Adaptive adjustment system and method for air circulation function of electronic industry clean workshop

Abstract

This invention discloses an adaptive adjustment system and method for air circulation function in cleanrooms for the electronics industry. The adjustment system includes a three-dimensional grid monitoring module, a data processing module, a visualization module, a computer vision module, an intelligent decision-making module, and an air circulation control module. The adjustment method includes three-dimensional grid sensor layout and monitoring, data processing, data visualization, image feature extraction, abnormal area early warning, air circulation parameter decision-making, adaptive adjustment execution, and closed-loop feedback optimization. By constructing a continuous spatial distribution model from real-time monitoring data of the three-dimensional grid sensors and converting it into a dynamic, visualized three-dimensional heat map, the correlation between environmental characteristics and air circulation control quantities is further established. Based on local pollution anomalies, airflow control parameters are output, which can significantly improve the system's adjustment accuracy for complex operating conditions and realize adaptive adjustment and continuous evolution of air circulation function.

Description

Technical Field

[0001] This invention relates to the field of control system technology, and in particular to an adaptive adjustment system and method for air circulation function in cleanrooms for the electronics industry. Background Technology

[0002] Cleanrooms in the electronics industry are production spaces specifically designed and constructed to meet the cleanliness requirements of modern high-precision electronic products. Their core objective is to control airborne particulate matter below a specific concentration, preventing it from settling on integrated circuit chips and other products, thus avoiding short circuits, performance degradation, or even product failure. Therefore, the cleanliness of cleanrooms in the electronics industry directly impacts the performance and yield of chips, integrated circuits, and other products. Cleanrooms in the electronics industry need to maintain ISO Class 5 or even higher cleanliness standards, and achieving this standard through traditional air circulation systems often relies on enormous airflow volumes and extremely high fan energy consumption.

[0003] Traditional air circulation systems in cleanrooms for the electronics industry are often set up according to fixed operating conditions or rely on simple sensors for monitoring and manual control. This approach has significant drawbacks: First, single-point monitoring cannot accurately reflect the distribution and dynamic changes of particulate matter concentration in three-dimensional space, making it difficult to detect localized pollution anomalies. Second, air circulation systems with fixed operating parameters cannot promptly and accurately adjust for changes in particulate matter concentration caused by factors such as equipment load variations, process differences, and varying personnel activity levels, often resulting in wasted fan energy. Third, air circulation systems with fixed operating parameters cannot quickly eliminate interference from sudden increases in particulate matter concentration or fluctuations in the production environment, potentially affecting the production quality of high-precision electronic products or even contaminating the production line. Fourth, traditional air circulation systems typically use the entire factory as a control unit, making it difficult to precisely match the cleanliness requirements of specific areas. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention applies for an adaptive adjustment system and method for air circulation function in cleanrooms for the electronics industry. Through three-dimensional grid monitoring, three-dimensional visualization analysis, computer vision feature extraction, and artificial intelligence decision-making, it achieves precise and adaptive adjustment of air circulation in cleanrooms for the electronics industry, solving problems such as insufficient monitoring accuracy, lack of adaptive capability, and delayed emergency response in traditional air circulation systems.

[0005] The first aspect of this application discloses an adaptive adjustment system for air circulation function in a cleanroom for the electronics industry, including a three-dimensional grid monitoring module, a data processing module, a visualization module, a computer vision module, an intelligent decision-making module, and an air circulation control module;

[0006] The three-dimensional grid monitoring module uses sensors arranged in a three-dimensional grid in the cleanroom of the electronics industry to acquire environmental data in real time, including data on changes in particulate matter concentration, wind speed and air pressure with space and time.

[0007] The data processing module transforms the environmental data acquired by the three-dimensional grid monitoring module into a continuous spatial distribution model, including a particulate matter concentration field model, a wind speed field model, and a pressure field model, according to the interpolation method. It also obtains the particulate matter concentration change rate field model by calculating the partial derivative of the particulate matter concentration field model with respect to time t.

[0008] The visualization module generates a visualized 3D heat map based on a continuous spatial distribution model and displays it to monitoring personnel through a display device.

