In existing technologies, there are no special requirements for liquid level control and fluid stability. This invention relies on the forcing force of multiple aluminum profiles and the strong straightness of the optical platform to ensure the stability of fluid flow. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the disclosure of this invention and, together with the description, serve to explain the principles of this disclosure; Figure 1 This is a schematic diagram of the dam-break experimental device based on assembly configuration recognition and visual monitoring provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the channel location provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the gate channel assembly provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the connection between the electromagnet and other components provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of the gate shape provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of the experimental process provided in the embodiments of the present invention; Figure 7 These are typical time-lapse images provided in embodiments of the present invention; wherein, (a) is the image at time 0s, (b) is the image at time 0.352s, (c) is the image at time 0.568s, (d) is the image at time 0.642s, (e) is the image at time 0.777s, (f) is the image at time 1.043s, (g) is the image at time 1.251s, and (h) is the image at time 1.459s; Figure 8 The top of the flexible plate provided in this embodiment of the invention x Directional displacement time-history curve; Figure 9 The top of the flexible plate provided in this embodiment of the invention y Directional displacement time-history curve; In the diagram: 100, Acrylic water tank; 110, First side wall; 120, Second side wall; 130, Third side wall; 140, Fourth side wall; 150, First channel; 160, Second channel; 170, Third channel; 180, Fourth channel; 190, Bottom wall; 200, Optical platform; 310, First rigid gate; 320, Second rigid gate; 330, Rigid connecting plate; 410, Aluminum profile for fixing the water tank; 420, Main aluminum profile; 430, Supporting aluminum profile; 500, Strong nylon rope; 600, Directional pulley; 700, Counterweight; 800, Electromagnet; 901, Flexible gate; 902, Flexible plate; 1010, Colored stickers; 1020, Camera; 1030, Processing computer. Detailed Implementation

[0018] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0019] Existing dam-break related devices often only allow for one type of experiment. The innovation of this invention lies in its ability to switch between multiple experiments without changing the basic environment. Related fluid experiments typically require changing camera parameters and tank structure configurations before video processing, while this invention provides automatic switching of camera parameters and post-processing program parameters based on the device configuration, reducing manpower. This device avoids fluid deformation and non-physical changes in flexible body shape caused by equipment intrusion. This invention ensures a level liquid level at the moment the electromagnet 800 is de-energized and the rope is taut, and ensures no vibration in the water tank when the first rigid gate 310 or the second rigid gate 320 is raised to its maximum height, further ensuring the stability of the fluid or free surface morphology.

[0020] Example 1.

[0021] like Figures 1-5 As shown, the dam-break experimental device based on assembly configuration recognition and visual monitoring provided in this embodiment of the invention includes: Aluminum profile support frame, used to fix and support various components; An acrylic water tank 100, fixed on an aluminum profile support frame, provides an experimental site for different dam-break experiments through different channels reserved on different side walls. The gate assembly, by setting a first rigid gate 310, a second rigid gate 320, a rigid connecting plate 330, and a flexible gate 901 fixedly connected to the rigid connecting plate 330, is used to simulate rigid dam failure experiments and flexible gate dam failure experiments. The detachable flexible plate 902 is used in conjunction with the first rigid gate 310 to simulate the impact of dam break flow on the flexible plate. A lifting system is used to control the opening of the first rigid gate 310 or the second rigid gate 320; Electromagnet 800, in conjunction with the lifting system, is used to attract and trigger the instantaneous release action of the lifting system before the experiment begins; The computer vision monitoring system includes a synchronous triggering and dynamic vision closed-loop control module based on the assembly configuration state vector. This module is used to identify the installation status of each component in different channels by acquiring static images, and to form an assembly configuration state vector by combining different colored dots 1010 and the initial position of the dots. Based on this state vector, the system automatically determines the current experimental mode and calls the corresponding camera frame rate, exposure time, initial ROI, dynamic ROI search range, color enhancement threshold, marker extraction threshold, connected component area threshold, centroid displacement mutation threshold, and weight release height verification parameters to verify whether the current experimental mode meets the instantaneous dam failure condition.

[0022] Example 2.

[0023] like Figures 1-5 As shown, the dam-break experimental device based on assembly configuration recognition and visual monitoring provided in this embodiment of the invention specifically includes: An aluminum profile support frame, consisting of an aluminum profile 410 for fixing the water tank, a main aluminum profile 420, and a supporting aluminum profile 430, is used to fix and support the various components. An acrylic water tank 100 is fixed to an aluminum profile 410 for fixing the water tank. The water tank includes a bottom wall 190, a first side wall 110, a second side wall 120, a third side wall 130, and a fourth side wall 140. The third side wall 130 and the fourth side wall 140 are reserved with a first channel 150, a second channel 160, and a third channel 170, respectively. The bottom wall 190 is reserved with a fourth channel 180. The gate assembly includes a first rigid gate 310 corresponding to the first channel 150, a second rigid gate 320 corresponding to the second channel 160, a rigid connecting plate 330 corresponding to the third channel 170, and a flexible gate 901 fixedly connected to the rigid connecting plate 330. The detachable flexible plate 902 is fixed inside the fourth channel 180; The lifting system includes a strong nylon rope 500, a directional pulley 600, and a five-kilogram counterweight 700. One end of the strong nylon rope 500 is connected to the top of the first rigid gate 310 or the second rigid gate 320, and the other end passes around the directional pulley 600 and is connected to the five-kilogram counterweight 700, which is used to control the opening of the first rigid gate 310 or the second rigid gate 320. Electromagnet 800, in conjunction with the lifting system, is used to attract and trigger the instantaneous release action of the lifting system before the experiment begins; A computer vision monitoring system includes colored dots 1010, a camera 1020, and a processing computer 1030; The colored dots 1010 come in multiple colors. Dots with the first color are attached to the flexible gate 901 or the flexible plate 902 to obtain the bottom displacement of the flexible gate 901 or the top displacement of the flexible plate 902. They are also used by the processing computer 1030 to identify the current experimental mode. Dots with the second color are attached to the top of the five-kilogram counterweight 700 and the top of the first rigid gate 310 or the second rigid gate 320 to identify the current experimental mode by the processing computer 1030. Dots with the third color are located on the wall of the acrylic water tank 100 for the conversion between pixel size and actual size.

