Atomizer parameter configuration and operation adjustment method and device and medium

Abstract

The invention relates to the technical field of atomizer control, in particular to a parameter configuration and operation adjustment method and device of an atomizer and a medium. Performing recursive discrete combination on each preset parameter extreme value of each different atomization index item based on the constraint judgment of the highest atomization effect index of the target atomization mode, outputting multiple groups of parameter configuration discrete combination solutions, and constructing an aerial fog model of each group of parameter configuration discrete combination solutions to perform anastomosis analysis, the optimal parameter configuration discrete combination solution is obtained to perform parameter configuration on the atomizer; a simulated atomization diffusion model is constructed through analogue simulation, approximation of a mean field is deduced according to aerial fog concentration random diffusion expression of the simulated atomization diffusion model, a simulated diffusion equation solution of the simulated atomization diffusion model is obtained, and operation adjustment is performed based on equation solution difference analysis of the simulated diffusion equation solution and an ideal diffusion equation solution. According to the invention, parameter configuration and operation adjustment can be accurately and reasonably adapted to the atomizer according to use requirements, and the atomization effect and quality are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of atomizer control, and particularly to a method, device, and medium for parameter configuration and operation adjustment of an atomizer. Background Art

[0002] An atomizer is a device that forms a mist by dispersing a liquid into tiny droplets and is widely used in medical, industrial, agricultural, and domestic fields. For example, in the medical field, atomizers are commonly used for drug delivery, where atomized medicinal liquids enter the respiratory tract to achieve a therapeutic effect; in the industrial field, atomizers are used in scenarios such as spraying, cooling, and humidity control. The core performance of an atomizer includes atomization particle size, atomization efficiency, and uniformity, and these parameters directly affect the usage effect and adaptability of the device. Therefore, it is particularly important to optimize the performance of the atomizer through reasonable parameter configuration and operation adjustment methods.

[0003] However, the parameter configuration of existing atomizers is usually preset before leaving the factory and cannot be dynamically adjusted according to the specific usage environment and demand status. Therefore, the fixed parameter configuration results in the inability to fully utilize the performance. Moreover, the existing adjustment methods for atomizers usually rely on manual operations and lack an accurate aerosol concentration diffusion analysis function. This not only increases the operation difficulty of adjusting according to the aerosol diffusion quality of the atomizer but also easily introduces human errors, affecting the stability and atomization effect of the device. At the same time, existing atomizers perform poorly when adapting to different media or complex working conditions, such as high-viscosity liquids or corrosive liquids, which easily leads to a decrease in atomization efficiency or an increase in device wear, increasing the maintenance and replacement costs of the atomizer. Therefore, there is an urgent need for a method with efficient and accurate parameter configuration and adjustment to improve and optimize the device performance and reliability of the atomizer. Summary of the Invention

[0004] The present invention overcomes the deficiencies of the prior art and provides a method, device, and medium for parameter configuration and operation adjustment of an atomizer.

[0005] To achieve the above object, the technical solution adopted by the present invention is as follows: In a first aspect of the present invention, a method for parameter configuration and operation adjustment of an atomizer is provided, including the following steps: S102: Obtain the target atomization requirement of the atomizer to sample the atomization state particles on multiple atomization time series. Through propagation prediction, resampling, and atomization state estimation of the state evolution of each atomization state particle, a time-series atomization state model is obtained, and based on the atomization decision logic, a similarity mapping of the atomization effect is performed on the time-series atomization state under the time-series atomization state model to obtain an expected atomization effect diagram; S104: Obtain the atomization mode function of the atomizer and the mode definition aggregation benchmark for different atomization effects. Starting from the root node of the atomization mode function, recursively split different atomization effects based on the mode definition aggregation benchmark, construct an atomization mode recognition decision tree, and traverse and recognize the branch tip nodes of the expected atomization effect diagram through the atomization mode recognition decision tree to obtain the target atomization mode; S106: Based on the constraint determination of the highest atomization effect index of the target atomization mode, recursively discretely combine each preset parameter extreme value of each different atomization index item, output multiple sets of parameter configuration discrete combination solutions, construct an aerosol model for each set of parameter configuration discrete combination solutions and perform coincidence analysis with the expected aerosol model one by one to obtain the best parameter configuration discrete combination solution and configure the parameters of the atomizer; S108: Perform simulation on the best parameter configuration discrete combination solution through the atomizer simulation model to construct a simulated atomization diffusion model, randomly express the state change of the continuously diffusing aerosol concentration according to the simulated atomization diffusion model and deduce the approximation of the mean field to obtain the solution of the simulated diffusion equation of the simulated atomization diffusion model, and perform operation adjustment on the atomizer based on the analysis of the difference between the solution of the simulated diffusion equation and the solution of the ideal diffusion equation.

[0006] More specifically, the step S102 specifically includes the following steps: Obtain the target atomization requirement of the atomizer, preset multiple atomization time series and necessary state samples on each atomization time series according to the target atomization requirement, and mark them as atomization state particles; Construct a state transition model for the target atomization requirement, and perform state evolution propagation prediction on the necessary atomization state particles on each atomization time series through the state transition model to generate new atomization state particles; Use the state transition model that generates new atomization state particles to update the weight of each atomization state particle and perform normalization processing to obtain multiple weight values, and resample the atomization state particles according to the normalized weight values to obtain a new set of atomization state particles; Calculate the estimated value of the atomization state that meets the target atomization requirement based on the new set of atomization state particles to obtain the atomization state estimated value; Repeat the above steps of state evolution propagation prediction, particle resampling, and calculation of atomization state estimated value of the atomization state particles to obtain the atomization state estimated value on each atomization time series and perform state deduction to obtain the time-series atomization state model corresponding to the target atomization requirement; Based on the atomization effects of the atomizer for different preset atomization states, an atomization decision logic is established. Calculate the state similarity between different preset atomization states and the time-series atomization state matrix expressed by the time-series atomization state model. Map the time-series atomization state matrix to the mapping space according to the KL divergence between the state similarity and the state similarity of the mapping space established by the atomization decision logic, and generate an expected atomization effect diagram.

