Optimization method and system for dry gas seal groove profile parameters

Abstract

The invention provides a dry gas seal groove profile parameter optimization method and system, and relates to the technical field of airtight structure optimizing.The method comprises the steps that the type and range of groove profile parameters are determined, parameter values are randomly generated and subjected to normalized combination, and an initial population is constructed; secondly, finite element simulation is conducted on each individual, the air film rigidity, the leakage rate and the wall attachment data are obtained, the rigidity-leakage ratio and the wall attachment freezing index are calculated, and a sensitive response index is constructed through parameter perturbation; then, designing a self-adaptive adjustment factor which is gradually reduced along with iteration, and constructing a first optimization index and a second optimization index; and finally, adopting a multi-objective genetic algorithm to perform selection, crossover and freezing guide variation on the population, performing iterative optimization until a preset number of times, and outputting two individual vectors with optimal indexes as final groove type parameters.

Description

Technical Field

[0001] This invention relates to the field of airtight structure optimization technology, specifically to a method and system for optimizing the parameters of a dry gas sealing groove. Background Technology

[0002] In the field of dry gas sealing technology, the formation and stability of the gas film between the sealing end faces are crucial to sealing performance. Especially in hydrogen media, due to the low density and high viscosity of hydrogen, the gas film exhibits a significant adhesion effect (i.e., the Coanda effect) at the micrometer scale. This means the gas does not flow in a traditional turbulent or laminar flow state, but rather remains in a state of unseparated boundary layer, with the gas film tightly adhering to the sealing end face. This physical characteristic makes the flow behavior between the sealing end faces more complex. Traditional groove design methods based on conventional flow assumptions cannot accurately describe the actual flow state of the gas film, resulting in a lack of specificity in the design and optimization of sealing structures. Existing spiral groove designs mostly rely on simplified flow models, failing to fully consider the impact of the adhesion effect on gas film stability, and thus failing to fully leverage the performance advantages of dry gas seals in low-density, high-viscosity gas media.

[0003] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for optimizing the parameters of a dry gas sealing groove, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A method for optimizing the parameters of a dry gas seal groove, comprising the following steps: Step 1: Determine the types and predetermined ranges of dry gas sealing groove parameters. For each parameter, generate several random parameter values ​​within its predetermined range. Arrange and combine the normalized random parameter values ​​of various parameters to generate multiple individual vectors to construct the initial population. Step 2: Perform finite element simulation on the dry gas sealing groove under each individual vector to obtain the gas film stiffness, leakage rate and wall adhesion data of the dry gas sealing groove, and calculate the stiffness-leakage ratio. Construct a wall adhesion freezing index based on the wall adhesion data. For each individual vector, perform micro-perturbation on various dry gas sealing groove type parameters, and construct a sensitive response index based on the stiffness-leakage ratio after micro-perturbation. Step 3: Design an adaptive adjustment factor that decreases with the number of iterations. Construct the first optimization index based on the rigid-leak ratio and the adaptive adjustment factor, and construct the second optimization index based on the wall-attachment freezing index and the adaptive adjustment factor. Step 4: Based on the first and second optimization indices, perform multi-objective genetic algorithm selection and crossover operations on the individual vectors in the initial population. Analyze the micro-perturbations of the crossover individual vectors to perform frozen guided mutation to obtain an iterative population. Update the iterative population to a new initial population and iterate until the preset number of iterations is reached. Select the individual vector with the largest stiffness-to-leakage ratio in the last iterative population as the final dry gas sealing groove type parameter.

[0006] Furthermore, the dry gas sealing groove parameters include inner radius, groove root radius, outer radius, helix angle, circumferential angle of the groove area, and circumferential angle of the dam area.

[0007] Furthermore, the stiffness-to-leakage ratio is the ratio of the gas film stiffness to the leakage rate; Based on the inner radius and the groove root radius, an annular area is selected within the dry gas sealing groove as the area to be analyzed. The wall-attached data includes the pressure gradient distribution and airflow radial velocity distribution of the area to be analyzed. The logic for constructing the wall-attachment freezing index is as follows: normalize the pressure gradient distribution and the radial velocity distribution of the airflow, integrate the pressure gradient distribution in the region to be analyzed to obtain the pressure driving energy, integrate the radial velocity of the airflow in the region to be analyzed to obtain the radial momentum release energy of the airflow, and divide the radial momentum release energy of the airflow by the pressure driving energy to obtain the wall-attachment freezing index.

