A method and system for optimizing the structure of an H-shaped photoacoustic cell based on dynamic modeling simulation

Abstract

The application belongs to the technical field of dynamic simulation modeling of H-type photoacoustic cell, and provides an H-type photoacoustic cell structure optimization method and system based on dynamic modeling simulation, which obtains H-type photoacoustic cell parameters by using a dynamic modeling simulation method of the H-type photoacoustic cell, calculates characteristic frequencies, quality factors and sound pressures, and constructs a multi-parameter small sample data set; takes the H-type photoacoustic cell parameters as size variables, takes the characteristic frequencies, quality factors and sound pressures as target performance labels, trains a random forest agent model based on the multi-parameter small sample data set, and obtains the trained random forest agent model; combines randomly generated photoacoustic cell sizes and the trained random forest agent model to perform performance prediction, obtains performance prediction results corresponding to each group of photoacoustic cell sizes, performs screening through a CDP constraint dominance criterion, and constructs a Pareto optimal solution set; makes a decision on the screened performance prediction results, and obtains the optimal structure size of the H-type photoacoustic cell. The optimization efficiency and accuracy are improved.

Description

Technical Field

[0001] This invention belongs to the field of dynamic simulation modeling technology for H-type photoacoustic cell, and particularly relates to a method and system for optimizing the structure of an H-type photoacoustic cell based on dynamic modeling and simulation. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] The H-type photoacoustic cell is a typical structure of resonant photoacoustic cells. Due to its simple structure, support for flowing gas measurements, and ease of miniaturization and integration, it has become the first choice for portable photoacoustic spectrometers. Its performance is evaluated by the sound pressure signal and the quality factor: the sound pressure determines the minimum detection limit, and the quality factor determines the resonance enhancement ratio. The core structure consists of a resonant cavity and a buffer cavity, and the choice of their geometry is crucial—reducing the cavity size can increase the laser energy density to enhance the sound pressure, but it will limit the optical path and reduce the quality factor.

[0004] The optimization and selection of structures often employs exhaustive simulation. Simulating photoacoustic cells is time-consuming, and due to constraints on energy density and optical path length within the cavity size, finding a photoacoustic cell that meets the expected characteristic frequency requirements demands significant time and computational resources. Traditional response surface methodology (RSM) predictions based on large amounts of structural simulation data result in a mathematical model that defaults to a second-order polynomial. Even using higher-order polynomials often fails to accurately represent the relationship between sound pressure level, quality factor, characteristic frequencies, and structural parameters. Furthermore, excessively high orders can cause abrupt changes in some polynomials during subsequent optimization processes. Since RSM relies heavily on statistics for mathematical model construction, it is inherently rigid and cannot incorporate physical feature enhancements. Traditional optimization algorithms select optimization points based on image reading, a method that is inefficient, inaccurate, and lacks rigor. Summary of the Invention

[0005] To address at least one of the technical problems mentioned above, this invention provides a method and system for optimizing the structure of an H-type photoacoustic cell based on dynamic modeling and simulation. This method not only reduces the workload of simulation but also improves optimization efficiency and accuracy, significantly enhancing the scientific validity and feasibility of optimizing the H-type photoacoustic cell.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: The first aspect of this invention provides a method for optimizing an H-shaped photoacoustic cell structure based on dynamic modeling and simulation, comprising the following steps: The parameters of the H-type photoacoustic cell were obtained by using the dynamic modeling and simulation method of the H-type photoacoustic cell, and the characteristic frequency, quality factor and sound pressure were calculated in batches to construct a multi-parameter small sample dataset. Using the parameters of the H-type photoacoustic cell as size variables and the characteristic frequency, quality factor, and sound pressure level as target performance labels, a random forest proxy model is trained based on a multi-parameter small sample dataset to obtain the trained random forest proxy model. Performance prediction is performed by combining randomly generated photoacoustic pool sizes and trained random forest surrogate models to obtain performance prediction results for each set of photoacoustic pool sizes. The Pareto optimal solution set is then constructed by filtering the results using the CDP constraint dominance criterion. By combining the Pareto optimal solution set, a decision is made based on the filtered performance prediction results to obtain the optimal structural size of the H-type photoacoustic cell.