[0007] More specifically, the steps of establishing an atomization decision logic based on the atomization effects of the atomizer for different preset atomization states, calculating the state similarity between different preset atomization states and the time-series atomization state matrix expressed by the time-series atomization state model, mapping the time-series atomization state matrix to the mapping space according to the KL divergence between the state similarity and the state similarity of the mapping space established by the atomization decision logic, and generating an expected atomization effect diagram are specifically as follows: Obtain the established atomization control strategies of the atomizer for different preset atomization states, and determine the corresponding atomization effects required by the atomizer for different preset atomization states according to the established atomization control strategies; Establish an atomization decision logic based on the corresponding atomization effects required by the atomizer for different preset atomization states as a criterion. Define different preset atomization states as the first state cluster, extract the time-series atomization state matrix through the time-series atomization state model, and define the time-series atomization state matrix as the second state cluster; Construct a mapping space based on the atomization effect, and establish a similarity matrix of the mapping space according to the atomization decision logic, defined as the first similarity matrix. Introduce the Gaussian distribution method to calculate the similarity matrix between the first state cluster and the second state cluster, defined as the second similarity matrix; Preset a KL divergence threshold, calculate the KL divergence between the first similarity matrix and the second similarity matrix, and map the time-series atomization state matrix into the mapping space based on the KL divergence until the KL divergence threshold is reached, obtaining the atomization effect diagram corresponding to the time-series atomization state that meets the target atomization requirements, defined as the expected atomization effect diagram.

[0008] More specifically, the step S104 specifically includes the following steps: Obtain the design goal of the atomizer, obtain the atomization mode function of the atomizer according to the design goal, and obtain the mode definition aggregation benchmark of the atomizer for different atomization effects through the established atomization control strategy; Preset the minimum Gini index, calculate the Gini index of the atomization mode function outputting different atomization effects based on the mode definition aggregation benchmark. Take the atomization mode function as the root node, and start from the root node to split different atomization effects into several sub-atomization effect sets using the splitting point where the Gini index reaches the minimum Gini index; Repeat the above splitting steps based on the Gini index to recursively split each of the sub-atomization effect sets until all the sub-atomization effect sets are recursively completed, generating an atomization pattern recognition decision tree; Generate recognition samples of the atomization effect based on the expected atomization effect diagram, and perform sample feature value extraction on the recognition samples according to the target atomization requirement to obtain the feature value range of the recognition samples. At the same time, extract the branch node pattern diagram of the atomization pattern recognition decision tree, and perform an overall traversal recognition of the atomization effect features of the atomization pattern recognition decision tree starting from each branch end node in the branch end node pattern diagram; If the current sample feature obtained by traversing and recognizing the recognition sample is within the feature value range, stop the traversal recognition operation at the branch end node or leaf node where the current recognition sample is located, and output the atomization pattern corresponding to the branch end node or leaf node at the stop traversal recognition position, obtaining the atomization pattern in which the atomizer realizes the expected atomization effect diagram and meets the target atomization requirement, which is marked as the target atomization pattern.

[0009] More specifically, the step S106 specifically includes the following steps: Obtain different atomization index items preset for the target atomization pattern by the atomizer and the preset parameter extreme value intervals of each atomization index item according to the design target of the atomizer; Extract the highest atomization evaluation value of the atomizer through the target atomization requirement, and retrieve the highest atomization effect index after the atomizer executes the target atomization pattern under the regulation of the highest atomization evaluation value based on the highest atomization evaluation value in the big data; Set atomization effect constraint conditions according to the highest atomization effect, and extract the current atomization state of the sequential atomization state model at each atomization time series. Perform a recursive discrete combination of each preset parameter in the preset parameter extreme value intervals of each atomization index item according to the current atomization state; During the recursive discrete combination, obtain the actual atomization effect index presented by the discrete combination of each preset parameter of each atomization index item corresponding to the current recursive node under the current atomization state; if the atomization effect constraint conditions cannot restrict the actual atomization effect index, retain the discrete combination under the current recursive node; If the atomization effect constraint conditions can restrict the actual atomization effect index, revoke the discrete combination under the current recursive node and backtrack to the previous recursive node for discrete combination of other preset parameters until all the discrete combination solutions of the preset parameters restricted by the atomization effect constraint conditions are excluded, and finally output multiple groups of parameter configuration discrete combination solutions; Construct an aerosol model for each group of parameter configuration discrete combination solutions based on the actual atomization effect index, which is defined as the actual aerosol model, and synchronously construct an aerosol model of the expected atomization effect diagram based on the target atomization requirement, which is defined as the expected aerosol model; Calculate the degree of fit between each actual aerosol model and the expected aerosol model one by one to obtain multiple degrees of fit. Only extract the discrete combination solution of the parameter configuration corresponding to the actual aerosol model with the maximum degree of fit, mark it as the optimal discrete combination solution of the parameter configuration, and configure the parameters of the atomizer based on the optimal discrete combination solution of the parameter configuration.

[0010] More specifically, the step S108 specifically includes the following steps: Obtain the initial design drawing of the atomizer, construct an atomizer simulation model according to the initial design drawing. After the parameter configuration is completed, use the atomizer simulation model to perform aerosol simulation on the optimal discrete combination solution of the parameter configuration in aerosol analysis software to obtain a number of simulated aerosol concentration data and the diffusion distribution sites of each simulated aerosol concentration data. Based on the diffusion distribution sites, analyze and calculate a number of simulated aerosol concentration data in aerosol analysis software to generate a simulated aerosol diffusion model, and extract the diffusion state sequence of the simulated aerosol diffusion model; represent the state change of the continuously diffusing aerosol concentration in the simulated aerosol diffusion model in the form of a stochastic process according to the diffusion state sequence to obtain the stochastic dynamics of the simulated aerosol diffusion. Introduce the stochastic differential method to describe the stochastic dynamics of the simulated aerosol diffusion and perform the mean field approximation treatment to obtain the probability density of the occurrence of the stochastic dynamics of the simulated aerosol diffusion. By deriving the distribution of the probability density, finally obtain the stochastic diffusion equation of the simulated aerosol diffusion model. Obtain the usage log of the atomizer, and obtain the ideal aerosol concentration data packet for the atomizer to run the optimal discrete combination solution of the parameter configuration to meet the target atomization requirement through the usage log. Solve the stochastic diffusion equation of the simulated aerosol diffusion model to obtain the solution of the simulated diffusion equation. Based on the ideal aerosol concentration data packet, construct an ideal aerosol diffusion model in aerosol analysis software. Repeat the above steps of stochastic state description and derivation to calculate the stochastic diffusion equation of the ideal aerosol diffusion model, and solve the stochastic diffusion equation of the ideal aerosol diffusion model to obtain the solution of the ideal diffusion equation. If the difference between the solution of the simulated diffusion equation and the solution of the ideal diffusion equation is within the preset diffusion equation solution difference interval, perform operation adjustment on the atomizer based on the upper limit difference of the preset diffusion equation solution difference interval.

