Mixing uniformity simulation analysis method and process parameter optimization method based on bionic wriggling

Abstract

The invention discloses a mixing uniformity simulation analysis method based on bionic wriggling and a process parameter optimization method, which are used for solving the problem of lack of quantitative evaluation and process optimization guidance in bidirectional wriggling mixing of a high-solid-content viscous system. In the simulation analysis method, a transient flow field model of bidirectional peristaltic motion is established by fusing axial and radial peristaltic motion, a solid-liquid two-phase mixing process of a high-solid-content viscous system is simulated, and a uniformity index is calculated based on flow field node density, so that the mixing effect is quantitatively evaluated and the mixing duration is determined. In the process optimization method, an integrated learning prediction model and a response surface optimization model are constructed based on multiple groups of process parameter-mixed duration simulation data, and an optimal process parameter combination is obtained through multi-objective optimization solution. According to the invention, accurate simulation, quantitative evaluation and process intelligent optimization of the peristaltic mixing process are realized. According to the invention, accurate simulation, quantitative evaluation and process intelligent optimization of the peristaltic mixing process are realized.

Description

Technical Field

[0001] This invention belongs to the field of biomimetic peristaltic mixing technology, specifically a simulation analysis method for mixing uniformity and a process parameter optimization method based on biomimetic peristalsis. Background Technology

[0002] High-solids-content, high-viscosity solid-liquid mixtures are widely used in production processes in industries such as military, aerospace, mining, and biotechnology, and their mixing quality directly determines the final product quality. Flexible mixing has become a core research direction in recent years for mixing high-solids-content, viscous systems. The animal digestive system is often considered a highly efficient mixing reactor, thus biomimetic peristaltic mixing flexible reactors have emerged as a new approach to solving engineering problems.

[0003] Current simulation studies on biomimetic intestinal flexible reactors have demonstrated that mimicking the peristaltic properties of the intestinal wall can enhance mass transfer and achieve efficient mixing in high-viscosity systems. However, these peristaltic simulation models are mostly designed for scenarios with no or low solid content, and there is a lack of quantitative research on the mixing uniformity within the flow field. Furthermore, high solid content and high fluid viscosity will cause differences in particle motion and fluid evolution, making existing mixing mechanisms potentially inapplicable to high-viscosity systems. Therefore, current simulations of flow fields and optimization of process parameters for peristaltic mixing still have the following technological gaps: (1) Intestinal peristalsis includes radial peristalsis and axial peristalsis. When the viscosity of the material is high, radial peristalsis plays a dominant role in mixing. Existing studies rarely take into account the coupling effect of radial and axial motion, and there is a lack of quantitative research on the uniformity of flow field mixing. (2) For peristaltic mixing processes, there is a lack of process parameter optimization methods for bidirectional peristaltic motion mixing processes, making it difficult to predict mixing time and optimize process parameters in order to meet the requirements of mixing uniformity.

[0004] Therefore, it is of great significance to develop a research method for simulation analysis and process optimization of hybrid homogeneity based on biomimetic peristalsis. Summary of the Invention

[0005] In view of this, the purpose of this invention is to provide a biomimetic peristalsis-based simulation analysis method for mixing uniformity and a process parameter optimization method. First, by integrating axial and radial peristaltic motion, a transient flow field model under bidirectional peristaltic motion is established to achieve quantitative simulation evaluation of mixing uniformity. Then, based on the simulation data, a prediction model for process parameters and mixing time is constructed, and the optimal combination of process parameters is obtained through multi-objective optimization, ultimately achieving process optimization with the goal of shortening mixing time and ensuring mixing uniformity.