[0009] The computer vision module establishes the correlation between the visualized 3D heat map and environmental features by employing convolutional neural network model technology; the environmental features include high particulate matter concentration areas, particulate matter sources, air pressure zone boundaries, airflow dead zones, and eddy regions;

[0010] The intelligent decision-making module uses deep neural network model technology to establish the correlation between the environmental characteristics and air circulation control quantities based on cleanliness control requirements and the principle of reducing energy consumption. The air circulation control quantities include the fan air volume and direction, the fan exhaust volume, and the air valve opening.

[0011] The air circulation control module adjusts the parameters of the air circulation equipment in the cleanroom of the electronics industry according to the air circulation control quantity.

[0012] Preferably, the sensors in the three-dimensional grid monitoring module include a particulate matter concentration sensor, a wind speed sensor, and a pressure sensor, which are used to acquire data on the spatial and temporal changes of particulate matter concentration, wind speed, and pressure, respectively. The three-dimensional grid divides the cleanroom of the electronics industry into multiple rectangular or polygonal areas in the horizontal direction and layers them according to height in the vertical direction. The sensors in each layer are distributed in an array, reducing the spacing between sensors in sensitive production environment areas. The sensitive production environment areas include areas where high-precision equipment operates, key process flow areas, areas with dense personnel activity, air pressure zone boundaries, areas with frequent historical pollution or areas where concentrations are prone to exceed standards, and local microenvironments with the highest cleanliness level.

[0013] Preferably, the visualization module adopts a differentiated visualization encoding method based on the visualized 3D heatmap, and extracts the environmental features from the visualized 3D heatmap through a computer vision module, including: adopting a color gradient visualization encoding method for the particulate matter concentration field model, assigning differentiated color marks according to the magnitude of the particulate matter concentration value, and highlighting high-concentration areas of particulate matter exceeding the threshold; adopting a vector arrow visualization encoding method for the wind speed field model, assigning the length and direction of the vector arrow according to the magnitude and direction of the wind speed, and then identifying and highlighting airflow dead zones and vortex areas through the computer vision module; adopting a contour line visualization encoding method for the pressure field model, drawing contour lines according to the spatial distribution of pressure, and then identifying and highlighting the boundaries of pressure zones through the computer vision module; adopting a color gradient visualization encoding method for the particulate matter concentration change rate field model, assigning differentiated color marks according to the magnitude of the particulate matter concentration change rate value, and then identifying and highlighting particulate matter generation sources through the computer vision module.

[0014] The second aspect of this application discloses an adaptive adjustment method for the air circulation function of a cleanroom in the electronics industry, comprising the following steps:

[0015] S1. Three-dimensional grid sensor deployment and monitoring: Sensors are deployed in the cleanroom of the electronics industry according to horizontal zoning and vertical layering to construct a three-dimensional grid monitoring network; environmental data within the cleanroom of the electronics industry is acquired in real time through the sensors, including data on particulate matter concentration changes with space and time c(x,y,z,t), wind speed changes with space and time w(x,y,z,t), and air pressure changes with space and time p(x,y,z,t); for sensitive areas of the production environment, the spacing between sensor deployments is reduced; the sensors transmit the collected environmental data to the data processing module via wireless communication;

[0016] S2. Data Processing: Based on the environmental data acquired by the sensors and the interpolation method, construct the particulate matter concentration field model C(x,y,z,t), the wind speed field model W(x,y,z,t), and the air pressure field model P(x,y,z,t). By calculating the partial derivative of the particulate matter concentration field model C(x,y,z,t) with respect to time t, construct the particulate matter concentration change rate field model V. c (x,y,z,t);

[0017] S3. Data Visualization: Visualizing the particulate matter concentration field model C(x,y,z,t) and the particulate matter concentration change rate field model V. cThe wind speed field model W(x,y,z,t) and the air pressure field model P(x,y,z,t) are converted into a three-dimensional visualization heat map, and the three-dimensional visualization heat map is used as a texture to map onto the three-dimensional model of the cleanroom in the electronics industry.

[0018] S4. Image feature extraction: Analyze the three-dimensional visualization heatmap using computer vision technology to extract environmental features, including high particulate matter concentration areas, particulate matter sources, air pressure zone boundaries, airflow dead zones, and eddy regions.