[0011] The second aspect of the present invention provides a parameter configuration and operation adjustment device for an atomizer. The parameter configuration and operation adjustment device includes a memory and a processor. The memory stores a parameter configuration and operation adjustment method program for an atomizer. When the parameter configuration and operation adjustment method program is executed by the processor, the steps of any one of the parameter configuration and operation adjustment methods are implemented.

[0012] In a third aspect of the present invention, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer detection program, which, when executed by at least one processor, implements the parameter configuration and operation tuning method of an atomizer as described in any one of the above.

[0013] The present invention solves the technical defects existing in the background art. The beneficial technical effects of the present invention are as follows: Through the propagation prediction, resampling, and atomization state estimation of the state evolution of the atomization state particles sampled for each target atomization requirement, a time-series atomization state model is obtained. Based on the atomization decision logic, a similarity mapping of the atomization effect is performed on the time-series atomization state under the time-series atomization state model to obtain an expected atomization effect diagram; based on the mode definition aggregation benchmark of the atomizer for different atomization effects, different atomization effects are recursively split starting from the root node of the atomization mode function to construct an atomization mode recognition decision tree, and the expected atomization effect diagram is traversed and recognized through the atomization mode recognition decision tree to obtain the target atomization mode; based on the constraint determination of the highest atomization effect index of the target atomization mode, each preset parameter extreme value of each different atomization index item is recursively discretely combined, and multiple groups of parameter configuration discrete combination solutions are output. The aerosol models of each group of parameter configuration discrete combination solutions are subjected to coincidence analysis to obtain the best parameter configuration discrete combination solution for parameter configuration of the atomizer; a simulation atomization diffusion model is constructed through simulation. According to the simulation atomization diffusion model, the state change of the continuously diffusing aerosol concentration is randomly expressed and the approximation of the mean field is deduced to obtain the simulation diffusion equation solution of the simulation atomization diffusion model. Based on the analysis of the difference between the simulation diffusion equation solution and the ideal diffusion equation solution, the operation of the atomizer is tuned. The present invention can accurately calculate and analyze the state of atomization use, the expected atomization effect, and the atomization mode required to achieve this effect according to the use requirements of the atomizer, so as to provide more accurate and reasonable parameter configuration for the atomizer and perform correct operation tuning, so that the atomization quality of the atomizer can better meet the atomization requirements of different users. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0015] Figure 1 Shows a first method flow chart of a parameter configuration and operation tuning method of an atomizer; Figure 2 Shows a second method flow chart of a parameter configuration and operation tuning method of an atomizer; Figure 3 The device structure diagram of a device for parameter configuration and operation adjustment of an atomizer is shown. Specific implementation manners

[0016] In order to be able to more clearly understand the above objects, features and advantages of the present invention, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners. It should be noted that, without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other.

[0017] In the following description, many specific details are set forth in order to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited by the specific embodiments disclosed below.

[0018] The first aspect of the present invention provides a method for parameter configuration and operation adjustment of an atomizer, as Figure 1 shown, including the following steps: S102: Obtain the target atomization requirement of the atomizer to sample atomization state particles on multiple atomization time series. Through propagation prediction, resampling and atomization state estimation of the state evolution of each of the atomization state particles, a time-series atomization state model is obtained, and based on the atomization decision logic, a similarity mapping of the atomization effect is performed on the time-series atomization state under the time-series atomization state model to obtain an expected atomization effect diagram; S104: Obtain the atomization mode function of the atomizer and the mode definition aggregation benchmark for different atomization effects. Based on the mode definition aggregation benchmark, different atomization effects are recursively split starting from the root node of the atomization mode function to construct an atomization mode recognition decision tree, and the expected atomization effect diagram is traversed and recognized at the branch tip nodes through the atomization mode recognition decision tree to obtain the target atomization mode; S106: Based on the constraint determination of the highest atomization effect index of the target atomization mode, perform recursive discrete combination on each preset parameter extreme value of each different atomization index item, output multiple groups of parameter configuration discrete combination solutions, construct an aerosol model for each group of parameter configuration discrete combination solutions and perform coincidence analysis with the expected aerosol model one by one to obtain the best parameter configuration discrete combination solution to configure the parameters of the atomizer; S108: Perform simulation on the best parameter configuration discrete combination solution through the atomizer simulation model to construct a simulated atomization diffusion model. According to the simulated atomization diffusion model, randomly express the state change of the aerosol concentration continuously diffusing and deduce the approximation of the mean field to obtain the solution of the simulated diffusion equation of the simulated atomization diffusion model, and perform operation adjustment on the atomizer based on the analysis of the difference between the solution of the simulated diffusion equation and the solution of the ideal diffusion equation.

[0019] More specifically, step S102 specifically includes the following steps: Obtain the target atomization requirement of the atomizer, preset multiple atomization time series according to the target atomization requirement and necessary state samples on each atomization time series, and label them as atomization state particles; Construct a state transition model for the target atomization requirement, and perform state evolution propagation prediction on the necessary atomization state particles on each atomization time series through the state transition model to generate new atomization state particles; Use the state transition model that generates new atomization state particles to update the weights of each atomization state particle and perform normalization processing to obtain multiple weight values, and resample the atomization state particles according to the normalized weight values to obtain a new set of atomization state particles; Calculate the estimated value of the atomization state that meets the target atomization requirement based on the new set of atomization state particles to obtain the atomization state estimated value; Repeat the steps of state evolution propagation prediction, particle resampling, and atomization state estimated value calculation of the above atomization state particles to obtain the atomization state estimated value on each atomization time series and perform state deduction to obtain the time-series atomization state model corresponding to the target atomization requirement; Based on the atomization effect of the atomizer for different preset atomization states, establish an atomization decision logic, calculate the state similarity between different preset atomization states and the time-series atomization state matrix expressed by the time-series atomization state model, and map the time-series atomization state matrix to the mapping space according to the KL divergence between the state similarity and the state similarity of the mapping space established by the atomization decision logic to generate an expected atomization effect diagram.