[0007] Furthermore, the dynamic modeling and simulation method for the H-shaped photoacoustic cell includes: Use MATLAB to connect to the COMSOL dialog box to construct a dynamic two-dimensional axisymmetric geometric model; Set the initial geometric parameters and environmental parameters: temperature and air pressure; The geometric model specifies the shape and location of the resonant cavity buffer cavity, omitting the external sound opening of the H-type photoacoustic cell and the air inlet / outlet structure. Thermodynamic system modules are used to generate mixed gas materials, which are then filled into the resonant cavity and buffer cavity. The physical field is set to pressure acoustics and the relevant basic conditions are set. Using parametric scanning, the pressure acoustic frequency domain and characteristic frequency are solved, and the sound pressure, characteristic frequency and quality factor are calculated. Finally, a multi-parameter dataset containing resonant cavity length, resonant cavity radius, buffer cavity length, buffer cavity radius, sound pressure, quality factor and characteristic frequency is constructed.

[0008] Furthermore, when constructing the dynamic two-dimensional axisymmetric geometric model, the heat source is set to exhibit a Gaussian distribution, as follows: , In the formula, Indicates the heat of the heat source. It refers to gas concentration. is the absorption cross section of the target gas, and N is the molecular number density under standard atmospheric conditions. It is the modulation coefficient corresponding to the laser modulation depth; It refers to the laser power. The waist radius is Here are the radial coordinates of the photoacoustic cell in cylindrical coordinates.

[0009] Furthermore, during the training of the random forest surrogate model, the input also includes the similarity area ratio and similarity volume ratio calculated based on the parameters of the H-type photoacoustic cell, which include the resonant cavity length, resonant cavity radius, buffer cavity length, and buffer cavity radius.

[0010] Furthermore, the formulas for calculating similar area ratios and similar volume ratios are as follows: , in, For similar area ratios, For similar volume ratios, The length of the resonant cavity. The length of the buffer chamber. Where is the radius of the resonant cavity. Where is the radius of the buffer cavity.

[0011] Furthermore, the size of the photoacoustic cell is randomly generated using the SPEA2 algorithm.

[0012] Furthermore, the setting of the total violation degree through the CDP constraint dominance criterion includes three constraints: Frequency constraint: The characteristic frequency deviation does not exceed a preset threshold; Length ratio constraint: The length ratio of the resonant cavity to the buffer cavity is within a preset range; Radius ratio constraint: The radius ratio of the resonant cavity to the buffer cavity is within a preset range; The violation score is 0 when each constraint is satisfied and 1 when it is not satisfied. The total violation score is the sum of the three constraints. The performance prediction value is selected based on the preset range of the total violation score.

[0013] A second aspect of the present invention provides a system for optimizing an H-shaped photoacoustic cell structure based on dynamic modeling and simulation, comprising: The dynamic modeling and simulation module is used to obtain the parameters of the H-type photoacoustic cell using the dynamic modeling and simulation method of the H-type photoacoustic cell, calculate the characteristic frequency, quality factor and sound pressure in batches, and construct a multi-parameter small sample dataset. The model training module is used to train the random forest proxy model based on a multi-parameter small sample dataset, using the parameters of the H-type photoacoustic pool as size variables and the feature frequency, quality factor and sound pressure as target performance labels. The model optimization module is used to combine the randomly generated photoacoustic pool size and the trained random forest proxy model to perform performance prediction, obtain the performance prediction results corresponding to each set of photoacoustic pool sizes, and construct the Pareto optimal solution set by filtering through the CDP constraint dominance criterion. The results output module is used to combine the Pareto optimal solution set to make a decision on the filtered performance prediction results and obtain the optimal structural size of the H-type photoacoustic cell.

[0014] A third aspect of the present invention provides a computer-readable storage medium.

[0015] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the above-described method for optimizing an H-type photoacoustic cell structure based on dynamic modeling and simulation.

[0016] A fourth aspect of the present invention provides a computer device.

[0017] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps in the above-described method for optimizing an H-type photoacoustic cell structure based on dynamic modeling and simulation.

[0018] Compared with the prior art, the beneficial effects of the present invention are: This invention proposes a method of using random forests to construct a large number of decision trees for training and prediction in optimizing the internal cavity structure of the H-type photoacoustic cell. It uses small sample data combined with traditional machine learning methods to summarize physical laws and control the prediction boundary. Then, it adopts the improved evolutionary algorithm SPEA2 with CDP constraint dominance criterion to reasonably limit the relationship between the structural dimensions of the photoacoustic cell, thus solving the problem that finding a photoacoustic cell that meets the expected characteristic frequency requirements requires a lot of time and computing resources.