[0019] S5. Abnormal Area Early Warning: For the particulate matter concentration field model C(x,y,z,t) and the particulate matter concentration change rate field model V... c Areas where the values ​​of (x,y,z,t) exceed a preset threshold are highlighted and an audible and visual alarm is triggered.

[0020] S6. Decision on air circulation parameters: The environmental characteristics are input into the intelligent decision module. Based on the cleanliness control requirements and the principle of reducing energy consumption, the air circulation control quantities are output, including the fan air volume and direction, the fan exhaust volume and the air valve opening.

[0021] S7. Adaptive adjustment execution: Adjust the parameters of the air circulation equipment in the cleanroom of the electronics industry according to the air circulation control quantity;

[0022] S8. Closed-loop feedback optimization: When the duration of the high particulate matter concentration area exceeds the preset duration T, closed-loop feedback optimization is triggered. Specifically, steps S1 to S4 are repeated, and the adjusted environmental characteristics are input into the intelligent decision-making module. The intelligent decision-making module is based on a deep neural network model and uses an incremental learning algorithm to iteratively optimize the correlation model between the environmental characteristics and the air circulation control quantity. The incremental learning algorithm updates the weights of the hidden layer of the neural network through real-time collected environmental data. The update frequency is automatically triggered after each adjustment cycle.

[0023] Compared with existing technologies, the beneficial effects of this invention are as follows: Addressing the problems of insufficient monitoring accuracy, lack of adaptive capability, and delayed emergency response in existing air circulation systems, this invention proposes an adaptive adjustment system and method for air circulation function in cleanrooms for the electronics industry. This adjustment system includes a three-dimensional grid monitoring module, a data processing module, a visualization module, a computer vision module, an intelligent decision-making module, and an air circulation control module. The adjustment method includes three-dimensional grid sensor layout and monitoring, data processing, data visualization, image feature extraction, abnormal area early warning, air circulation parameter decision-making, adaptive adjustment execution, and closed-loop feedback optimization. A continuous spatial distribution model is constructed using real-time monitoring data from the three-dimensional grid sensors, enabling monitoring of particulate matter distribution within the cleanroom. High-precision prediction of airflow conditions effectively reduces quality risks caused by pollution blind spots such as pollution sources, dead air zones, and eddy current areas. The visualization module transforms particulate matter concentration fields and wind speed fields into dynamic, visualized 3D heat maps. Combined with computer vision technology to extract environmental features, technicians can intuitively locate pollution sources and abnormal airflow points, improving operation and maintenance efficiency. Artificial intelligence technology establishes a correlation between environmental features and air circulation control quantities, outputting airflow control parameters based on local pollution anomalies. Compared with traditional artificial intelligence technology that outputs airflow control parameters based on single-point monitoring data, this significantly improves the system's accuracy in adjusting complex operating conditions and achieves adaptive continuous evolution through a closed-loop feedback optimization mechanism. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of the adaptive adjustment system for air circulation in a cleanroom for the electronics industry, as shown in an embodiment of the present invention.

[0025] Figure 2 This is a flowchart of the adaptive adjustment method for air circulation function in a cleanroom for the electronics industry, as shown in an embodiment of the present invention.

[0026] Figure labels: 1-3D grid monitoring module, 11-particulate matter concentration sensor, 12-wind speed sensor, 13-barrier sensor, 2-data processing module, 3-visualization module, 4-computer vision module, 5-intelligent decision-making module, 6-air circulation control module. Detailed Implementation

[0027] The embodiments of the present invention will be described in more detail below with reference to the accompanying drawings and reference numerals, so that those skilled in the art can implement them after reading this specification. It should be understood that the specific embodiments described herein are only for explaining the present invention and are not intended to limit the present invention.

[0028] The first aspect of this application discloses as follows: Figure 1The adaptive adjustment system for air circulation in a cleanroom for the electronics industry shown includes a three-dimensional grid monitoring module 1, a data processing module 2, a visualization module 3, a computer vision module 4, an intelligent decision-making module 5, and an air circulation control module 6.