[0006] To achieve the above objectives, the present invention provides the following technical solution: This invention first proposes a simulation analysis method for the homogeneity of mixing based on biomimetic peristalsis, comprising the following steps: S1: Establish a finite volume model for transient simulation of a bidirectional peristaltic hybrid flow field; set the model inlet and outlet as pressure inlet boundaries and pressure outlet boundaries respectively, and use a UDF function to control the periodic change of the inlet pressure boundary conditions to simulate axial peristaltic motion; on this basis, adopt a no-slip wall condition for the wall, divide the radial peristaltic motion parts by region partitioning, and use a UDF function to control the periodic change of the dynamic mesh velocity in the wall region to simulate radial peristaltic motion. S2: The transient simulation finite volume model of the flow field is set to use the Euler multiphase flow model to simulate a high solid content viscous solid-liquid two-phase system, and a turbulence model is set. The flow field in the computational domain must satisfy the mass conservation equation, momentum conservation equation and component transport equation. S3: Solve the transient simulation finite volume model of the flow field, extract the density data of each node at different times in the flow field, and calculate the mixing homogeneity index of the flow field at the corresponding time based on the density data. ; S4: Based on the aforementioned mixing uniformity index The time required to reach a preset uniformity standard over time is determined and used as the mixing duration. .

[0007] Furthermore, in step one, the inlet pressure boundary condition used to simulate axial creep motion follows the following formula: in: It is the inlet pressure; It refers to the magnitude of the periodic pressure changes; It is the frequency of periodic pressure changes; It is the current time; It is atmospheric pressure; This is the current time.

[0008] Furthermore, in step one, the velocity of the moving mesh on the wall used to simulate radial creep motion follows the following formula: in: yes The curve of velocity along the axial direction as a function of time; It is the periodic variation amplitude of radial peristalsis; It is the periodic variation frequency of radial peristalsis; This is the current time.

[0009] Furthermore, in step two, the mass conservation equation is expressed as: in: Density; For time; , and Spatial coordinates; , and For speed in corresponding , and Velocity components in the axial direction; The momentum conservation equation is expressed as: in: It is the velocity vector of the flow; It is the speed of the moving mesh; It is a volume force; It is the fluid density; It's pressure. It's viscosity; ; The component transport equation is expressed as: in: It is the first Mass fraction of each component; It is the velocity vector of the flow; It is the speed of the moving mesh; express Mass diffusion coefficient of the component.

[0010] Furthermore, in step three, the mixing uniformity index The calculation method is as follows: First, calculate the standard deviation of the nodal density in the flow field. : Then calculate the mixing uniformity index. : in: For a certain moment Node density; This is the theoretical density of the solid-liquid mixture; The number of nodes in the flow field; Standard deviation; The standard error value represents the mixing uniformity index of the flow field.

[0011] Furthermore, in step four, the mixing time is determined. The method is as follows: The uniformity index The starting time point corresponding to a series of consecutive time steps that first reach and remain below a preset threshold is determined as the mixed duration. The preset threshold is related to the ideal uniformity index 0.

[0012] This invention also proposes a process parameter optimization method based on biomimetic peristalsis, comprising the following steps: Step 1: Obtain mixed data under multiple combinations of different process parameters and construct a mixed dataset. Each set of mixed data includes the process parameter combination and its corresponding mixing time. The mixing duration This was obtained through the mixing homogeneity simulation analysis method described above; Step 2: Construct an ensemble learning model and train it to establish a system for calculating the time required for process parameter combinations and mixed durations. Predictive relationship model; Step 3: Build to minimize blending time A process parameter optimization model is developed with the objective of obtaining the optimal combination of process parameters by performing a search optimization within the constraints of the process parameters based on a response surface-guided search strategy.

[0013] Furthermore, step two includes: standardizing the process parameters in the mixed data and constructing a response surface feature set including linear terms, interaction terms, and quadratic terms as input features of the ensemble learning model.

[0014] Furthermore, in step three, the process parameter optimization model is a multi-objective optimization model, and the optimization objective includes minimizing the mixing time. To ensure process stability, optimization methods include response surface-guided search strategies, and output Pareto optimal solution sets.

[0015] Furthermore, the combination of process parameters includes: the pressure variation range of axial creep. and frequency Radial creep amplitude of wall movement and frequency And the number of radially creeping parts.

