A method for rapid detection of water surface velocity in large rivers based on CFD computational fluid dynamics and video image recognition

Abstract

The present invention relates to a method for monitoring the flow rate of large rivers based on the combination of CFD computational fluid dynamics and video image recognition, and belongs to the field of hydrological measurement technology. The present invention mainly combines the shape of the water surface line formed by the bridge pier foundation and the water surface through video recognition and CFD computational fluid dynamics simulation. Among them, the CFD model needs to first obtain the size of the bridge pier foundation near the site, as well as the riverbed topography upstream and downstream, and then perform CFD modeling, and then perform numerical simulation of VOF multiphase flow at different flow rates. After the simulation is completed, the shape characteristics of the boundary line formed by the bridge pier foundation and the water surface at different flow rates are obtained; in addition, the video image of the real-world bridge pier foundation and the water surface is obtained through a video camera, and the shape and angle of the boundary line are identified through an embedded computer, and matched with the result calculated by the CFD model, so that the water flow velocity of all the bridge pier foundations in the river section can be quickly obtained, and more rapid and accurate flow velocity data can be provided for the next step of flow deduction.

Description

Technical Field

[0001] The invention relates to a method for quickly detecting the surface flow velocity of large rivers based on the combination of CFD computational fluid dynamics and video image recognition, and belongs to the technical field of hydrological testing. Background Art

[0002] In the water conservancy and environmental industries that are generally known, in the past, the common method for measuring river flow was to use lead fish to measure the flow velocity at 3 or 5 points in the river, and then use the flow relationship curve to measure the flow. This traditional method not only consumes a lot of manpower and energy, but the accuracy of flow measurement will produce large errors when the river section changes due to scouring and siltation.

[0003] The existing video flow measurement technology simply uses the video image frame change technology to give the surface flow velocity of the river, and then uses the flow velocity curve from the river bottom to the water surface to infer the flow of the river. This existing algorithm can achieve flow measurement to a certain extent, but there are two obvious defects: 1. The shape of the river flow velocity curve is very different in different river sections, riverbed terrain, and river channel slopes, but the existing video flow measurement algorithm does not correct the flow velocity curve under different conditions in the river section. 2. The existing video image flow measurement method often requires huge computing power, the calculation time is too long, and there is no good timeliness. 3. In addition, since the traditional video flow measurement device simply uses the image frame change to reverse the flow velocity, the flow velocity is easily affected by the wind speed and the accuracy is not high enough. Summary of the invention

[0004] The present invention is based on a method for rapidly detecting the surface flow velocity of large rivers based on the combination of CFD computational fluid dynamics and video image recognition, mainly through the combination of video water surface line shape recognition and CFD computational fluid dynamics simulation. The CFD model first needs to obtain the size of the bridge pier foundation near the site, as well as the riverbed topography upstream and downstream, and then perform CFD modeling, and then perform numerical simulation of VOF multiphase flow at different flow rates. After the simulation is completed, the shape of the boundary line formed by the bridge pier foundation and the water surface at different flow rates is obtained; in addition, the video image of the real-world bridge pier foundation and the water surface is obtained through a video camera, and the shape and angle characteristics of the boundary line are identified by an embedded computer, and matched with the results calculated by the CFD model, so that the water flow velocity of all the bridge pier foundations in the river section can be quickly obtained, and more rapid and accurate flow velocity data can be provided for the next step of flow deduction.

[0005] The technical problem solved by the present invention is: to solve the problem that the previous video flow measurement device requires a large computing power, the calculation results are affected by the wind, and the flow velocity inversion result has low accuracy. By identifying the shape angle of the boundary line between the bridge pier foundation and the water surface, combined with CFD numerical simulation, the flow velocity of several points on the river section can be quickly inferred, and a method for quickly, accurately and conveniently obtaining the flow velocity of the river section is proposed. The shape angle characteristics of the boundary line of the bridge pier foundation and the water surface near the hydrological station are identified by using an embedded computer and a video camera device, and the shape angle characteristics simulated by CFD are matched to achieve rapid, convenient, wide adaptability and high-precision measurement of the flow of the river.

