A gas-solid two-phase flow pattern identification method, system, computer equipment and medium

Abstract

The present invention belongs to the technical field of multiphase flow pattern detection, and discloses a method, system, computer equipment and medium for identifying the flow pattern of a gas-solid two-phase flow. The method of the present invention applies the fractional Fourier transform (FRFT) signal processing method based on the optimal transformation order to the flow pattern identification of the gas-solid two-phase flow for the first time. By processing the signal samples, the signal energy is highly concentrated, so that the signal characteristics are more prominent, and the detection and analysis capabilities of the gas-solid two-phase flow signal are improved. In addition, the present invention uses the minnow optimization BFO algorithm to calculate the optimal transformation order of the gas-solid two-phase flow signal sample and its corresponding maximum amplitude. The present invention applies the fractional Fourier transform (FRFT) optimized by the minnow optimization BFO algorithm to the flow pattern identification of the gas-solid two-phase flow, which can realize the effective identification of the two flow patterns of laminar flow and dispersed flow in the gas-solid two-phase flow in the horizontal pneumatic conveying pipeline.

Description

Technical Field

[0001] The present invention belongs to the technical field of multiphase flow pattern detection, and in particular relates to a gas-solid two-phase flow pattern identification method, system, computer equipment and medium. Background Art

[0002] Gas-solid two-phase flow is a complex random process that is widely present in the fields of energy, chemical industry, pharmaceuticals, agriculture and mining. In order to reduce equipment wear, reduce environmental pollution, and improve the safety and efficiency of industrial production processes, pneumatic conveying technology has been gradually applied to the measurement of gas-solid two-phase flow parameters. Accurately identifying the flow pattern of gas-solid two-phase flow (mainly divided into laminar flow and dispersed flow) plays an important role in calculating key flow parameters such as gas velocity, particle diameter and solid concentration of gas-solid two-phase flow, and helps prevent blockage and wear in pneumatic conveying pipelines, and improve the stability and reliability of the system.

[0003] At present, the existing gas-solid two-phase flow pattern identification methods have certain limitations in identifying two-phase flow patterns, such as unsatisfactory results when processing nonlinear and non-stationary signals, high sensitivity to noise, and complexity of operation steps. Such limitations make it impossible to accurately realize the effective identification of laminar flow and dispersed flow in gas-solid two-phase flow.

[0004] Fractional Fourier Transform (FRFT) is a powerful time-frequency analysis tool that effectively avoids cross-term interference when processing complex signals. Due to the flexible transformation characteristics of FRFT in the time-frequency domain, more and more studies have applied FRFT to the field of signal processing in recent years. Although the FRFT method based on the optimal fractional order has achieved ideal results in specific fields, its application in gas-solid two-phase flow signal processing is still rare. Summary of the invention

[0005] The purpose of the present invention is to propose a method for identifying flow patterns of gas-solid two-phase flow, which uses a FRFT signal processing method based on the optimal fractional order to extract the flow characteristics of particles in a horizontal pneumatic conveying pipeline, and based on the minnow optimization BFO algorithm, finds the optimal transformation order p and its corresponding maximum amplitude MA of the signal sample after FRFT processing, thereby facilitating the effective identification of laminar flow and dispersed flow in the gas-solid two-phase flow in the horizontal pneumatic conveying pipeline.

[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical scheme: a method for identifying flow patterns of gas-solid two-phase flow, comprising the following steps: step 1. collecting particle flow signals of different flow patterns of gas-solid two-phase flow, and preprocessing the signals to remove noise interference; decomposing the processed signals into multiple time series signal samples; step 2. using a signal processing method based on fractional Fourier transform FRFT for each time series signal sample, by selecting the transformation order, converting the signal sample from the time domain to the fractional order domain, thereby extracting the frequency domain characteristics of the signal; step 3. using the minnow optimization BFO algorithm to find the optimal transformation order of each signal sample in the time series after FRFT processing and its corresponding maximum amplitude MA; Step 4. Preset the parameters corresponding to the signal samples extracted from the time series by different flow types after FRFT processing The value range of MA is used as the basis for identifying different flow patterns of gas-solid two-phase flow, thereby realizing the identification of gas-solid two-phase flow pattern.

