Anti-static cleanroom suit micro-pollution blocking regulation system based on electrostatic distribution prediction

Abstract

This invention relates to the field of computer simulation technology, specifically to a micro-contamination blocking and control system for antistatic cleanroom garments based on electrostatic distribution prediction. The system includes an electrostatic data acquisition module that collects various data from the working environment and the surface of the cleanroom garment to construct first electrostatic monitoring characteristic parameters; an electrostatic simulation module comprising a simulation unit and a prediction unit, wherein the simulation unit analyzes the variation patterns of the collected first electrostatic monitoring characteristic parameters and their joint probability distribution, combining an electrostatic physical model and environmental parameters through multidimensional data space clustering analysis to obtain second electrostatic monitoring characteristic parameters; the prediction unit processes the second electrostatic monitoring characteristic parameters based on a mathematical model and simulation methods to obtain electrostatic distribution prediction results; a micro-contamination prediction module dynamically predicts and divides the electrostatic distribution prediction results based on the collected micro-contamination data to obtain a set of micro-contamination distribution areas; and micro-contamination is blocked within these areas based on cleanroom garment data and the first electrostatic monitoring characteristic parameters.

Description

Technical Field

[0001] This invention relates to the field of computer simulation technology, specifically to a micro-contamination blocking and control system for antistatic cleanroom garments based on electrostatic distribution prediction. Background Technology

[0002] Static electricity, a common physical effect, is prevalent in industrial production, power systems, and aerospace, posing hazards to equipment safety, personnel protection, and system reliability. For example, in substations, transient currents generated by static electricity induced in the human body or equipment can interfere with electronic equipment or cause safety accidents; in pneumatic conveying systems, the accumulation of static electricity caused by particle friction can reduce conveying efficiency or create fire hazards. Traditional static electricity monitoring methods mostly rely on finite element simulation or fixed sensors, which suffer from low computational efficiency, poor real-time performance, and insufficient adaptability. Especially in complex electromagnetic environments or dynamic operating conditions, they struggle to accurately capture static electricity characteristics and achieve effective early warning.

[0003] Electrostatic discharge can not only damage sensitive equipment, but also easily attract tiny particles from the air, causing pollution problems. In highly clean environments, even minor contamination can lead to reduced product quality, equipment failure, or even process interruption. This is especially true in the semiconductor and precision manufacturing industries, where micro-contamination is considered a major risk factor.

[0004] In recent years, with the rapid development of technologies such as machine learning and neural networks, electrostatic monitoring and prediction methods have gradually evolved towards intelligence and multi-dimensionality. Methods based on state-space models, grey relational analysis, and deep learning can integrate multi-source monitoring data to uncover the correlation between electrostatic parameters and system performance. Combining field-circuit co-simulation and closed-loop iterative optimization can improve the model's adaptability to dynamic environments. Electric field strength and potential distribution can be calculated using three-dimensional simulation models and the finite element method. However, existing technologies still face challenges such as insufficient extraction of electrostatic signal features and inadequate prediction accuracy due to multi-factor coupling, particularly in real-time monitoring, cross-scenario generalization, and closed-loop feedback mechanisms, requiring further breakthroughs. How to construct an efficient and accurate electrostatic prediction and micro-contamination blocking system remains a key challenge in current research.

[0005] To address this, a micro-contamination blocking and control system for antistatic cleanroom garments based on electrostatic distribution prediction is proposed. Summary of the Invention

[0006] The purpose of this invention is to provide a micro-contamination blocking and control system for antistatic cleanroom garments based on electrostatic distribution prediction. The system comprises: an electrostatic data acquisition module that collects data from the working environment and the surface of the cleanroom garment to construct first electrostatic monitoring characteristic parameters; an electrostatic simulation module including a simulation unit and a prediction unit; a simulation unit that analyzes the variation patterns and joint probability distribution of the collected first electrostatic monitoring characteristic parameters using a multidimensional data space clustering analysis method, combined with an electrostatic physical model and environmental parameters, to obtain second electrostatic monitoring characteristic parameters; a prediction unit that processes the second electrostatic monitoring characteristic parameters based on a mathematical model and simulation method to obtain electrostatic distribution prediction results; a micro-contamination prediction module that dynamically predicts and divides the electrostatic distribution prediction results based on the collected micro-contamination data to obtain a set of micro-contamination distribution areas; and a system that blocks micro-contamination within these areas based on cleanroom garment data and the first electrostatic monitoring characteristic parameters.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A micro-contamination blocking and control system for antistatic cleanroom garments based on electrostatic distribution prediction includes:

[0009] The electrostatic data acquisition module collects various data from the working environment and the surface of the cleanroom garment in real time through several types of sensors, and constructs the first electrostatic monitoring characteristic parameters;

[0010] Furthermore, the electrostatic data acquisition module includes several types of sensor arrays, including a charge detection unit, an environmental parameter sensing unit, and an equipment status sensing unit. The charge detection unit collects the amount and distribution of charge on the surface of the cleanroom garment and the equipment shell in real time. The environmental parameter sensing unit monitors the temperature, humidity, air pressure, gas flow rate, and concentration of airborne particles in the working area. The equipment status sensing unit captures the motion acceleration and vibration frequency of the operating equipment through inertial measurement elements.