[0029] The three-dimensional grid monitoring module 1 uses sensors arranged in a three-dimensional grid within the cleanroom of the electronics industry to acquire real-time environmental data, including data on changes in particulate matter concentration, wind speed, and air pressure over space and time. Specifically, the sensors in the three-dimensional grid monitoring module 1 include a particulate matter concentration sensor 11, a wind speed sensor 12, and an air pressure sensor 13, which are used to acquire data on changes in particulate matter concentration, wind speed, and air pressure over space and time, respectively. The three-dimensional grid divides the cleanroom into multiple rectangular or polygonal areas horizontally and layers them vertically, with sensors arranged in an array within each layer. For conventional production areas, the horizontal spacing ranges from 5m to 10m, and the vertical layer spacing ranges from 3m to 5m. For environmentally sensitive areas, the sensor spacing is reduced according to the sensitivity, with a horizontal spacing range of 1m to 5m and a vertical layer spacing range of 1m to 3m. These environmentally sensitive areas include areas where high-precision equipment operates, key process flow areas, areas with high personnel activity, air pressure zone boundaries, areas with frequent historical pollution or where concentrations easily exceed standards, and local microenvironments with the highest cleanliness level. The initial sensor deployment scheme is based on the simulation results of the computational fluid dynamics model of a cleanroom in the electronics industry; the sensor deployment scheme during continuous operation is adjusted according to the environmental characteristics acquired by the computer vision module 4.

[0030] The data processing module 2 converts the environmental data acquired by the three-dimensional grid monitoring module 1 into a continuous spatial distribution model according to the interpolation method, including a particulate matter concentration field model, a wind speed field model, and a pressure field model. It then obtains the particulate matter concentration change rate field model by calculating the partial derivative of the particulate matter concentration field model with respect to time t.

[0031] The visualization module 3 generates a visualized 3D heat map based on a continuous spatial distribution model and displays it to monitoring personnel through a display device. The computer vision module 4 establishes a correlation between the visualized 3D heat map and environmental features by employing convolutional neural network model technology; the environmental features include high particulate matter concentration areas, particulate matter sources, pressure zone boundaries, airflow dead zones, and eddy regions. In specific implementation, the visualization module 3 adopts a differentiated visualization encoding method based on the visualized 3D heat map, and extracts the environmental features from the visualized 3D heat map through the computer vision module 4. This includes: adopting a color gradient visualization encoding method for the particulate matter concentration field model, assigning differentiated color marks according to the magnitude of the particulate matter concentration value, and highlighting high-concentration areas of particulate matter exceeding the threshold; adopting a vector arrow visualization encoding method for the wind speed field model, assigning the length and direction of the vector arrow according to the magnitude and direction of the wind speed, and then identifying and highlighting airflow dead zones and vortex areas through the computer vision module 4; adopting a contour line visualization encoding method for the pressure field model, drawing contour lines according to the spatial distribution of air pressure, and then identifying and highlighting the boundaries of pressure zones through the computer vision module 4; and adopting a color gradient visualization encoding method for the particulate matter concentration change rate field model, assigning differentiated color marks according to the magnitude of the particulate matter concentration change rate value, and then identifying and highlighting particulate matter generation sources through the computer vision module 4.

[0032] The intelligent decision-making module 5 employs deep neural network model technology to establish the correlation between environmental features and air circulation control quantities based on cleanliness control requirements and the principle of reducing energy consumption. These air circulation control quantities include fan airflow and direction, fan exhaust volume, and valve opening. Specifically, the input layer of the deep neural network model is a 20-dimensional environmental feature vector, including: three-dimensional coordinates of high-concentration particulate matter areas, particulate matter concentration values, and area type labels (5 dimensions); three-dimensional coordinates of particulate matter sources, particulate matter concentration change rate, and source type labels (5 dimensions); coordinates of pressure zone boundaries, pressure gradient values, and zone numbers (5 dimensions); and center coordinates, wind speed vector, and eddy intensity index of airflow dead zones or vortex areas (5 dimensions). The output layer of the deep neural network model is a (3k+m+n)-dimensional air circulation control quantity matrix, specifically including: a three-dimensional vector of airflow from k fans, exhaust volume from m fans, and opening degrees of n valves. The intelligent decision-making module, based on cleanliness control requirements and the principle of reducing energy consumption, uses the following loss function:

[0033] L = w1·L c +w2·L e (1)

[0034] Among them, L cFor the loss of cleanliness, L e For energy loss; w1 and w2 are both weighting coefficients. When the duration of the high concentration area of ​​particulate matter exceeds the preset duration T, w1 = 1 and w2 = 0. Otherwise, it automatically switches to the mode of w1 = 0.8 and w2 = 0.2.