[0007] In addition, based on the above-mentioned gas-solid two-phase flow type identification method, the present invention also proposes a gas-solid two-phase flow type identification system that is compatible with the gas-solid two-phase flow type identification method, which adopts the following technical solutions: a preprocessing and signal decomposition module, which is used to collect particle flow signals of different flow types of gas-solid two-phase flow, and preprocess the signals to remove noise interference; decompose the processed signals into multiple time series signal samples; a fractional-order Fourier transform module, which is used to use a FRFT-based signal processing method for each time series signal sample, and convert the signal sample from the time domain to the fractional-order domain by selecting the transformation order, thereby extracting the frequency domain characteristics of the signal; an optimization module, which is used to use the minnow optimization BFO algorithm to find the optimal transformation order for each signal sample in the time series after FRFT processing. and its corresponding maximum amplitude MA; and a flow pattern recognition module, which is used to preset the parameters corresponding to the signal samples extracted from the time series by different flow patterns after FRFT processing The value range of MA is used as the basis for identifying different flow patterns of gas-solid two-phase flow, thereby realizing the identification of gas-solid two-phase flow pattern.

[0008] Based on the gas-solid two-phase flow pattern identification method, the present invention also proposes a computer device, which includes a memory and one or more processors. The memory stores executable code, and when the processor executes the executable code, it is used to implement the gas-solid two-phase flow pattern identification method mentioned above.

[0009] Based on the gas-solid two-phase flow pattern identification method, the present invention also proposes a computer-readable storage medium on which a program is stored. When the program is executed by a processor, it is used to implement the gas-solid two-phase flow pattern identification method mentioned above.

[0010] The present invention has the following advantages: As mentioned above, the present invention relates to a method for identifying the flow pattern of a gas-solid two-phase flow. The method applies the fractional Fourier transform FRFT signal processing method based on the optimal transformation order to the flow pattern identification of the gas-solid two-phase flow for the first time. By processing the signal samples, the signal energy is highly concentrated, so that the signal characteristics are more prominent, and the detection and analysis capabilities of the gas-solid two-phase flow signal are improved. In addition, the present invention uses the minnow optimization BFO algorithm to calculate the optimal transformation order and the corresponding maximum amplitude of the gas-solid two-phase flow signal sample. By applying the fractional Fourier transform FRFT optimized by the minnow optimization BFO algorithm to the flow pattern identification of the gas-solid two-phase flow, the effective identification of the two flow patterns of laminar flow and dispersed flow in the gas-solid two-phase flow in the horizontal pneumatic conveying pipeline can be achieved, which provides a new perspective for extracting the signal characteristics of the gas-solid two-phase flow, and obtains the general method steps for identifying the laminar flow and dispersed flow in the gas-solid two-phase flow. At the same time, the process of identifying the flow pattern of the gas-solid two-phase flow is simplified. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 Flow chart of the gas-solid two-phase flow pattern identification method in an embodiment of the present invention.

[0012] Figure 2 Schematic diagram of the annular electrostatic sensor used in the embodiment of the present invention.

[0013] Figure 3 Schematic diagram of the experimental platform constructed in the embodiment of the present invention.

[0014] Figure 4 Schematic diagram of laminar flow photographs taken by a high-speed camera in an embodiment of the present invention.

[0015] Figure 5 Schematic diagram of a dispersed flow photograph taken by a high-speed camera in an embodiment of the present invention.

[0016] Figure 6 Schematic diagram of the time-frequency domain plane and a set of coordinate systems rotated by an angle α relative to the original coordinate system.

[0017] Figure 7 Schematic diagram of the gas-solid two-phase flow signal sample after FRFT processing at the optimal order.

[0018] Figure 8 Schematic diagram of laminar voltage signal in an embodiment of the present invention.

[0019] Fig. 9 Schematic diagram of a dispersed flow voltage signal in an embodiment of the present invention.

[0020] Fig.10 Schematic diagram of laminar flow and dispersed flow signal samples collected in an embodiment of the present invention.

[0021] Fig.11 Schematic diagram of laminar flow signal IMFs after EMD processing in an embodiment of the present invention.

[0022] Fig.12 Schematic diagram of dispersed flow signal IMFs after EMD processing in an embodiment of the present invention.

[0023] Fig.13 Schematic diagram of a gas-solid two-phase flow sample after FRFT processing in an embodiment of the present invention.

[0024] Fig.14 It is a box plot of the optimal transformation order of laminar flow and dispersed flow signal samples after FRFT processing in an embodiment of the present invention.

[0025] Fig.15 It is a box plot of the maximum amplitude of the laminar flow and dispersed flow signal samples after FRFT processing in an embodiment of the present invention.

[0026] Fig.16 This is a scatter plot of the gas-solid two-phase flow signal sample after FRFT processing in an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The present invention is further described in detail below with reference to the accompanying drawings and specific implementation methods: This embodiment describes a gas-solid two-phase flow pattern identification method to achieve effective identification of laminar flow and dispersed flow in gas-solid two-phase flow in a horizontal pneumatic conveying pipeline.