[0024] The camera 1020 is used to record the experimental process. The processing computer 1030 is configured with a pattern recognition module and a visual parameter adaptive switching module. The pattern recognition module is used to collect static images before the experiment starts. It can automatically determine the current experimental mode based on the installation status of the flexible gate 901 and the position of the stickers. The visual parameter adaptive switching module is used to automatically call the corresponding camera frame rate, exposure time, initial ROI, dynamic ROI search range, color enhancement threshold, marker extraction threshold, connected component area threshold, centroid displacement mutation threshold, and weight release height verification parameters based on the recognition results. In another example, the first sidewall 110 and the second sidewall 120 are arranged opposite each other from left to right, and the third sidewall 130 and the fourth sidewall 140 are arranged opposite each other from front to back and located between the first sidewall 110 and the second sidewall 120. The bottom of the acrylic water tank 100 is provided with an optical platform 200 to ensure that the bottom surface of the water tank is completely horizontal and the initial free liquid surface is horizontal. The first channel 150, the second channel 160 and the third channel 170 are located on the third side wall 130 and the fourth side wall 140 and are arranged opposite to each other. The fourth channel 180 is located on the bottom wall 190. The first channel 150 corresponds to the first rigid gate 310, the second channel 160 corresponds to the second rigid gate 320, and the third channel 170 corresponds to the rigid connecting plate 330 connected to the flexible gate 901. In another example, the first channel 150 is disposed at half the height of the third side wall 130 or the fourth side wall 140, for installing the first rigid gate 310 to conduct rigid dam break tests or dam break flow impact flexible plate tests. The second channel 160 and the third channel 170 adopt a coupled stepped design. The second channel 160 is used to install the second rigid gate 320, and the third channel 170 is used to install the rigid connecting plate 330 and the flexible gate 901 to conduct a dam-break test of the flexible gate. The bottom wall 190 is provided with a fourth channel 180, which is used to install the flexible plate 902 to conduct a dam-break flow impact test of the flexible plate.

[0025] In another example, the first rigid gate 310 is used for rigid dam-break experiments and dam-break flow impact on flexible plates experiments; the second rigid gate 320 is used for flexible gate dam-break experiments; the bottom of the rigid connecting plate 330 has a stepped structure corresponding to the stepped design of the third channel 170, and is fixedly connected to the flexible gate 901 with screws, and is waterproofed with waterproof tape to transfer the upstream hydrostatic pressure to the gate when the second rigid gate 320 is opened. With the above arrangement, the same acrylic water tank 100 does not need to be replaced in different modes; only the internal components of the channel need to be replaced. For example, the first rigid gate 310 has a relatively thick preset thickness, which is configured such that when bearing the maximum design hydrostatic pressure, its maximum structural deflection is less than a preset safety threshold, so as to maintain the watertightness of the fit with the first channel 150.

[0026] For example, depending on the experiment, the top of the first rigid gate 310 or the second rigid gate 320 is made with an opening, and a strong nylon rope 500 is wrapped around the opening and connected to a five-kilogram counterweight 700 through two directional pulleys 600. In the initial stage, the five-kilogram counterweight 700 is attracted by the electromagnet 800 through the power of the electromagnet 800, and comes into contact with the supporting aluminum profile 430 that supports the weight of the five-kilogram counterweight 700.

[0027] Two directional pulleys 600 and an electromagnet 800 are mounted on the main aluminum profile 420; In one exemplary embodiment, the six sides of the acrylic water tank 100 are fixed by an aluminum profile 410 for fixing the water tank; The aluminum profile 410 for fixing the water tank, together with the main aluminum profile 420 and the supporting aluminum profile 430 for supporting the weight of a 5 kg counterweight 700, form an aluminum profile support frame. The aluminum profile 410 for fixing the water tank, the main aluminum profile 420, and the supporting aluminum profile 430 for supporting the weight of the five-kilogram counterweight 700 are preferably connected by angle steel to improve the overall rigidity and limit the deformation of the acrylic water tank 100.

[0028] In an exemplary embodiment, sealing silicone grease is applied between the first rigid gate 310 and the first channel 150, between the second rigid gate 320 and the second channel 160, and between the third channel 170 and the flexible gate 901, to increase water tightness while reducing the frictional force during extraction. A hydrophobic agent is coated on the water-facing surfaces of the first rigid gate 310 and the second rigid gate 320 to reduce droplet dragging and adhesion during instantaneous gate opening, thereby reducing background noise in visual monitoring from the source.

