Aeration pipeline water-gas two-phase identification method

Abstract

The invention relates to an aeration pipeline water-gas two-phase flow pattern identification method, which comprises the following steps that an aeration pipeline system consisting of a variable frequency water pump, a variable frequency air pump, a high-speed camera and a transparent pipeline is built, the variable frequency water pump is matched with an aeration device to form water-gas two-phase flow, and the high-speed camera shoots the flowing state of the water-gas two-phase flow; the method comprises the steps that a YOLOv5 model is used for classifying based on image characteristics, classification software is constructed through operation such as data enhancement, and the software precision is smaller than 0.886, the model parameter quantity (mAP50 is 0.892) and the average precision mean value (mAP0.5-0.95 is 0.775); a method for calculating the Reynolds number and the on-way head loss coefficient of the aeration pipeline under different flow pattern conditions is provided, and a method for calculating the on-way head loss of the aeration pipeline is established. The method not only solves the identification problem of the water-gas two-phase flow pattern in the aeration irrigation system, but also has important significance in water-gas two-phase flow hydraulic force calculation.

Description

Technical Field

[0001] The invention relates to a method for identifying a water-gas two-phase flow pattern in an aeration irrigation system, in particular to a method for identifying a water-gas two-phase flow pattern in an aeration pipeline. Background Art

[0002] In modern agriculture, water-saving irrigation technology is gaining increasing attention. Aeration irrigation combined with water-saving irrigation technology is a highly efficient irrigation method. It introduces air into the irrigation water to improve soil aeration, thereby promoting crop root growth, improving irrigation efficiency and crop yield. Aeration irrigation also adapts to the development trend of precision agriculture. Through precise control of parameters such as aeration volume and irrigation water volume, more precise irrigation can be achieved according to the needs of different crops, different growth stages, and different soil types, providing technical support for the development of smart agriculture. The identification of water-gas two-phase flow patterns in aeration irrigation is of great significance to the hydraulic design of irrigation pipe networks. Correctly identifying the water-gas two-phase flow pattern can help adjust parameters such as aeration volume to achieve a more uniform mixing of water and gas, thereby optimizing irrigation uniformity and improving irrigation effectiveness. It can also control parameters such as pressure and flow in the irrigation system, thereby improving the utilization efficiency of the irrigation system. However, the current research technology only focuses on aeration irrigation to provide high irrigation efficiency and crop yield, but does not conduct further theoretical analysis on the state identification, flow rate, pressure and hydraulics of the water-gas two-phase flow pattern in the aeration pipe during aeration irrigation, and does not deduce how to efficiently add air to make the gas-water ratio in the aeration pipe more scientific and reasonable. Therefore, the development of a method for identifying the water-gas two-phase flow pattern in the aeration pipe has very important theoretical and practical significance. Summary of the Invention

[0003] In view of the problems and defects in the above-mentioned prior art, the purpose of the present invention is to provide a method for identifying water-gas two-phase flow patterns in an aeration pipe, so as to overcome the problem of identifying water-gas two-phase flow patterns in an aeration irrigation system.

[0004] To achieve the above object, the present invention adopts the following technical solution: a method for identifying water-gas two-phase flow patterns in an aeration pipeline, comprising the following steps:

[0005] 1) Build an aeration pipe system consisting of a variable-frequency water pump, a variable-frequency air pump, a high-speed camera, a flow meter, a pressure sensor, a PLC controller, a PC, and a transparent pipe. The variable-frequency water pump provides different pressure heads. The variable-frequency air pump cooperates with the water pump to form a water-gas two-phase flow under different flow rates and aeration volumes. The high-speed camera is used to capture the water-gas two-phase flow pattern in the transparent pipe, recording the flow state under different working conditions and obtaining several sets of continuous water-gas two-phase flow pattern images.

[0006] 2) The high-speed camera sends the captured water-gas two-phase flow state images to a PC, and uses the YOLOv5 model to classify the image characteristics under different flow patterns of the water-gas two-phase flow, and establishes water-gas two-phase flow pattern classification software.