[0035] The air circulation control module 6 adjusts the parameters of the air circulation equipment in the cleanroom of the electronics industry according to the air circulation control parameters. The air circulation equipment includes a fan and a damper.

[0036] The second aspect of this application discloses, as follows: Figure 2 The adaptive adjustment method for air circulation function in a cleanroom for the electronics industry, as shown, includes the following steps:

[0037] S1. Three-dimensional grid sensor deployment and monitoring: Sensors are deployed in the cleanroom of the electronics industry according to horizontal zoning and vertical layering to construct a three-dimensional grid monitoring network. The sensors include a particulate matter concentration sensor 11, a wind speed sensor 12, and a pressure sensor 13. In specific implementation, the sensors are integrated into non-critical parts of the production equipment as much as possible, while ensuring that the sensor measurement ports are flush with or exposed to the equipment surface to avoid data deviation caused by equipment structure obstruction. The sensors acquire environmental data in the cleanroom of the electronics industry in real time, including particulate matter concentration data c(x,y,z,t) changing with space and time, wind speed data w(x,y,z,t) changing with space and time, and pressure data p(x,y,z,t) changing with space and time. For sensitive areas of the production environment, the spacing between sensor deployments is reduced. The sensors transmit the collected environmental data to the data processing module via wireless communication.

[0038] S2. Data Processing: Based on the environmental data acquired by sensors and interpolation methods, construct particulate matter concentration field model C(x,y,z,t), wind speed field model W(x,y,z,t), and air pressure field model P(x,y,z,t). By calculating the partial derivative of the particulate matter concentration field model C(x,y,z,t) with respect to time t, construct the particulate matter concentration change rate field model V. c (x,y,z,t). In practice, Kriging interpolation is used for particulate matter concentration data to improve sensitivity to areas with high particulate matter concentrations or sources of particulate matter generation; spline interpolation is used for wind speed and air pressure data to ensure smooth changes in the wind speed and air pressure fields.

[0039] S3. Data Visualization: Visualizing the particulate matter concentration field model C(x,y,z,t) and the particulate matter concentration change rate field model V. cThe wind speed field model W(x,y,z,t) and the air pressure field model P(x,y,z,t) are converted into a three-dimensional visualization heat map, and the three-dimensional visualization heat map is used as a texture to map onto the three-dimensional model of the cleanroom in the electronics industry.

[0040] S4. Image Feature Extraction: Analyze the three-dimensional visualization heatmap using computer vision technology to extract environmental features, including high particulate matter concentration areas, particulate matter sources, air pressure zone boundaries, airflow dead zones, and eddy regions.

[0041] S5. Abnormal Area Early Warning: For the particulate matter concentration field model C(x,y,z,t) and the particulate matter concentration change rate field model V... c Areas where the values ​​of (x,y,z,t) exceed a preset threshold are highlighted and an audible and visual alarm is triggered.

[0042] S6. Decision on air circulation parameters: The environmental characteristics are input to the intelligent decision module 5. Based on the cleanliness control requirements and the principle of reducing energy consumption, the air circulation control quantities are output, including the fan air volume and direction, the fan exhaust volume and the opening degree of the air valve.

[0043] S7. Adaptive Adjustment Execution: Adjust the parameters of the air circulation equipment in the cleanroom of the electronics industry according to the air circulation control quantity.