[0029] In one exemplary embodiment, a 5 kg counterweight 700 is held at a preset height by two electromagnets 800. After power is cut off, it falls freely and, via a directional pulley 600 and a strong nylon rope 500, rapidly moves either the first rigid gate 310 or the second rigid gate 320 upwards. By limiting the length of the strong nylon rope 500, the rope can be prevented from losing tension after the 5 kg counterweight 700 touches the ground, thus preventing the first rigid gate 310 or the second rigid gate 320 from falling back. By changing the release height of the 5 kg counterweight 700, the opening speed and lifting stroke of the first rigid gate 310 or the second rigid gate 320 can be adjusted to meet the instantaneous opening requirements under different initial water depths and experimental scales. This also improves the reproducibility of experiments across different batches.

[0030] In another example, the second rigid gate 320 adopts a variable cross-section design, with its bottom thickness being greater than its upper thickness, to accommodate the non-through independent structure of the second channel 160 and the third channel 170, prevent interference caused by the lifting process of the first rigid gate 310 or the second rigid gate 320, and ensure pressure-bearing rigidity.

[0031] In another example, the bottom stepped structure of the rigid connecting plate 330 includes a load-bearing surface, a vertical contact surface, and a fixing surface. The load-bearing surface abuts against the bottom surface of the third channel 170 to bear its own weight and the weight of the flexible gate 901. The vertical contact surface forms a sliding fit with the side wall of the third channel 170. The fixing surface is in contact with the flexible gate 901. The contact between the fixing surface and the flexible gate 901 requires waterproof sealing treatment. This treatment adopts a multi-layer composite sealing structure: the inner layer is a cured sealing layer formed by filling the screw gaps with water-resistant adhesive, and the outer layer is a flexible covering layer formed by wrapping waterproof tape. The direction of tape wrapping corresponds to the insertion direction of the rigid plate to prevent the tape from being peeled off by shear force during the insertion of the rigid connecting plate 330 into the third channel 170.

[0032] In another example, the first rigid gate 310 and the second rigid gate 320 are coated with an acrylic hydrophobic agent on the water side to prevent fluid splashing on the gates caused by instantaneous opening of the gates. The first rigid gate 310 and the first channel 150, the second rigid gate 320 and the second channel 160, and the flexible plate 902 connected to the flexible gate 901 and the third channel 170 are all coated with sealing silicone grease to further improve water tightness.

[0033] In another example, the sticker with the first color is attached to the bottom of the flexible gate 901 during the flexible gate dam failure test, and to the top of the flexible plate 902 during the dam failure flow impact test; the camera 1020 is a monocular camera, positioned directly opposite the sticker.

[0034] Another example is that the computer vision monitoring system also includes a synchronous triggering and dynamic visual closed-loop control module based on the assembly configuration state vector.

[0035] Configuration markings that can be recognized by camera 1020 are provided near the first rigid gate 310, the second rigid gate 320, the rigid connecting plate 330, the flexible gate 901, the flexible plate 902, and the first channel 150, the second channel 160, the third channel 170, and the fourth channel 180. Before the experiment begins, the processing computer 1030 acquires static images, identifies the installation status of each component in the first channel 150, the second channel 160, the third channel 170, and the fourth channel 180, and combines the dots with the first color, the dots with the second color, and the initial position of the dots to form an assembly configuration state vector. The colored dots 1010 have multiple colors. The dots with the first color are attached to the flexible gate 901 or the flexible plate 902 to obtain the bottom displacement of the flexible gate 901 or the top displacement of the flexible plate 902. They are also used by the processing computer 1030 to identify the current experimental mode. The dots with the second color are attached to the top of the five-kilogram counterweight 700 and the top of the first rigid gate 310 or the second rigid gate 320 to be used by the processing computer 1030 to identify the current experimental mode. The processing computer 1030 automatically determines the current experimental mode based on the state vector and calls the corresponding camera frame rate, exposure time, initial ROI, dynamic ROI search range, color enhancement threshold, marker extraction threshold, connected component area threshold, centroid displacement mutation threshold, and weight release height verification parameters. At the start of the experiment, after confirming that the assembly configuration state vector, weight release height, closed state of flexible gate 901 or flexible plate 902, and trigger-ready state of camera 1020 all meet the preset conditions, the processing computer 1030 controls the electromagnet 800 to be de-energized and simultaneously triggers the camera 1020 to acquire data. The actual start-up time of the first rigid gate 310 or the second rigid gate 320 at the moment the electromagnet 800 is de-energized is taken as the time zero point. At the same time, the actual gate opening time is calculated through the second-colored stickers on the top of the first rigid gate 310 or the second rigid gate 320 to verify whether it meets the instantaneous dam failure condition.

[0036] During the experiment, the processing computer 1030 predicted the position of the current frame's patch point based on the centroid of the patch point in the previous frame, dynamically generated the ROI, and comprehensively scored the candidate connected components according to color difference value, area, shape roundness, distance from the predicted position, and consistency of motion direction, thereby eliminating splashed droplets, bubble obstruction, and reflective false targets. It can be seen that this invention elevates the combination of multiple known components to a closed-loop control system encompassing assembly configuration recognition, synchronous trigger control, pattern-constrained visual tracking, and gate opening time verification.

[0037] Working principle.

[0038] In a classic rigid dam-break experiment, a first rigid gate 310 is inserted into the first channel 150, while neither the flexible gate 901 nor the flexible plate 902 is installed. After injecting a preset water depth upstream, the computer 1030 identifies and disconnects the power supply to the electromagnet 800, causing a 5 kg counterweight 700 to fall and momentarily pull up the first rigid gate 310. The camera 1020 then begins synchronous recording, forming a classic rigid dam-break flow and acquiring the recorded video. If it is necessary to measure the sidewall pressure, a pressure sensor can be attached to the first sidewall 110 or the second sidewall 120.