[0008] Furthermore, a preset disturbance amplitude is set. For any dry gas sealing groove type parameter in any individual vector, a disturbance is applied within a predetermined range according to the preset disturbance amplitude, and the stiffness-to-leakage ratio after the disturbance is obtained. The difference between the stiffness-to-leakage ratio before and after the disturbance is calculated. The square of the ratio of the stiffness-to-leakage ratio difference to the preset disturbance amplitude is calculated as the response sensitivity of the dry gas sealing groove type parameter. All dry gas sealing groove type parameters of the individual vector are traversed. The response sensitivity of all dry gas sealing groove type parameters is first accumulated and then squared to obtain the sensitivity response index of the individual vector.

[0009] Furthermore, the logic for designing an adaptive adjustment factor that decreases with the number of iterations is as follows: obtain a preset number of iterations, divide the current number of iterations by the preset number of iterations to obtain an iteration ratio, and process the iteration ratio based on the Sigmoid function to obtain the adaptive adjustment factor under the current iteration round.

[0010] Furthermore, the logic for constructing the first optimization index is as follows: for any individual vector, calculate the difference between 1 and its adaptive adjustment factor in the current iteration round, and then calculate the product of this difference and its rigid-leakage ratio in the current iteration round, which serves as its first optimization index in the current iteration round.

[0011] Furthermore, the logic for constructing the second optimization index is as follows: for any individual vector, calculate the product between its adaptive adjustment factor and the wall-freezing index in the current iteration round, and use it as its second optimization index in the current iteration round.

[0012] Furthermore, the logic for freezing-guided variation is as follows: preset the threshold of sensitive response index, the variation range, and the response sensitivity threshold of each dry gas sealing groove type parameter; If the sensitivity response index of an individual vector is greater than the sensitivity response index threshold, for any dry gas seal groove type parameter in the individual vector, a random number is randomly generated within its variation range. 1 is added to the random number, and then multiplied by the dry gas seal groove type parameter to obtain the variation result of the dry gas seal groove type parameter. If the sensitivity response index of an individual vector is not greater than the sensitivity response index threshold, the response sensitivity of any dry gas sealing groove type parameter in that individual vector is compared one by one with the corresponding response sensitivity threshold. The dry gas sealing groove type parameters whose response sensitivity is less than the corresponding response sensitivity threshold are called frozen parameters, and the remaining dry gas sealing groove type parameters are called non-frozen parameters. For any non-frozen parameter, a random number is randomly generated within its variation range. 1 is added to the random number, and then multiplied by the dry gas sealing groove type parameter. If the variation result is not within the predetermined range of the corresponding non-frozen parameter, it is mutated again to obtain the variation result of the non-frozen parameter. For the frozen parameter, a random number is selected within its predetermined range as the variation result of the frozen parameter.

[0013] The present invention further provides an optimization system for dry gas seal groove type parameters. The system is used to implement the optimization method for the dry gas seal groove type parameters, specifically including: The population generation module is used to determine the types and predetermined ranges of parameters for dry gas sealing grooves. For each parameter, several random parameter values ​​are generated within its predetermined range. The random parameter values ​​after normalization of various parameters are arranged and combined to generate multiple individual vectors to construct the initial population. The simulation analysis module is used to perform finite element simulation of the dry gas sealing groove under each individual vector to obtain the gas film stiffness, leakage rate and wall adhesion data of the dry gas sealing groove, and calculate the stiffness-leakage ratio. Based on the wall adhesion data, a wall adhesion freezing index is constructed. For each individual vector, various dry gas sealing groove type parameters are micro-perturbed, and a sensitive response index is constructed by combining the micro-perturbed stiffness-leakage ratio. The target construction module is used to design an adaptive adjustment factor that decreases with the number of iterations. The first optimization index is constructed based on the rigid-leak ratio and the adaptive adjustment factor, and the second optimization index is constructed based on the wall-attachment freezing index and the adaptive adjustment factor. The algorithm optimization module is used to perform multi-objective genetic algorithm selection and crossover operations on individual vectors in the initial population based on the first and second optimization indices. After crossover, the individual vectors are analyzed for micro-perturbation to perform frozen guided mutation to obtain an iterative population. The iterative population is updated to a new initial population and iterated until a preset number of iterations is reached. The individual vector with the largest stiffness-to-leakage ratio in the last iterative population is selected as the final dry gas sealing groove type parameter.

