Air purification system for clean workshop

Abstract

The invention discloses an air purification system for a clean workshop. The air purification system comprises a self-adaptive control unit, an air quality monitoring module, an air flow control assembly, an environmental parameter adjusting assembly, a self-optimization filtering assembly, a human-computer interaction interface and predictive maintenance system, an energy efficiency optimization module and a remote diagnosis and early warning system. On the basis of intelligent technologies such as machine learning, reinforcement learning and a data fusion algorithm, real-time monitoring, dynamic optimization and energy efficiency management of the air quality of the clean workshop are achieved. Based on real-time data and historical data, an algorithm is used for carrying out self-adaptive optimization on the air purification system, and manual intervention is reduced. And the running state of the system is dynamically adjusted through the energy efficiency optimization module, so that the energy consumption is reduced while the cleanliness is ensured. The system can diagnose equipment faults in advance, reduce downtime and prolong the service life of equipment. And due to a man-machine interaction interface and a remote monitoring function, an operator can know the state of the system at any time and carry out regulation and control.

Description

Technical Field

[0001] The invention relates to the technical field of air purification, and in particular to an air purification system for a clean workshop. Background Art

[0002] With the development of high-tech industry, medical equipment manufacturing and electronic products industry, the environmental quality requirements for clean workshops are getting higher and higher. Traditional air purification systems have problems such as single monitoring means, low adjustment efficiency, high energy consumption and inconvenient maintenance, which are difficult to meet the needs of modern clean workshops for high efficiency, energy saving and intelligent management. In addition, how to use intelligent means to monitor and optimize the operating status of the air purification system in real time, reduce energy consumption while ensuring cleanliness, has become a hot spot and difficulty in current research. Summary of the invention

[0003] The purpose of the present invention is to provide a clean workshop air purification system.

[0004] To achieve the above object, the present invention is implemented according to the following technical solutions:

[0005] The clean room air purification system of the present invention comprises an adaptive control unit, an air quality monitoring module, an air flow control component, an environmental parameter adjustment component, a self-optimizing filter component, a human-machine interaction interface and a predictive maintenance system, an energy efficiency optimization module and a remote diagnosis and early warning system. The adaptive control unit predicts and adjusts various parameters of the clean room environment through a machine learning algorithm based on real-time data and historical data. The air quality monitoring module collects air quality data in real time by arranging sensors in the clean room, and transmits the data to the adaptive control unit to optimize the operation of the system in real time. The self-optimizing filter component dynamically adjusts the operating state of the filter system in combination with the adaptive control unit algorithm and the air quality data. The environmental parameter adjustment component dynamically adjusts the operating state of the HVAC system by analyzing the environmental parameters in real time. The air flow control component optimizes and controls the air flow of the air duct. The human-machine interaction interface and the predictive maintenance system provide an intuitive monitoring interface for operators, and diagnose potential problems of the equipment in advance based on the adaptive control unit and big data. The energy efficiency optimization module optimizes and manages energy consumption in combination with the data of the adaptive control unit and big data analysis. The remote diagnosis and early warning system understands the operating state of the air purification system in real time through intelligent remote monitoring, and performs remote fault diagnosis and early warning.

[0006] Furthermore, the machine learning algorithm of the adaptive control unit is based on a reinforcement learning algorithm, a deep neural network algorithm and a data fusion algorithm. The sensors of the air quality monitoring module include an M2.5 sensor, a CO2 sensor, a temperature and humidity sensor, an air flow velocity sensor, a particle concentration sensor and a microbial detection sensor. Based on the data, the air pollution trend is predicted, the air treatment strategy is automatically adjusted, and the pollution source changes are promptly responded to. The self-optimizing filter component monitors the load of the filter and adjusts the fan speed according to the filter status and air quality. The environmental parameter adjustment component automatically adjusts the HVAC system according to the current environmental changes in the clean room. The air flow control component analyzes the air flow path to avoid local air stagnation or pollution diffusion, and adjusts the air supply and exhaust volume according to the real-time monitoring of air quality. The remote diagnosis and early warning system checks the air quality, equipment status and maintenance status of the clean room in real time through the APP or PC terminal, predicts potential equipment failures or air quality problems, and sends alarm information in advance. The energy efficiency optimization module automatically adjusts the system operation status to reduce energy consumption while maintaining the air quality required for the clean room. The human-computer interaction interface and predictive maintenance system predict equipment failure or performance degradation based on historical data and real-time data, and perform maintenance in advance.