[0016] The beneficial effects of this invention are as follows: This invention presents a biomimetic peristaltic mixing homogeneity simulation analysis method. By integrating axial and radial peristaltic motions, a finite volume model for transient simulation of the flow field under bidirectional peristaltic conditions is constructed. Employing an Eulerian multiphase flow model, it achieves a realistic simulation of the mixing process of high-solids-content viscous solid-liquid systems under biomimetic peristaltic conditions. The invention utilizes a defined mixing homogeneity index... This invention enables a quantitative assessment of mixing effects, filling the gap in existing research regarding the quantitative characterization of mixing effects in viscous systems with high solids content. The method provides an effective numerical analysis tool for accurately predicting mixing time and gaining a deeper understanding of the bidirectional peristaltic mixing mechanism, and lays a reliable simulation foundation for subsequent process parameter optimization.

[0017] This invention presents a biomimetic peristaltic process parameter optimization method. By employing an ensemble learning model to uncover the complex mapping relationship between process parameters and mixing time, it achieves high-precision mixing time prediction based on simulation data. By combining response surface methodology and multi-objective optimization algorithms, it efficiently seeks optimization in a multi-dimensional parameter space, simultaneously optimizing mixing time and process stability, and outputting a Pareto optimal solution set. This method transforms simulation data into actionable process decisions, realizing a closed loop from mechanistic analysis to intelligent optimization, significantly improving the design efficiency and reliability of peristaltic mixing processes. Attached Figure Description

[0018] To make the objectives, technical solutions, and beneficial effects of this invention clearer, the following figures are provided for illustration: Figure 1 This is a flowchart of the method for transient simulation analysis and process parameter optimization of bidirectional peristaltic mixed flow field according to the present invention; Figure 2 The flowchart shows the simulation analysis method for mixing uniformity based on biomimetic peristalsis according to the present invention. Figure 3 This is a schematic diagram of the boundary conditions for a bidirectional creep simulation model; Figure 4 This is a diagram showing the pressure variation in the flow field under axial creep at the inlet. Figure 5 This is a schematic diagram of radial creep on the wall. Figure 6 A flowchart for multi-objective mixing duration prediction and process parameter optimization. Detailed Implementation

[0019] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can better understand and implement the present invention. However, the embodiments described are not intended to limit the present invention.

[0020] This embodiment, based on a biomimetic peristalsis-based process parameter optimization method, aims to address the shortcomings of existing technologies in simulating bidirectional peristaltic mixing of high-solids-content viscous systems, as well as the lack of quantitative evaluation and process optimization guidance. For example... Figure 1 As shown, the process parameter optimization method based on biomimetic peristalsis in this embodiment includes the following steps.

[0021] Step 1: Obtain mixed data under multiple combinations of different process parameters and construct a mixed dataset. Each set of mixed data includes the process parameter combination and its corresponding mixing time. The mixing duration The results were obtained through a biomimetic peristalsis-based simulation analysis method for the uniformity of mixing.

[0022] Specifically, multiple simulations are performed by changing multiple combinations of key process parameters. Each combination of process parameters... At least including: the magnitude of pressure changes used to simulate axial creep. and frequency and the amplitude of wall motion used to simulate radial creep. ,frequency And the number of radial creep regions. For each combination of process parameters A biomimetic peristalsis-based simulation analysis method for mixing uniformity was used to obtain the corresponding mixing time. Collect a sufficient number of sample pairs. This constitutes a hybrid dataset. The hybrid dataset should cover the possible range of parameter values ​​to reflect its complex relationship with the duration of the hybrid dataset.

[0023] like Figure 2 As shown in the figure, the biomimetic peristalsis-based simulation analysis method for mixing uniformity includes the following steps.

[0024] S1: Establish a finite volume model for transient simulation of a bidirectional peristaltic hybrid flow field. The model inlet and outlet are set as pressure inlet and pressure outlet boundaries, respectively. A periodic variation of the inlet pressure boundary conditions is controlled using a UDF function to simulate axial peristaltic motion. Based on this, a no-slip wall condition is adopted for the wall surface. Radial peristaltic motion is segmented by region partitioning. A periodic variation of the dynamic mesh velocity in the wall region is controlled by applying a UDF function to the radial peristaltic region to simulate radial peristaltic motion.