[0020] It should be noted that the atomization effect is one of the indicators for evaluating the performance of an atomizer, and it is also a crucial benchmark for the parameter configuration, adjustment, and operation calibration of the atomizer. The atomization effect determines the quality of the atomizer product. Therefore, it is necessary to define the expected atomization effect for the atomizer to provide a reference for parameter configuration. The atomization effect is closely dependent on the actual atomization requirements of the atomizer to determine. However, the existing methods for configuring the parameters of the atomizer are difficult to accurately determine the required expected atomization effect based on the analysis of the atomization scenarios, behaviors, or states such as the action of the target atomization requirements, making the definition of the expected atomization effect have greater ambiguity and abstraction. Eventually, there are large errors in the parameter configuration of the atomizer, affecting the use quality of the atomizer. Therefore, this method first collects a series of state samples at each atomization time series according to the target atomization requirements. These state samples cover the state objects served by different target atomization requirements, such as the physiological physical state of patients, the room air humidity state, or the industrial processing cooling state, etc. The atomization state particles represent the possible initial states and distributions. Then, by constructing a state transition model of the target atomization requirements, the state evolution propagation prediction of these atomization state particles is carried out, that is, according to the state of the previous moment and the control input, the state of the particles at the current moment is predicted using the state transition model, so that the atomization state particles diffuse in the state space of the target atomization requirements, simulating the potential evolution process of the state system required by the target atomization requirements, helping each particle explore the state space, and making the subsequent state expression reflected by the target atomization requirements more accurate.

[0021] It should be noted that the higher the matching degree between the atomization state particles and the observation, the greater the weight. By weighting the atomization state particles, the particles can be concentrated in the most likely state region, reflecting the influence of the current timing requirement of the target atomization demand on the particle state. Through resampling, the atomization state particles will be more concentrated in the high-weight region, enabling the particle filter to effectively reflect the main trend of the target atomization demand state and improving the accuracy of state estimation for different target atomization demands. The newly generated set of atomization state particles can then be used to calculate the estimated value of the atomization state that meets the target atomization demand, thereby obtaining a relatively accurate estimate for the target atomization demand. This estimated value synthesizes the contributions of all atomization state particles and reflects the most likely scenarios, behaviors, or effects, etc. of a series of atomization service states at the current moment. Finally, by repeating the above series of steps, the state estimation for each atomization time series is obtained, and based on this, the temporal atomization state model corresponding to the target atomization demand can be deduced. This model reflects the state premise of the target atomization demand and is a parameter configuration standard for the atomizer to operate according to the target atomization demand. Finally, the expected atomization effect is defined through the state analysis of this temporal atomization state model. Through this method, the temporal state of the services required by the atomizer for different target atomization demands or objects can be described, and the expected atomization effect can be planned according to the required atomization state, improving the clarity and visualization of the atomizer's interpretation of different target atomization demand states, ensuring the accuracy rate of the definition of the expected atomization effect, and effectively improving the reliability and credibility of the atomizer parameter configuration.

[0022] More specifically, based on the atomization effects of the atomizer for different preset atomization states, an atomization decision logic is established. The state similarity between the different preset atomization states and the temporal atomization state matrix expressed by the temporal atomization state model is calculated. According to the KL divergence between the state similarity and the state similarity in the mapping space established by the atomization decision logic, the temporal atomization state matrix is mapped to the mapping space to generate an expected atomization effect diagram, which specifically includes the following steps: Obtain the established atomization control strategies of the atomizer for different preset atomization states, and determine the corresponding atomization effects required by the atomizer for different preset atomization states according to the established atomization control strategies; Based on the criterion that the atomizer requires corresponding atomization effects for different preset atomization states, an atomization decision logic is established. Define different preset atomization states as the first state cluster, extract the temporal atomization state matrix through the temporal atomization state model, and define the temporal atomization state matrix as the second state cluster; Construct a mapping space based on the atomization effect, and establish a similarity matrix of the mapping space according to the atomization decision logic, defined as the first similarity matrix. Introduce the Gaussian distribution method to calculate the similarity matrix between the first state cluster and the second state cluster, defined as the second similarity matrix; Preset the KL divergence threshold, calculate the KL divergence between the first similarity matrix and the second similarity matrix, and map the time-series atomization state matrix into the mapping space based on the KL divergence until the KL divergence threshold is reached, obtaining the atomization effect diagram corresponding to the time-series atomization state that meets the target atomization requirement, which is defined as the expected atomization effect diagram.

[0023] It should be noted that usually in the production and manufacturing of atomizers, a corresponding ideal atomization effect is assigned according to different atomization states to control the operation of the atomizer, so as to ensure that the target atomization requirement can be met to the greatest extent. Therefore, in this method, an atomization decision logic is established according to the atomization effect that the atomizer needs to apply corresponding to different preset atomization states, so as to form a decision logic relationship reference for assigning corresponding atomization effects according to the atomization state; then, by calculating the state similarity between different preset atomization states and the time-series atomization state matrix included in the time-series atomization state model, a similarity matrix, that is, the second similarity matrix, will be obtained at this time. This second similarity matrix is used to measure the approximation degree between each time-series atomization state included in the time-series atomization state model and the preset atomization state defined in advance by the atomizer, making the atomization effect assigned to each time-series atomization state of the time-series atomization state model more accurate in the follow-up, and further improving the definition reliability of the expected atomization effect. After determining the approximation degree between the two, the corresponding atomization effect can be mapped for each time-series atomization state according to the expression of the second similarity matrix. In this method, a mapping space for all atomization effects is constructed. This mapping space is the mapping carrier for all time-series atomization states, which can cover all time-series atomization states included in the time-series atomization state model, ensuring that each time-series atomization state can be mapped to a corresponding atomization effect, so that the atomization effect definition of the atomizer adapts to the time-series state changes of different target atomization requirements. And a similarity mapping benchmark for this mapping space, that is, the first similarity matrix, is established according to the atomization decision logic. The time-series atomization state matrix is mapped into the mapping space based on the KL divergence between the first similarity matrix and the second similarity matrix. The KL divergence is a measure of the difference between two distributions. The smaller the KL divergence, the more accurate the mapping rate of the atomization effect of each time-series atomization state can be ensured.

[0024] It should be noted that through this method, the corresponding expected atomization effect can be mapped for each atomization state output by the time-series atomization state model of the target atomization requirement based on the atomization effect definition logic preset by the atomizer, so that the operation control of the atomizer can be realized strictly around the atomization quality specified by the target atomization requirement, providing a reliable positioning basis for the subsequent mode matching of the atomizer to meet the target atomization requirement, and ensuring the accuracy rate of the atomizer parameter configuration.