[0007] The purification method of the clean room air purification system of the present invention comprises the following steps:

[0008] S1: Install a variety of air quality sensors in the clean room to monitor and collect air quality data in real time, and transmit the collected sensor data to the adaptive control unit in real time through the Internet of Things technology;

[0009] S2: The adaptive control unit uses historical data and real-time data, adopts data fusion algorithm, integrates data from different sources into a unified analysis platform, and predicts the trend of clean room air quality based on machine learning algorithm, determines the future air pollution trend, and automatically adjusts the operation strategy of the air purification system;

[0010] S3: Use sensors to monitor the air flow status in real time, adjust the air supply and exhaust volume according to the changes in air quality, and dynamically adjust the air flow direction in the air duct according to the analysis of the air flow control component to ensure that the air flows evenly in the clean room;

[0011] S4: The environmental parameter adjustment component automatically adjusts the operating status of the HVAC system based on real-time data. If it detects that the temperature or humidity in the clean room exceeds the set range, the HVAC system automatically adjusts the air conditioning temperature or humidity level. The environmental parameter adjustment module adjusts the fan speed according to data such as temperature, humidity and air quality to ensure the smoothness and cleanliness of air flow;

[0012] S5: The sensor monitors the pressure difference and flow rate changes of the filter, and the adaptive control unit adjusts the working state of the filtration system based on the air quality data. If the concentration of air pollutants increases, the fan speed is automatically increased;

[0013] S6: Energy efficiency optimization module monitors the energy consumption of each device in the clean room in real time, analyzes the relationship between the energy consumption of the device and environmental data, automatically adjusts the operating status of the system, optimizes energy use, and intelligently adjusts the operating timing and intensity of the equipment according to the system operating status to ensure that the clean room always maintains the required air quality while reducing energy waste;

[0014] S7: Monitor the operating status of key equipment through sensors and predict the risk of equipment failure. The prediction model compares historical data with real-time data to identify potential equipment failures in advance, and issues alarms through the remote diagnosis and early warning system to notify operators to conduct inspections and maintenance.

[0015] S8: Through the remote diagnosis and early warning system, operators can view the air quality, equipment operating status, and maintenance records of the clean room in real time through the APP or PC. When equipment abnormalities occur, the system automatically sends alarm information to the operator through the early warning mechanism and provides preliminary fault diagnosis suggestions. Operators can take timely measures according to the system prompts.

[0016] The beneficial effects of the present invention are:

[0017] The present invention is a clean room air purification system. Compared with the prior art, the present invention is based on intelligent technologies such as machine learning, reinforcement learning and data fusion algorithms to achieve real-time monitoring, dynamic optimization and energy efficiency management of clean room air quality. Based on real-time data and historical data, the air purification system is adaptively optimized using algorithms to reduce manual intervention. The system operating state is dynamically adjusted through the energy efficiency optimization module to reduce energy consumption while ensuring cleanliness. The system can diagnose equipment failures in advance, reduce downtime and increase the service life of the equipment. The human-computer interaction interface and remote monitoring function enable operators to understand the system status at any time and make adjustments. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a system structure principle block diagram of the present invention. DETAILED DESCRIPTION

[0019] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments. The illustrative embodiments and descriptions of the present invention are used to explain the present invention but are not intended to limit the present invention.

[0020] like Figure 1As shown: The clean workshop air purification system of the present invention includes an adaptive control unit, an air quality monitoring module, an air flow control component, an environmental parameter adjustment component, a self-optimizing filter component, a human-computer interaction interface and a predictive maintenance system, an energy efficiency optimization module and a remote diagnosis and early warning system;

[0021] 1. Adaptive control unit: Based on reinforcement learning (RL) and deep neural network (DNN), it analyzes the air quality data collected in real time. RL optimizes the operation strategy through the reward function, while DNN is used to predict air quality trends and provide support for optimization decisions.

[0022] 2. Air quality monitoring module: Use a variety of sensors (such as PM2.5, CO2 concentration, microbial detection, etc.) to achieve real-time monitoring of multiple parameters in the air. The sensor data is transmitted to the control unit through the Internet of Things and denoised through the Kalman filter to ensure the accuracy and consistency of the data.

[0023] 3. Air flow control component: Adopt fluid mechanics model and reinforcement learning algorithm to optimize air flow distribution path, dynamically adjust air supply and exhaust volume, and avoid pollutant accumulation. Reinforcement learning automatically selects the optimal control strategy based on air flow distribution state.

[0024] 4. Environmental parameter adjustment component: Real-time optimization of environmental parameters such as temperature and humidity is performed through support vector machines (SVM) and neural networks to maintain constant environmental conditions in the clean room. This module is linked with the HVAC system to ensure a balance between air quality and energy efficiency.

[0025] 5. Self-optimizing filter component: Use adaptive filtering algorithm to adjust the filter operation status according to the air pollution concentration, monitor the filter load and dynamically adjust the fan speed to extend the filter life.

[0026] 6. Energy efficiency optimization module: Combining air quality data and energy consumption data, using energy efficiency optimization algorithm, dynamically adjust the operating status of each device to ensure that the system achieves the set air quality with the lowest energy consumption.

[0027] 7. Human-machine interaction interface and predictive maintenance system: The interface displays the system operation status in real time and provides maintenance suggestions. The remaining useful life (RUL) of key equipment is predicted through the LSTM model to identify equipment failures in advance.

[0028] 8. Remote diagnosis and early warning system: Based on cloud management, it provides remote monitoring and alarm services. It uses intelligent algorithms to diagnose equipment status and send maintenance reminders to operators.