[0025] Specifically, such as Figure 3 As shown, a geometric model of the flow field computational domain is established based on the structure of an actual peristaltic mixing reactor. The flow field computational domain is typically a cylinder to simulate the mixing pipe. At the inlet end face, it needs to be divided into independent solid phase inlet region and liquid phase inlet region using the surface segmentation function to simulate the separate feeding of solid and liquid phases.

[0026] After establishing the geometric model, the computational domain is meshed using an unstructured mesh, specifically a tetrahedral mesh. To improve computational accuracy, local mesh refinement is applied in regions with significant flow field gradient changes, such as the solid inlet, liquid inlet, outlet, and near the pipe wall. After the mesh quality is checked and approved, boundary conditions are defined.

[0027] like Figure 4As shown, based on the actual peristaltic mixing operating parameters, the peristaltic tube inlet is given as the pressure inlet boundary, and the total temperature is given as a constant operating temperature. Axial peristaltic motion is simulated by controlling the periodic changes in total pressure through a user-defined function (UDF). The outlet is the pressure outlet boundary, with pressure equal to atmospheric pressure. And backflow is allowed. In this embodiment, the inlet pressure boundary condition used to simulate axial creep motion follows the following formula: in: It is the inlet pressure; It refers to the magnitude of the periodic pressure changes; It is the frequency of periodic pressure changes; It is the current time; It is atmospheric pressure; This is the current time.

[0028] like Figure 5 As shown, the pipe wall boundary type is set to wall, a no-slip wall condition is applied, a standard wall function is used to process the boundary, and the radial creeping motion is segmented by region partitioning. For this part of the wall, a dynamic mesh function is enabled, and user-defined functions (UDFs) are applied to these parts to control the dynamic mesh movement and simulate the compression and expansion of the wall. Specifically, in this embodiment, the dynamic mesh velocity of the wall used to simulate radial creeping motion follows the following formula: in: yes The curve of velocity along the axial direction as a function of time; It is the periodic variation amplitude of radial peristalsis; It is the periodic variation frequency of radial peristalsis; This is the current time.

[0029] Specifically, the method for constructing a user-defined function (UDF) is as follows: based on the characteristics of axial creep, a custom function is constructed that shows the periodic change of inlet pressure over time; based on the characteristics of radial creep, a custom function is constructed that shows the periodic change of wall region movement velocity over time; and the custom functions that show the periodic change of inlet pressure over time and wall region movement velocity over time are integrated into a single UDF file.

[0030] S2: The transient simulation finite volume model of the flow field is set to use the Euler multiphase flow model to simulate a high solid content viscous solid-liquid two-phase system, and a turbulence model is set. The flow field in the computational domain must satisfy the mass conservation equation, momentum conservation equation and component transport equation.

[0031] Since this is a high-solids-content viscous system, the Euler multiphase flow model is selected in the solver settings to simulate the interaction between the solid and liquid phases. In this embodiment, the first phase is set as a continuous liquid phase and the second phase as a discrete solid phase, and their material properties (such as density and viscosity) are defined respectively. Considering that the flow field may contain both laminar and turbulent regions, the k-omega model, which is suitable for simulating near-wall flow, is selected as the turbulence model. The component transport model is activated to track the evolution of the mixture concentration.

[0032] The governing equations include the mass conservation equation (continuity equation), the momentum conservation equation, and the component transport equation. The momentum conservation equation needs to consider the influence of the moving mesh. The pressure-velocity coupled solution algorithm uses the Phase Coupled SIMPLE model, suitable for multiphase flow. The pressure term is discretized using a second-order upwind model, while the other equations can use a first-order upwind model to enhance computational stability. Appropriate time steps and maximum iteration steps are set for transient calculations.