[0028] Example 1 Figure 1 As shown, the gas-solid two-phase flow pattern identification method in this embodiment includes the following steps: Step 1. Collecting particle flow signals of different flow patterns of gas-solid two-phase flow, and preprocessing the signals to remove noise interference; decomposing the processed signals into multiple time series signal samples.

[0029] In this embodiment, in order to collect the particle flow signals of different flow patterns of gas-solid two-phase flow, the following method is used: Figure 2 The annular electrostatic sensor shown in FIG. Figure 3 The experimental platform shown is used for flow pattern identification of gas-solid two-phase flow.

[0030] Based on the characteristics of Geldart D particles such as large particle size, high density and large inertial force, the present invention selects Geldart D particles with a diameter of 2-5 mm to simulate the flow state of gas-solid two-phase flow in the pipeline.

[0031] In order to visualize the particle flow and shield the interference of the external electric field, a transparent polyvinyl chloride tube was used in this experiment. The thickness, inner diameter and length of the tube were 5 mm, 100 mm and 3 m respectively. A ring-shaped electrostatic sensor was installed on the outer wall of the tube. Figure 2 and Figure 3 As shown, the sensor is a ring electrode made of copper with a thickness of 1 mm and a width of 15 mm.

[0032] In order to generate a stable airflow, a blower is installed at the inlet of the duct, such as Figure 3 As shown. A high-speed camera is used to record the movement of particles in the pipeline. In addition, a recovery device is installed at the outlet of the pipeline to achieve the recycling of particles.

[0033] When particles flow in a horizontal pipe, they become charged due to the collision and friction between the particles and the friction between the particles and the pipe wall. When the particles pass through the sensing area of ​​the annular electrostatic sensor, the electrodes generate induced current signals, which are converted into voltage signals by the signal processing unit. Subsequently, these signals are collected by the data acquisition board at a sampling frequency of 20kHz. Finally, the collected data is transmitted to the computer for storage and analysis.

[0034] The signal processing unit and the data acquisition board are different functional modules of the data acquisition system. The signal processing unit is the front-end part, responsible for amplifying and filtering the signals collected by the sensor to ensure the validity and quality of the signals. The data acquisition board is the core part, responsible for digitizing the processed signals and transmitting them to the computer to ensure that the signals can be stored and analyzed.

[0035] By adjusting the material barrel control valve and the blower speed, two flow patterns, laminar flow and dispersed flow, were simulated. Specifically, when the air flow velocity and material feed rate were low, the particles accumulated at the bottom of the pipe, forming a layered and uniform flow pattern. At this stage, the particle flow velocity was highest at the center of the pipe, and decreased significantly near the pipe wall, gradually forming a laminar flow, such as Figure 4 Shown in the middle circle.

[0036] As the feed rate and air velocity increase, the particles gradually reach the minimum pickup velocity. Finally, the particles are suspended in the pipe and randomly distributed throughout the pipe, forming a dispersed flow with random trajectories, such as Figure 5 Shown in the middle circle.

[0037] In the embodiment of the present invention, for each flow type, the measurement time of the particle flow signal is set to 10 seconds. In this embodiment, the voltage signals of the laminar flow and the dispersed flow are measured, such as Figure 8 and Fig. 9 shown.

[0038] like Figure 8 and Fig. 9 As shown in Figure 2, the original time domain signals of the two flow types show obvious differences. However, due to the time-frequency coupling and non-stationary characteristics of the gas-solid two-phase flow signal, only using time domain analysis is not enough to capture its frequency domain characteristics.

[0039] Therefore, the present invention adopts the FRFT method to perform a more comprehensive and accurate analysis of the gas-solid two-phase flow signal.

[0040] In order to better analyze the signal characteristics and simplify the calculation, 50 0.2 second signal samples were extracted from laminar flow and dispersed flow, respectively. Specifically, each signal sample contains 4000 sampling points, and the voltage amplitude ranges from -5V to 5V.

[0041] In this embodiment, the laminar flow and dispersed flow signal samples collected are as follows: Fig.10 shown.

[0042] like Fig.10 The (a1), (a2), ... on the left side represent the 1st, 2nd, ... laminar flow signal samples respectively. Fig.10 The (b1), (b2), ... etc. on the right side represent the 1st, 2nd, ... dispersed flow signal samples respectively.

[0043] Step 2. For each time series signal sample, a signal processing method based on fractional Fourier transform (FRFT) is used. By selecting the transformation order, the signal sample is converted from the time domain to the fractional order domain, thereby extracting the frequency domain characteristics of the signal.