[0025] More specifically, the step S104 specifically includes the following steps: Obtain the design objectives of the atomizer, obtain the atomization mode function of the atomizer according to the design objectives, and obtain the mode definition aggregation benchmark for different atomization effects of the atomizer through a predefined atomization control strategy; Preset the minimum Gini index, calculate the Gini index of the atomization mode function outputting different atomization effects based on the mode definition aggregation benchmark, take the atomization mode function as the root node, and start from the root node to split different atomization effects into several sub-atomization effect sets using the splitting point where the Gini index reaches the minimum Gini index; Repeat the above splitting steps based on the Gini index to recursively split each of the sub-atomization effect sets until all the sub-atomization effect sets are recursively completed, generating an atomization mode recognition decision tree; Generate recognition samples of atomization effects based on the expected atomization effect diagram, perform sample feature value extraction on the recognition samples according to the target atomization requirements to obtain the feature value range of the recognition samples, and at the same time extract the branch node pattern diagram of the atomization mode recognition decision tree, and perform an overall traversal and recognition of the atomization effect features of the atomization mode recognition decision tree starting from each branch end node pattern diagram in the branch end node pattern diagram; If the current sample feature of the traversal and recognition of the recognition sample is within the feature value range, stop the traversal and recognition operation at the branch end node or leaf node where the current recognition sample is located, and output the atomization mode corresponding to the branch end node or leaf node at the stop traversal and recognition position, obtaining the atomization mode of the atomizer to achieve the expected atomization effect diagram that meets the target atomization requirements, marked as the target atomization mode.

[0026] It should be noted that some existing atomizers are highly intelligent and integrated with multiple atomization mode functions for users to switch and use under different requirements. The atomization effects presented by each atomization mode function are inconsistent. For example, in the high-temperature moisturizing mode, the aerosol particle size, working temperature, and atomization efficiency are improved, so the atomization effect will be significantly enhanced. The parameter configuration and operation adjustment of the atomizer are controlled on the premise of mode switching. Therefore, it is necessary to locate the atomization mode that meets the target atomization demand based on the expected atomization effect, so that there is a reliable planning basis for the parameter configuration and operation adjustment of the atomizer. In this regard, this method first calculates the Gini index of the atomization mode function outputting different atomization effects based on the mode definition aggregation benchmark of the atomizer for different atomization effects. Using the characteristics of this Gini index to divide all atomization effects for each atomization mode function can achieve the classification and distinction of the atomization effects that the atomization mode function can correspondingly achieve, making the subsequent recognition of the expected atomization effect diagram more accurate and efficient, and reducing the error rate of classification recognition. Through continuous recursive splitting, an atomization mode recognition decision tree can be formed. This atomization mode recognition decision tree has a global recognition architecture with the atomization mode function as the main root node connecting the branches and endings corresponding to one or more atomization effects, and has strong distinguishability. Therefore, the expected atomization effect diagram can be directly imported into the atomization mode recognition decision tree as an identification sample for traversal recognition of each branch and ending. Among them, if the current sample feature of the traversal recognition of the identification sample is within the feature value range, it means that the atomization effect of the current branch and ending or leaf node matches any one of the expected atomization effects in the expected atomization effect diagram, and the atomization mode connected by the branch and ending or leaf node is the atomization mode that meets the expected atomization effect diagram, and the recognition result is obtained. Through this method, a mode recognition decision tree can be constructed by recursively splitting the mode definition aggregation benchmark of the atomization mode function presenting the corresponding atomization effect. Through this mode recognition decision tree, the expected atomization effect diagram can be recognized more efficiently and accurately, improving the rate of parameter configuration and operation adjustment of the atomizer, reducing unnecessary recognition operation steps, enhancing the matching accuracy of the atomization mode that meets the target atomization demand, and improving the reliability of the parameter configuration of the atomizer.

[0027] More specifically, the step S106 specifically includes the following steps: Obtain different atomization index items preset for the target atomization mode by the atomizer and the preset parameter extreme value intervals of each atomization index item according to the design goal of the atomizer; Extract the highest atomization evaluation value of the atomizer through the target atomization demand, and retrieve the highest atomization effect index after the atomizer executes the target atomization mode under the regulation of the highest atomization evaluation value based on the highest atomization evaluation value in the big data; Set the atomization effect constraint conditions according to the highest atomization effect, extract the current atomization state of the time-series atomization state model at each atomization time series, and perform a recursive discrete combination of each preset parameter in the preset parameter extreme value interval of each atomization index item according to the current atomization state; When performing the recursive discrete combination, obtain the actual atomization effect index presented by the discrete combination between each preset parameter of each atomization index item corresponding to the current recursive node under the current atomization state; if the atomization effect constraint conditions cannot restrict the actual atomization effect index, retain the discrete combination under the current recursive node; If the atomization effect constraint conditions can restrict the actual atomization effect index, revoke the discrete combination under the current recursive node and backtrack to the previous recursive node for discrete combination of other preset parameters until all discrete combination solutions of the preset parameters restricted by the atomization effect constraint conditions are excluded, and finally output multiple groups of parameter configuration discrete combination solutions; Construct an aerosol model for each group of parameter configuration discrete combination solutions based on the actual atomization effect index, defined as the actual aerosol model, and simultaneously construct an aerosol model of the expected atomization effect diagram based on the target atomization requirement, defined as the expected aerosol model; Calculate the degree of coincidence between each actual aerosol model and the expected aerosol model one by one to obtain multiple degrees of coincidence, only extract the parameter configuration discrete combination solution corresponding to the actual aerosol model with the maximum degree of coincidence, marked as the optimal parameter configuration discrete combination solution, and configure the parameters of the atomizer based on the optimal parameter configuration discrete combination solution.

[0028] It should be noted that the atomization index items include spray volume, spray particle size, spray efficiency, and spray range. The parameters of the atomizer include voltage, current, power, and operating temperature. Since different atomization modes have specific atomization indexes, for example, the atomization index in the moisturizing atomization mode needs to ensure uniform atomization volume and large atomization range. For each atomization index, there is a corresponding parameter value range and configuration interval. For each atomization index, appropriate parameters need to be selected from the interval for configuration to ensure that the atomization index requirements of the target atomization mode can be achieved. If the parameter values of each index of the atomizer in the interval are inappropriate, it may cause the atomizer to fail to achieve the atomization effect of the required target atomization mode. At the same time, there are several parameter values included in the parameter value range of each atomization index. Therefore, there are various combinations of each index parameter under different atomization modes to achieve different atomization effects. Therefore, this method first obtains the highest atomization effect index after the atomizer executes the target atomization mode based on the highest atomization evaluation value specified by the target atomization requirement. The highest atomization effect index is an effect constraint for the combination of each parameter value of each atomization index. It is necessary to ensure that the corresponding parameter combination can be retained only when the constraint condition of the highest atomization effect index is satisfied. In other words, only when the discrete combination of each parameter value is not restricted by this constraint condition can it be proved that this set of discrete combination parameters can achieve the effect of the target atomization mode. That is, if the atomization effect constraint condition cannot restrict the actual atomization effect index, it means that the parameter discrete combination at the current recursive node can achieve the effect of the target atomization mode, and the discrete combination under the current recursive node is retained. On the contrary, if it can be restricted, it means that the parameter discrete combination at the current recursive node is difficult to achieve the effect of the target atomization mode. Therefore, it is necessary to revoke the discrete combination under the current recursive node and backtrack to the previous recursive node for re-discretization of other preset parameters. In this way, the discrete combinations that do not meet the target atomization effect can be screened out, thereby greatly improving the screening efficiency and accuracy of the combination of each parameter value of each atomization index to form an effect, and avoiding missing possible parameter configuration combinations that can achieve the effect of the target atomization mode for each atomization index.