[0033] Specifically, in this embodiment, the mass conservation equation is expressed as: Introducing the velocity divergence sign and assuming the fluid is incompressible, the fluid density remains constant. Therefore, equation (1) can be written as: in: Density; For time; , and Spatial coordinates; , and For speed in corresponding , and The velocity component in the axial direction.

[0034] The momentum conservation equation is expressed as: in: It is the velocity vector of the flow; It is the speed of the moving mesh; It is a volume force; It is the fluid density; It's pressure. It's viscosity; .

[0035] The component transport equation is expressed as: in: It is the first Mass fraction of each component; It is the velocity vector of the flow; It is the speed of the moving mesh; Indicates components Net yield in a chemical reaction process; These are source terms that cannot be included in convection-diffusion and reaction; express Component diffusion flux. According to Fick's law, diffusion flux is expressed as: in: express Mass diffusion coefficient of the component; express The thermal diffusivity of the components. From the component transport equation, we know that the diffusion flux is related to temperature and concentration. In this model, the temperature remains constant and equal everywhere; therefore, the temperature term can be removed. Furthermore, the density in incompressible fluids does not change with time. Based on the above analysis, the component transport equation can be rewritten as: in: It is the first Mass fraction of each component; It is the velocity vector of the flow; It is the speed of the moving mesh; express Mass diffusion coefficient of the component.

[0036] S3: Solve the transient simulation finite volume model of the flow field, extract the density data of each node at different times in the flow field, and calculate the mixing homogeneity index of the flow field at the corresponding time based on the density data. .

[0037] During transient calculations, the flow field state is monitored. At regular time steps, the instantaneous density values ​​of each grid node (N in number) within the flow field are extracted. Calculate its relationship with the average density. Standard deviation Then divide by the square root of the number of nodes to obtain the standard error value. This is the mixing uniformity index. In this embodiment, the mixing uniformity index... The calculation method is as follows: First, calculate the standard deviation of the nodal density in the flow field. : Then calculate the mixing uniformity index. : in: For a certain moment Node density; This is the theoretical density of the solid-liquid mixture; The number of nodes in the flow field; Standard deviation; The standard error value represents the mixing uniformity index of the flow field.

[0038] Specifically, the mixing uniformity index The smaller the value, the more uniform the distribution of the solid and liquid phases in the flow field. Its theoretical lower limit is 0, which represents complete uniformity.

[0039] S4: Based on the aforementioned mixing uniformity index The time required to reach a preset uniformity standard over time is determined and used as the mixing duration. .

[0040] In this embodiment, the mixing time is determined. The method is as follows: The uniformity index The starting time point corresponding to a series of consecutive time steps that first reach and remain below a preset threshold is determined as the mixed duration. The preset threshold is related to the ideal uniformity index 0.

[0041] Specifically, plot the mixing uniformity index. Over time The changing curve. A threshold representing sufficient mixing is set. When... When the value decreases from the initial state and reaches the threshold for the first time, and remains below the threshold for several consecutive time steps, this moment is considered the moment when the target mixing uniformity is achieved. The time elapsed from the start of the calculation (t=0) to this moment is defined as the mixing time under this condition. .

[0042] Specifically, this embodiment uses the 95% rule to determine the mixing time, that is, when the mixing uniformity index... From the start of the calculation to the material mixing uniformity index The time required to reach ±95% of the final stable uniformity index is considered, while the ideal uniformity index is 0. Therefore, in this embodiment, five consecutive node values ​​that satisfy a uniformity index of 0.05 are determined as the mixing time.

[0043] Step 2: Construct an ensemble learning model and train it to establish a system for handling process parameter combinations and mixed durations. The predictive relationship model is described. In this embodiment, the process parameters in the mixed dataset are standardized to eliminate the influence of different parameter dimensions. To better fit the nonlinear relationship, features are constructed from the standardized parameters to generate a new response surface feature set as the input features of the ensemble learning model. The response surface feature set includes linear terms from the mixed dataset, interaction terms between process parameters, and quadratic terms for each parameter.