[0044] Fractional Fourier transform (FRFT) is a time-frequency domain transformation method that rotates the signal in the time-frequency domain plane by introducing an adjustable fractional order. Compared with traditional signal processing methods, fractional Fourier transform provides a more flexible framework, which enables time-frequency domain analysis to be adjusted according to different transformation orders, and exhibits superior performance when processing time-frequency domain coupling and non-stationary signals. In addition, by selecting an appropriate fractional order, FRFT can highly concentrate the signal energy, forming a narrow and sharp peak, making the signal features more prominent and improving the detection and analysis capabilities of the signal.

[0045] Due to these advantages, FRFT is widely used in signal processing, phase recovery, data compression and image reconstruction.

[0046] The present invention uses a signal processing method based on FRFT to process gas-solid two-phase flow signals. Processing; gas-solid two-phase flow signal In the transformation order The fractional Fourier transform of Defined by: .

[0047] in, Represents the kernel function, which plays a key role in fractional Fourier transform and is defined as follows: .

[0048] Where α represents the rotation angle in FRFT; represents an imaginary unit; Indicates the position of the signal on the time axis; Represents the transition domain between the time domain and the frequency domain; Represents an integer.

[0049] represents the Dirac function, which means that There is a transient impulse signal at , and it is 0 at other locations; It means in There is a transient impulse signal at the location and it is 0 at other locations.

[0050] The relationship between the rotation angle α and the transformation order p is as follows: .

[0051] when , FRFT is equivalent to the original time domain signal, that is, the gas-solid two-phase flow signal ;when When , FRFT becomes the traditional Fourier transform; when , which stands for Inverse Fourier Transform.

[0052] These conclusions further prove that FRFT can effectively characterize the time domain and frequency domain characteristics of the signal. In order to reduce the complexity of the calculation, the present invention normalizes the transformation order of FRFT to the interval of [0.5, 1.5].

[0053] Signal along the coordinate system The distribution of Figure 6 shown.

[0054] The coordinate system Relative to the original coordinate system Rotated by angle α. Respectively represent the time axis and frequency axis in the original coordinate system, They respectively represent the new coordinate axes obtained after the time axis and frequency axis are rotated by an angle α.

[0055] Step 3. Use the BFO algorithm to find the optimal transformation order for each signal sample in the time series after FRFT processing. and its corresponding maximum amplitude MA.

[0056] Specifically, the transformation order corresponding to the maximum amplitude after FRFT transformation is defined as the optimal transformation order; the minnow optimization algorithm BFO is used to search for the optimal transformation order and maximum amplitude of FRFT.

[0057] The BFO algorithm for minnow optimization is a novel meta-heuristic method that balances the development and exploration strategies by simulating the reproductive behavior of minnows, effectively avoiding the risk of falling into local optimality.

[0058] The BFO algorithm finds the optimal solution by minimizing the objective function, as shown in the following formula: .

[0059] in, represents the optimal transformation order, represents the voltage amplitude, and It represents the objective function of the BFO algorithm. Represents the amplitude of the signal samples extracted from the time series after FRFT transformation.

[0060] like Figure 7 A sample of the gas-solid two-phase flow signal after FRFT processing at the optimal order is shown.

[0061] Depend on Figure 7 It can be seen that after FRFT processing, the energy of the signal samples extracted from the time series is highly concentrated and a narrow peak is generated, which is beneficial to the detection and analysis of the signal.

[0062] The following is a method to use the minnow optimization BFO algorithm to find the optimal transformation order of each signal sample in the time series after FRFT processing. And its corresponding maximum amplitude MA process: Step 3.1. Initialize the population.

[0063] The BFO algorithm is used to find the optimal transformation order p of the fractional Fourier transform FRFT; the population F consists of n solutions, each of which corresponds to a possible transformation order, expressed as: .

[0064] in, represents the i-th transformation order, , n is the population size, indicating the number of solutions searched.

[0065] The initial value of is randomly generated by the following formula: .

[0066] in, and They represent the lower and upper bounds of the transformation order. In this embodiment, the lower bound The value is 0.5, the upper bound The value is 1.5. Represents a random number between 0 and 1.

[0067] Step 3.2. Construct a fitness function.

[0068] The fitness function is used to evaluate the amplitude of FRFT under different transformation orders p; by calculating the amplitude corresponding to each transformation order and using it as a criterion, the optimal transformation order is found by optimizing the fitness function.

[0069] Fitness function The expression is as follows: .

[0070] in, Indicates the transformation order The corresponding amplitude, .

[0071] Step 3.3. Search and development.

[0072] The BFO algorithm optimizes the transformation order by simulating the search and development strategies of fish schools. During the search process, the step size J is adjusted to achieve the transition from global to local search, thereby exploring the search space and finding the optimal solution.

[0073] The specific formula of this process is as follows: .