[0007] 3) Based on the pipeline flow data detected by the electromagnetic flowmeter and the pipeline pressure data detected by the pressure sensor and hydraulic theory analysis, the aeration pipeline pressure calculation formula is established:

[0008] Aeration pipe pressure calculation formula: H0′=H w +4.009Q A

[0009] Where H0′ is the pressure head of the aeration pipe, m; H w is the pure water pipe pressure head, m; Q A is the aeration volume, 10 -3 *m 3 / s.

[0010] 4) The Reynolds number calculation method of the aeration pipeline is obtained based on the pipeline flow data detected by the electromagnetic flowmeter, the pipeline pressure data detected by the pressure sensor, and the hydraulic theory analysis.

[0011] The calculation formula for the Reynolds number of water-gas two-phase flow is:

[0012] Where Re′ is the Reynolds number of the aeration pipe; λ′ is the head loss coefficient along the aeration pipe.

[0013] 5) Establish a calculation formula for head loss along the aeration pipeline

[0014] The calculation formula for the head loss along the aeration pipeline is:

[0015] Among them, H′ f is the head loss along the aeration pipe, m; λ′ is the head loss coefficient along the aeration pipe, l is the pipe length, m; d is the pipe diameter, m; v′ is the water flow velocity in the aeration pipe, m 3 / s;Q a is the aeration volume, 10 -3 *m 3 / s;V w is the flow velocity of pure water pipeline, m 3 / s; g is the acceleration due to gravity, m / s 2 .

[0016] The water-gas two-phase flow pattern classification software establishes classification standards for image characteristics under different states and performs data enhancement, dividing the water-gas two-phase flow patterns into three categories: bubble flow, large bubble flow, and plug flow. The classification software based on Yolov5 water-gas two-phase flow patterns has an accuracy of 0.886, a model parameter mAP50 of 0.892, and an average precision mean mAP0.5-0.95 of 0.775.

[0017] The calculation formula of the head loss coefficient along the aeration pipeline is:

[0018] λ′=λ w +0.018Q a

[0019] Where λ′ is the head loss coefficient along the aeration pipe, m; λ w is the head loss coefficient along the pure water pipeline, m; Q a is the aeration volume, 10 -3 *m 3 / s.

[0020] The calculation formula of the head loss coefficient and Re' along the bubble flow pattern is: R e ′=(279λ′) 4 R 2 =0.986; the calculation formula of the head loss coefficient along the large bubble flow pattern and Re' is: R 2 =0.999; the calculation formula of the plug flow type along the head loss coefficient and Re' is: R 2 =0.999.

[0021] Compared with the prior art, the present invention has the following advantages:

[0022] The present invention proposes a method for calculating the Reynolds number and the head loss coefficient along the aeration pipeline under different flow conditions, and establishes a method for calculating the head loss along the aeration pipeline. This method not only solves the problem of identifying the water-gas two-phase flow pattern in the aeration irrigation system, but also has important significance for the hydraulic calculation of the water-gas two-phase flow. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 Schematic diagram of the aeration pipeline system structure of the aeration pipeline water-gas two-phase identification method of the present invention;

[0024] Figure 2 Diagram of different types of water-gas two-phase flow patterns;

[0025] Figure 3 Effect diagram of different image enhancement methods;

[0026] Figure 4Flowchart for constructing water-gas two-phase flow pattern detection model;

[0027] Figure 5 Comparison chart of classification software test results;

[0028] Figure 6 Aeration pipe pressure analysis diagram. DETAILED DESCRIPTION

[0029] The water-gas two-phase identification method of the aeration pipeline is further described in detail below with reference to the accompanying drawings and embodiments provided by the inventor.