[0044] S8. Closed-loop feedback optimization: When the duration of the high particulate matter concentration area exceeds a preset duration T, closed-loop feedback optimization is triggered. Specifically, steps S1 to S4 are repeated, and the adjusted environmental characteristics are input into the intelligent decision-making module. The intelligent decision-making module, based on a deep neural network model, uses an incremental learning algorithm to iteratively optimize the correlation model between the environmental characteristics and the air circulation control quantity. The incremental learning algorithm updates the weights of the hidden layer of the neural network through real-time collected environmental data, and the update frequency is automatically triggered after each adjustment cycle. In specific implementation, the preset duration T is 10 minutes, and the goal of the iterative optimization is to reduce the fan energy consumption by no less than 15% while meeting the cleanliness level requirements. The learning rate of the incremental learning algorithm is 0.01.

[0045] Therefore, by constructing a continuous spatial distribution model using real-time monitoring data from a three-dimensional grid sensor, high-precision prediction of particulate matter distribution and airflow conditions within the factory can be achieved, effectively reducing quality risks caused by pollution blind spots such as pollution sources, airflow dead zones, and eddy areas. The visualization module transforms particulate matter concentration fields and wind speed fields into dynamic, visualized three-dimensional heat maps. Combined with computer vision technology for extracting environmental features, this allows technicians to intuitively locate pollution sources and abnormal airflow points, improving operational and maintenance efficiency. Furthermore, by establishing a correlation between environmental features and air circulation control parameters using artificial intelligence technology, and outputting airflow control parameters based on local pollution anomalies, this significantly improves the system's accuracy in adjusting to complex operating conditions compared to traditional AI technologies that output airflow control parameters based on single-point monitoring data. Moreover, the closed-loop feedback optimization mechanism enables adaptive and continuous evolution.

[0046] The above describes one or more embodiments of the present invention in a relatively specific and detailed manner, but it should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this patent should be determined by the appended claims.

Claims

1. An adaptive regulating system for air circulation function in a clean room of electronic industry, characterized in that, The system comprises a three-dimensional grid monitoring module, a data processing module, a visualization module, a computer vision module, an intelligent decision-making module, and an air circulation control module. The three-dimensional grid monitoring module is arranged in the form of a three-dimensional grid in the electronic industry clean workshop to obtain environmental data in the electronic industry clean workshop in real time, including particle concentration, wind speed, and air pressure data varying with space and time. The data processing module converts the environmental data obtained by the stereoscopic grid monitoring module into continuous spatial distribution models, including a particulate matter concentration field model, a wind speed field model and an air pressure field model, according to an interpolation method, and obtains a particulate matter concentration change rate field model by performing partial derivative calculation on the particulate matter concentration field model with respect to time t ​ The three-dimensional grid monitoring module comprises a particle concentration sensor, a wind speed sensor, and an air pressure sensor. The computer vision module establishes the correlation between the visualized three-dimensional heat map and the environmental features by adopting a convolutional neural network model technology; the environmental features include a high-concentration area of particulate matter, a particulate matter generation source, a pressure partition boundary, a dead air zone, and a vortex zone; the intelligent decision module establishes the correlation between the environmental features and the air circulation control quantity by adopting a deep neural network model technology based on the cleanliness control requirement and the principle of reducing energy consumption; the air circulation control quantity includes the fan supply air volume and direction, the fan exhaust air volume, and the air valve opening degree; the air circulation control module adjusts the parameters of the air circulation equipment of the electronic industry cleanroom according to the air circulation control quantity; wherein the input layer of the deep neural network model is a 20-dimensional environmental feature vector, specifically including: three-dimensional coordinates of the high-concentration area of particulate matter, a particulate matter concentration value, and an area type label, a total of 5 dimensions; three-dimensional coordinates of the particulate matter generation source, a particulate matter concentration change rate, and a source type label, a total of 5 dimensions; coordinates of the pressure partition boundary, a pressure gradient value, and a partition number, a total of 5 dimensions; center coordinates of the dead air zone or vortex zone, a wind speed vector, and a vortex intensity index, a total of 5 dimensions; the output layer of the deep neural network model is a (3 k + m + n ) dimensional air circulation control quantity matrix, specifically including: k a three-dimensional vector of the fan supply air volume, m an exhaust air volume of the fan, n an opening degree of the air valve; the deep neural network model adopts a loss function as shown in the following formula based on the cleanliness control requirement and the principle of reducing energy consumption: L = w 1· L c + w 2· L e wherein, L c is a loss of cleanliness, L e is a loss of energy consumption; w 1 and w 2 are weight coefficients, when the particulate matter high concentration area duration exceeds a preset length of time T , w 1=1 and w 2=0, otherwise automatically switch to w 1=0.8 and w 2=0.2 normal mode.