[0011] Furthermore, the collected sensor data is processed to construct the first electrostatic monitoring characteristic parameters, which include environmental monitoring characteristic parameters and cleanroom garment monitoring characteristic parameters. The environmental monitoring characteristic parameters include the temperature, humidity and concentration of airborne particles in the working environment, while the cleanroom garment monitoring characteristic parameters include surface-to-surface resistance, surface-to-grounding resistance, surface potential and fabric capacitance.

[0012] The electrostatic simulation module includes a simulation unit and a prediction unit. The simulation unit analyzes the variation patterns and joint probability distribution of the first electrostatic monitoring characteristic parameters collected, and combines the electrostatic physical model and environmental parameters with a multidimensional data space clustering analysis method to obtain the second electrostatic monitoring characteristic parameters. The prediction unit processes the second electrostatic monitoring characteristic parameters based on a mathematical model and simulation method to obtain the electrostatic distribution prediction results.

[0013] Furthermore, the environmental monitoring characteristic parameters and cleanroom garment monitoring characteristic parameters are divided into several corresponding monitoring characteristic parameter pairs according to the collection time; the multidimensional data spatial clustering analysis method includes data dimensionality reduction, cluster analysis, and dynamic correlation analysis;

[0014] Furthermore, the simulation unit employs a multidimensional data space clustering analysis method to perform dimensionality reduction and clustering analysis on the monitoring feature parameter pairs, and performs dynamic correlation analysis based on the time series to obtain the joint probability distribution of the feature parameters of environmental monitoring feature parameters and cleanroom garment monitoring feature parameters. Through Monte Carlo simulation under environmental parameter constraints, the monitoring feature parameter pairs and the joint probability distribution of feature parameters are mapped to a three-dimensional feature space to generate a second electrostatic monitoring feature parameter containing spatial gradient distribution and risk probability.

[0015] Furthermore, based on finite element numerical simulation and stochastic process theory, the prediction unit establishes a partial differential equation model of electrostatic diffusion that includes the second environmental monitoring characteristic parameters. The second electrostatic monitoring characteristic parameters are transformed into a gridded simulation input electrostatic diffusion partial differential equation model through a discretization method. An implicit difference algorithm is used to iteratively solve the electrostatic potential field distribution. The boundary conditions are dynamically adjusted in combination with environmental parameters, and the predicted electrostatic distribution results including spatial potential gradient, charge density and discharge probability are output.

[0016] The micro-contamination prediction module dynamically predicts and divides the electrostatic distribution prediction results based on the collected micro-contamination data to obtain a set of micro-contamination distribution areas; and blocks micro-contamination within the area based on cleanroom garment data and the first electrostatic monitoring characteristic parameters.

[0017] Furthermore, the collected micro-pollution data includes air particle monitoring data, air microbial monitoring data, surface pollution monitoring data, and chemical pollution monitoring data;

[0018] Furthermore, the micro-pollution prediction module uses spatial overlay analysis to fuse micro-pollution data and electrostatic distribution prediction results on a unified spatial grid based on the weighting rules of pollution concentration gradient and discharge probability, thereby obtaining a pollution-electrostatic coupling index. The pollution-electrostatic coupling index is then segmented using an adaptive threshold to dynamically divide a set of micro-pollution distribution areas containing different degrees of micro-pollution severity.

[0019] Furthermore, the wearable status monitoring unit acquires data on the antistatic cleanroom garment in the working environment in real time during the working state, and constructs cleanroom garment data, including static data and dynamic data. The static data includes material dielectric constant, surface resistivity, wear fit and local friction coefficient, while the dynamic data includes cleanroom garment positioning data, electrostatic decay time, surface potential and discharge current.

[0020] Furthermore, the collected cleanroom garment data is processed and cleaned to construct a micro-contamination blocking and control vector.

[0021] Furthermore, based on the location of the antistatic cleanroom garment in the set of micro-contamination distribution areas, data of the corresponding micro-contamination distribution areas are obtained, including spatial potential gradient, charge density, discharge probability, and contamination-electrostatic coupling index, and a micro-contamination blocking and control evaluation vector is constructed.

[0022] Furthermore, the difference between the real-time micro-pollution blocking and control evaluation vector and the corresponding micro-pollution blocking and control evaluation vector in the electrostatic-free state in historical data is calculated, and the micro-pollution blocking and control vector is adjusted according to the difference.

[0023] Furthermore, after obtaining the difference, the micro-pollution blocking control vector is corrected online based on the difference and mapping rules; through a closed-loop feedback mechanism, the charge state of the cleanroom garment surface and environmental sensing data are continuously read, and new differences are calculated for correction, so as to achieve coordinated blocking of pollution diffusion paths and electrostatic risk areas.