[0044] Specifically, the ensemble learning model is trained to establish a system from process parameter features to mixing time. A regression prediction model is developed. The dataset is proportionally divided into training and test sets. The model is trained using the training set, and its hyperparameters are tuned using cross-validation. The model performance is evaluated on the test set using the coefficient of determination. and mean square error As an evaluation metric, it is essential to ensure that the model has sufficient predictive accuracy and generalization ability.

[0045] Step 3: Build to minimize blending time A process parameter optimization model is proposed with the objective of minimizing the mixing time. Based on a response surface methodology-guided search strategy, the model performs optimization within the constraints of the process parameters to obtain the optimal combination of process parameters. The process parameter optimization model is a multi-objective optimization model, with the optimization objective also including minimizing the mixing time. To ensure process stability, the optimization method includes a response surface-guided search strategy, outputting a Pareto optimal solution set. In this embodiment, the process parameter combination includes: the pressure variation amplitude of axial creep. and frequency Radial creep amplitude of wall movement and frequency And the number of radially creeping parts.

[0046] like Figure 6 As shown, in this embodiment, the optimization objective is to find a set of process parameters. Under the premise of meeting the actual equipment and process constraints, achieve the following: 1) Minimize the mixing time. 2) Ensuring process stability. A well-trained ensemble learning model is used as a high-precision surrogate model for the objective function. Combining the constructed response surface feature idea, a multi-objective optimization algorithm is employed to search within the parameter design space under the guidance of the surrogate model. The multi-objective optimization algorithm explores a large number of parameter combinations and quickly predicts their mixing duration through the surrogate model, ultimately converging to a set of Pareto optimal solutions. The Pareto optimal solution cannot simultaneously improve both the mixing duration and process stability objectives.

[0047] The Pareto optimal solution set obtained from the analysis can be used to select a final solution based on actual production needs. For example, if the goal is to minimize the mixing time, then the solution on the Pareto front is selected. The solution with the minimum value is selected; if the robustness of the process is of greater concern, a solution with more moderate parameter values ​​and gradual changes may be chosen. The final optimal combination of process parameters can be directly used to guide the operation settings of actual peristaltic mixing equipment.

[0048] Specifically, in this embodiment, the specific operations of the process parameter optimization method based on biomimetic peristalsis include the following:

[0049] I. Constructing User-Defined Functions (UDFs): (1) Based on the characteristics of axial creep, a custom function is constructed to make the inlet pressure change periodically with time; (2) Based on the characteristics of radial creep, a custom function is constructed to make the wall region movement velocity change periodically with time; (3) Integrate the custom functions of (1) and (2) into a single UDF file.

[0050] 2. Establish a bidirectional peristaltic mixing geometric model of the pipe shape. Divide the inlet surface into two parts: solid phase inlet and liquid phase inlet. Based on the structural characteristics of the model, select tetrahedral mesh type to divide the mesh. Local refinement is performed in the solid and liquid phase inlet, outlet and nearby flow field regions, as well as in the area near the peristaltic wall. Output a case file.

[0051] 3. Open Fluent software and import the case file into Fluent software.

[0052] 4. Check the grid to verify whether the grid divided in step S1 is correct;

[0053] 5. Import custom function UDF to simulate radial and axial creep motion.

[0054] VI. Select the correct model to simulate the bidirectional creeping mixing of high solid viscosity systems: Use the Eulerian multiphase flow model in the multiphase flow model and the standard k-omega model in the turbulence model.

[0055] 7. Set material properties: Set the material properties for both solid and liquid phases.

[0056] 8. Phase selection to realistically simulate the mixing of high solid viscosity systems: Set the first phase as a liquid material and the second phase as a solid material in the phase.

[0057] 9. Set boundary conditions and simulate the bidirectional peristaltic mixing process under these preset conditions: Based on the actual peristaltic mixing operating parameters, the peristaltic tube inlet is given as the pressure inlet boundary, the total temperature is given, and a user-defined function is called to control the periodic change of the total pressure to simulate axial peristaltic motion. This can be written as: in: It is the inlet pressure; It refers to the magnitude of the periodic pressure changes; It is the frequency of periodic pressure changes; It is the current time; It is atmospheric pressure; This is the current time.