[0039] For example, in a flexible gate dam failure experiment, a second rigid gate 320 is installed in the second channel 160, and a rigid connecting plate 330 and a flexible gate 901 are installed in the third channel 170. Colored stickers 1010 are affixed to the bottom of the flexible gate 901. After the experiment begins, the processing computer 1030 identifies the current experimental mode and changes various parameters. Then, the power supply to the electromagnet 800 is disconnected, and the camera 1020 begins synchronous recording. A five-kilogram counterweight 700 causes the second rigid gate 320 to open instantaneously, while the flexible gate 901 undergoes continuous large deformation after the upstream hydrostatic pressure is released. This allows for the observation of the progressive failure process of the flexible gate 901. The processing computer 1030 then performs computer vision technology steps to obtain fluid-structure interaction data of the displacement response at the bottom of the flexible gate 901.

[0040] For example, in the experiment of dam-break flow impacting a flexible plate, a first rigid gate 310 is installed in the first channel 150, a flexible plate 902 is inserted into the fourth channel 180, and a sticker with a first color is pasted on the top of the flexible plate 902. After the experiment begins, the processing computer 1030 identifies the current experimental mode and changes various parameters, then disconnects the power supply to the electromagnet 800, and the camera 1020 begins synchronous recording. The first rigid gate 310 opens instantaneously, and the dam-break flow advances downstream and impacts the flexible plate 902, thereby realizing the displacement phenomenon of the flexible plate 902 under the action of impact load. The processing computer 1030 performs computer vision technology steps to obtain fluid-structure interaction data of the displacement response of the top of the flexible plate 902.

[0041] In an exemplary embodiment, after the flexible plate 902 is placed into the fourth channel 180, hot melt adhesive is used to fill the gap between the two sides of the flexible plate 902 and the wall of the fourth channel 180. After the hot melt adhesive cures, it works together with the bottom wall 190 of the acrylic water tank 100 and the thick wall of the fourth channel 180 to form a rigid constraint boundary that is completely fixed to the bottom of the flexible plate 902. This rigid constraint not only effectively limits the translational and rotational degrees of freedom of the root of the flexible plate 902 under transient dam-break impact loads, preventing root pull-out instability caused by local stress concentration or instantaneous hydrodynamic pressure; but also ensures that the physical experimental model is equivalent to the ideal cantilever beam dynamic model in terms of boundary conditions, effectively eliminating the nonlinear interference of base slip on the structural dynamic response, and providing highly consistent and clear boundary input parameters for the subsequent fluid-structure interaction numerical simulation model.

[0042] In one exemplary embodiment, such as Figures 1 to 5 As shown, at the start of the flexible plate 902 dam-break experiment, the second rigid gate 320 is inserted into the second channel 160. Because the length of the first channel 150 is controlled, even without filling the gaps in the first channel 150, the fluid will not touch the first channel 150 when the flexible plate 902 dam breaks. A hole is drilled in the top of the second rigid gate 320, extending beyond the height of the acrylic water tank 100. A strong nylon rope 500 is routed around the hole and connected to a five-kilogram counterweight 700 via two directional pulleys 600. By applying sealing silicone grease to the bottom and sides of the second rigid gate 320, water tightness is ensured, and the water volume is more easily controlled. Simultaneously, the five-kilogram counterweight 700 is fixed at a certain height by an electromagnet 800, making the height fixed and standardizing the gate lifting speed, thus greatly enhancing repeatability. At the start of the experiment, the power was disconnected, and the electromagnet 800 and the five-kilogram counterweight 700 separated instantly. The five-kilogram counterweight 700 then fell freely. After the strong nylon rope 500 was tightened, the gate was lifted instantly.

[0043] In an exemplary embodiment, the instantaneous lifting of the first rigid gate 310 or the second rigid gate 320 is determined based on the gate opening time threshold standard proposed by Lauber and Hager. The dimensionless time of opening of the first rigid gate 310 or the second rigid gate 320 is... ,in, The initial water depth upstream. It is the acceleration due to gravity. For the gate opening time, To consider the dimensionless time of the Froude similarity criterion, when the first rigid gate 310 or the second rigid gate 320 Less than the critical value At that time, the first rigid gate 310 or the second rigid gate 320 can be considered to be opened instantaneously.

[0044] Example 3.

[0045] For example, the dam-break experiment principle based on assembly configuration recognition and visual monitoring provided in this embodiment of the invention is as follows: Figure 6 As shown; Specifically, the dam-break experiment method based on assembly configuration recognition and visual monitoring includes the following steps: S1. Affix colored dots 1010 to the flexible gate 901 or flexible plate 902, gate assembly, and 5 kg counterweight 700 to be tested; select the appropriate channel, gate assembly, and flexible gate 901 or flexible plate 902 for installation according to the type of experiment. S2, fix the five-kilogram counterweight 700 with the electromagnet 800 to keep the gate assembly in the closed state; inject water to a preset height into the acrylic water tank 100; S3 processes the static images acquired by the computer 1030, automatically determines the current experimental mode based on the gate installation status and the arrangement position of the colored dots 1010, and calls the corresponding camera frame rate, exposure time, initial ROI, dynamic ROI search range, color enhancement threshold, marker extraction threshold, connected component area threshold, centroid displacement mutation threshold, and weight release height verification parameters. After the processing computer 1030 confirms that the assembly configuration, weight release height, gate closure status, and camera 1020 trigger status all meet the preset conditions, it controls the electromagnet 800 to be powered off and simultaneously triggers the camera 1020 to collect data.