[0019] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0020] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0021] Figure 1 This is a flowchart of an H-type photoacoustic cell structure optimization method based on dynamic modeling and simulation provided by an embodiment of the present invention; Figure 2 This is a flowchart of the dynamic modeling process of the H-type photoacoustic cell provided in an embodiment of the present invention; Figure 3 This is the frequency domain simulation curve of the sound pressure signal and frequency of CH4 in the dynamic modeling simulation of the H-type photoacoustic cell provided in this embodiment of the invention; Figure 4 This is a model network distribution diagram established based on an example of dynamic modeling, simulation, and structural optimization scheme for an H-type photoacoustic cell provided in an embodiment of the present invention. Figure 5 This is a flowchart of the random forest algorithm provided in the embodiments of the present invention; wherein, (a) is the processing step of the random forest algorithm, and (b) is the internal details of the random forest algorithm; Figure 6This is a flowchart of the improved SPEA2 algorithm provided in an embodiment of the present invention; Figure 7 This is the CDP algorithm flow provided in the embodiments of the present invention; Figure 8 The Pareto optimal solution front distribution of the SPEA2 algorithm results provided in this embodiment of the invention. Detailed Implementation

[0022] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0023] It should be noted that the following detailed description is illustrative and intended to provide further explanation of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0024] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0025] Example 1 like Figure 1 As shown in the figure, this embodiment provides a method for optimizing the H-shaped photoacoustic cell structure based on dynamic modeling and simulation, including the following steps: Step 1: Establish a two-dimensional axisymmetric finite element model of the H-type photoacoustic cell, randomly sample the parameters of the H-type photoacoustic cell, calculate the characteristic frequency, quality factor and sound pressure in batches, and construct a multi-parameter small sample dataset; Current finite element simulations often employ COMSOL's thermoviscous acoustic interface for optical-thermal-acoustic coupling calculations. However, while this common approach binds heat and sound together, its computational complexity is excessive, and boundary conditions become unclear due to the simulated heat flow, leading to simultaneous sound pressure vibrations in both the buffer cavity and the resonant cavity. Most current simulations ignore the spatial distribution characteristics of the input laser beam, which exhibits Gaussian beam patterns. Network settings often use default refinement or standard networks, treating the resonant cavity and buffer cavity equally, which wastes computational resources and results in some networks exceeding the resonant cavity radius, causing significant simulation errors. Directly using COMSOL for simulation requires cumbersome parameter and structural modifications, and large-scale parametric scans can consume excessive memory, potentially causing software crashes.

[0026] In this embodiment, when using COMSOL and MATLAB Livelink programming for joint simulation, 2D axisymmetric modeling is set up in the finite element simulation by combining the centrally symmetric structural characteristics of the H-type photoacoustic cell, setting structural parameters, and establishing a global basic cavity model.

[0027] like Figure 2 As shown, this embodiment takes the simulation modeling method and structural optimization of the H-type photoacoustic cell for CH4 as an example, and specifically includes the following steps: Step 201: Establish a two-dimensional axisymmetric geometric model; In this embodiment, when establishing a two-dimensional axisymmetric geometric model, the specific steps include setting the resonant cavity in the center, placing two buffer cavities on both sides, constructing the size and graphic coordinate relationship, and setting a probe at a position slightly above the center to simulate a microphone; selecting a pressure acoustic interface at the physical interface position, selecting a thermoviscous acoustic boundary layer impedance at the resonant cavity position, and setting a heat source, which can achieve the same effect as the thermoviscous acoustic interface calculation method, and minimize the impact of vibration caused by heat diffusion from the buffer cavity.

[0028] In this embodiment, the accuracy of the simulation structure is closely related to the network settings. A common method is to use the default refined network partitioning. However, this embodiment proposes a user-defined network partitioning method to ensure both computational accuracy and speed. The network partitioning is as follows: Figure 3 As shown, the largest element size in the network should be smaller than the resonant cavity radius. In the buffer cavity section, the network size is set to a predefined standard, with a maximum element size of 2.5mm. Refining the network in the resonant cavity section ensures computational accuracy; a refinement order of 3 ensures precision while preventing insufficient computational memory due to an excessive number of elements. The network is set to a free tetrahedral network. Compared to hexahedral networks, tetrahedral networks are more adaptable to complex geometries and are often used for free mesh generation.