[0024] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0025] 1. Dynamic correlation analysis further captures the abrupt changes and coupling evolution of parameters over time, enhancing the sensitivity to sudden electrostatic disturbances; Monte Carlo simulation under environmental parameter constraints can fully quantify input uncertainties and approximate the joint probability distribution through a large number of random samples, enabling the prediction results to have reliable risk probability assessment; the gradient distribution and risk probability generated after mapping the monitoring feature pairs and their joint probability distribution to the three-dimensional feature space can not only intuitively reflect the spatial hotspot layout, but also support refined decision-making, which is conducive to the deployment of anti-static strategies tailored to local conditions.

[0026] 2. By overlaying and analyzing multi-source data on airborne particles, microorganisms, surface and chemical pollution with electrostatic distribution prediction results in the same spatial grid, multi-level spatial coupling can be achieved, intuitively revealing the spatial distribution patterns of pollution and electrostatics. Based on the weighting rules of pollution concentration gradient and discharge probability, a pollution-electrostatic coupling index can be constructed, which can quantify the interaction strength between the two and provide a quantitative basis for risk assessment. By segmenting the pollution-electrostatic coupling index through adaptive thresholds, different levels of pollution severity can be reflected, and the distribution areas of micro-pollution can be accurately delineated.

[0027] 3. By calculating the difference between the real-time micro-pollution blocking control evaluation vector and the corresponding region under historical electrostatic-free conditions, the system can quantify the extent to which the current blocking effect deviates from the ideal state. Based on this difference, the driving update mechanism can adaptively optimize the blocking control vector, so that the strategy continuously fits the current pollution distribution and electrostatic charge state, significantly improving the system's response speed and sensitivity. By using the gradient descent algorithm to iteratively optimize the blocking strategy combination, it can efficiently converge to the optimal electrostatic discharge and pollution path blocking measures, minimizing the decision-making time. Attached Figure Description

[0028] Figure 1 A schematic diagram of the micro-contamination blocking and control system for antistatic cleanroom garments based on electrostatic distribution prediction provided in an embodiment of the present invention;

[0029] Figure 2 This is a flowchart for obtaining the second electrostatic monitoring characteristic parameters provided in an embodiment of the present invention;

[0030] Figure 3 This is a schematic diagram of the structure for obtaining a set of micro-pollution distribution areas provided in an embodiment of the present invention. Detailed Implementation

[0031] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] By predicting static hotspots and areas of static accumulation, potential electrostatic discharge risks can be identified in advance, providing a basis for designing targeted protective measures. This is particularly important in fields with high environmental cleanliness requirements, such as cleanrooms, electronics manufacturing, semiconductors, and biomedicine, because electrostatic discharge can not only damage sensitive equipment but also easily attract fine particles from the air, leading to pollution problems.

[0033] Microcontamination refers to extremely fine particles, microorganisms, organic or inorganic particulate matter, and other pollutants present in the environment. These pollutants have a wide range of sources, including shed skin flakes, clothing fibers, and dust generated during processing. Due to electrostatic adsorption, particles from the surrounding air are more likely to accumulate at electrostatic hotspots. To improve protection against electrostatic contamination during production, a semiconductor manufacturing company introduced the microcontamination blocking and control system for antistatic cleanroom garments based on electrostatic distribution prediction, as provided in this invention. Customized antistatic cleanroom garments were then manufactured. The system structure is as follows: Figure 1 As shown, the specific implementation method is as follows:

[0034] The electrostatic data acquisition module collects various data from the working environment and the surface of the cleanroom garment in real time through several types of sensors, and constructs the first electrostatic monitoring characteristic parameters.

[0035] Furthermore, the electrostatic data acquisition module includes several types of sensor arrays, including a charge detection unit, an environmental parameter sensing unit, and an equipment status sensing unit. The charge detection unit collects the amount and distribution of charge on the surface of the cleanroom garment and the equipment shell in real time. The environmental parameter sensing unit monitors the temperature, humidity, air pressure, gas flow rate, and concentration of airborne particles in the working area. The equipment status sensing unit captures the motion acceleration and vibration frequency of the operating equipment through inertial measurement elements.

[0036] Furthermore, the collected sensor data is processed to construct the first electrostatic monitoring characteristic parameters. These parameters include environmental monitoring characteristic parameters and cleanroom garment monitoring characteristic parameters. The environmental monitoring characteristic parameters include the working environment's temperature, humidity, and airborne particle concentration. The cleanroom garment monitoring characteristic parameters include surface-to-surface resistance, surface-to-grounding resistance, surface potential, and fabric capacitance. Table 1 shows some of the collected data.

[0037] Table 1. Partial Data Collection

[0038]

[0039] By collecting electrostatic accumulation, electrical, and motion data from three types of sensing units—charge detection unit, environmental parameter sensing unit, and equipment status sensing unit—the system can comprehensively and accurately collect the electrostatic charge distribution, resistance characteristics, and dynamic operating conditions of the cleanroom garment surface, working environment, and equipment shell, significantly improving the sensitivity and reliability of electrostatic anomaly detection.