[0046] The computer 1030 processes static images by identifying whether the installed static image is the first rigid gate 310 or the second rigid gate 320 to make a preliminary judgment on the experimental category. If it is the first rigid gate 310, it identifies whether there is a top attachment point of the flexible plate 902. If not, it is a rigid dam failure; if so, it is a dam failure flow impacting the flexible plate 902. If it is the second rigid gate 320, it simultaneously detects the bottom attachment point of the flexible gate 901. If both are present, it is a flexible gate dam failure experiment.

[0047] After the device is identified, the processing computer 1030 acquires the experimental mode information, adjusts various parameters, and before the experiment, the processing computer 1030 re-checks whether the conditions for starting the target experiment are met. Depending on the gate installation, the monitoring of the closed state should be different.

[0048] S4, the processing computer 1030 sequentially performs frame cutting, marker enhancement, marker extraction, continuity filtering, coordinate transformation and displacement calculation to obtain the displacement time history curve of the flexible gate 901 or the flexible plate 902, and outputs fluid-structure interaction characteristic data for fluid dynamics analysis or numerical simulation verification.

[0049] For example, in step S4, the frame-slicing step involves extracting and saving the video file frame by frame as JPG images under the control of a computer program. The program obtains the video's frame rate; the experimental video is sliced ​​into an image sequence at 120 frames per second. The marker enhancement step includes removing connected regions with an area of ​​less than 4 pixels, and in the HSV color space, treating pixels with a brightness V below 0.35 as dark regions, and in the LAB color space, treating pixels with an a channel greater than 6 as reddish regions, and enhancing pixels that simultaneously meet the conditions of dark regions and reddish regions.

[0050] For example, the marker extraction step involves extracting the pixel coordinates of the marker in the image coordinate system; The marker extraction step is based on the RGB difference method. The image after the marker enhancement step has stable lighting and strong contrast between the red dots and the background, making it suitable for the RGB difference method.

[0051] ; In the formula, for The value of the red channel at the point. for The value of the green channel at the point. for The value of the blue channel at the point. The threshold is set to 40 for the first time. If no red dot is detected, the threshold is set to 20 for the second time. If no red dot is detected, the frame is marked as missing. For example, the marker extraction step defines the detection region ROI and sets the minimum area threshold of the red dot region to 5 pixels. Regions smaller than this threshold are removed as noise. The marker extraction involves automatically scanning the JPG file under the target file, locating the red dots. The red dots have regular shapes and the red-covered pixel blocks are relatively concentrated. The mean value is used as the centroid, and then the centroid coordinates are output.

[0052] The continuous filtering step is based on the principle of motion continuity between consecutive frames. It applies an area threshold and a centroid displacement abrupt change threshold to the candidate connected components to filter out false targets caused by bubble occlusion, splashing droplets and background reflection. The coordinate transformation step is to convert pixel coordinates into actual coordinates based on the ratio of pixel distance to actual distance at a reference point outside the water tank. The displacement calculation involves plotting a displacement time-history curve and outputting fluid-structure interaction characteristic data for fluid dynamics analysis or numerical simulation verification.

[0053] Example 4.

[0054] In another embodiment of the present invention, the computer vision monitoring system includes colored dots 1010, a camera 1020, and a processing computer 1030; the camera 1020 is used to acquire information from the colored dots 1010 and to send the acquired information to the processing computer 1030. The processing computer 1030 also includes a synchronous triggering and dynamic visual closed-loop control module based on the assembly configuration state vector; colored dots 1010 (dots with a first color) are attached to the bottom of the flexible gate 901 or the top of the flexible plate 902. Under the action of water, the bottom of the flexible gate 901 or the top of the flexible plate 902 is displaced, and the displacement is recorded by the camera 1020 and imported into the processing computer 1030. The computer vision technology process used in the computer 1030 is processed step by step in the following steps: Step 1: The processing computer 1030 first divides the video frame by frame into a sequence of JPG images, and then performs label enhancement and label extraction on the images. In this embodiment, the colored dots 1010 use dots with a first color, namely red dots, and use blue dots as reference points to determine the ratio of the actual size to the pixel size; Step 2: In the marker enhancement stage, the program preferentially removes connected components with an area of ​​less than 4 pixels first, then identifies dark red targets in the HSV and LAB dual color spaces, and performs a brightening operation on pixels that meet the conditions of low brightness and reddish a-channel. This process is beneficial for restoring pixel information that has been obscured by water, covered by shadows, or weakened by refraction. Step 3: In the label extraction stage, the program preferably calculates based on the RGB difference method. The program first uses 40 as the initial threshold. If no target is detected, the threshold is lowered to 20 for a second search. The program also limits the Area of ​​Interest (ROI) and removes red connected components with an area smaller than 5 pixels to avoid misidentifying background noise as target points. Step 4: In the continuity filtering stage, the program compares the candidate centroid of the current frame with the target centroid of the previous frame. If the area of ​​the candidate connected components is abnormal or the centroid displacement shows a non-physical abrupt change relative to adjacent frames, the candidate target is determined to be a bubble, splashing droplet, or reflective noise and is filtered out. Through this step, even in the transient stage where gas-liquid two-phase mixing is strong, the continuity of flexible structure trajectory extraction can still be maintained. Step 5: In the coordinate transformation and displacement calculation stage, the program converts the pixel coordinates of the marker points into actual coordinates based on the ratio of the pixel distance to the actual distance of the external reference point of the water tank, and then draws the map. x Directional displacement time history curve y The directional displacement time history curve and the total displacement time history curve are output as data for fluid-structure interaction analysis and numerical simulation verification.