[0030] Example 1: Aeration Pipe System with a Method for Identifying Water and Gas Phases in an Aeration Pipe

[0031] Figure 1 A schematic diagram of the aeration pipeline system structure for a method for identifying water and gas phases in an aeration pipeline according to the present invention is provided. The system comprises a variable-frequency water pump 1, an analog converter 2, a PLC controller 3, a PC 4, a check valve 5, a gas flowmeter 6, a variable-frequency air pump 7, an electromagnetic flowmeter 8, a pressure sensor 9, and a high-speed camera 10. The PLC controller controls the variable-frequency water pump and the variable-frequency air pump to provide the system with target pressure head and aeration volume. The pressure sensor collects pipeline pressure data in real time, the gas flowmeter collects aeration volume data in real time, and the electromagnetic flowmeter collects pipeline flow data in real time. The PC records pipeline pressure, aeration volume, and flow data in real time via the analog converter. The high-speed camera records water flow pattern data under different operating conditions.

[0032] Example 2: A method for identifying water and gas phases in an aeration pipeline

[0033] 1. After the test device of the embodiment is turned on, the variable frequency water pump provides different pressure heads. The aeration device cooperates with the water pump to form a water-gas two-phase flow under conditions of different flow rates and aeration amounts. A high-speed camera is used to capture the water-gas two-phase flow pattern in the transparent pipe, recording the flow state under different working conditions, and obtaining several sets of continuous water-gas two-phase flow pattern images.

[0034] 2. The high-speed camera sends the water-gas two-phase flow state images captured to a PC. The YOLO v5 model is used to classify the image characteristics under different flow patterns of the water-gas two-phase flow, and water-gas two-phase flow pattern classification software is established.

[0035] 3. Water-gas two-phase flow classification software: Based on the image characteristics under different working conditions, the following classification standards are established to divide the images into 3 categories: Figure 2 . Figure 2a. Bubble flow b. Large bubble flow c. Plug flow. The LabelImg object detection and annotation tool was used to manually annotate the test tube region in the image. The image and annotation results were used as the dataset for training the YOLOv5 network. The training and validation sets were divided in a 6:3:1 ratio. Before training and testing, the collected images were augmented by adding Gaussian noise, random brightness changes, and increasing contrast. Figure 3 .

[0036] The water-gas two-phase flow classification model is built on the YOLOv5 version. The construction process is as follows: Figure 4 As shown, the results of the water-gas two-phase flow pattern classification software finally established are shown in Table 1.

[0037] Table 1 Quality results of water-gas two-phase classification model

[0038] type Precision Recall mAP@50 mAP0.5-0.95 All 0.863 0.886 0.892 0.775 Bubble Flow 0.885 0.865 0.946 0.779 Large bubble flow 0.932 0.844 0.939 0.799 plug flow 0.787 0.941 0.904 0.747

[0039] The overall precision was 0.863, indicating that 86.3% of the model's positive predictions were correct. Recall (R): The model achieved an overall recall of 0.886, meaning it correctly identified 88.6.4% of all true positives. Large Bubble Flow had the lowest recall, at 0.826. mAP50: The mean average precision at an IoU threshold of 0.50 was 0.892 overall. This metric indicates the model's accuracy in predicting object centers. Large Bubble Flow had the lowest mAP50, at 0.815. Overall performance demonstrates a good balance between precision and recall.

[0040] Input 50 newly taken water-gas two-phase flow images into the improved model and analyze the model's image classification results, such as Figure 6 The detection results show that 50 targets were detected in 50 water-gas two-phase flow images, with no missed detections. This shows that the water-gas two-phase flow recognition and classification model established based on the YOLOv5 model can meet the actual water-gas two-phase flow detection and classification task requirements, and performs well in scenarios such as changing light intensity, image blur, and target rotation.

[0041] 4. Reynolds number for water-gas two-phase flow: Based on the hydraulic pipe flow theory, the Reynolds number is one of the important parameters for determining the flow pattern. Its calculation method is as follows:

[0042]

[0043] Where R e is the Reynolds number, v is the flow velocity in the pipe, m / s; d is the pipe diameter, m; v is the viscosity coefficient of the water flow, m2 / s.

[0044] The pipeline flow formula is

[0045]

[0046] Where Q is the pipe flow rate, m3 / s; A is the cross-sectional area, m 2 ;μ c is the pipe flow coefficient; g is the acceleration of gravity, m / s2; H0 is the pipe pressure head, m.