[0058] The outlet is a pressure outlet boundary, with pressure equal to atmospheric pressure, and backflow is allowed. A no-slip wall condition is applied to the wall surface, and a standard wall function is used to handle the boundary. Radial creep motion is segmented through region partitioning, and user-defined functions are applied to these segments to control the dynamic mesh motion, simulating the compression and relaxation of the wall surface. This can be written as: in: yes The curve of velocity along the axial direction as a function of time; It is the periodic variation amplitude of radial peristalsis; It is the periodic variation frequency of radial peristalsis; This is the current time.

[0059] 10. Selection of solution method and determination of solution accuracy: The pressure-velocity coupling adopts the Phase Coupled SIMPLE mode, the pressure correction equation adopts the second-order upwind mode, and the difference scheme of other equations adopts the first-order upwind mode.

[0060] 11. Initialize the boundary and set the iteration parameters, then perform the calculation.

[0061] 12. Quantitative Simulation Results: By extracting the density of each node in the flow field at different times... Calculate its relationship with the average density Standard deviation Then divide by the square root of the number of nodes to obtain the standard error value. That is, the Segregation Index (SI): in: For a certain moment Node density; This is the theoretical density of the solid-liquid mixture; The number of nodes in the flow field; Standard deviation; The standard error value represents the mixing uniformity index of the flow field.

[0062] This embodiment further uses the 95% rule to determine the mixing time based on the uniformity index of the simulation, that is, the time taken from the start of the calculation of the uniformity index to the material uniformity index reaching ±95% of the final stable uniformity index. Since the ideal uniformity index is 0, this invention determines the mixing time as five consecutive node values ​​that satisfy the uniformity index of 0.05.

[0063] Thirteen, input the process parameters under different operating conditions and the corresponding simulation-derived mixed duration quantization data into the process parameter optimization model that combines ensemble learning and response surface model.

[0064] Fourteen, standardize the original process data to eliminate the influence of dimensions, and construct a set of response surface features, including linear terms, interaction terms, and quadratic terms.

[0065] 15. Training and performance evaluation of ensemble learning models, where the performance evaluation of the models relies on the R2 and MSE metrics.

[0066] 16. Construct the quadratic response surface model matrix and fit the response surface model coefficients.

[0067] 17. Define the optimization objective as minimizing the mixing time and ensuring process stability, and set parameter boundary constraints.

[0068] 18. Response surface-guided optimization search: By identifying promising regions, response surface information is integrated into a multi-starting point optimization framework, and multi-objective optimization based on ensemble learning is performed.

[0069] 19. Based on Pareto front analysis, extract the Pareto optimal solution set from the optimization results, and output the optimal combination of process parameters based on engineering requirements.

[0070] The above-described embodiments are merely preferred embodiments provided to fully illustrate the present invention, and the scope of protection of the present invention is not limited thereto. Equivalent substitutions or modifications made by those skilled in the art based on the present invention are all within the scope of protection of the present invention. The scope of protection of the present invention is defined by the claims.

Claims

1. A simulation analysis method for the homogeneity of mixtures based on biomimetic peristalsis, characterized in that: Includes the following steps: S1: Establish a finite volume model for transient simulation of the flow field under bidirectional creeping motion; set the model inlet and outlet as pressure inlet boundaries and pressure outlet boundaries respectively, and use UDF functions to control the periodic changes of the inlet pressure boundary conditions to simulate axial creeping motion; on this basis, adopt no-slip wall conditions for the wall surface, divide the radial creeping motion parts by region partitioning, and use UDF functions to control the periodic changes of the dynamic mesh velocity in the wall region to simulate radial creeping motion. S2: The transient simulation finite volume model of the flow field is set to use the Euler multiphase flow model to simulate a high solid content viscous solid-liquid two-phase system, and a turbulence model is set. The flow field in the computational domain must satisfy the mass conservation equation, momentum conservation equation and component transport equation. S3: Solve the transient simulation finite volume model of the flow field, extract the density data of each node at different times in the flow field, and calculate the mixing homogeneity index of the flow field at the corresponding time based on the density data. ; S4: Based on the aforementioned mixing uniformity index The time required to reach a preset uniformity standard over time is determined and used as the mixing duration. .