[0006] In order to solve the above technical problems, the technical solution proposed by the present invention is: a method for quickly detecting the surface flow velocity of large rivers based on the combination of CFD computational fluid dynamics and video image recognition. The specific steps of the monitoring method are as follows:

[0007] Step 1: First, for the hydrological station or the station that needs to be monitored or measured, obtain the size of the bridge pier foundation near it and the riverbed topography upstream and downstream of the river section;

[0008] Step 2: Use SolidWork or CAD tools to build a three-dimensional physical model of the upstream and downstream river sections and bridge pier foundations; perform meshing of the three-dimensional model and build a CFD numerical model;

[0009] Step 3: Perform CFD numerical simulation using the VOF (Volume of Fluid) water-air multiphase open channel flow model, where the VOF model uses the following equation to define the volume ratio of water and air:

[0010]

[0011] α q represents the volume proportion of the fluid in a discrete unit q, is the density of the q phase, S αq is the source term, defaults to zero, is the quality of the transition from one phase q to phase p, represents the flow rate of phase q, is the conversion quality from phase p to phase q, is the Laplace operator;

[0012] Step 4: Through the VOF model simulation results in step 3, a position surface with a phase q to phase p value of 0.5 is obtained, which is the water surface position. The shape and angle characteristics of the boundary line between the water surface and the bridge pier foundation under different water depths and different flow rates can be obtained through the water surface position. ;

[0013] Step 5: By performing curve fitting on the shape of the water surface boundary line, a CFD model is established to simulate the flow velocity and the shape and angle characteristics of the water surface boundary line. The mutual relationship among them represents the simulated flow velocity at the bridge pier foundation, Represents the shape and angle characteristics of the simulated water surface boundary line. A, B, and C are constants of the fitted equation:

[0014]

[0015] Step 6: Add the shape angle features from step 4 , and the corresponding flow rate To correspond, to form A characteristic table is input into the embedded computer as items to be matched;

[0016] Step 7, the video image recognition device includes a video camera, an embedded computer, and a solar panel; turn on the camera of the video image recognition device, adjust the camera field of view to the bridge pier foundation, and obtain a video image of the interface area formed by the bridge pier foundation and the water surface;

[0017] Step 8: Identify the shape and angle features of the video boundary image through the water boundary algorithm built into the embedded computer ;

[0018] Step 9: Corresponding to the shape angle features in step 6, by comparison Characteristic table, quickly obtain the flow velocity of large rivers at the bridge pier foundation ;

[0019] Step 10, repeating the velocity characteristics at a bridge pier foundation obtained in steps 7 to 9, matching the velocity at different pier foundations of the entire river section, and obtaining the velocity distribution at different positions on the entire river section;

[0020] Step 11, using the flow velocity obtained in step 10, the flow of the entire river section can be inferred using the cross-sectional area integral method commonly used in hydrology.

[0021] Preferably, it includes two parts: CFD numerical calculation and video image recognition of water surface boundary line.

[0022] Preferably, the video image recognition device includes a video camera, an embedded computer, and a solar panel, wherein the solar panel can be powered by a battery in the field for use by an edge computing computer.

[0023] Preferably, video water surface line shape recognition and CFD computational fluid dynamics simulation are combined, wherein the CFD model must first obtain the size of the bridge pier foundation near the site and the riverbed topography upstream and downstream, and then perform CFD modeling, and then perform numerical simulation of VOF multiphase flow at different flow rates. After the simulation is completed, the shape of the boundary line formed by the bridge pier foundation and the water surface at different flow rates is obtained; in addition, video images of the real-world bridge pier foundation and the water surface are obtained through a video camera, and the shape and angle characteristics of the boundary line are identified by an embedded computer and matched with the results calculated by the CFD model, so that the water flow velocity of all bridge pier foundations in the river section can be quickly obtained, and more rapid and accurate flow velocity data can be provided for the next step of flow deduction.