[0047] Combined with formula (2), the formula for the duct flow rate is:

[0048]

[0049] Where: v is the pipe flow velocity, m / s; μ c is the pipe flow coefficient; g is the acceleration of gravity, m / s 2 ; H0 is the main pipe pressure head, m.

[0050] The water flow in the pipe is turbulent. According to the results of Nicolaz test, when Re<10 6 And when it belongs to the turbulent smooth region, Nicolaz formula is more suitable, that is,

[0051]

[0052] Where λ is the resistance coefficient along the path.

[0053] The calculation formula for the loss along the pressurized pipeline is:

[0054]

[0055] Among them, H f is the head loss along the way, m; λ is the head loss coefficient along the way, m; l is the pipe length, m, d is the pipe diameter, m; v is the water velocity, m / s; g is the acceleration of gravity, m / s 2 .

[0056] Combining (1)-(5), we can get the calculation method of the Reynolds number of the aeration pipe:

[0057]

[0058] Where Re′ is the Reynolds number of the aeration pipe; λ′ is the head loss coefficient along the aeration pipe.

[0059] 5. Perform linear regression analysis on the head loss coefficient along the aeration pipeline, the head loss coefficient along the pure water pipeline, and the aeration volume, and calculate the head loss coefficient along the aeration pipeline as follows:

[0060] λ′=λ w +0.018Q a R 2 =0.934

[0061] Among them, λ′ is the head loss coefficient along the aeration pipe, λ w is the head loss coefficient along the pure water pipeline, Q a is the aeration volume, 10 -3 *m 3 / s.

[0062] 6. Obtain the formula for calculating the Reynolds number and head loss along the bubbly flow:

[0063] R e ′=(279λ′) 4 R 2 =0.986

[0064] The calculation formula of Reynolds number and head loss along the large bubble flow is obtained:

[0065] R 2 =0.999

[0066] The calculation formula of plug flow Reynolds number and head loss along the flow path is obtained:

[0067] R 2 =0.999

[0068] 7. The present invention collects the pressure 50cm behind the aeration position. The pressure fluctuation of the aeration pipe is greater than that of the pure water condition. At the same time, the aeration volume can increase the pipe pressure of the pure water condition to a certain extent. By performing a multivariate linear regression analysis on the main pipe pressure, aeration volume, and aeration pipe pressure

[0069] H0′=H w +4.009Q A R 2 =0.967

[0070] Where H0′ is the pressure head of the aeration pipe, m; H w Pipeline pressure head, m; Q A is the aeration volume, 10 -3 *m 3 / s.

[0071] To verify the accuracy of the above analysis results, the relationship between the three was verified by measuring the actual aeration pipeline pressure head corresponding to a certain gas pressure head under different pipeline pressures. The calculation results are shown in Table 2. The maximum relative error between the measured value and the calculated value is 2.21%, and the minimum is 1.15%.

[0072] Table 2 Verification table of the relationship between pipeline pressure head, air pipe pressure and aeration pipeline pressure head

[0073] Pipeline pressure head (m) Tracheal pressure (m) Calculated value Measured value error 11 11.35 13.30 13.10 0.0153 17 18.05 19.83 19.40 0.0221 22 23.63 25.27 24.98 0.0115

[0074] 8. Linear regression of the aeration pipe water flow velocity, aeration volume and pure water pipe flow velocity can be obtained to obtain the aeration pipe water flow velocity calculation formula.

[0075] v′=V w -0.043Q a R 2 =0.999

[0076] Where v′ is the water flow velocity in the aeration pipe, m 3 / s;Q a is the aeration volume, 10 -3 *m 3 / s;V w is the flow rate of pure water pipeline.