[0029] Step 202: Set initial geometric parameters and environmental parameters: temperature and air pressure; In this embodiment, the physical interface needs to be initially set with the following parameters: temperature 293.15K, absolute pressure 1 atm, rigid boundary wall, no slippage mechanical conditions, and isothermal thermal conditions.

[0030] Step 203: Geometric modeling defines the shape and location of the resonant cavity and buffer cavity, and divides the model network; In this embodiment, the propagation of the acoustic signal in the photoacoustic cell follows the wave equation, but its propagation process generates viscous and thermal losses. To improve the accuracy of the analysis and calculation, this embodiment proposes to set a boundary layer to partition the network. The rigid wall parts of the photoacoustic cell structure itself, namely the outer wall of the resonant cavity, the outer wall of the buffer cavity, and the laser window, are all set as boundary layers. Since the window of the photoacoustic cell itself is relatively small in this design, a rigid wall is chosen to represent it.

[0031] Step 204: The thermodynamic system generates a mixed gas material, which is then filled into the resonant cavity and the buffer cavity; In this embodiment, the material inside the photoacoustic cell is a mixture of CH4 gas at a concentration of 800 ppm and other gases, as shown in Table 1. Table 1 Specific Composition of Mixed Gas Material

[0032] In the global definition, select the thermodynamic system, set the properties of the mixed gas CH4, H2, CH4, C2H4, C2H2, C2H6, CO2, O2, N2, and add it as a new material that matches the actual standard gas.

[0033] The calculation process for the heat source parameters in the simulated acoustic interface is as follows: According to the expression for the heat source's heat... Where N is the molecular number density of the target gas. That is, the target gas absorption cross section, and I is the laser intensity; Based on the characteristics of a Gaussian beam, the light intensity can be expressed as: , Where I(r) represents the change in light intensity at the radial position of the photoacoustic cell. Let be the radial distance of the photoacoustic cell in cylindrical coordinates. It is the laser power. Let the waist radius be the expression. The peak light intensity is at the waist position. It is the radial spatial attenuation of the laser; The intensity of the Gaussian beam at Rayleigh distance position satisfies the condition , Using the heat source heat formula, combined with its definition The expression for the heat source heat H can be calculated as follows: , In the formula, It refers to gas concentration. is the absorption cross section of the target gas, and N is the molecular number density under standard atmospheric conditions. It is the modulation coefficient corresponding to the modulation depth of the laser; by substituting this variable into the expression for the heat source, the Gaussian distribution of the heat source can be presented in mathematical terms, making the simulation closer to the actual situation and improving the calculation speed.

[0034] In this embodiment, the size parameters are set as follows: resonant cavity length 10-100mm, radius 1-4.8mm; buffer cavity length 5-50mm, radius 5-20mm.

[0035] Step 205: Set up the physical field for pressure acoustics and set the basic conditions; Step 206: The study is set to the frequency domain and characteristic frequency; parametric scanning and cyclic simulation are used to calculate sound pressure, characteristic frequency, and quality factor; a 7-element dataset is generated; In this embodiment, parametric scanning is used to solve for the pressure acoustic frequency domain and characteristic frequencies. The results of a single simulation are as follows: Figure 4 The simulation was repeated 240 times to generate a multi-parameter dataset including resonant cavity length L1, resonant cavity radius R1, buffer cavity length L2, buffer cavity radius R2, sound pressure level P, quality factor Q, and characteristic frequency f.

[0036] In this embodiment, the pressure acoustic two-dimensional simulation modeling method not only presents the process of thermal change into sound, but also simplifies its calculations, solving the problem that boundary conditions become unclear due to the simulated flow of heat. Using this dynamic simulation method, the intensity expression and spatial distribution relationship of the Gaussian beam are utilized, also addressing the current problem of ignoring the spatial characteristics of the Gaussian beam itself in simulations. Furthermore, pressure acoustics allows for the addition of more external conditions, such as background noise, internal friction and reflection, and simulation of non-thermally viscous regions. The network settings differ from commonly used default thinning or conventional networks; the use of free tetrahedral networks and thinning saves computational resources for the resonant cavity and buffer cavity, reducing simulation errors caused by some networks exceeding the resonant cavity radius.