[0029] It should be noted that after the parameter values of each atomization index are screened out through discrete combination and the discrete combination corresponding to the target atomization mode can be achieved, it is difficult to intuitively judge whether the screened discrete combination parameters can actually achieve the required atomization effect by traditional methods. It is necessary to conduct actual atomization experiments one by one to verify each group of discrete combination parameters, which will consume a large amount of manpower and material resources, and the rate is low, and there may be large human intervention errors. Therefore, in this method, an aerosol model corresponding to each group of screened parameter configuration discrete combinations is constructed to reflect the actual atomization presentation of the discrete combination solution, and each actual aerosol model is compared and calculated with the expected aerosol model of the expected atomization effect diagram one by one. The discrete combination solution of the parameter configuration corresponding to the actual aerosol model with the highest degree of coincidence is the best solution that can maximize the atomization effect of the target atomization mode. The atomizer is configured based on the best parameter configuration discrete combination solution. Through this method, the value of each parameter configuration of the atomization index required under the target atomization mode can be globally discretely combined and quickly and accurately screened to ensure that the parameter configuration can maximize the effect specified by the target atomization mode, and systematic verification can be carried out for each discrete combination parameter configuration, so as to obtain the best configuration parameters to configure the atomizer, replacing the cumbersome steps of traditional manual experiment verification and unnecessary operations, saving time and effort, ensuring the accuracy rate and operation reliability of the atomizer parameter configuration, and improving the performance of the atomizer.

[0030] More specifically, the step S108, as Figure 2 shown, specifically includes the following steps: S202: Obtain the initial design drawing of the atomizer, construct an atomizer simulation model according to the initial design drawing. After the parameter configuration is completed, use the atomizer simulation model to perform atomization simulation on the best parameter configuration discrete combination solution in the aerosol analysis software to obtain a number of simulated aerosol concentration data and the diffusion distribution sites of each simulated aerosol concentration data; S204: Analyze and calculate a number of simulated aerosol concentration data based on the diffusion distribution sites in the aerosol analysis software to generate a simulated atomization diffusion model, and extract the diffusion state sequence of the simulated atomization diffusion model; Represent the state change of the continuously diffusing aerosol concentration in the simulated atomization diffusion model in the form of a stochastic process according to the diffusion state sequence to obtain the stochastic dynamics of the simulated atomization diffusion; S206: Introduce the stochastic differential method to describe the stochastic dynamics of the simulated atomization diffusion and perform the mean field approximation treatment to obtain the probability density of the occurrence of the stochastic dynamics of the simulated atomization diffusion. By deriving the distribution of the probability density, the stochastic diffusion equation of the simulated atomization diffusion model is finally obtained; S208: Obtain the usage log of the atomizer, obtain the ideal aerosol concentration data packet that meets the target atomization requirement through the usage log to obtain the discrete combination solution of the optimal parameter configuration for the atomizer operation, solve the stochastic diffusion equation of the simulated atomization diffusion model, and obtain the solution of the simulated diffusion equation; S210: Based on the ideal aerosol concentration data packet, construct an ideal atomization diffusion model in the aerosol analysis software, repeat the above steps of stochastic description and derivation to calculate the stochastic diffusion equation of the ideal atomization diffusion model, solve the stochastic diffusion equation of the ideal atomization diffusion model, and obtain the solution of the ideal diffusion equation; S212: If the difference between the solution of the simulated diffusion equation and the solution of the ideal diffusion equation is within the preset diffusion equation solution difference interval, perform operation adjustment on the atomizer based on the upper limit difference of the preset diffusion equation solution difference interval.

[0031] It should be noted that after screening out the discrete combination solution of the optimal parameter configuration whose atomization indexes of the target atomization mode meet the target atomization requirement, it can be configured on the atomizer for the actual operation of the target atomization mode. However, if there are operation errors or improper operations in the atomizer, it will cause the aerosol concentration conversion and aerosol diffusion of the atomizer to fail to achieve the expected effect, resulting in an unsatisfactory atomization quality. Therefore, it is necessary to adjust the operation of the atomizer to ensure the best atomization effect. For this reason, this method conducts a simulation on the atomizer after configuring the discrete parameter solution of the optimal parameter configuration to observe the actual atomization performance of the atomizer. The simulated aerosol concentration data and the diffusion distribution sites of each simulated aerosol concentration data are the data manifestations of the atomization effect of the atomizer. Then, use the aerosol analysis software to perform diffusion analysis and calculation on the simulated aerosol concentration data based on the diffusion distribution sites, so as to generate a simulated atomization diffusion model to visually represent the actual diffusion quality of the aerosol conversion of the atomizer after configuring the discrete combination solution of the optimal parameter configuration. The aerosol diffusion quality of the simulated atomization diffusion model is the basis for judging whether the operation of the atomizer is accurate. If the simulated atomization diffusion model is unreasonable, it means that there are improper phenomena in the operation of the atomizer after configuring the optimal parameter solution, and further adjustment is made based on this.