[0074] in, represents the position of the i-th solution in the t-th iteration; Indicated in The position of the i-th solution in the iteration; J represents the step length and jump value of the fish, which describes the movement rate of the fish when escaping or approaching the target, and gradually decreases with the algorithm iteration; represents a random solution in the fish school search process, Represents the optimal solution found in the current iteration; It is a random factor used to adjust the direction and distance that the fish school moves toward the target solution.

[0075] It is a predefined threshold used to select the current search direction.

[0076] like , the current solution is based on the random solution Update; if , then based on the optimal solution renew.

[0077] Step 3.4. Avoid local optimum.

[0078] Escape behavior is introduced into the BFO algorithm, and random jumping and a global search strategy based on the center of gravity of the fish school are used near the local optimum to improve the ability to search for the optimal transformation order p; the specific implementation process is as follows: I. Random jumping strategy.

[0079] when When the solution is located within the search range Random updates within: .

[0080] II. Global search strategy based on the centroid of fish school.

[0081] when When the solution position is based on the optimal solution and fish center of gravity renew: .

[0082] Among them, the center of gravity of the fish school .

[0083] Step 3.5. Solutions with low fitness will be eliminated, and their probability of being eliminated is Calculated by the following formula: .

[0084] in, is the objective function of the solution, that is, the transformation order The corresponding amplitude.

[0085] Step 3.6. Termination.

[0086] When the optimal fractional order that maximizes the FRFT amplitude is found When , the BFO algorithm terminates; it is expressed as: .

[0087] in, It represents the optimal transformation order corresponding to the maximum amplitude MA after FRFT processing.

[0088] In addition, the present invention considers the influence of population size and iteration number on the BFO algorithm.

[0089] Specifically, the present invention uses two indicators, standard deviation SD and relative error RE, to determine the values ​​of two parameters, population size and number of iterations, of the BFO algorithm; definitions of SD and RE are shown in the following formula.

[0090] .

[0091] .

[0092] in, represents the optimal transformation order p obtained after the signal samples extracted from the time series are processed by FRFT, Represents the average value of the optimal transformation order obtained after the signal is processed by FRFT , represents the standard deviation, Indicates relative error.

[0093] Table 1 shows the calculation results of SDx and REx% under different parameter combinations (population size and number of iterations).

[0094] Table 1 Calculation results of SDx and REx% under different parameter combinations.

[0095] .

[0096] When the population size is 100 and the number of iterations is 50, both SDx and REx% reach the minimum value, indicating that the data quality is optimal at this time. Therefore, the population size and the number of iterations of the BFO algorithm of the present invention are set to 100 and 50 respectively.

[0097] Step 4. Preset the parameters corresponding to the signal samples extracted from the time series by different flow types after FRFT processing The value range of MA is used as the basis for identifying different flow patterns of gas-solid two-phase flow, thereby realizing the identification of gas-solid two-phase flow pattern.

[0098] After the gas-solid two-phase flow signal samples are processed by the BFO-FRFT method, the optimal transformation orders p of laminar flow and dispersed flow are distributed around 1, such as Fig.14 As shown. For laminar flow, the value range of p is 0.9971 to 1.0037, showing a narrow and relatively symmetrical distribution; while for dispersed flow, the value range of p is 0.995 to 1.0065, showing a wide and irregular distribution. Since the two flow types partially overlap in the optimal transformation order, it is difficult to distinguish them simply by the optimal transformation order. In order to effectively identify the two flow types, the maximum amplitude MA corresponding to the optimal transformation order p after the signal sample is processed by FRFT should be further considered, as shown in Fig.15 As shown in the figure, the MA of laminar flow is between 10.3890 and 18.6541, with a relatively concentrated distribution. The MA of dispersed flow signal samples is between 18.0429 and 45.2909, with a significantly wider distribution range and more obvious volatility.

[0099] In order to make a more intuitive comparison of the parameters of the two flow types, the present invention provides a scatter plot of the gas-solid two-phase flow signal sample after FRFT processing, such as Fig.16 The results show that based on the above parameters The two parameters of and MA can be used to obtain the judgment criteria for identifying the laminar flow and dispersed flow of gas-solid two-phase flow in horizontal pneumatic conveying pipelines: when the optimal transformation order is in the range of 0.9971 to 1.0037, and the MA value is in the range of 10.3890 to 18.6541, it can be identified as laminar flow; when the optimal transformation order is in the range of 0.995 to 1.0065, and the MA value is in the range of 18.0429 to 18.6541, it can be considered to be in a transition state between laminar flow and dispersed flow; when the optimal transformation order is in the range of 0.995 to 1.0065, and the MA value is in the range of 18.6541 to 45.2909, it can be identified as dispersed flow, as shown in Table 2.