[0014] Compared with the prior art, the beneficial effects of the present invention are: This invention transforms the abstract wall-attachment effect into a quantifiable optimization objective by constructing a wall-attachment freezing index. This enables precise characterization and control of the gas film adhesion stability in the sealed dam area, allowing the optimized groove parameters to effectively match the flow behavior of the gas film under the wall-attachment effect. This significantly enhances the adsorption stability of the gas film at the sealing end face. Furthermore, this invention introduces an adaptive adjustment factor. In the early stages of optimization, it encourages the algorithm to explore high stiffness-to-leakage ratio regions within a broad parameter space. In the later stages, it guides the algorithm to refine the selection of robust individuals with strong anti-disturbance capabilities. Through a layered threshold freezing-guided mutation strategy, it avoids the algorithm getting trapped in local optima and ultimately ensures that the optimized groove parameters balance high stiffness-to-leakage ratio and high wall-attachment freezing index. This significantly improves the operational reliability and engineering practicality of dry gas seals in low-density, high-viscosity gas media. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the overall method flow of the present invention; Figure 2 This is a geometric model of a dry gas sealing groove type according to the present invention; Figure 3 This is a schematic diagram of the overall system structure of the present invention. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0017] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0018] Example: Please see Figure 1 The present invention provides a technical solution: A method for optimizing the parameters of a dry gas seal groove, comprising the following steps: Step 1: Determine the types and predetermined ranges of dry gas sealing groove parameters. For each parameter, generate several random parameter values ​​within its predetermined range. Arrange and combine the normalized random parameter values ​​of various parameters to generate multiple individual vectors to construct the initial population. Furthermore, the dry gas seal groove parameters include the inner radius, groove root radius, outer radius, helix angle, circumferential angle of the groove area, and circumferential angle of the dam area. The predetermined range of the gas seal groove parameters can be determined by consulting relevant literature or according to the client's requirements. Since the dimensions of each parameter are inconsistent, to eliminate dimensional differences, after generating random parameter values, the randomly generated parameter values ​​are subjected to maximum and minimum normalization processing, wherein the selected maximum and minimum values ​​are the maximum and minimum values ​​within the predetermined range.

[0019] Step 2: Perform finite element simulation on the dry gas sealing groove under each individual vector to obtain the gas film stiffness, leakage rate and wall adhesion data of the dry gas sealing groove, and calculate the stiffness-leakage ratio. Construct a wall adhesion freezing index based on the wall adhesion data. For each individual vector, perform micro-perturbation on various dry gas sealing groove type parameters, and construct a sensitive response index based on the stiffness-leakage ratio after micro-perturbation. Finite element simulation of the dry gas sealing groove under each individual vector can be performed using existing technology. Essentially, it is to solve the Reynolds equation for compressible gas that describes the flow law of micron-level gas film. First, the geometric model of the dry gas sealing groove can be drawn according to the dry gas sealing groove type parameters corresponding to the individual vector. By accurately converting the individual vector parameters generated in step 1 into a periodic sector-shaped computational domain geometric model containing a single groove and sealing weir and performing fine meshing, the pressure boundary, periodic boundary, and physical properties considering gas viscosity and compressibility are set at the inner and outer diameters in the finite element software. The pressure distribution and velocity distribution of the entire gas film are obtained through iterative solution. Finally, the opening force and gas film stiffness are obtained by integrating the pressure field, the leakage rate is obtained by integrating the flow rate, and the pressure gradient and radial velocity distribution are extracted from a specific annular region (between the inner radius and the groove root radius) to construct a wall freezing index that quantifies the wall adhesion effect.

[0020] This is existing technology, and you can refer to the content of "Research on Dry Gas Sealing Performance and Structural Optimization for Helium Compressors" from Lanzhou University of Technology. It will not be elaborated here.

[0021] Please see Figure 2 , Figure 2 This is a geometric model of a dry gas sealing groove type according to the present invention; in, Let the inner radius be , The outer radius is The radius of the groove root, The helix angle, The circumferential angle of the groove area, The circumferential angle of the dam area is defined. The parameters of each type of dry gas sealing groove are clearly defined in the field of dry gas sealing groove design, and will not be elaborated here.