[0040] The electrostatic simulation module includes a simulation unit and a prediction unit. The simulation unit analyzes the variation patterns and joint probability distribution of the first electrostatic monitoring characteristic parameters collected, and combines the electrostatic physical model and environmental parameters with a multidimensional data space clustering analysis method to obtain the second electrostatic monitoring characteristic parameters. The prediction unit processes the second electrostatic monitoring characteristic parameters based on a mathematical model and simulation method to obtain the electrostatic distribution prediction results.

[0041] Furthermore, the environmental monitoring characteristic parameters and cleanroom garment monitoring characteristic parameters are divided into several corresponding monitoring characteristic parameter pairs according to the collection time; the multidimensional data spatial clustering analysis method includes data dimensionality reduction, cluster analysis, and dynamic correlation analysis;

[0042] Furthermore, the simulation unit employs a multidimensional data space clustering analysis method to perform dimensionality reduction and clustering analysis on the monitoring feature parameter pairs, and performs dynamic correlation analysis based on the time series to obtain the joint probability distribution of the feature parameters of environmental monitoring feature parameters and cleanroom garment monitoring feature parameters. Through Monte Carlo simulation under environmental parameter constraints, the monitoring feature parameter pairs and the joint probability distribution of feature parameters are mapped to a three-dimensional feature space to generate a second electrostatic monitoring feature parameter containing spatial gradient distribution and risk probability.

[0043] Furthermore, the environmental monitoring feature parameters and the cleanroom garment monitoring feature parameters are first divided into several corresponding monitoring feature parameter pairs according to the collection time, so as to ensure that a one-to-one mapping relationship between the environment and the garment status is obtained at the same time point or within the same time window. This time series segmentation method based on sliding window or time-by-time pairing is a common practice in time series data mining, which helps with subsequent correlation analysis and feature comparison.

[0044] Furthermore, in multidimensional data space clustering analysis methods, the first step is to reduce the dimensionality of high-dimensional monitoring feature parameter pairs. A common approach is to combine principal component analysis with K-means clustering, using iterative feedback to optimize the dimensionality reduction space so that the low-dimensional representation can retain the maximum variance information while enhancing cluster separation. Subsequently, clustering analysis is performed, using subspace clustering algorithms to discover local groups in the dimensionality-reduced subspace, or density-based methods such as DBSCAN to identify non-convex cluster structures. Finally, through dynamic correlation analysis, cluster labels obtained in continuous time windows are matched, and combined with dynamic time warping or cross-correlation analysis methods, the coupling evolution of environmental monitoring feature parameters and cleanroom garment monitoring feature parameters over time is captured.

[0045] Furthermore, based on the above multidimensional clustering results, the simulation unit obtains the joint probability distribution of environmental monitoring characteristic parameters and cleanroom garment monitoring characteristic parameters through the cluster center trajectory and similarity measurement of the time series. This process is equivalent to constructing a distribution estimation model of multivariate random variables in the feature space, which can quantify the joint variation characteristics of the two under different working conditions.

[0046] Furthermore, under the constraints of environmental parameters (such as temperature, humidity, and airflow rate), Monte Carlo simulation is used to perform large-scale random sampling of the monitoring feature parameter pairs and their joint probability distribution. By repeatedly generating sample vectors that conform to this joint distribution and inputting them into the electrostatic simulation model, the samples generated by Monte Carlo are mapped to a three-dimensional feature space. The electrostatic potential gradient and corresponding risk probability are calculated at each grid point, forming a second electrostatic monitoring feature parameter that includes the spatial gradient distribution of the potential and the risk probability. This makes the high-dimensional complex data intuitive and operable in visualization and control decision-making. Table 2 shows the second electrostatic monitoring feature parameters for some locations. The overall process is as follows: Figure 2 As shown.

[0047] Table 2. Characteristic parameters of the second electrostatic monitoring

[0048] Three-dimensional feature space location Spatial gradient of electric potential (V / cm) Risk probability (0.8,-0.2,1.5) 0.50 0.06 (0.9,-0.3,1.6) 0.55 0.05 (-0.5,1.2,0.1) 0.14 0.02 (1.1,-0.5,1.9) -0.23 0.04

[0049] Dynamic correlation analysis further captures the abrupt changes and coupling evolution of parameters over time, enhancing sensitivity to sudden electrostatic disturbances; Monte Carlo simulation under environmental parameter constraints can fully quantify input uncertainties and approximate the joint probability distribution through a large number of random samples, enabling the prediction results to have reliable risk probability assessment; the gradient distribution and risk probability generated after mapping the monitoring feature pairs and their joint probability distribution to the three-dimensional feature space can not only intuitively reflect the spatial hotspot layout, but also support refined decision-making, which is conducive to the deployment of anti-static strategies tailored to local conditions.