[0055] To verify the experimental effectiveness of this device, a related experiment—the dam-break flow impact on a flexible plate—was conducted using this device.

[0056] In this experiment, a water column 0.1m wide and 0.4m high was generated. A fourth channel 180 was installed at a distance of 0.345m from the third sidewall 130 to mount a flexible plate 902. A dam-break simulation of the flexible plate 902 was then performed, and typical moment images were obtained. Figure 7 ,in, Figure 7 Figure (a) shows the image at time 0s. Figure 7 Image (b) is the image at time 0.352s. Figure 7 Image (c) is the image at 0.568s. Figure 7 Image (d) is the image at time 0.642s. Figure 7 Image (e) shows the image at time 0.777s. Figure 7 Image (f) is the image at time 1.043s. Figure 7 Image (g) is the image at time 1.251s. Figure 7 The middle (h) image is the image at time 1.459s; Taking the location of the red dot at 0s as the origin, the horizontal direction to the right is... x The positive direction of the axis, vertically downwards is y The top of the flexible plate 902, obtained by capturing the red dot in the positive axis direction. x The directional displacement time history curve is shown below. Figure 8 Flexible panel 902 top y The directional displacement time history curve is shown below. Figure 9 After the experiment began, the flexible plate 902 began to bend upon contact with the water flow. The water then flowed over the flexible plate 902 and impacted the third sidewall 130. Upon impact, the flexible plate 902 vibrated slightly and experienced a partial return displacement to its initial position. As the water flow on the flexible plate 902 increased, it continued to bend to the left. After impacting the third sidewall 130, some of the water fell back and rolled inwards. As more water increased on the left side of the flexible plate 902, it was gradually lifted, then returned to its original position, and after reaching its highest point, began to bend to the right. The experimental process was clear, and the red dot was easily captured. The top of the flexible plate 902 was obtained. Directional displacement time history curve and top of flexible plate 902 The directional displacement time-history curve is consistent with reality. This invention can meet the requirements of laboratory simulation of dam-break flow impacting flexible plate 902, flexible plate 902 dam failure, and classic rigid dam failure.

[0057] As can be seen from the above embodiments, a single experimental platform can cover three typical dam-break scenarios, improving the integration and practical efficiency of the experimental system; the gate opening power input is constant and clear, with high experimental repeatability, making it suitable for conducting parameter comparison experiments; the flexible structure response acquisition method is non-contact, without introducing additional sensor intrusion, which is more conducive to maintaining the authenticity of the flow field; the adaptive switching of pattern recognition and visual parameters reduces manual costs and improves image processing efficiency; the synergy between hydrophobicity and visual algorithms resists visual interference, making the extraction of continuous large deformation trajectories of flexible structures more stable.

[0058] This invention provides a multi-mode dam-break experimental device and method based on assembly configuration recognition and adaptive visual monitoring. Through the coordinated control of the instantaneous opening of a rigid gate, assembly configuration recognition, adaptive switching of visual parameters, and continuous deformation extraction process, it achieves stable acquisition of the continuous large deformation trajectory of a flexible structure. This invention integrates fluid-structure interaction experiments and computer vision technology. While ensuring the two-dimensionality of the experiment, it captures details such as liquid splashing and watertightness. Computer vision technology only requires simple marking on the target and recording the experimental process to obtain the true motion trajectory of the flexible gate 901, without the need for additional sensors. This minimizes the impact of equipment intrusion on the experimental results, resulting in higher accuracy in flexible gate dam-break experiments and effectively improving the three-dimensional representation of dam-break flows in this experiment.

[0059] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by those skilled in the art within the scope of the technology disclosed in the present invention and within the spirit and principles of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A dam-break experimental device based on assembly configuration recognition and visual monitoring, characterized in that, The device includes: Aluminum profile support frame, used to fix and support various components to ensure liquid level stability; An acrylic water tank (100) is fixed on an aluminum profile support frame and provides an experimental site for different dam break experiments through different channels reserved on different side walls; The gate assembly includes a first rigid gate (310), a second rigid gate (320), a rigid connecting plate (330), and a flexible gate (901) fixedly connected to the rigid connecting plate (330), for simulating rigid dam failure experiments and flexible gate dam failure experiments; The flexible plate (902) is used in conjunction with the first rigid gate (310) to simulate the impact of dam break flow on the flexible plate. A lifting system is used to control the opening of the first rigid gate (310) or the second rigid gate (320); An electromagnet (800), in conjunction with the lifting system, is used to attract and trigger the instantaneous release action of the lifting system before the start of the experiment; The computer vision monitoring system includes a synchronous triggering and dynamic vision closed-loop control module based on the assembly configuration state vector. This module is used to identify the installation status of each component in different channels by acquiring static images, and to form an assembly configuration state vector by combining different colored dots (1010) and the initial position of the dots. Based on this state vector, the system automatically determines the current experimental mode and calls the corresponding camera frame rate, exposure time, initial ROI, dynamic ROI search range, color enhancement threshold, marker extraction threshold, connected component area threshold, centroid displacement mutation threshold, and weight release height verification parameters to verify whether the current experimental mode meets the instantaneous dam failure condition.