[0022] Based on the inner radius and the groove root radius, an annular area is selected within the dry gas sealing groove as the area to be analyzed. The wall-attached data includes the pressure gradient distribution and airflow radial velocity distribution of the area to be analyzed. The logic for constructing the wall-attachment freezing index is as follows: normalize the pressure gradient distribution and the radial velocity distribution of the airflow, integrate the pressure gradient distribution in the region to be analyzed to obtain the pressure driving energy, integrate the radial velocity of the airflow in the region to be analyzed to obtain the radial momentum release energy of the airflow, and divide the radial momentum release energy of the airflow by the pressure driving energy to obtain the wall-attachment freezing index.

[0023] The annular region between the inner radius and the groove root radius is selected as the region to be analyzed. Based on the center of the dry gas sealing groove, circles are drawn with the inner radius and the groove root radius respectively. The annular region enclosed by the two circles with the radius between the inner radius and the groove root radius is selected as the region to be analyzed.

[0024] The main purpose is to accurately capture the influence of the Coanda effect on the stability of the gas film in dry gas seals, thereby achieving an optimization effect that goes beyond simply pursuing the stiffness-to-leakage ratio to taking into account flow stability. In the dry gas seal structure, the inner radius is the outlet boundary on the sealing medium side, and the groove root radius is the termination position of the spiral groove. The area between the two is usually called the sealing dam area—this is the necessary channel for the gas film to leak from the dynamic pressure generation area to the low-pressure side, and it is also the area where the Coanda effect is most obvious. When low-density, high-viscosity gases such as hydrogen flow through this narrow annular dam area, if the Coanda effect is strong, the airflow will adhere tightly to the end face wall, forming a stable boundary layer with a uniform pressure gradient distribution and smooth radial velocity. Conversely, if the Coanda effect is weak, the airflow may separate, vortex, or even flow backward, resulting in turbulent pressure gradient and radial velocity fluctuations. Therefore, by extracting the pressure gradient distribution (reflecting the pressure energy driving the gas film flow) and the radial velocity distribution of the airflow (reflecting the momentum release of the gas film flow) within the selected area, and constructing the wall-freezing index (i.e., the ratio of radial momentum release energy to pressure driving energy), the adhesion stability of the gas film in the dam area can be quantified. Finally, by incorporating this index into the optimization objective, the algorithm will not only select groove types with high stiffness-to-leakage ratios, but also prioritize parameter combinations that allow the gas film to "freeze" and adhere in the dam area, ensuring stable flow, thereby effectively improving the operational reliability of the dry gas seal under low-density gas conditions.

[0025] The wall-attachment freezing index specifically reflects the flow stability and anti-disturbance capability of the gas film due to the wall-attachment effect in the sealing dam area (the annular region between the inner radius and the groove root radius) of a dry gas seal. This index characterizes the firmness of the gas film's adhesion to the end-face wall by quantifying the ratio of radial momentum release energy to pressure driving energy. A higher ratio indicates less momentum loss during radial flow and more effective conversion of pressure energy into stable adhesion flow, making the gas film less prone to separation from the wall or turbulent disturbances. Conversely, a lower ratio means that a significant portion of the pressure energy is consumed in driving the airflow to detach from the wall, forming eddies or backflow, placing the gas film in a state of instability. Therefore, this index essentially transforms the abstract wall-attachment effect into a quantifiable optimization objective, enabling the optimization algorithm to select groove parameter combinations that allow the gas film to "freeze" its adhesion and stabilize flow in key sealing areas while pursuing a high stiffness-to-leakage ratio, thereby improving the operational reliability of the dry gas seal under low-density gas conditions.

[0026] Furthermore, a preset disturbance amplitude is used. Since the applied disturbance is micro-disturbance, the disturbance amplitude is usually less than 2% of the predetermined range. Since this application normalizes all dry gas sealing groove parameters, the disturbance amplitude can be directly represented by a decimal less than 0.02. In this embodiment, the disturbance amplitude is set to 1%, which is 0.01. For any dry gas sealing groove parameter in any individual vector, a disturbance is applied within its predetermined range according to the preset disturbance amplitude, and the stiffness-leakage ratio after the disturbance is obtained. The difference between the stiffness-leakage ratio before and after the disturbance is calculated. The square of the ratio of the stiffness-leakage ratio difference to the preset disturbance amplitude is calculated as the response sensitivity of the dry gas sealing groove parameter. All dry gas sealing groove parameters of the individual vector are traversed, and the response sensitivity of all dry gas sealing groove parameters is first accumulated and then squared to obtain the sensitivity response index of the individual vector.