[0100] Table 2 Criteria for identifying laminar flow and dispersed flow.

[0101] .

[0102] By using the FRFT signal processing method based on the optimal transformation order, the gas-solid two-phase flow signal shows significant energy aggregation characteristics in the fractional domain, and the signal characteristics are more prominent, which is helpful for the identification of flow types. In addition, the present invention uses the BFO algorithm to calculate the optimal transformation order of the gas-solid two-phase flow signal sample and its corresponding maximum amplitude. The combination of the two is conducive to the effective identification of the two flow types of laminar flow and dispersed flow in the gas-solid two-phase flow in the horizontal pneumatic conveying pipeline.

[0103] The following is a comparative analysis of the results of gas-solid two-phase flow signals processed by the method of the present invention (hereinafter referred to as BFO-FRFT) and the results processed by the traditional EMD method to verify the effectiveness of the BFO-FRFT method proposed in the present invention.

[0104] Empirical mode decomposition (EMD) is an adaptive time series signal decomposition method. Unlike traditional methods that require predefined basis functions, EMD decomposes signals into intrinsic mode functions (IMFs) based on the time scale characteristics of the signal, which makes it particularly suitable for analyzing nonlinear and non-stationary signals. In this experiment, the EMD signal processing method was used to process the gas-solid two-phase flow signal, and the processing results were compared with the BFO-FRFT method to verify the effectiveness of the BFO-FRFT method.

[0105] After EMD processing, the gas-solid two-phase flow signal will be decomposed into multiple IMFs, each of which represents an oscillation component in a specific frequency band. The results of EMD processing of laminar and dispersed flow signals are as follows: Fig.11 and 12 shown.

[0106] It can be seen from the result graph that the oscillation range of laminar flow is mainly concentrated in IMF4-IMF8; the oscillation range of dispersed flow is mainly distributed in IMF3-IMF8, and the oscillation amplitude is relatively large. This phenomenon shows that the laminar flow signal is mainly affected by the low-frequency component, while the dispersed flow signal contains a higher proportion of high-frequency components. However, different flow type signals show significant differences in frequency characteristics after EMD processing. The same order of IMF may correspond to different frequency bands and oscillation modes in laminar flow and dispersed flow. This complexity makes it challenging to effectively identify different flow types through quantitative analysis methods.

[0107] In contrast, after FRFT processing, the energy of the gas-solid two-phase flow signal is obviously concentrated, and there are significant differences in peak amplitude at different transformation orders, such as Fig.13 As shown. By calculating the optimal transformation order and its corresponding maximum amplitude, FRFT can provide quantitative indicators for flow pattern identification, thereby quantifying the characteristic differences of different flow patterns. Compared with the EMD method, the FRFT method in the present invention simplifies the identification process of gas-solid two-phase flow patterns and can more effectively identify the flow patterns of gas-solid two-phase flow.

[0108] in, Fig.13 (a1), (a2), (a3), ... on the left side of the middle represent the first, second, third, ... results of the laminar signal sample after FRFT processing; Fig.13 (b1), (b2), (b3), ... on the right side of the middle represent the 1st, 2nd, 3rd, ... result diagrams of the dispersed flow signal samples after FRFT processing.

[0109] By summarizing the optimal transformation order p of all signal samples after FRFT processing of laminar flow and dispersed flow and the value range of the corresponding maximum amplitude MA, the identification rules of the two flow types can be obtained. Among them, the overlapping part of the amplitude of the laminar flow and dispersed flow signal samples after FRFT processing can be regarded as the transition state from laminar flow to dispersed flow.

[0110] Example 2 This Example 2 describes a gas-solid two-phase flow pattern identification system, which is based on the same inventive concept as the gas-solid two-phase flow pattern identification method described in the above Example 1. Specifically, the gas-solid two-phase flow pattern identification system includes: a preprocessing and signal decomposition module, which is used to collect particle flow signals of different flow patterns of gas-solid two-phase flow, and preprocess the signals to remove noise interference; decompose the processed signals into multiple time series signal samples; a fractional-order Fourier transform module, which is used to use a FRFT-based signal processing method for each time series signal sample, and convert the signal sample from the time domain to the fractional-order domain by selecting the transformation order, so as to extract the frequency domain characteristics of the signal; an optimization module, which is used to use the minnow optimization BFO algorithm to find the optimal transformation order for each signal sample in the time series after FRFT processing. and its corresponding maximum amplitude MA; and a flow pattern recognition module, which is used to preset the parameters corresponding to the signal samples extracted from the time series by different flow patterns after FRFT processing The value range of MA is used as the basis for identifying different flow patterns of gas-solid two-phase flow, thereby realizing the identification of gas-solid two-phase flow pattern.