[0032] It should be noted that due to the randomness and irregularity of the diffusion of aerosol particles in the air environment, the traditional operation tuning method of the atomizer cannot capture the law and directionality of the dynamic diffusion of the aerosol, making it difficult to evaluate whether the aerosol conversion and diffusion quality presented by the simulated atomization diffusion model is ideal. This may lead to a large deviation in the operation tuning of the atomizer, which is not conducive to achieving the best atomization effect of the target atomization mode. Therefore, this method extracts the diffusion state sequence of the simulated atomization diffusion model, and represents the state change of the continuously diffusing aerosol concentration in the simulated atomization diffusion model in the form of a stochastic process based on the diffusion state sequence, so as to obtain the stochastic dynamics of the simulated atomization diffusion. Then, the stochastic differential method is introduced to describe this stochastic dynamics, and the dynamics of the stochastic process expressing the aerosol concentration diffusion is described in the form of a stochastic differential equation. An average field approximation treatment is performed to further show the evolution process of the stochastic dynamic aggregation or diffusion movement of aerosol particles. This process will form a probability density distribution. By deriving this probability density distribution, the simulated atomization diffusion model can be transformed into a stochastic diffusion equation representation. On the one hand, this avoids the calculation error rate of the traditional method for the simulated atomization diffusion model and improves the accuracy of the stochastic analysis of the aerosol dynamic diffusion in the model. On the other hand, it reduces the cumbersome model disassembly operation steps and improves the efficiency of the atomizer operation tuning. If the difference between the solution of the simulated diffusion equation and the solution of the ideal diffusion equation is within the preset diffusion equation solution difference interval, it indicates that the actual aerosol concentration diffusion performance of the atomizer is difficult to meet the ideal state, that is, the improper operation of the atomizer leads to unqualified aerosol diffusion effect. Therefore, the operation of the atomizer is tuned based on the upper limit difference of the preset diffusion equation solution difference interval. Through this method, the aerosol concentration diffusion performance under the actual operation of the atomizer after configuring the optimal parameter solution can be accurately simulated and analyzed, and whether the diffusion effect is ideal is analyzed in the form of a diffusion equation, so as to accurately and efficiently tune the operation state of the atomizer, improve the operation stability and accuracy of the atomizer, and ensure the aerosol diffusion conversion quality.

[0033] The second aspect of the present invention provides a parameter configuration and operation tuning device for an atomizer, as Figure 3 shown. The parameter configuration and operation tuning device includes a memory 31 and a processor 32. A parameter configuration and operation tuning method program for an atomizer is stored in the memory 31. When the parameter configuration and operation tuning method program is executed by the processor 32, the steps of any one of the parameter configuration and operation tuning methods are implemented.

[0034] The third aspect of the present invention provides a computer-readable storage medium. The computer-readable storage medium stores a computer detection program. When the computer detection program is executed by at least one processor, the steps of any one of the parameter configuration and operation tuning methods for an atomizer are implemented.

[0035] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A method for parameter configuration and operation adjustment of an atomizer, characterized in that: The following steps are involved: S102: Obtaining the target atomization demand of the atomizer to sample atomization state particles on multiple atomization time series, obtaining a time series atomization state model by performing propagation prediction, resampling and atomization state estimation of state evolution for each atomization state particle, and performing atomization effect similarity mapping of the time series atomization state under the time series atomization state model based on the atomization decision logic to obtain an expected atomization effect diagram; S104: obtaining an atomization mode function of the atomizer and a mode definition collection benchmark for different atomization effects, recursively splitting different atomization effects from the root node of the atomization mode function based on the mode definition collection benchmark, constructing an atomization mode recognition decision tree, and traversing and identifying branch and trunk nodes of an expected atomization effect graph through the atomization mode recognition decision tree to obtain a target atomization mode; S106: recursively and discretely combine each preset parameter extreme value of each different atomization index item based on the constraint judgment of the highest atomization effect index of the target atomization mode, output multiple sets of parameter configuration discrete combination solutions, construct an aerosol model of each set of parameter configuration discrete combination solutions and perform matching analysis with the expected aerosol model one by one, obtain the optimal parameter configuration discrete combination solution to configure the parameters of the atomizer; S108: The discrete combination solution of the optimal parameter configuration is simulated through the atomizer simulation model to construct a simulated atomization diffusion model, and the state change of the continuous diffusion of the aerosol concentration is randomly expressed according to the simulated atomization diffusion model and the approximation of the mean field is derived to obtain the simulated diffusion equation solution of the simulated atomization diffusion model. The operation of the atomizer is adjusted based on the difference analysis of the solution of the simulated diffusion equation and the solution of the ideal diffusion equation.

2. A parameter configuration and operation adjustment method for an atomizer according to claim 1, characterized in that: The step S102 specifically includes the following steps: Obtaining a target atomization requirement of the atomizer, presetting a plurality of atomization time series and necessary state samples on each atomization time series according to the target atomization requirement, and marking them as atomization state particles; Constructing a state transition model for target atomization requirements, and using the state transition model to predict the propagation of state evolution of necessary atomization state particles in each atomization time series, to generate new atomization state particles; The weight of each atomized state particle is updated by using the state transition model for generating new atomized state particles and normalized to obtain a plurality of weight values, and the atomized state particles are resampled according to the normalized weight values ​​to obtain a new atomized state particle set; Calculating an estimated value of an atomization state that meets a target atomization requirement based on the new atomization state particle set to obtain an estimated value of the atomization state; Repeat the above steps of state evolution propagation prediction of atomization state particles, particle resampling, and atomization state estimation value calculation to obtain the atomization state estimation value on each atomization time series and perform state deduction to obtain the time series atomization state model corresponding to the target atomization demand; An atomization decision logic is established based on the atomization effect of the atomizer for different preset atomization states, and the state similarity between the temporal atomization state matrix expressed by the different preset atomization states and the temporal atomization state model is calculated. The temporal atomization state matrix is ​​mapped to the mapping space according to the KL divergence between the state similarity and the state similarity of the mapping space established by the atomization decision logic to generate an expected atomization effect diagram.

3. A parameter configuration and operation adjustment method for an atomizer according to claim 2, characterized in that: The method establishes an atomization decision logic based on the atomizer for the atomization effects of different preset atomization states, calculates the state similarity between different preset atomization states and the time-series atomization state matrix expressed by the time-series atomization state model, maps the time-series atomization state matrix to the mapping space according to the KL divergence between the state similarity and the state similarity of the mapping space established by the atomization decision logic, and generates an expected atomization effect map, specifically including the following steps: Obtaining a predetermined atomization control strategy of the atomizer for different preset atomization states, and determining a corresponding atomization effect that the atomizer needs to apply for the different preset atomization states according to the predetermined atomization control strategy; Atomization decision logic is established based on the atomizer's corresponding atomization effects required under different preset atomization states as a criterion, different preset atomization states are defined as a first state cluster, a sequential atomization state matrix is ​​extracted through a sequential atomization state model, and the sequential atomization state matrix is ​​defined as a second state cluster; A mapping space is constructed based on the atomization effect, and a similarity matrix of the mapping space is established according to the atomization decision logic, which is defined as a first similarity matrix. A Gaussian distribution method is introduced to calculate the similarity matrix between the first state cluster and the second state cluster, which is defined as a second similarity matrix. A KL divergence threshold is preset, the KL divergence between the first similarity matrix and the second similarity matrix is ​​calculated, and the temporal atomization state matrix is ​​mapped to the mapping space based on the KL divergence until the KL divergence threshold is reached, and an atomization effect graph corresponding to the temporal atomization state that meets the target atomization requirement is obtained, which is defined as the expected atomization effect graph.