[0037] Step 2: Using the parameters of the H-type photoacoustic cell as size variables, and the characteristic frequency, quality factor, and sound pressure level as target performance labels, train the random forest surrogate model based on a multi-parameter small sample dataset to obtain the trained random forest surrogate model. In this embodiment, the resonant cavity length L1, resonant cavity radius R1, buffer cavity length L2, and buffer cavity radius R2 are selected as input size variables; sound pressure P, quality factor Q, and characteristic frequency f are selected as target performance labels.

[0038] To ensure accurate extraction of physical laws from small sample datasets and improve prediction accuracy, additional calculations were performed for similar area ratios (ar) and similar volume ratios (vr). , The final model input is 6-dimensional. Due to the large differences in the dimensions of parameters such as frequency and sound pressure, normalization preprocessing was performed on the dataset to make the features of different dimensions and orders of magnitude more balanced, and to simultaneously take into account the three output performance indicators.

[0039] In this embodiment, 200 decision trees are established, with a depth not exceeding 10 layers, and the surrogate model training structure is as follows. Figure 5 As shown. Simultaneously, based on R on the training set... 2 The coefficient of determination is used to perform in-train fitting.

[0040] Step 3: Combine the randomly generated photoacoustic pool size with the trained random forest surrogate model to perform performance prediction, obtain the performance prediction results corresponding to each group of photoacoustic pool sizes, filter them through CDP constraint dominance, and construct the Pareto optimal solution set; For the optimization of photoacoustic cell, higher sound pressure level (SPL) and quality factor result in better performance. This embodiment proposes a method combining the trained RF proxy model with the improved SPEA2 (Strength Pareto Evolutionary Algorithm 2) algorithm to optimize SPL and quality factor. The improved SPEA2 algorithm is constructed, and its core flowchart is shown in the figure below. Figure 6 Specifically, it includes the following steps: Step 301: Randomly generate the size of the photoacoustic cell; In this embodiment, the SPEA2 algorithm generates a random initial population, combines the designed basic four-dimensional variables—resonant cavity length L1, resonant cavity radius R1, buffer cavity length L2, and buffer cavity radius R2—and calculates the similarity area ratio and similarity volume ratio required by the random forest surrogate model based on different samples. The population size is set to 150, the elite set to 100, the maximum number of iterations to 250, the crossover probability to 0.8, and the mutation probability to 0.2, generating 400 photoacoustic cell sizes.

[0041] Step 302: Input all random photoacoustic cell sizes into the trained RF proxy model to predict the performance results, including predicted values ​​for sound pressure level, quality factor, and characteristic frequency. Step 303: Filter by CDP constraint dominance and construct the Pareto optimal solution set; To ensure that its size and performance predictions meet the constraints of the H-type photoacoustic cell, the SPEA2 was improved by incorporating CDP constraint dominance. The steps of the CDP constraint dominance criterion are as follows: Figure 7By setting the total violation value cv, target constraints are carried out in three aspects according to actual requirements: feature frequency constraint, length ratio constraint, and radius ratio constraint. If the prediction result meets the corresponding constraints, the total violation value cv remains unchanged; if it does not meet the constraints, the v value is incremented by 1. Finally, the cv values ​​of the three constraints are added together. The larger the cv value, the more severe the violation. The introduction of this mechanism ensures that feasible solutions are always superior to infeasible solutions. As a further implementation, the characteristic frequency constraint is that the characteristic frequency deviation does not exceed a preset threshold, such as: ; The length ratio constraint is that the ratio of the length of the resonant cavity to the length of the buffer cavity is within a preset range, such as: ; The radius ratio constraint is that the ratio of the resonant cavity to the buffer cavity radius is within a preset range, such as: ; The smallest cv value, which satisfies Pareto dominance when the objective variable is negative, is placed into the elite set Archive. The population size, elite set, maximum number of iterations, crossover probability, and mutation probability are set. The fitness of individuals is calculated, and the optimal non-dominated solution is selected to fill the external archive. If the external archive is full, a better solution is selected to fill it. A tournament is held to select the parent generation for crossover and mutation, and a new generation is created. The above process is repeated until the set number of iterations is reached. The non-dominated solutions are selected according to the requirements of maximum sound pressure P and quality factor Q to obtain the Pareto optimal solution set.

[0042] It should be noted that, since the set condition SPEA2 is aimed at minimization, the target performance label takes a negative value.

[0043] Step 4: Based on the Pareto optimal solution set, make a decision on the filtered performance prediction results to obtain the optimal structural size of the H-type photoacoustic cell.