[0050] The prediction unit establishes a partial differential equation model of electrostatic diffusion that includes environmental monitoring characteristic parameters based on finite element numerical simulation and stochastic process theory. The second electrostatic monitoring characteristic parameter is transformed into a gridded simulation through discretization and input into the electrostatic diffusion partial differential equation model. The electrostatic potential field distribution is solved iteratively using an implicit difference algorithm. The boundary conditions are dynamically adjusted in combination with environmental parameters, and the predicted electrostatic distribution results including spatial potential gradient, charge density and discharge probability are output.

[0051] By incorporating environmental monitoring characteristic parameters into the electrostatic diffusion partial differential equation model and combining it with finite element numerical simulation, the system can accurately characterize the electrostatic potential field distribution under complex geometric and multi-physics coupling conditions. Modeling the diffusion equation using stochastic process theory fully reflects the randomness and uncertainty brought about by environmental disturbances and material heterogeneity. Iterative solving using an implicit difference algorithm achieves unconditional stability, supports rapid simulations with large time steps, and significantly improves computational efficiency and numerical robustness. The final high-precision prediction results of spatial potential gradient, charge density, and discharge probability provide a scientific basis for the refined deployment of micro-pollution blocking strategies.

[0052] The micro-contamination prediction module dynamically predicts and divides the electrostatic distribution prediction results based on the collected micro-contamination data to obtain a set of micro-contamination distribution areas; and blocks micro-contamination within the area based on cleanroom garment data and the first electrostatic monitoring characteristic parameters.

[0053] Furthermore, the collected micro-pollution data includes air particle monitoring data, air microbial monitoring data, surface pollution monitoring data, and chemical pollution monitoring data;

[0054] Furthermore, the micro-pollution prediction module uses spatial overlay analysis to fuse micro-pollution data with electrostatic distribution prediction results on a unified spatial grid based on a weighted allocation rule of pollution concentration gradient and discharge probability, obtaining a pollution-electrostatic coupling index. This index is then segmented using an adaptive threshold, dynamically dividing the distribution area into sets of micro-pollution regions containing different degrees of micro-pollution severity. The specific process is as follows: Figure 3 As shown.

[0055] Furthermore, the micro-pollution monitoring data and electrostatic distribution prediction results are resampled to the same grid cell, and then fused according to the pollution concentration gradient, the discharge probability of the electrostatic distribution prediction results, and the corresponding weights, as shown in the formula:

[0056]

[0057] Among them, K ij The coupling index is represented by Q, which represents the type of microcontamination data, and ω represents the coupling index. q This represents the weighting coefficient for the q-th type of micro-contamination data. The max||C represents the Euclidean norm of the pollution gradient of the grid cell at position (i,j) for the q-th type of micro-contamination data. q || represents the maximum gradient of the q-th micro-contamination data across all grid cells, ω represents the weighting coefficient of the discharge probability, and R ij This represents the discharge probability of the grid cell at position (i,j).

[0058] Furthermore, to distinguish between high-coupling and low-coupling regions, an adaptive threshold segmentation algorithm is adopted to dynamically calculate the threshold based on the mean and standard deviation of the coupling index in the neighborhood of each grid cell. The adaptive threshold can overcome the failure of the global threshold in a non-uniform background and maintain the segmentation accuracy.

[0059] Furthermore, 4-neighbor (or 8-neighbor) connectivity detection is performed to identify several connected sub-regions, each of which is a micro-contamination distribution area.

[0060] By overlaying and analyzing multi-source data on airborne particles, microorganisms, surface and chemical pollution with electrostatic distribution prediction results in the same spatial grid, multi-level spatial coupling can be achieved, intuitively revealing the co-occurrence patterns and spatial distribution laws of pollution and electrostatics. Based on the weighting rules of pollution concentration gradient and discharge probability, a pollution-electrostatic coupling index can be constructed, which can quantify the interaction strength between the two and provide a quantitative basis for risk assessment. By segmenting the pollution-electrostatic coupling index through adaptive thresholds, different pollution severity levels can be reflected, and the distribution areas of micro-pollution can be accurately delineated.

[0061] Furthermore, the wearable status monitoring unit acquires data on the antistatic cleanroom garment in the working environment in real time during the working state, and constructs cleanroom garment data, including static data and dynamic data. The static data includes material dielectric constant, surface resistivity, wear fit and local friction coefficient, while the dynamic data includes cleanroom garment positioning data, electrostatic decay time, surface potential and discharge current.

[0062] Furthermore, the collected cleanroom garment data is processed and cleaned to construct a micro-contamination blocking and control vector.

[0063] Furthermore, the collected cleanroom garment data undergoes time-series synchronization and denoising, outlier detection, and feature normalization. The cleaned static and dynamic features are concatenated into a high-dimensional vector, which is then mapped to a low-dimensional control subspace using PCA or Autoencoder to obtain a micro-contamination blocking control vector, facilitating online calculation and strategy matching.