[0027] For any individual vector, the formula for calculating the response sensitivity is: in, Let be the response sensitivity of the i-th dry gas seal groove type parameter. Let be the difference in stiffness-to-leakage ratio before and after the disturbance of the parameters of the i-th dry gas seal groove. Let represent the disturbance amplitude of the i-th dry gas seal groove type parameter, where i is the index of the dry gas seal groove type parameter.

[0028] The formula for the sensitivity response index is: in, As a sensitive response indicator, The number of types of individual vector dry gas seal groove parameters.

[0029] The response sensitivity reflects the severity of fluctuations in the stiffness-to-leakage ratio (SRR) of a sealing performance index caused by a small change in a single dry gas seal groove parameter. A higher value indicates that a small perturbation in this parameter leads to a significant change in the SRR, meaning the parameter is highly sensitive around its current value. The sensitivity response index, on the other hand, is a comprehensive quantification of the response sensitivity of all parameters in an individual vector. A higher value indicates that the overall performance of this individual is extremely sensitive to small fluctuations in multiple groove parameters. This index is constructed and introduced into the optimization process. In the iteration of the genetic algorithm, an adaptive adjustment factor dynamically guides the search direction—when an individual's sensitivity response index is too high, the optimization algorithm tends to suppress excessive mutation.

[0030] Step 3: Design an adaptive adjustment factor that decreases with the number of iterations. Construct the first optimization index based on the rigid-leak ratio and the adaptive adjustment factor, and construct the second optimization index based on the wall-attachment freezing index and the adaptive adjustment factor. Furthermore, the logic for designing an adaptive adjustment factor that decreases with the number of iterations is as follows: obtain a preset number of iterations, divide the current number of iterations by the preset number of iterations to obtain an iteration ratio, and process the iteration ratio based on the Sigmoid function to obtain the adaptive adjustment factor under the current iteration round.

[0031] The formula for calculating the adaptive adjustment factor is: in, As a phase regulation factor, The current iteration number is T, where T is the preset iteration number. The adaptive adjustment factor generates a value that increases with the number of iterations by inputting the ratio of the current iteration number to the preset iteration number (i.e., the iteration ratio) into the Sigmoid function. The adaptive adjustment factor can adjust the focus of the algorithm at different times, allowing the algorithm to select the most appropriate index for optimization at different stages.

[0032] Furthermore, the logic for constructing the first optimization index is as follows: For any individual vector, calculate the difference between 1 and its adaptive adjustment factor in the current iteration round, and then calculate the product of this difference and its rigid-leak ratio in the current iteration round. This product serves as the first optimization index in the current iteration round. The adaptive adjustment factor is a number less than 1 and greater than 0. The first optimization index is essentially an optimization of the performance index rigid-leak ratio. This invention has two optimization indices. In conventional genetic algorithms, firstly, each optimization index is independent of the others, and secondly, the performance effect of optimization is equivalent in each iteration. The first optimization index is essentially an optimization of the performance index rigid-leak ratio. However, this invention aims to ensure that the genetic algorithm takes into account different missions at different stages. The rigid-leak ratio is the most basic and important indicator of airtightness. In the early stage of optimization, this invention aims to efficiently select individuals with a large rigid-leak ratio. The characteristic of the adaptive adjustment factor in the early stage is to select individual vectors with corresponding small performance. Therefore, by subtracting the adaptive adjustment factor from 1, individuals meeting the conditions are selected in the early stage. The situation is reversed in the later stage.

[0033] Furthermore, the logic for constructing the second optimization index is as follows: for any individual vector, calculate the product between its adaptive adjustment factor and the wall-freezing index in the current iteration round, and use it as its second optimization index in the current iteration round.

[0034] The second optimization metric is essentially an optimization of the wall-frozen metric. Unlike the rigid-leak ratio, the wall-frozen metric only has meaning under a stable structure. If the rigid-leak ratio does not meet the requirements, optimizing the wall-frozen metric is meaningless. Therefore, it is necessary to increase the screening weight of the wall-frozen metric in the later stages of the algorithm. Thus, the product between the adaptive adjustment factor and the wall-frozen metric is directly used as the second optimization metric in the current iteration round.