2. The dam-break experimental device based on assembly configuration recognition and visual monitoring according to claim 1, characterized in that, The aluminum profile support frame consists of an aluminum profile (410) for fixing the water tank, a main aluminum profile (420), and a supporting aluminum profile (430) for supporting the counterweight (700); An acrylic water tank (100) is fixed to an aluminum profile (410) for fixing the water tank, including a bottom wall (190), a first side wall (110), a second side wall (120), a third side wall (130), and a fourth side wall (140); wherein the third side wall (130) and the fourth side wall (140) are reserved with a first channel (150), a second channel (160), and a third channel (170), and the bottom wall (190) is reserved with a fourth channel (180).

3. The dam-break experimental device based on assembly configuration recognition and visual monitoring according to claim 2, characterized in that, The gate assembly includes a first rigid gate (310) corresponding to the first channel (150), a second rigid gate (320) corresponding to the second channel (160), a rigid connecting plate (330) corresponding to the third channel (170), and a flexible gate (901) fixedly connected to the rigid connecting plate (330). The flexible plate (902) is fixed inside the fourth channel (180); The lifting system includes a strong nylon rope (500), a directional pulley (600), and a counterweight (700). One end of the strong nylon rope (500) is connected to the top of the first rigid gate (310) or the second rigid gate (320), and the other end passes around the directional pulley (600) and connects to the counterweight (700) to control the opening of the first rigid gate (310) or the second rigid gate (320). The computer vision monitoring system also includes colored markers (1010), a camera (1020), and a processing computer (1030); the colored markers (1010) include: The first color sticker is attached to the flexible gate (901) or the flexible plate (902) to obtain the bottom displacement of the flexible gate (901) or the top displacement of the flexible plate (902), and is also used by the processing computer (1030) to identify the current experimental mode. A second-colored sticker is attached to the top of the counterweight (700) and the top of the first rigid gate (310) or the second rigid gate (320) for the computer (1030) to identify the current experimental mode. The dots with a third color are located on the wall of the acrylic water tank (100) for converting between the image size and the actual size; The synchronous triggering and dynamic visual closed-loop control module based on the assembly configuration state vector is configured with a pattern recognition module and a visual parameter adaptive switching module. The pattern recognition module automatically determines the current experimental mode by acquiring static images, gate installation status and sticker placement position before the experiment starts. The visual parameter adaptive switching module automatically calls the corresponding camera frame rate, exposure time, initial ROI, dynamic ROI search range, color enhancement threshold, marker extraction threshold, connected component area threshold, centroid displacement mutation threshold, and weight release height verification parameters based on the recognition results; and verifies whether the current experimental mode meets the instantaneous dam failure condition.

4. The dam-break experimental device based on assembly configuration recognition and visual monitoring according to claim 3, characterized in that, An optical platform (200) is provided at the bottom of the acrylic water tank (100) to ensure that the bottom surface of the water tank is horizontal and the initial free liquid surface is horizontal. The first channel (150) is located at half the height of the third side wall (130) or the fourth side wall (140) for installing the first rigid gate (310) to conduct rigid dam break tests or dam break flow impact flexible plate tests.

5. The dam-break experimental device based on assembly configuration recognition and visual monitoring according to claim 4, characterized in that, The second channel (160) and the third channel (170) adopt a coupled stepped shape. The second channel (160) is used to install the second rigid gate (320), and the third channel (170) is used to install the rigid connecting plate (330) and the flexible gate (901) for conducting flexible gate dam failure experiments. The bottom wall (190) is provided with a fourth channel (180) for installing the flexible plate (902) for conducting dam failure flow impact on the flexible plate experiments.

6. The dam-break experimental device based on assembly configuration recognition and visual monitoring according to claim 5, characterized in that, Sealing silicone grease is applied between the first rigid gate (310) and the first channel (150), between the second rigid gate (320) and the second channel (160), and between the third channel (170) and the flexible gate (901) and the rigid connecting plate (330); a hydrophobic agent is applied to the water-facing surfaces of the first rigid gate (310) and the second rigid gate (320); By changing the release height of the counterweight (700), the opening speed and lifting stroke of the first rigid gate (310) or the second rigid gate (320) can be adjusted to meet the instantaneous opening requirements under different initial water depths and different experimental scales. ;in, The initial water depth upstream. It is the acceleration due to gravity. For the gate opening time, To consider the dimensionless time of the Froude similarity criterion.