[0077] 9. The calculation formula for the head loss along the aeration pipeline is as follows:

[0078] This invention proposes a method for identifying water-gas two-phase flow patterns in aeration pipes. This method uses a variable-frequency water pump and an aeration device to generate water-gas two-phase flow. A high-speed camera captures the flow state, and classification software is constructed based on image characteristics using the YOLOv5 model. The invention proposes a method for calculating the Reynolds number and head loss coefficient along the aeration pipe under different flow patterns, and establishes a method for calculating head loss along the aeration pipe. This method not only solves the problem of identifying water-gas two-phase flow patterns in aerated irrigation systems but also has important implications for hydraulic calculations of water-gas two-phase flow.

Claims

1. A method for identifying water-gas two-phase flow patterns in an aeration pipeline, characterized by: It includes the following steps: 1) Build an aeration pipeline system consisting of a variable-frequency water pump, a variable-frequency air pump, a high-speed camera, a flow meter, a pressure sensor, a PLC controller, a PC, and a transparent pipe. The variable-frequency water pump provides different pressure heads. The variable-frequency air pump cooperates with the water pump to form a water-gas two-phase flow under different flow rates and aeration rates. The high-speed camera is used to capture the water-gas two-phase flow pattern in the transparent pipe, recording the flow state under different working conditions and obtaining several sets of continuous water-gas two-phase flow pattern images. 2) Use the YOLO v5 model to classify the image characteristics of different water-gas two-phase flow patterns and establish water-gas two-phase flow pattern classification software; 3) Based on the pipeline flow data detected by the electromagnetic flowmeter and the pipeline pressure data detected by the pressure sensor and hydraulic theory analysis, the aeration pipeline pressure calculation formula is established: Aeration pipe pressure calculation formula: H0′=H w +4.009Q a Where H0′ is the pressure head of the aeration pipe, m; H w is the pure water pipe pressure head, m; Q a is the aeration volume, 10 -3 *m 3 / s; 4) The Reynolds number calculation method of the aeration pipeline is obtained based on the pipeline flow data detected by the electromagnetic flowmeter, the pipeline pressure data detected by the pressure sensor, and the hydraulic theory analysis. The calculation formula for the Reynolds number of water-gas two-phase flow is: Where Re′ is the Reynolds number of the aeration pipe; λ′ is the head loss coefficient along the aeration pipe; 5) Establish a calculation formula for head loss along the aeration pipeline The calculation formula for the head loss along the aeration pipeline is: Among them, H′ f is the head loss along the aeration pipe, m; λ′ is the head loss coefficient along the aeration pipe, l is the pipe length, m; d is the pipe diameter, m; v′ is the water flow velocity in the aeration pipe, m 3 / s;Q a is the aeration volume, 10 -3 *m 3 / s;V w is the flow velocity of pure water pipeline, m 3 / s; g is the acceleration due to gravity, m / s 2 .

2. The method for identifying water-gas two-phase flow patterns in an aeration pipeline according to claim 1, characterized in that: The water-gas two-phase flow type classification software establishes classification standards for image characteristics under different states and performs data enhancement, dividing the water-gas two-phase flow types into three categories: bubble flow, large bubble flow, and plug flow. The established water-gas two-phase flow type classification software based on Yolov5 has an accuracy of 0.886, a model parameter mAP50 of 0.892, and an average precision mean mAP0.5-0.95 of 0.

775.

3. The method for identifying water-gas two-phase flow patterns in an aeration pipeline according to claim 1, characterized in that: The calculation formula of the head loss coefficient along the aeration pipeline is: λ′=λ w +0.018Q a Where λ′ is the head loss coefficient along the aeration pipe, m; λ w is the head loss coefficient along the pure water pipeline, m; Q a is the aeration volume, 10 -3 *m 3 / s.

4. A method for identifying water-gas two-phase flow patterns in an aeration pipeline according to claim 2, characterized in that: The head loss coefficient along the bubble flow pattern is related to R e Calculation formula of ′: R e ′=(279λ′) 4 R 2 =0.986; the head loss coefficient along the large bubble flow pattern is related to R e Calculation formula of ′: R 2 =0.999; the head loss coefficient along the plug flow pattern is related to R e Calculation formula of ′: R 2 =0.999.

Images