[0035] The technical effect of this adaptive adjustment factor is as follows: In the early stage of optimization, the first optimization index (focusing on stiffness-to-leakage ratio) dominates, encouraging the algorithm to search for high-performance regions in a broad parameter space; while in the later stage of optimization, the algorithm has already locked a batch of candidate regions with excellent stiffness-to-leakage ratios through the early exploration; at this time, it is necessary to pursue more refined indices (wall-attachment freezing index), so the adaptive adjustment factor makes the second optimization index (focusing on wall-attachment freezing) dominant, guiding the algorithm to shift from simply pursuing performance to refining and stabilizing individuals with stable wall-attachment effects.

[0036] Step 4: Based on the first and second optimization indices, perform multi-objective genetic algorithm selection and crossover operations on the individual vectors in the initial population. Analyze the micro-perturbations of the crossover individual vectors to perform frozen guided mutation to obtain an iterative population. Update the iterative population to a new initial population and iterate until the preset number of iterations is reached. Select the individual vector with the largest stiffness-to-leakage ratio in the last iterative population as the final dry gas sealing groove type parameter.

[0037] The selection and crossover operation of individual vectors in the initial population based on the first and second optimization indices is a conventional technique of multi-objective genetic algorithms. Specifically, all individual vectors are sorted in a non-dominated manner based on maximizing the first and second optimization indices to obtain multiple non-dominated sorting layers. In each non-dominated sorting layer, individuals are sorted by crowding. Individuals with higher non-dominated sorting results are selected in a tournament manner. Specifically, two individual vectors are selected each time. If the two individual vectors are in different non-dominated sorting layers, the individual in the lower non-dominated sorting layer is selected. If the two individuals are in the same non-dominated sorting layer, the individual with greater crowding is selected.

[0038] Crossing is an existing technology, specifically: a crossover ratio is preset in advance, and dry gas sealing groove parameters are randomly selected according to the crossover ratio for crossover.

[0039] The logic for freeze-guided mutation is as follows: preset the sensitivity response index threshold, the mutation range, and the sensitivity threshold for each dry gas sealing groove parameter; these thresholds can be set using existing technologies, and the sensitivity response index threshold can be set based on the distribution statistics of the sensitivity response index of all individuals in the initial population - usually the upper quartile of all statistics is selected as the sensitivity response index threshold. The range of variation can be referenced with industry standards; this application takes... .

[0040] The response sensitivity threshold for each dry gas seal groove type parameter can be set by consulting relevant literature or according to industry standards, or by requesting experts in the field to set it according to the specific requirements of the client. If the sensitivity response index of an individual vector is greater than the sensitivity response index threshold, for any dry gas seal groove type parameter in that individual vector, a random number is randomly generated within its variation range. 1 is added to the random number, and then multiplied by the dry gas seal groove type parameter to obtain the variation result of the dry gas seal groove type parameter. If the sensitivity response index of an individual vector is not greater than the sensitivity response index threshold, the response sensitivity of any dry gas sealing groove type parameter in that individual vector is compared one by one with the corresponding response sensitivity threshold. Dry gas sealing groove type parameters with response sensitivity less than the corresponding response sensitivity threshold are called frozen parameters, and the remaining dry gas sealing groove type parameters are called non-frozen parameters. For any non-frozen parameter, a random number is randomly generated within its variation range. The variation range specifically refers to... Add the random number to 1, and then multiply it by the dry gas sealing groove type parameter to obtain the variation result of the non-freezing parameter. If the variation result is not within the predetermined range of the corresponding non-freezing parameter, then mutate again. For a frozen parameter, a random number is selected within its predetermined range as the result of the mutation of that frozen parameter.

[0041] It should be noted that the mutation method used in this application is conventional random mutation, which is the prior art, that is, a preset mutation range, and for each parameter, a mutation operation is performed to randomly select values ​​within the preset mutation range. Furthermore, this application determines the sensitivity of each parameter through sensitivity. For insensitive parameters, the variation range is expanded to the entire feasible region (predetermined range) to break the freeze and prevent the algorithm from getting trapped in a local optimum due to overprotection.