7. A dam-break experimental method based on assembly configuration recognition and visual monitoring, characterized in that, This method is implemented using the dam-break experimental device based on assembly configuration recognition and visual monitoring as described in any one of claims 1-6, and the method includes the following steps: S1, attach colored dots (1010) to the flexible gate (901) or flexible plate (902), gate assembly, and counterweight (700) to be tested; select the appropriate channel, gate assembly and flexible gate (901) or flexible plate (902) according to the type of experiment for installation; S2, fix the counterweight (700) with an electromagnet (800) to keep the gate assembly in the closed state; inject water to a preset height into the acrylic water tank (100); S3, the computer (1030) collects static images, automatically determines the current experimental mode based on the gate installation status and the position of the colored dots (1010), and calls the corresponding camera frame rate, exposure time, initial ROI, dynamic ROI search range, color enhancement threshold, marker extraction threshold, connected region area threshold, centroid displacement mutation threshold and weight release height verification parameters. The processing computer (1030) makes a preliminary judgment on the experimental category by identifying whether the first rigid gate (310) or the second rigid gate (320) is installed in the static image; if it is the first rigid gate (310), it identifies whether there is a top attachment point of the flexible plate (902); if there is no attachment point, it is a rigid dam failure; if there is an attachment point, it is a dam failure flow impacting the flexible plate (902); if it is the second rigid gate (320), it simultaneously detects the bottom attachment point of the flexible gate (901); if both are present, it is a flexible gate dam failure experiment. After identification, the processing computer (1030) acquires the experimental mode information and adjusts the parameters; before the experiment, the processing computer (1030) re-monitors whether the conditions for starting the target experiment are met. Depending on the different gate installations, the monitoring of the closed state should be different. After the processing computer (1030) confirms that the assembly configuration, weight release height, gate closing status and camera (1020) waiting to be triggered all meet the preset conditions, it controls the electromagnet (800) to be de-energized and synchronously triggers the camera (1020) to collect data. S4, the processing computer (1030) sequentially performs frame cutting, marker enhancement, marker extraction, continuity filtering, coordinate transformation and displacement calculation to obtain the displacement time history curve of the flexible gate (901) or flexible plate (902), and outputs fluid-structure interaction characteristic data for fluid dynamics analysis or numerical simulation verification.

8. The dam-break experiment method based on assembly configuration recognition and visual monitoring according to claim 7, characterized in that, In step S3, the experimental modes include: In the classic rigid dam-break experiment, the first rigid gate (310) is inserted into the first channel (150), and neither the flexible gate (901) nor the flexible plate (902) is installed. After injecting a preset water depth upstream, the processing computer (1030) identifies and disconnects the power supply of the electromagnet (800), causing the counterweight (700) to fall and instantly lift the first rigid gate (310). The camera (1020) starts recording synchronously, forming a classic rigid dam-break flow and acquiring the recorded video. In the flexible gate dam failure experiment, the second rigid gate (320) is installed in the second channel (160), the rigid connecting plate (330) and the flexible gate (901) are installed in the third channel (170), and the first color sticker is pasted on the bottom of the flexible gate (901). After the experiment starts, the processing computer (1030) identifies the current experimental mode and changes various parameters. Then the power supply of the electromagnet (800) is disconnected, the camera (1020) starts to record synchronously, the counterweight (700) drives the second rigid gate (320) to open instantaneously, and the flexible gate (901) undergoes continuous large deformation after the upstream hydrostatic pressure is released, realizing the observation of the progressive failure process of the flexible gate (901). The processing computer (1030) performs computer vision technology steps to obtain the fluid-structure interaction data of the displacement response at the bottom of the flexible gate (901). In the experiment of impacting a flexible plate by a dam-break flow, a first rigid gate (310) is installed in the first channel (150), a flexible plate (902) is inserted into the fourth channel (180), and a first-color sticker is pasted on the top of the flexible plate (902). After the experiment starts, the processing computer (1030) identifies the current experimental mode and changes various parameters. Then the power supply of the electromagnet (800) is disconnected, the camera (1020) starts recording synchronously, the first rigid gate (310) opens instantaneously, the dam-break flow moves downstream and impacts the flexible plate (902), realizing the displacement of the flexible plate (902) under the impact load. The processing computer (1030) performs computer vision technology steps to obtain fluid-structure interaction data of the displacement response of the top of the flexible plate (902).

9. The dam-break experimental method based on assembly configuration recognition and visual monitoring according to claim 7, characterized in that, In step S4, the frame cutting involves extracting and saving the video file frame by frame as JPG images under the control of a computer program; the program will obtain the frame rate of the video, and the experimental video will be divided into image sequences by frame. The mark enhancement includes removing connected regions with an area of ​​less than 4 pixels, and in the HSV color space, treating pixels with a brightness V below 0.35 as dark regions, and in the LAB color space, treating pixels with an a channel greater than 6 as reddish regions, and brightening and enhancing pixels that simultaneously meet the conditions of dark regions and reddish regions. The marker extraction involves extracting the pixel coordinates of the marker points in the image coordinate system; the marker extraction step is based on the RGB difference method. The core formula of the RGB difference method is: ； In the formula, for The value of the red channel at the point. for The value of the green channel at the point. for The value of the blue channel at the point. The threshold is set to 40 for the first time. If no red dot is detected, the threshold is set to 20 for the second time. If no red dot is detected, the frame is marked as missing. The marker extraction step defines the detection region ROI and sets the minimum area threshold of the red dot region to 5 pixels. Regions smaller than this threshold are removed as noise. The marker extraction involves automatically scanning the JPG file under the target file, locating the red dots. The red dots have regular shapes and the red-covered pixel blocks are relatively concentrated. The mean value is used as the centroid, and then the centroid coordinates are output.

10. The dam-break experimental method based on assembly configuration recognition and visual monitoring according to claim 9, characterized in that, The continuity filtering is based on the principle of motion continuity between consecutive frames. It applies an area threshold and a centroid displacement abrupt change threshold to the candidate connected components to filter out false targets caused by bubble occlusion, splashing droplets and background reflection. The coordinates are converted into the ratio of pixel distance to actual distance based on the external reference point of the water tank, thus converting the pixel coordinates into actual coordinates; The displacement calculation involves plotting a displacement time-history curve and outputting fluid-structure interaction characteristic data for fluid dynamics analysis or numerical simulation verification.

Images