[0064] By integrating static and dynamic electrical and positioning data of cleanroom garments in real time, the system can not only comprehensively characterize the antistatic performance of antistatic cleanroom garments in actual working environments, but also generate precise micro-contamination blocking and control vectors after data cleaning and high-dimensional vectorization, significantly improving the model's sensitivity to the risk of electrostatic and particle coupling.

[0065] Furthermore, based on the location of the antistatic cleanroom garment in the set of micro-contamination distribution areas, data of the corresponding micro-contamination distribution areas are obtained, including spatial potential gradient, charge density, discharge probability, and contamination-electrostatic coupling index, and a micro-contamination blocking and control evaluation vector is constructed.

[0066] Furthermore, the difference between the real-time micro-pollution blocking and control evaluation vector and the corresponding micro-pollution blocking and control evaluation vector in the electrostatic-free state in historical data is calculated, and the micro-pollution blocking and control vector is adjusted according to the difference.

[0067] Furthermore, by using a closed-loop feedback mechanism to match the surface charge state of the cleanroom garment with changes in environmental parameters in real time, and by employing a gradient descent algorithm to optimize the combination of blocking strategies, the synergistic blocking of contamination diffusion paths and electrostatic risk areas is achieved.

[0068] Furthermore, based on the location of the antistatic cleanroom garment within the set of micro-contamination distribution areas, the spatial potential gradient, charge density, discharge probability, and contamination-electrostatic coupling index within the grid cell are obtained. A micro-contamination blocking and control evaluation vector is constructed by vector splicing. This evaluation vector integrates a large amount of data, providing a comprehensive state description for subsequent difference calculation and control optimization.

[0069] Furthermore, after obtaining the difference, the micro-pollution blocking and control vector is corrected online based on the difference and mapping rules. The system continuously reads the charge state of the cleanroom garment surface and environmental sensor data through a closed-loop feedback mechanism, and calculates new differences for correction. The mapping rules refer to finding the corrected data in the historical data based on the real-time micro-pollution blocking and control evaluation vector and making corrections. The search method can be based on the similarity between data, the similarity of events, and the similarity of data in space.

[0070] By calculating the difference between the real-time micro-pollution blocking control evaluation vector and the corresponding region under historical electrostatic undisturbed conditions, the system can quantify the extent to which the current blocking effect deviates from the ideal state. This difference-based driven update mechanism can adaptively optimize the blocking control vector, ensuring that the strategy continuously conforms to the normal pollution distribution and electrostatic charge state, significantly improving the system's response speed and sensitivity. By using the gradient descent algorithm to iteratively optimize the blocking strategy combination, it can efficiently converge to the optimal electrostatic discharge and pollution path blocking measures, minimizing the decision-making time.

[0071] The antistatic cleanroom garment micro-contamination blocking and control system based on electrostatic distribution prediction provided by this invention achieves accurate data acquisition by the electrostatic acquisition module and constructs the first electrostatic monitoring characteristic parameters, realizing data preprocessing; the electrostatic simulation module realizes the construction of the electrostatic model through data simulation and realizes the prediction of electrostatic distribution through the prediction unit, providing data support for subsequent micro-contamination blocking; the micro-contamination prediction module combines the electrostatic distribution prediction results with the acquired micro-contamination data, divides the micro-contamination distribution area, and performs micro-contamination blocking, improving the accuracy of blocking and shortening the blocking time.

[0072] Example 2:

[0073] To prevent static electricity and micro-contamination, an electronic component company introduced the antistatic cleanroom garment micro-contamination blocking and control system based on static electricity distribution prediction provided by this invention. The specific implementation method is as follows:

[0074] The electrostatic data acquisition module collects various data from the working environment and the surface of the cleanroom garment in real time through several types of sensors, and constructs the first electrostatic monitoring characteristic parameters.

[0075] Furthermore, the electrostatic data acquisition module includes several types of sensor arrays, including a charge detection unit, an environmental parameter sensing unit, and an equipment status sensing unit. The charge detection unit collects the amount and distribution of charge on the surface of the cleanroom garment and the equipment shell in real time. The environmental parameter sensing unit monitors the temperature, humidity, air pressure, gas flow rate, and concentration of airborne particles in the working area. The equipment status sensing unit captures the motion acceleration and vibration frequency of the operating equipment through inertial measurement elements.

[0076] Table 3. Partial Data Collection

[0077]

[0078] Furthermore, the collected sensor data is processed to construct the first electrostatic monitoring characteristic parameters. These parameters include environmental monitoring characteristic parameters and cleanroom garment monitoring characteristic parameters. The environmental monitoring characteristic parameters include the working environment's temperature, humidity, and airborne particle concentration. The cleanroom garment monitoring characteristic parameters include surface-to-surface resistance, surface-to-grounding resistance, surface potential, and fabric capacitance. Table 3 shows some of the collected data.