4. The parameter configuration and operation adjustment method of an atomizer according to claim 1, characterized in that: The step S104 specifically includes the following steps: Obtaining a design goal of the atomizer, obtaining an atomization mode function of the atomizer according to the design goal, and obtaining a mode definition collection benchmark of the atomizer for different atomization effects through a predetermined atomization control strategy; A minimum Gini index is preset, and the Gini index of different atomization effects output by the atomization mode function is calculated based on the mode definition collection benchmark. The atomization mode function is taken as the root node, and the different atomization effects are split into several sub-atomization effect sets using the Gini index to reach the splitting point of the minimum Gini index starting from the root node; Repeat the above-mentioned splitting step based on the Gini index to recursively split each of the sub-atomization effect sets until all sub-atomization effect sets are recursively completed, and generate an atomization pattern recognition decision tree; Generate an identification sample of the atomization effect based on the expected atomization effect diagram, and perform sample feature value selection on the identification sample according to the target atomization requirement to obtain a feature value range of the identification sample, and extract a branch node pattern diagram of the atomization pattern recognition decision tree, and use each branch terminal node in the branch terminal node pattern diagram as a starting point to perform overall traversal recognition of the atomization effect features of the atomization pattern recognition decision tree; If the current sample feature identified by the identification sample traversal is within the feature value range, the traversal and identification operation is stopped at the branch terminal node or leaf node where the current identification sample is located, and the atomization mode corresponding to the branch terminal node or leaf node where the traversal and identification position is stopped is output, so as to obtain the atomization mode in which the atomizer realizes the expected atomization effect diagram and meets the target atomization requirements, which is marked as the target atomization mode.

5. The parameter configuration and operation adjustment method of an atomizer according to claim 1, characterized in that: The step S106 specifically includes the following steps: According to the design goal of the atomizer, different atomization index items preset by the atomizer for the target atomization mode and preset parameter extreme value ranges of each atomization index item are obtained; Extracting the highest atomization evaluation value of the atomizer according to the target atomization demand, and retrieving the highest atomization effect index of the atomizer after executing the target atomization mode under the highest atomization evaluation value in the big data based on the highest atomization evaluation value; Atomization effect constraint conditions are set according to the highest atomization effect, and the current atomization state of the time series atomization state model in each atomization time series is extracted, and each preset parameter in the preset parameter extreme value interval of each atomization index item is recursively and discretely combined one by one according to the current atomization state; When performing a recursive discrete combination, obtain the actual atomization effect index presented by the discrete combination of each preset parameter of each atomization index item under the current atomization state corresponding to the current recursive node; if the atomization effect constraint condition cannot constrain the actual atomization effect index, retain the discrete combination under the current recursive node; If the atomization effect constraint condition can constrain the actual atomization effect index, the discrete combination under the current recursive node is canceled and the previous recursive node is traced back to perform discrete combinations of other preset parameters until all discrete combination solutions of preset parameters constrained by the atomization effect constraint condition are eliminated, and finally multiple sets of parameter configuration discrete combination solutions are output; Based on the actual atomization effect index, an aerosol model with discrete combination solutions for each set of parameter configurations is constructed, which is defined as the actual aerosol model. Simultaneously, an aerosol model with expected atomization effect diagram is constructed based on the target atomization demand, which is defined as the expected aerosol model. The degree of fit between each actual aerosol model and the expected aerosol model is calculated one by one to obtain multiple degrees of fit. Only the discrete combination solution of parameter configuration corresponding to the actual aerosol model with the maximum degree of fit is extracted and marked as the optimal discrete combination solution of parameter configuration. The parameters of the atomizer are configured based on the discrete combination solution of the optimal parameter configuration.

6. The parameter configuration and operation adjustment method of an atomizer according to claim 1, characterized in that: The step S108 specifically includes the following steps: Obtaining an initial design drawing of the atomizer, constructing an atomizer simulation model according to the initial design drawing, and after parameter configuration is completed, using the atomizer simulation model to perform atomization simulation on the discrete combination solution of the optimal parameter configuration in the aerosol analysis software to obtain a number of simulated aerosol concentration data and diffusion distribution sites of each simulated aerosol concentration data; Based on the diffusion distribution site, a number of simulated aerosol concentration data are analyzed and calculated in the aerosol analysis software to generate a simulated atomization diffusion model, and a diffusion state sequence of the simulated atomization diffusion model is extracted; according to the diffusion state sequence, the state change expressing the continuous diffusion of aerosol concentration in the simulated atomization diffusion model is expressed in the form of a random process to obtain the random dynamics of the simulated atomization diffusion; A stochastic differential method is introduced to describe the random dynamics of the simulated atomization diffusion and a mean field approximation process is performed to obtain a probability density of the random dynamics of the simulated atomization diffusion, and a random diffusion equation of the simulated atomization diffusion model is finally obtained by deriving the distribution of the probability density; Obtain the usage log of the atomizer, obtain the optimal parameter configuration discrete combination solution of the atomizer operation to achieve the ideal aerosol concentration data packet of the target atomization requirement by using the log, solve the random diffusion equation of the simulated atomization diffusion model, and obtain the solution of the simulated diffusion equation; Based on the ideal aerosol concentration data packet, an ideal atomization diffusion model is constructed in the aerosol analysis software, the above-mentioned state random description and derivation steps are repeated to calculate the random diffusion equation of the ideal atomization diffusion model, and the random diffusion equation of the ideal atomization diffusion model is solved to obtain the solution of the ideal diffusion equation; If the difference between the simulated diffusion equation solution and the ideal diffusion equation solution is within a preset diffusion equation solution difference interval, the atomizer is calibrated based on the upper limit difference of the preset diffusion equation solution difference interval.

7. A parameter configuration and operation adjustment device for an atomizer, characterized in that: The parameter configuration and operation adjustment device includes a memory and a processor. The memory stores a parameter configuration and operation adjustment method program for an atomizer. When the parameter configuration and operation adjustment method program is executed by the processor, the parameter configuration and operation adjustment method steps as described in any one of claims 1 to 6 are implemented.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer detection program, and when the computer detection program is executed by at least one processor, it implements the parameter configuration and operation adjustment method of an atomizer according to any one of claims 1 to 6.

Images