[0111] It should be noted that, in the gas-solid two-phase flow pattern identification system, the implementation process of the functions and effects of each functional module is specifically described in the implementation process of the corresponding steps in the method in the above embodiment 1, and will not be repeated here.

[0112] Embodiment 3 Embodiment 3 of the present invention relates to a computer device, which includes a memory and one or more processors. An executable code is stored in the memory. When the processor executes the executable code, the steps of the gas-solid two-phase flow pattern identification method in the above embodiment 1 are implemented.

[0113] In this embodiment, the computer device is any device or apparatus with data processing capability, which will not be described in detail here.

[0114] Embodiment 4 This embodiment 4 describes a computer-readable storage medium having a program stored thereon. When the program is executed by a processor, the program is used to implement the steps of the gas-solid two-phase flow pattern identification method in the above-mentioned embodiment 1.

[0115] The computer-readable storage medium may be an internal storage unit of any device or apparatus with data processing capabilities, such as a hard disk or memory, or an external storage device of any device with data processing capabilities, such as a plug-in hard disk, a smart media card (SMC), an SD card, a flash card, etc., equipped on the device.

[0116] Of course, the above description is only a preferred embodiment of the present invention, and the present invention is not limited to the above embodiments. It should be noted that all equivalent substitutions and obvious deformation forms made by any technician familiar with the field under the guidance of this specification fall within the essential scope of this specification and should be protected by the present invention.

Claims

1. A method for identifying flow pattern of gas-solid two-phase flow, characterized in that: The steps include: Step 1. Collect particle flow signals of different flow patterns of gas-solid two-phase flow, and preprocess the signals to remove noise interference; decompose the processed signals into multiple time series signal samples; Step 2. For each time series signal sample, a signal processing method based on fractional Fourier transform (FRFT) is used to convert the signal sample from the time domain to the fractional order domain by selecting the transformation order, thereby extracting the frequency domain characteristics of the signal; Step 3. Use the BFO algorithm to find the optimal transformation order p and the corresponding maximum voltage amplitude MA of each signal sample in the time series after FRFT processing; Step 4. Preset the value ranges of parameters p and MA corresponding to the signal samples extracted from the time series of different flow types after BFO-FRFT processing as the judgment basis for identifying different flow types of gas-solid two-phase flow, thereby realizing gas-solid two-phase flow type identification.

2. The gas-solid two-phase flow pattern identification method according to claim 1, characterized in that: In the step 2, the gas-solid two-phase flow signal f(t) is processed using a signal processing method based on FRFT; the fractional Fourier transform X of the gas-solid two-phase flow signal f(t) at the transformation order p is p (u) is defined by the following formula: Among them, K p (t,u) represents the kernel function, which is defined as follows: Among them, α represents the rotation angle in FRFT; j represents the imaginary unit; t represents the position of the signal on the time axis; u represents the transition domain between the time domain and the frequency domain; n represents an integer; δ(tu) represents the Dirac function, which means that there is an instantaneous impulse signal at t=u and it is 0 at other positions; δ(t+n) means that there is an instantaneous impulse signal at t=-u and it is 0 at other positions; The relationship between the rotation angle α and the transformation order p is as follows: When α=0, FRFT is equivalent to the original time domain signal, that is, the gas-solid two-phase flow signal f(t); when α=π / 2, FRFT becomes the traditional Fourier transform; when α=π, it represents the inverse Fourier transform.

3. The gas-solid two-phase flow pattern identification method according to claim 2, characterized in that: In the step 3, the transformation order corresponding to the maximum amplitude after FRFT transformation is defined as the optimal transformation order; and the minnow optimization algorithm BFO is used to search for the optimal transformation order and maximum amplitude of FRFT.

4. The gas-solid two-phase flow pattern identification method according to claim 3, characterized in that: In step 3, the BFO algorithm seeks the optimal solution by minimizing the objective function, as shown in the following formula: (p * ,u * )=argmin(-|X p (u)| 2 ); Among them, p * represents the optimal transformation order, u * represents the voltage amplitude, and -|X p (u)| 2 represents the objective function of the BFO algorithm, X p (u) represents the amplitude of the signal sample extracted from the time series after FRFT transformation.

5. The gas-solid two-phase flow pattern identification method according to claim 4, characterized in that: In step 3, the process of using the minnow optimization BFO algorithm to find the optimal transformation order p and the corresponding maximum amplitude MA of each signal sample in the time series after FRFT processing is as follows: Step 3.