[0079] The electrostatic simulation module includes a simulation unit and a prediction unit. The simulation unit analyzes the variation patterns and joint probability distribution of the first electrostatic monitoring characteristic parameters collected, and combines the electrostatic physical model and environmental parameters with a multidimensional data space clustering analysis method to obtain the second electrostatic monitoring characteristic parameters. The prediction unit processes the second electrostatic monitoring characteristic parameters based on a mathematical model and simulation method to obtain the electrostatic distribution prediction results.

[0080] Furthermore, the environmental monitoring characteristic parameters and cleanroom garment monitoring characteristic parameters are divided into several corresponding monitoring characteristic parameter pairs according to the collection time; the multidimensional data spatial clustering analysis method includes data dimensionality reduction, cluster analysis, and dynamic correlation analysis;

[0081] Furthermore, the simulation unit employs a multidimensional data space clustering analysis method to perform dimensionality reduction and clustering analysis on the monitoring feature parameter pairs, and performs dynamic correlation analysis based on the time series to obtain the joint probability distribution of the feature parameters of environmental monitoring feature parameters and cleanroom garment monitoring feature parameters. Through Monte Carlo simulation under environmental parameter constraints, the monitoring feature parameter pairs and the joint probability distribution of feature parameters are mapped to a three-dimensional feature space to generate a second electrostatic monitoring feature parameter containing spatial gradient distribution and risk probability.

[0082] Furthermore, based on finite element numerical simulation and stochastic process theory, the prediction unit establishes a partial differential equation model of electrostatic diffusion that includes the second environmental monitoring characteristic parameters. The second electrostatic monitoring characteristic parameters are transformed into a gridded simulation input electrostatic diffusion partial differential equation model through a discretization method. An implicit difference algorithm is used to iteratively solve the electrostatic potential field distribution. The boundary conditions are dynamically adjusted in combination with environmental parameters, and the predicted electrostatic distribution results including spatial potential gradient, charge density and discharge probability are output.

[0083] The micro-contamination prediction module dynamically predicts and divides the electrostatic distribution prediction results based on the collected micro-contamination data to obtain a set of micro-contamination distribution areas; and blocks micro-contamination within the area based on cleanroom garment data and the first electrostatic monitoring characteristic parameters.

[0084] Furthermore, the collected micro-pollution data includes air particle monitoring data, air microbial monitoring data, surface pollution monitoring data, and chemical pollution monitoring data;

[0085] Table 4. Pollution-electrostatic coupling indices for some grid cells

[0086] Grid cell number Pollution-electrostatic coupling index (normalized value) Grid-00223 0.75 Grid-01254 0.41 Grid-14983 0.92 Grid-03024 0.80

[0087] Furthermore, the micro-pollution prediction module uses spatial overlay analysis to fuse micro-pollution data with electrostatic distribution prediction results on a unified spatial grid based on a weighted allocation rule of pollution concentration gradient and discharge probability, obtaining a pollution-electrostatic coupling index. This index is then segmented using an adaptive threshold, dynamically dividing the distribution area into sets of micro-pollution regions containing different degrees of severity. Table 4 shows the pollution-electrostatic coupling index for some grid cells.

[0088] Furthermore, the wearable status monitoring unit acquires data on the antistatic cleanroom garment in the working environment in real time during the working state, and constructs cleanroom garment data, including static data and dynamic data. The static data includes material dielectric constant, surface resistivity, wear fit and local friction coefficient, while the dynamic data includes cleanroom garment positioning data, electrostatic decay time, surface potential and discharge current.

[0089] Furthermore, the collected cleanroom garment data is processed and cleaned to construct a micro-contamination blocking and control vector.

[0090] Furthermore, based on the location of the antistatic cleanroom garment within the set of micro-contamination distribution areas, data on the corresponding micro-contamination distribution areas are obtained, including spatial potential gradient, charge density, discharge probability, and contamination-electrostatic coupling index, and a micro-contamination blocking and control evaluation vector is constructed; among which, data such as spatial potential gradient and charge density can be obtained through the first electrostatic monitoring characteristic parameter.

[0091] Furthermore, the difference between the real-time micro-pollution blocking and control evaluation vector and the corresponding micro-pollution blocking and control evaluation vector in the electrostatic-free state in historical data is calculated, and the micro-pollution blocking and control vector is adjusted according to the difference.

[0092] Furthermore, after obtaining the difference, the micro-pollution blocking control vector is corrected online based on the difference and mapping rules; through a closed-loop feedback mechanism, the charge state of the cleanroom garment surface and environmental sensing data are continuously read, and new differences are calculated for correction, so as to achieve coordinated blocking of pollution diffusion paths and electrostatic risk areas.