[0042] Please see Figure 3 The present invention further provides an optimization system for dry gas seal groove parameters, the system being used to implement the optimization method for dry gas seal groove parameters, specifically including: The population generation module is used to determine the types and predetermined ranges of parameters for dry gas sealing grooves. For each parameter, several random parameter values ​​are generated within its predetermined range. The random parameter values ​​after normalization of various parameters are arranged and combined to generate multiple individual vectors to construct the initial population. The simulation analysis module is used to perform finite element simulation of the dry gas sealing groove under each individual vector to obtain the gas film stiffness, leakage rate and wall adhesion data of the dry gas sealing groove, and calculate the stiffness-leakage ratio. Based on the wall adhesion data, a wall adhesion freezing index is constructed. For each individual vector, various dry gas sealing groove type parameters are micro-perturbed, and a sensitive response index is constructed by combining the micro-perturbed stiffness-leakage ratio. The target construction module is used to design an adaptive adjustment factor that decreases with the number of iterations. The first optimization index is constructed based on the rigid-leak ratio and the adaptive adjustment factor, and the second optimization index is constructed based on the wall-attachment freezing index and the adaptive adjustment factor. The algorithm optimization module is used to perform multi-objective genetic algorithm selection and crossover operations on individual vectors in the initial population based on the first and second optimization indices. After crossover, the individual vectors are analyzed for micro-perturbation to perform frozen guided mutation to obtain an iterative population. The iterative population is updated to a new initial population and iterated until a preset number of iterations is reached. The individual vector with the largest stiffness-to-leakage ratio in the last iterative population is selected as the final dry gas sealing groove type parameter.

[0043] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0044] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0045] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0046] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that cannot be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for optimizing the parameters of a dry gas seal groove, characterized in that, The specific steps include: Step 1: Determine the types and predetermined ranges of dry gas sealing groove parameters. For each parameter, generate several random parameter values ​​within its predetermined range. Arrange and combine the normalized random parameter values ​​of various parameters to generate multiple individual vectors to construct the initial population. Step 2: Perform finite element simulation on the dry gas sealing groove under each individual vector to obtain the gas film stiffness, leakage rate and wall adhesion data of the dry gas sealing groove, and calculate the stiffness-leakage ratio. Construct a wall adhesion freezing index based on the wall adhesion data. For each individual vector, perform micro-perturbation on various dry gas sealing groove type parameters, and construct a sensitive response index based on the stiffness-leakage ratio after micro-perturbation. Step 3: Design an adaptive adjustment factor that decreases with the number of iterations. Construct the first optimization index based on the rigid-leak ratio and the adaptive adjustment factor, and construct the second optimization index based on the wall-attachment freezing index and the adaptive adjustment factor. Step 4: Based on the first and second optimization indices, perform multi-objective genetic algorithm selection and crossover operations on the individual vectors in the initial population. Perform micro-perturbation analysis on the crossover individual vectors to perform frozen guided mutation to obtain an iterative population. Update the iterative population to a new initial population and iterate until the preset number of iterations is reached. Select the individual vector with the largest stiffness-to-leakage ratio in the last iterative population as the final dry gas sealing groove type parameter.

2. The method for optimizing the parameters of a dry gas sealing groove according to claim 1, characterized in that: The parameters of the dry gas sealing groove include the inner radius, groove root radius, outer radius, helix angle, circumferential angle of the groove area, and circumferential angle of the dam area.

3. The method for optimizing the parameters of a dry gas seal groove according to claim 2, characterized in that: The stiffness-to-leakage ratio is the ratio of the film stiffness to the leakage rate. Based on the inner radius and the groove root radius, an annular area is selected within the dry gas sealing groove as the area to be analyzed. The wall-attached data includes the pressure gradient distribution and airflow radial velocity distribution of the area to be analyzed. The logic for constructing the wall-attachment freezing index is as follows: normalize the pressure gradient distribution and the radial velocity distribution of the airflow, integrate the pressure gradient distribution in the region to be analyzed to obtain the pressure driving energy, integrate the radial velocity of the airflow in the region to be analyzed to obtain the radial momentum release energy of the airflow, and divide the radial momentum release energy of the airflow by the pressure driving energy to obtain the wall-attachment freezing index.

4. The method for optimizing the parameters of a dry gas seal groove according to claim 1, characterized in that, The logic for constructing the sensitive response index is as follows: A preset disturbance amplitude is used. For any dry gas sealing groove type parameter in any individual vector, a disturbance is applied within its predetermined range according to the preset disturbance amplitude. The stiffness-to-leakage ratio after the disturbance is obtained. The difference between the stiffness-to-leakage ratio before and after the disturbance is calculated. The square of the ratio of this stiffness-to-leakage ratio difference to the preset disturbance amplitude is calculated as the response sensitivity of the dry gas sealing groove type parameter. All dry gas sealing groove type parameters of the individual vector are iterated through. The response sensitivity of all dry gas sealing groove type parameters is first accumulated and then squared to obtain the sensitive response index of the individual vector.