1. Initialize the population; The BFO algorithm is used to find the optimal transformation order p of the fractional Fourier transform FRFT; the population F consists of n solutions, each of which corresponds to a possible transformation order, expressed as: F={p1,p2,…p n }; Among them, p i represents the i-th transformation order, i=1,…n, n is the population size, indicating the number of solutions searched; p i The initial value of is randomly generated by the following formula: p i =l+(u-l)·r; Among them, l and u represent the lower and upper bounds of the transformation order, and r represents a random number between 0 and 1; Step 3.

2. Construct fitness function; The fitness function is used to evaluate the amplitude of FRFT under different transformation orders p; by calculating the amplitude corresponding to each transformation order and using it as a criterion, the optimal transformation order is found by optimizing the fitness function; The fitness function Fitness expression is as follows: Fitness=[f(p1),f(p2),…f(p n )]; Among them, f(p i ) represents the transformation order p i The corresponding amplitude, i=1,…n; Step 3.

3. Search and development; The BFO algorithm optimizes the transformation order by simulating the search and development strategies of fish schools. During the search process, the step size J is adjusted to achieve the transition from global to local search, thereby exploring the search space and finding the optimal solution. The specific formula of this process is as follows: Among them, p i t represents the position of the i-th solution in the t-th iteration; p i t+1 represents the position of the i-th solution in the t+1th iteration; J represents the step length of the fish; p + represents a random solution in the fish school search process; p * represents the optimal solution found in the current iteration; δ is a random factor used to adjust the direction and distance of the fish school moving toward the target solution; P is a predefined threshold used to select the current search direction; if r≤P, the current solution is based on the random solution p + Update; if r>P, then based on the optimal solution p * renew; Step 3.

4. Avoid local optimum; The escape behavior is introduced into the BFO algorithm, and random jumps and a global search strategy based on the center of gravity of the fish school are used near the local optimum to improve the ability to search for the optimal transformation order p. The specific implementation process is as follows: I. Random jump strategy; When r>0.5, the solution position is randomly updated within the search range [l,u]: p i t+1 =l+(ul)·δ; II. Global search strategy based on the center of gravity of the fish school; When r≤0.5, the solution position is based on the optimal solution p * Update of fish school center of gravity M: p i t+1 =J·p i t +(p * -J·M)·δ; Among them, the center of gravity of the fish school Step 3.

5. Solutions with low fitness will be eliminated, and the probability of being eliminated is d(p i ) is calculated by the following formula: Among them, f(p i ) is the objective function of the solution, that is, the transformation order p i The corresponding amplitude; Step 3.

6. Termination; When the optimal fractional order p that maximizes the FRFT amplitude is found * When , the BFO algorithm terminates; it is expressed as: p * =argmaxf(p i ); Among them, p * It represents the optimal transformation order corresponding to the maximum amplitude MA after FRFT processing.

6. The gas-solid two-phase flow pattern identification method according to claim 5, characterized in that: In step 2, in order to reduce the computational complexity, the transformation order of FRFT is normalized to the interval of [0.5, 1.5].

7. The gas-solid two-phase flow pattern identification method according to claim 5, characterized in that: In step 3, the standard deviation and relative error are used to determine the values ​​of the two parameters of the BFO algorithm, namely, the population size and the number of iterations. The definitions of the standard deviation and the relative error are shown in the following formula: Among them, x i represents the optimal transformation order p obtained after the signal samples extracted from the time series are processed by FRFT, Represents the average value of the optimal transformation order obtained after the signal is processed by FRFT σ represents the standard deviation, ε r Indicates relative error.

8. A gas-solid two-phase flow pattern identification system, characterized in that: Includes the following modules: The preprocessing and signal decomposition module is used to collect the particle flow signals of different flow patterns of gas-solid two-phase flow, and preprocess the signals to remove noise interference; decompose the processed signals into multiple time series signal samples; The fractional Fourier transform module is used to apply the FRFT-based signal processing method to each time series signal sample, and convert the signal sample from the time domain to the fractional order domain by selecting the transformation order, so as to extract the frequency domain characteristics of the signal; The optimization module is used to use the minnow optimization BFO algorithm to find the optimal transformation order p and the corresponding maximum voltage amplitude MA of each signal sample in the time series after FRFT processing; And the flow pattern recognition module is used to preset the value ranges of parameters p and MA corresponding to the signal samples extracted from the time series with different flow patterns after BFO-FRFT processing, which serves as the judgment basis for identifying different flow patterns of gas-solid two-phase flow, thereby realizing gas-solid two-phase flow pattern recognition.

9. A computer device comprising a memory and one or more processors, wherein the memory stores executable code, characterized in that: When the processor executes the executable code, it is used to implement the gas-solid two-phase flow pattern identification method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a program stored thereon, characterized in that: When the program is executed by a processor, it is used to implement the gas-solid two-phase flow pattern identification method as described in any one of claims 1 to 7.

Images