[0093] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A micro-contamination blocking and control system for antistatic cleanroom garments based on electrostatic distribution prediction, characterized in that, include: The electrostatic data acquisition module collects various data from the working environment and the surface of the cleanroom garment in real time through several types of sensors, and constructs the first electrostatic monitoring characteristic parameters. The electrostatic simulation module includes a simulation unit and a prediction unit. The simulation unit uses a multidimensional data space clustering analysis method to perform dimensionality reduction and clustering analysis on the first electrostatic monitoring characteristic parameters. Based on the cluster center trajectory and similarity metric of the time series, it obtains the joint probability distribution of the characteristic parameters of environmental monitoring and cleanroom garment monitoring. Through large-scale random sampling using Monte Carlo simulation under environmental parameter constraints, it maps the monitoring characteristic parameter pairs and the joint probability distribution of the characteristic parameters to a three-dimensional feature space, generating a second electrostatic monitoring characteristic parameter that includes spatial gradient distribution and risk probability. The prediction unit processes the second electrostatic monitoring characteristic parameter based on mathematical models and simulation methods to obtain the electrostatic distribution prediction results. The multidimensional data space clustering analysis method includes data dimensionality reduction, clustering analysis, and dynamic correlation analysis. The micro-contamination prediction module dynamically predicts and divides the electrostatic distribution prediction results based on the collected micro-contamination data to obtain a set of micro-contamination distribution areas; and blocks micro-contamination within the area based on cleanroom garment data and the first electrostatic monitoring characteristic parameters.

2. The antistatic cleanroom garment micro-contamination blocking and control system based on electrostatic distribution prediction according to claim 1, characterized in that: The electrostatic data acquisition module includes several types of sensor arrays, including a charge detection unit, an environmental parameter sensing unit, and an equipment status sensing unit. The charge detection unit collects the amount and distribution of charge on the surface of the cleanroom garment and the outer shell of the equipment in real time. The environmental parameter sensing unit monitors the temperature, humidity, air pressure, gas flow rate, and concentration of airborne particles in the working area. The equipment status sensing unit captures the motion acceleration and vibration frequency of the operating equipment through inertial measurement elements. The collected sensor data is processed to construct the first electrostatic monitoring characteristic parameters, which include environmental monitoring characteristic parameters and cleanroom garment monitoring characteristic parameters. The environmental monitoring characteristic parameters include the temperature, humidity and concentration of airborne particles in the working environment, while the cleanroom garment monitoring characteristic parameters include the surface resistance to surface, the surface resistance to grounding point, the surface potential and the fabric capacitance.

3. The antistatic cleanroom garment micro-contamination blocking and control system based on electrostatic distribution prediction according to claim 1, characterized in that: The prediction unit establishes a partial differential equation model of electrostatic diffusion that includes the second environmental monitoring characteristic parameters based on finite element numerical simulation and stochastic process theory. The second electrostatic monitoring characteristic parameters are transformed into a gridded simulation through discretization and input into the electrostatic diffusion partial differential equation model. An implicit difference algorithm is used to iteratively solve the electrostatic potential field distribution. The boundary conditions are dynamically adjusted in combination with environmental parameters, and the predicted electrostatic distribution results including spatial potential gradient, charge density and discharge probability are output.

4. The antistatic cleanroom garment micro-contamination blocking and control system based on electrostatic distribution prediction according to claim 1, characterized in that: The collected micro-pollution data includes air particle monitoring data, air microbial monitoring data, surface pollution monitoring data, and chemical pollution monitoring data; The micro-pollution prediction module uses spatial overlay analysis to fuse micro-pollution data and electrostatic distribution prediction results on a unified spatial grid based on the weighting rules of pollution concentration gradient and discharge probability, thereby obtaining a pollution-electrostatic coupling index. The pollution-electrostatic coupling index is segmented by an adaptive threshold, dynamically dividing a set of micro-pollution distribution areas containing different degrees of micro-pollution severity.

5. The antistatic cleanroom garment micro-contamination blocking and control system based on electrostatic distribution prediction according to claim 1, characterized in that: The wearable status monitoring unit acquires real-time data of the antistatic cleanroom garment in the working environment under working conditions, and constructs cleanroom garment data, including static data and dynamic data. The static data includes material dielectric constant, surface resistivity, wear fit and local friction coefficient, while the dynamic data includes cleanroom garment positioning data, electrostatic decay time, surface potential and discharge current. Data processing and cleaning were performed on the collected cleanroom garment data to construct a micro-contamination blocking and control vector.

6. The antistatic cleanroom garment micro-contamination blocking and control system based on electrostatic distribution prediction according to claim 1, characterized in that: Based on the location of the antistatic cleanroom garment in the set of micro-contamination distribution areas, the data of the corresponding micro-contamination distribution areas are obtained, including spatial potential gradient, charge density, discharge probability, and contamination-electrostatic coupling index, and a micro-contamination blocking and control evaluation vector is constructed. Calculate the difference between the real-time micro-pollution blocking and control evaluation vector and the corresponding micro-pollution blocking and control evaluation vector in the electrostatic-free state in historical data, and adjust the micro-pollution blocking and control vector according to the difference; After obtaining the difference, the micro-pollution blocking control vector is corrected online based on the difference and mapping rules. Through a closed-loop feedback mechanism, the charge state of the cleanroom garment surface and environmental sensing data are continuously read, and new differences are calculated for correction, so as to achieve coordinated blocking of pollution diffusion paths and electrostatic risk areas.

Images