5. The method for optimizing the parameters of a dry gas seal groove according to claim 4, characterized in that, The logic for designing an adaptive adjustment factor that decreases with the number of iterations is as follows: obtain a preset number of iterations, divide the current number of iterations by the preset number of iterations to obtain the iteration ratio, and process the iteration ratio based on the Sigmoid function to obtain the adaptive adjustment factor under the current iteration round.

6. The method for optimizing the parameters of a dry gas seal groove according to claim 5, characterized in that: The logic for constructing the first optimization index is as follows: For any individual vector, calculate the difference between 1 and its adaptive adjustment factor in the current iteration round, and then calculate the product of this difference and its rigid-leak ratio in the current iteration round, which is used as its first optimization index in the current iteration round.

7. The method for optimizing the parameters of a dry gas seal groove according to claim 6, characterized in that, The logic for constructing the second optimization index is as follows: for any individual vector, calculate the product between its adaptive adjustment factor and the wall-freezing index in the current iteration round, and use it as its second optimization index in the current iteration round.

8. The method for optimizing the parameters of a dry gas seal groove according to claim 4, characterized in that, The logic for freeze-guided variation is as follows: preset the sensitivity response index threshold, variation range, and response sensitivity threshold for each dry gas seal groove type parameter; If the sensitivity response index of an individual vector is greater than the sensitivity response index threshold, for any dry gas seal groove type parameter in the individual vector, a random number is randomly generated within its variation range. 1 is added to the random number, and then multiplied by the dry gas seal groove type parameter to obtain the variation result of the dry gas seal groove type parameter. If the sensitivity response index of an individual vector is not greater than the sensitivity response index threshold, the response sensitivity of any dry gas sealing groove type parameter in that individual vector is compared one by one with the corresponding response sensitivity threshold. The dry gas sealing groove type parameters whose response sensitivity is less than the corresponding response sensitivity threshold are called frozen parameters, and the remaining dry gas sealing groove type parameters are called non-frozen parameters. For any non-frozen parameter, a random number is randomly generated within its variation range. 1 is added to the random number, and then multiplied by the dry gas sealing groove type parameter. If the variation result is not within the predetermined range of the corresponding non-frozen parameter, it is mutated again to obtain the variation result of the non-frozen parameter. For the frozen parameter, a random number is selected within its predetermined range as the variation result of the frozen parameter.

9. A system for optimizing the parameters of a dry gas seal groove, characterized in that: The system is used to implement the method for optimizing the parameters of the dry gas seal groove as described in any one of claims 1-8, specifically including: The population generation module is used to determine the types and predetermined ranges of parameters for dry gas sealing grooves. For each parameter, several random parameter values ​​are generated within its predetermined range. The random parameter values ​​after normalization of various parameters are arranged and combined to generate multiple individual vectors to construct the initial population. The simulation analysis module is used to perform finite element simulation of the dry gas sealing groove under each individual vector to obtain the gas film stiffness, leakage rate and wall adhesion data of the dry gas sealing groove, and calculate the stiffness-leakage ratio. Based on the wall adhesion data, a wall adhesion freezing index is constructed. For each individual vector, various dry gas sealing groove type parameters are micro-perturbed, and a sensitive response index is constructed by combining the micro-perturbed stiffness-leakage ratio. The target construction module is used to design an adaptive adjustment factor that decreases with the number of iterations. The first optimization index is constructed based on the rigid-leak ratio and the adaptive adjustment factor, and the second optimization index is constructed based on the wall-attachment freezing index and the adaptive adjustment factor. The algorithm optimization module is used to perform multi-objective genetic algorithm selection and crossover operations on individual vectors in the initial population based on the first and second optimization indices. After crossover, the individual vectors are subjected to micro-perturbation analysis to perform frozen guided mutation to obtain an iterative population. The iterative population is updated to a new initial population and iterated until a preset number of iterations is reached. The individual vector with the largest stiffness-to-leakage ratio in the last iterative population is selected as the final dry gas sealing groove type parameter.

Images