[0044] In this embodiment, when making decisions based on the screened performance prediction results, the optimal structural size of the H-shaped photoacoustic cell can be obtained by using the ideal point method to find a compromise solution. The core step is to determine the ideal point and the worst point, with the aim of maximizing both sound pressure level and Q value. The normalized inferiority, i.e., the deviation from the ideal point, is then calculated, with 0 representing perfect and 1 representing worst. The weighted Euclidean distance is calculated, and the solution with the smallest distance is selected as the final optimized size.

[0045] In this embodiment, the specific results are as follows: Figure 8 As shown, the final optimization results are: resonant cavity radius 2.7mm, resonant cavity length 34.5mm, buffer cavity radius 13.7mm, buffer cavity length 15.9mm, length ratio 2.17, and radius ratio 5.1.

[0046] In this embodiment, the ideal point method is used to find a compromise solution, effectively solving the optimization problem of lacking a basis for selecting optimization dimensions based on Pareto images, and making up for the shortcomings of the SPEA2 evolutionary algorithm. This algorithm demonstrates excellent noise robustness and global nonlinear mapping capability at the optimization level of the H-shaped photoacoustic cell, thus successfully presenting a Parato front curve that far surpasses the traditional response surface method under multidimensional physical constraints. It not only improves the efficiency of finite element simulation but also incorporates physical priors into the optimization process, overcoming the problem of excessive reliance on statistics in the deep response surface method, and has broad prospects and high engineering application value.

[0047] Example 2 This embodiment provides a system for optimizing the structure of an H-shaped photoacoustic cell based on dynamic modeling and simulation, including: The dynamic modeling and simulation module is used to obtain the parameters of the H-type photoacoustic cell using the dynamic modeling and simulation method of the H-type photoacoustic cell, calculate the characteristic frequency, quality factor and sound pressure in batches, and construct a multi-parameter small sample dataset. The model training module is used to train the random forest proxy model based on a multi-parameter small sample dataset, using the parameters of the H-type photoacoustic pool as size variables and the feature frequency, quality factor and sound pressure as target performance labels. The model optimization module is used to combine the randomly generated photoacoustic pool size and the trained random forest proxy model to perform performance prediction, obtain the performance prediction results corresponding to each set of photoacoustic pool sizes, and construct the Pareto optimal solution set by filtering through the CDP constraint dominance criterion. The results output module is used to combine the Pareto optimal solution set to make a decision on the filtered performance prediction results and obtain the optimal structural size of the H-type photoacoustic cell.

[0048] It should be noted that the specific implementation of the H-type photoacoustic cell structure optimization method system based on dynamic modeling and simulation in this embodiment of the invention is similar to the specific implementation of the H-type photoacoustic cell structure optimization method based on dynamic modeling and simulation in this embodiment of the invention. For details, please refer to the description in the method section. In order to reduce redundancy, it will not be repeated here.

[0049] Example 3 This embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in the above-described method for optimizing an H-type photoacoustic cell structure based on dynamic modeling and simulation.

[0050] Example 4 This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps in the above-described method for optimizing an H-type photoacoustic cell structure based on dynamic modeling and simulation.

[0051] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0052] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0053] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0054] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0055] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0056] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for optimizing an H-shaped photoacoustic cell structure based on dynamic modeling and simulation, characterized in that, Includes the following steps: The parameters of the H-type photoacoustic cell were obtained by using the dynamic modeling and simulation method of the H-type photoacoustic cell, and the characteristic frequency, quality factor and sound pressure were calculated in batches to construct a multi-parameter small sample dataset. Using the parameters of the H-type photoacoustic cell as size variables and the characteristic frequency, quality factor, and sound pressure level as target performance labels, a random forest surrogate model is trained based on a multi-parameter small sample dataset to obtain the trained random forest surrogate model. Performance prediction is performed by combining randomly generated photoacoustic pool sizes and trained random forest surrogate models to obtain performance prediction results for each set of photoacoustic pool sizes. The Pareto optimal solution set is then constructed by filtering the results using the CDP constraint dominance criterion. By combining the Pareto optimal solution set, a decision is made based on the filtered performance prediction results to obtain the optimal structural size of the H-type photoacoustic cell.