[0024] Beneficial effects:

[0025] The present invention takes into account the cumbersome procedures and large workload of the original traditional flow measurement scheme, as well as the low credibility of the flow measurement results of the new simple video flow measurement device due to the influence of wind. Considering the existing technology, since most video flow measurement equipment only uses simple image algorithms to reversely measure the flow, the credibility of the measurement results is mostly not well guaranteed, and the accuracy is not high. By identifying the shape angle of the boundary line between the bridge pier and the water surface, combined with CFD numerical simulation, the flow velocity at several points on the river section can be quickly reversed, and a method for quickly, accurately and conveniently obtaining the flow velocity of the river section is proposed. The shape angle characteristics of the boundary line of the bridge pier foundation and the water surface near the hydrological station are identified by using an embedded computer and a video camera device, and the shape angle characteristics simulated by CFD are matched to achieve rapid, convenient, widely adaptable and high-precision measurement of the river flow.

[0026] The present invention mainly combines video water surface line shape recognition and CFD computational fluid dynamics simulation. The CFD model first needs to obtain the size of the bridge pier foundation near the site and the riverbed topography upstream and downstream, and then perform CFD modeling, and then perform numerical simulation of VOF multiphase flow at different flow rates. After the simulation is completed, the shape of the boundary line formed by the bridge pier foundation and the water surface at different flow rates is obtained ( Figure 1 In addition, by acquiring video images of real-world bridge piers and water surfaces through video cameras, and identifying features such as the shape and angle of the boundary line through an embedded computer, and matching them with the results calculated by the CFD model, the water flow velocity at all bridge piers in the river section can be quickly obtained, providing faster and more accurate flow velocity data for the next step of flow calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The present invention will be further described below in conjunction with the accompanying drawings.

[0028] Figure 1 Flow chart of the method of the present invention

[0029] Figure 2 The method principle flow chart of the present invention is DETAILED DESCRIPTION

[0030] Example 1

[0031] In order to make the purpose, features and advantages of the present invention more obvious and easy to understand, the following is a detailed description of the implementation case and the accompanying drawings.

[0032] like Figure 1 As shown, Figure 1 The present invention is a flowchart of the method of the present invention.

[0033] A method for quickly detecting the surface flow velocity of large rivers based on the combination of CFD computational fluid dynamics and video image recognition, including CFD numerical calculation and video image recognition of water surface boundary line; the specific steps of the monitoring method are as follows:

[0034] Step 1: First, for the hydrological station or the station that needs to be monitored (flow measurement), obtain the size of the bridge pier foundation near it and the riverbed topography upstream and downstream of the river section;

[0035] Step 2: Use SolidWork or CAD tools to build a three-dimensional physical model of the upstream and downstream river sections and bridge pier foundations; perform meshing of the three-dimensional model and build a CFD numerical model;

[0036] Step 3: Perform CFD numerical simulation using the VOF (Volume of Fluid) water-air multiphase open channel flow model, where the VOF model uses the following equation to define the volume ratio of water and air:

[0037]

[0038] α q represents the volume proportion of the fluid in a discrete unit q, is the density of the q phase, S αq is the source term, defaults to zero, is the quality of the transition from one phase q to phase p, represents the flow rate of phase q, is the conversion quality from phase p to phase q, is the Laplace operator;

[0039] Step 4: Through the VOF model simulation results in step 3, a position surface with a phase q to phase p value of 0.5 is obtained, which is the water surface position. The shape and angle characteristics of the boundary line between the water surface and the bridge pier foundation under different water depths and different flow rates can be obtained through the water surface position. ;

[0040] Step 5: By performing curve fitting on the shape of the water surface boundary line, a CFD model is established to simulate the flow velocity and the shape and angle characteristics of the water surface boundary line. The mutual relationship among them represents the simulated flow velocity at the bridge pier foundation, Represents the shape and angle characteristics of the simulated water surface boundary line. A, B, and C are constants of the fitted equation:

[0041]

[0042] Step 6: Add the shape angle features from step 4 , and the corresponding flow rate To correspond, to form A characteristic table is input into the embedded computer as items to be matched;