2. The biomimetic peristalsis-based simulation analysis method for mixing uniformity according to claim 1, characterized in that: In step one, the inlet pressure boundary condition used to simulate axial creep motion follows the following formula: in: It is the inlet pressure; It refers to the magnitude of the periodic pressure changes; It is the frequency of periodic pressure changes; It is the current time; It is atmospheric pressure; This is the current time.

3. The method for simulation analysis of mixing uniformity based on biomimetic peristalsis according to claim 1, characterized in that: In step one, the velocity of the moving mesh on the wall used to simulate radial creep motion follows the following formula: in: yes The curve of velocity along the axial direction as a function of time; It is the periodic variation amplitude of radial peristalsis; It is the periodic variation frequency of radial peristalsis; This is the current time.

4. The method for simulation analysis of mixing uniformity based on biomimetic peristalsis according to claim 1, characterized in that: In step two, the mass conservation equation is expressed as: in: Density; For time; , and Spatial coordinates; , and For speed in corresponding , and Velocity components in the axial direction; The momentum conservation equation is expressed as: in: It is the velocity vector of the flow; It is the speed of the moving mesh; It is a volume force; It is the fluid density; It's pressure. It's viscosity; It is a gradient operator; It is the Laplace operator; The component transport equation is expressed as: in: It is the first Mass fraction of each component; It is the velocity vector of the flow; It is the speed of the moving mesh; express Mass diffusion coefficient of the component.

5. The method for simulation analysis of mixing uniformity based on biomimetic peristalsis according to claim 1, characterized in that: In step three, the mixing uniformity index The calculation method is as follows: First, calculate the standard deviation of the nodal density in the flow field. : Then calculate the mixing uniformity index. : in: For a certain moment Node density; This is the theoretical density of the solid-liquid mixture; The number of nodes in the flow field; Standard deviation; The standard error value represents the mixing uniformity index of the flow field.

6. The method for simulation analysis of mixing uniformity based on biomimetic peristalsis according to claim 1, characterized in that: In step four, the mixing time is determined. The method is as follows: The uniformity index The starting time point corresponding to a series of consecutive time steps that first reach and remain below a preset threshold is determined as the mixed duration. The preset threshold is related to the ideal uniformity index 0.

7. A method for optimizing process parameters based on biomimetic peristalsis, characterized in that: Includes the following steps: Step 1: Obtain mixed data under multiple combinations of different process parameters and construct a mixed dataset. Each set of mixed data includes the process parameter combination and its corresponding mixing time. The mixing duration Obtained by the mixing homogeneity simulation analysis method as described in any one of claims 1-6; Step 2: Construct an ensemble learning model and train it to establish a system for calculating the time required for process parameter combinations and mixed durations. Predictive relationship model; Step 3: Build to minimize blending time A process parameter optimization model is developed with the objective of obtaining the optimal combination of process parameters by performing a search optimization within the constraints of the process parameters based on a response surface-guided search strategy.

8. The process parameter optimization method based on biomimetic peristalsis according to claim 7, characterized in that: Step two includes: standardizing the process parameters in the mixed data and constructing a response surface feature set including linear terms, interaction terms, and quadratic terms as input features of the ensemble learning model.

9. The process parameter optimization method based on biomimetic peristalsis according to claim 7, characterized in that: In step three, the process parameter optimization model is a multi-objective optimization model, and the optimization objective includes minimizing the mixing time. To ensure process stability, optimization methods include response surface-guided search strategies, and output Pareto optimal solution sets.

10. The process parameter optimization method based on biomimetic peristalsis according to claim 7, characterized in that: The combination of process parameters includes: the pressure variation amplitude of axial creep. and frequency Radial creep amplitude of wall movement and frequency And the number of radially creeping parts.

Images