2. The H-type photoacoustic cell structure optimization method based on dynamic modeling and simulation as described in claim 1, characterized in that, The dynamic modeling and simulation method for the H-type photoacoustic cell includes: Use MATLAB to connect to the COMSOL dialog box to construct a dynamic two-dimensional axisymmetric geometric model; Set the initial geometric parameters and environmental parameters: temperature and air pressure; The geometric model specifies the shape and location of the resonant cavity buffer cavity, omitting the external sound opening of the H-type photoacoustic cell and the air inlet / outlet structure. Thermodynamic system modules are used to generate mixed gas materials, which are then filled into the resonant cavity and buffer cavity. The physical field is set to pressure acoustics and the relevant basic conditions are set. Using parametric scanning, the pressure acoustic frequency domain and characteristic frequency are solved, and the sound pressure, characteristic frequency and quality factor are calculated. Finally, a multi-parameter dataset containing resonant cavity length, resonant cavity radius, buffer cavity length, buffer cavity radius, sound pressure, quality factor and characteristic frequency is constructed.

3. The H-type photoacoustic cell structure optimization method based on dynamic modeling and simulation as described in claim 2, characterized in that, When constructing a dynamic two-dimensional axisymmetric geometric model, the heat source is set to exhibit a Gaussian distribution, as shown below: ， In the formula, Indicates the heat of the heat source. It refers to gas concentration. is the absorption cross section of the target gas, and N is the molecular number density under standard atmospheric conditions. It is the modulation coefficient corresponding to the laser modulation depth; It is the laser power. The waist radius is This represents the radial distance of the photoacoustic cell in cylindrical coordinates.

4. The H-type photoacoustic cell structure optimization method based on dynamic modeling and simulation as described in claim 1, characterized in that, When training the random forest surrogate model, the input also includes calculating the similarity area ratio and similarity volume ratio based on the parameters of the H-type photoacoustic cell, which include the resonant cavity length, resonant cavity radius, buffer cavity length, and buffer cavity radius.

5. The H-type photoacoustic cell structure optimization method based on dynamic modeling and simulation as described in claim 4, characterized in that, The formulas for calculating similar area ratios and similar volume ratios are as follows: ， in, For similar area ratios, For similar volume ratios, The length of the resonant cavity. The length of the buffer chamber. Where is the radius of the resonant cavity. Where is the radius of the buffer cavity.

6. The H-type photoacoustic cell structure optimization method based on dynamic modeling and simulation as described in claim 1, characterized in that, The size of the photoacoustic cell is randomly generated using the SPEA2 algorithm.

7. The H-type photoacoustic cell structure optimization method based on dynamic modeling and simulation as described in claim 1, characterized in that, The total violation degree is set through the CDP constraint dominance criterion, which includes three constraints: Frequency constraint: The characteristic frequency deviation does not exceed a preset threshold; Length ratio constraint: The length ratio of the resonant cavity to the buffer cavity is within a preset range; Radius ratio constraint: The radius ratio of the resonant cavity to the buffer cavity is within a preset range; The violation score is 0 when each constraint is satisfied and 1 when it is not satisfied. The total violation score is the sum of the three constraints. The performance prediction value is selected based on the preset range of the total violation score.

8. A system for optimizing the structure of an H-shaped photoacoustic cell based on dynamic modeling and simulation, characterized in that, include: The dynamic modeling and simulation module is used to obtain the parameters of the H-type photoacoustic cell using the dynamic modeling and simulation method of the H-type photoacoustic cell, calculate the characteristic frequency, quality factor and sound pressure in batches, and construct a multi-parameter small sample dataset. The model training module is used to train the random forest proxy model based on a multi-parameter small sample dataset, using the parameters of the H-type photoacoustic pool as size variables and the feature frequency, quality factor and sound pressure as target performance labels. The model optimization module is used to combine the randomly generated photoacoustic pool size and the trained random forest proxy model to perform performance prediction, obtain the performance prediction results corresponding to each set of photoacoustic pool sizes, and construct the Pareto optimal solution set by filtering through the CDP constraint dominance criterion. The results output module is used to combine the Pareto optimal solution set to make a decision on the filtered performance prediction results and obtain the optimal structural size of the H-type photoacoustic cell.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the H-type photoacoustic cell structure optimization method based on dynamic modeling and simulation as described in any one of claims 1-7.

10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the H-type photoacoustic cell structure optimization method based on dynamic modeling and simulation as described in any one of claims 1-7.

Images