[0043] Step 7, the video image recognition device includes a video camera, an embedded computer, and a solar panel; turn on the camera of the video image recognition device, adjust the camera field of view to the bridge pier foundation, and obtain a video image of the interface area formed by the bridge pier foundation and the water surface;

[0044] Step 8: Identify the shape and angle features of the video boundary image through the water boundary algorithm built into the embedded computer ;

[0045] Step 9: Same as the shape angle feature in step 6 To correspond, by comparison Characteristic table, quickly obtain the flow velocity of large rivers at the bridge pier foundation ;

[0046] Step 10, repeating the velocity characteristics at a bridge pier foundation obtained in steps 7 to 9, matching the velocity at different pier foundations of the entire river section, and obtaining the velocity distribution at different positions on the entire river section;

[0047] Step 11, using the flow velocity obtained in step 10, the flow of the entire river section can be inferred using the cross-sectional area integral method commonly used in hydrology.

[0048] The video image recognition device includes a video camera, an embedded computer, and a solar panel, wherein the solar panel can be powered by a battery in the field for use by an edge computing computer.

[0049] Through the combination of video water surface line shape recognition and CFD computational fluid dynamics simulation, the CFD model must first obtain the size of the bridge pier foundation near the site and the riverbed topography upstream and downstream, and then perform CFD modeling, followed by numerical simulation of VOF multiphase flow at different flow rates. After the simulation is completed, the shape of the boundary line formed by the bridge pier foundation and the water surface at different flow rates is obtained; in addition, video images of the real-world bridge pier foundation and the water surface are obtained through a video camera, and the shape and angle characteristics of the boundary line are identified by an embedded computer and matched with the results calculated by the CFD model. The water flow velocity of all bridge piers in the river section can be quickly obtained, and more rapid and accurate flow velocity data can be provided for the next step of flow calculation.

[0050] like Figure 2 As shown, Figure 2 The following is a bottom-level logic flow chart of the method of the present invention.

[0051] A method for rapid detection of water surface velocity monitoring in large rivers based on the combination of CFD computational fluid dynamics and video image recognition includes two parts: CFD numerical calculation and video image recognition of water surface boundary line; the specific underlying logic of the method is as follows:

[0052] Step 1: Complete the preliminary work required for CFD numerical simulation calculation, including the size of the bridge pier foundation and the riverbed topography upstream and downstream of the river section;

[0053] Step 2: Use the VOF model to complete CFD numerical simulations under different water depths and flow velocities, and obtain the corresponding characteristic table of the boundary angle α between the bridge pier foundation and the water surface and the flow velocity V under the simulation conditions (including the flow velocity distribution from the river bottom to the river surface);

[0054] Step 3, obtaining the boundary line α1 between the water surface and the bridge pier foundation in the real world through a video imaging device;

[0055] Step 4, completing the matching of α, α1 and flow velocity V, the water surface flow velocity at the bridge pier foundation and the flow velocity distribution from the river bottom to the river surface can be quickly obtained;

[0056] Step 5, repeat steps 3-4 to complete the flow velocity V at different bridge pier foundations;

[0057] Step 6: Obtain more accurate river cross-sectional flow through the area integration method commonly used in hydrology.

[0058] The present invention is not limited to the specific technical solutions described in the above embodiments, and all technical solutions formed by equivalent replacement are within the protection scope required by the present invention.