2. A method for adaptive regulation of air circulation function in a clean room of the electronic industry, characterized in that, The three-dimensional grid divides the electronic industry clean workshop into multiple rectangular or polygonal areas in the horizontal direction and is layered according to height in the vertical direction. S1, three-dimensional grid sensor arrangement and monitoring: arranging sensors in the electronic industry clean workshop according to horizontal partition and vertical stratification, constructing a three-dimensional grid monitoring network; acquiring environmental data in the electronic industry clean workshop in real time through sensors, including data of particle concentration changing with space and time c ( The production environment sensitive area includes an area where high-precision equipment operates, a key process area, a personnel activity intensive area, an air pressure partition boundary, a historical pollution frequent area or an area where concentration is prone to exceed the standard, and a local microenvironment with the highest cleanliness level. ), data of wind speed changing with space and time w ( The visualization module generates a three-dimensional visualization heat map according to a continuous spatial distribution model and displays the heat map to monitoring personnel through a display device. ), and data of air pressure changing with space and time p ( The three-dimensional visualization heat map adopts a differential visualization coding method, and the computer vision module extracts environmental features from the three-dimensional visualization heat map, including a color gradient visualization coding method for a particle concentration field model, differential color labels according to particle concentration values, and highlighted display of high-concentration particle areas exceeding a threshold value. ); for the production environment sensitive area, reducing the arrangement interval of sensors; the sensors transmit the collected environmental data to the data processing module through wireless communication mode; S2. Data Processing: Based on the environmental data acquired by the sensors and the interpolation method, construct a particulate matter concentration field model. C ( x, The system is applied to the electronic industry clean workshop air circulation function adaptive adjustment system of claim 1 and comprises the following steps. Wind speed field model W ( x, y, z, t ) and pressure field model P ( x, y, z, t ); By analyzing the particulate matter concentration field model C ( x, y, z, t ) to perform time t Partial derivative calculations were used to construct a field model of particulate matter concentration change rate. V c ( y, z, t x, y, z, t ); S3, data visualization: converting the particulate concentration field model C x, y, z, t , the particulate concentration rate of change field model V c x, y, z, t , the wind speed field model W x, y, , and the air pressure field model P z, t into a three-dimensional visualization heat map and mapping the three-dimensional visualization heat map as a texture onto a three-dimensional model of an electronic industry cleanroom;​​​​ x, y, z, t S5, abnormal area early warning: for the particulate matter concentration field model C ( x, y, z, t ) and the particulate matter concentration rate of change field model V c ( x, y, z, t x, y, z, t S4, image feature extraction: using computer vision technology to analyze the three-dimensional visualization heat map and extract environmental features, including high-concentration particle areas, particle sources, air pressure partition boundaries, airflow dead angles, and vortex areas. x, y, z, t x, y, z, t ) The area whose numerical value exceeds the preset threshold is highlighted, and an audible and light alarm is triggered; S6, decision of air circulation parameters: input the environment features into the intelligent decision module, output air circulation control amount including fan supply air volume and direction, fan exhaust air volume and air valve opening degree based on cleanliness control requirement and principle of reducing energy consumption; S7, adaptive adjustment execution: perform parameter adjustment on air circulation equipment of electronic industry clean workshop according to the air circulation control amount; S8. Closed-loop feedback optimization: When the duration of the high concentration area of ​​particulate matter exceeds a preset time... T When the closed-loop feedback optimization is triggered, specifically: repeat steps S1 to S4, input the adjusted environmental features into the intelligent decision-making module, the intelligent decision-making module is based on a deep neural network model, and uses an incremental learning algorithm to iteratively optimize the correlation model between the environmental features and the air circulation control quantity. The incremental learning algorithm updates the weights of the hidden layer of the neural network through real-time collected environmental data, and the update frequency is automatically triggered after each adjustment cycle.

Images