Claims

1. A method for rapid detection of water surface velocity in large rivers based on CFD computational fluid dynamics and video image recognition, characterized in that : Step 1: First, for the hydrological station or the station where flow measurement is required, obtain the size of the bridge pier foundation near it and the size of the riverbed terrain; Step 2: Establish physical models of upstream and downstream river sections and bridge pier foundations, perform mesh generation, and establish a CFD numerical model; Step 3: Use the VOF open channel flow model to simulate the shape and angle characteristics of the water surface boundary line formed by the water surface and the bridge pier foundation under different water depths and different flow rates; specifically, use the VOF water-air multiphase flow open channel flow model to perform CFD numerical simulation, where the VOF model uses the following equation to define the volume ratio of water and air: ； α q Represents the fluid in a discrete unit q The volume percentage of yes q The density of the phase, S αq is the source term, defaults to zero, One of the phases q To p The conversion quality, represent q The flow rate of the phase, It is one of the p To q The conversion quality, is the Laplace operator; Through the VOF model simulation results, a phase q To p The position plane with a value of 0.5 is the water surface position. The shape and angle characteristics of the boundary line between the water surface and the bridge pier foundation under different water depths and flow rates can be obtained through the water surface position. ; Step 4, forming a characteristic table of the shape and angle features in step 3 and the corresponding flow velocity at the bridge pier foundation, and inputting the table into the embedded computer; Step 5, turn on the camera, adjust the camera field of view to the bridge pier foundation, and obtain a video image of the interface area formed by the bridge pier foundation and the water surface; Step 6: Using the water boundary algorithm built into the embedded computer, the shape and angle features of the video boundary image are identified and matched with the boundary shape and angle simulated by CFD in step 3 to quickly obtain the flow velocity of the large river at the bridge pier foundation; Step 7: Using the velocity characteristics obtained in step 6, flow inversion of the entire river section is performed.

2. The method for rapid detection of water surface velocity of large rivers based on CFD computational fluid dynamics and video image recognition according to claim 1 is characterized in that The specific steps of the monitoring method are as follows: Step 1: First, for the hydrological station or the station that needs to be monitored or measured, obtain the size of the bridge pier foundation near it and the riverbed topography upstream and downstream of the river section; Step 2: Use SolidWork or CAD tools to build a three-dimensional physical model of the upstream and downstream river sections and bridge pier foundations; perform meshing of the three-dimensional model and build a CFD numerical model; Step 3, CFD numerical simulation is performed using the VOF water-air multiphase open channel flow model, where the VOF model uses the following equation to define the volume ratio of water and air: ； α q Represents the fluid in a discrete unit q The volume proportion of yes q The density of the phase, S αq is the source term, defaults to zero, One of the phases q To p The conversion quality, represent q The flow rate of the phase, It is one of the p To q The conversion quality, is the Laplace operator; Through the VOF model simulation results, a phase q To p The position plane with a value of 0.5 is the water surface position. The shape and angle characteristics of the boundary line between the water surface and the bridge pier foundation under different water depths and flow rates can be obtained through the water surface position. ; By fitting the shape of the water surface boundary line, a CFD model is established to simulate the shape and angle characteristics of the flow velocity and the water surface boundary line. The mutual relationship among them represents the simulated flow velocity at the bridge pier foundation, Represents the shape and angle characteristics of the simulated water surface boundary line, A. B and C are constants of the fitted equation: Step 4: Add the shape angle features from step 3 , and the corresponding flow rate To correspond, to form A characteristic table is input into the embedded computer as items to be matched; Step 5, the video image recognition device includes a video camera, an embedded computer, and a solar panel; Turning on the camera of the video image recognition device, adjusting the camera field of view to the bridge pier foundation, and acquiring a video image in the interface area formed by the bridge pier foundation and the water surface; Step 6: Identify the shape and angle features of the video boundary image through the water boundary algorithm built into the embedded computer ; Will Same as the shape angle feature in step 6 To correspond, by comparison Characteristic table, quickly obtain the flow velocity of large rivers at the bridge pier foundation ; Step 7, repeat the velocity characteristics at a bridge pier foundation obtained in steps 5 and 6, match the velocity at different pier foundations of the entire river section, and obtain the velocity distribution at different positions on the entire river section; The flow of the entire river section can be inferred using the cross-sectional area integral method commonly used in hydrology.

3. The method for rapid detection of water surface velocity of large rivers based on the combination of CFD computational fluid dynamics and video image recognition according to claim 1 is characterized in that: It includes two parts: CFD numerical calculation and video image recognition of water surface boundary line.

4. The method for rapid detection of water surface velocity of large rivers based on the combination of CFD computational fluid dynamics and video image recognition according to claim 1 is characterized in that: The video image recognition device includes a video camera, an embedded computer, and a solar panel, wherein the solar panel can be powered by a battery in the field for use by an edge computing computer.

Images