[0029] The adaptive control unit predicts and adjusts various parameters of the clean room environment through a machine learning algorithm based on real-time data and historical data. The air quality monitoring module collects air quality data in real time by arranging sensors in the clean room, and transmits the data to the adaptive control unit to optimize the operation of the system in real time. The self-optimizing filtering component dynamically adjusts the operating state of the filtering system in combination with the adaptive control unit algorithm and air quality data. The environmental parameter adjustment component dynamically adjusts the operating state of the HVAC system by analyzing environmental parameters in real time. The air flow control component optimizes the air flow in the air duct. The human-computer interaction interface and predictive maintenance system provide operators with an intuitive monitoring interface and diagnose potential equipment problems in advance based on the adaptive control unit and big data. The energy efficiency optimization module optimizes management by combining the data of the adaptive control unit and big data to analyze energy consumption. The remote diagnosis and early warning system understands the operating state of the air purification system in real time through intelligent remote monitoring, and performs remote fault diagnosis and early warning.

[0030] Furthermore, the machine learning algorithm of the adaptive control unit is based on a reinforcement learning algorithm, a deep neural network algorithm and a data fusion algorithm. The sensors of the air quality monitoring module include an M2.5 sensor, a CO2 sensor, a temperature and humidity sensor, an air flow velocity sensor, a particle concentration sensor and a microbial detection sensor. Based on the data, the air pollution trend is predicted, the air treatment strategy is automatically adjusted, and the pollution source changes are promptly responded to. The self-optimizing filter component monitors the load of the filter and adjusts the fan speed according to the filter status and air quality. The environmental parameter adjustment component automatically adjusts the HVAC system according to the current environmental changes in the clean room. The air flow control component analyzes the air flow path to avoid local air stagnation or pollution diffusion, and adjusts the air supply and exhaust volume according to the real-time monitoring of air quality. The remote diagnosis and early warning system checks the air quality, equipment status and maintenance status of the clean room in real time through the APP or PC terminal, predicts potential equipment failures or air quality problems, and sends alarm information in advance. The energy efficiency optimization module automatically adjusts the system operation status to reduce energy consumption while maintaining the air quality required for the clean room. The human-computer interaction interface and predictive maintenance system predict equipment failure or performance degradation based on historical data and real-time data, and perform maintenance in advance.

[0031] The purification method of the clean room air purification system of the present invention comprises the following steps:

[0032] S1: Install a variety of air quality sensors in the clean room to monitor and collect air quality data in real time, and transmit the collected sensor data to the adaptive control unit in real time through the Internet of Things technology;

[0033] S2: The adaptive control unit uses historical data and real-time data, and adopts a data fusion algorithm to integrate data from different sources into a unified analysis platform. It predicts the trend of clean room air quality based on machine learning algorithms, determines future air pollution trends, and automatically adjusts the operation strategy of the air purification system. The data fusion algorithm removes noise and improves data accuracy by fusing data from different sensors. Assume that we have data from multiple sensors and fuse them as follows:

[0034]

[0035] Where: The estimated current state (filtered value); The estimated state at the previous moment (prior value), K k : gain; z k : The observed value of the sensor; H k : observation matrix;

[0036] Based on the multi-dimensional data collected by sensors, a deep neural network (DNN) is used to predict air quality trends, especially the changes in PM2.5 and CO2 concentrations. The machine learning algorithm is:

[0037]

[0038] Where: Predicted air quality output (such as PM2.5 concentration); W 1 , W 2 : The weight matrix of the neural network, x t : Input data at the current moment (such as temperature, humidity, CO2 concentration, etc.); h t-1 : The hidden layer output at the previous moment; b 1 , b 2 : Bias term.

[0039] S3: Use sensors to monitor the air flow status in real time, adjust the air supply and exhaust volume according to the changes in air quality, and dynamically adjust the airflow direction in the air duct according to the analysis of the air flow control component to ensure that the air flows evenly in the clean room; dynamically adjust the airflow direction in the air duct using fluid mechanics models and reinforcement learning algorithms to analyze air flow. Air flow control involves optimizing the air flow path through the air duct. Use a fluid mechanics model to analyze air flow to determine the appropriate wind speed and airflow direction. The fluid mechanics model is as follows:

[0040]

[0041]

[0042] Where: v: air velocity; ρ: air density; t: time; p: pressure; μ: dynamic viscosity coefficient of air; f: volume force (such as gravity, force provided by the fan);

[0043] The reinforcement learning algorithm can dynamically adjust the air flow to optimize the air quality. Based on the environmental state (air quality data), the agent (algorithm) selects different wind speeds and air supply strategies to achieve the best effect. The reinforcement learning algorithm is as follows:

[0044]

[0045] Where: Q(s t , a t ): state-action value function, indicating that in state s t Take action a t The value of; α: learning rate; r t : The reward of the current step, which is related to air quality improvement or energy efficiency optimization; γ: Discount factor, which controls the impact of future rewards, max a′ Q(s t+1 , a′): in the next state s t+1 Next, choose the value of the optimal action.

[0046] S4: The environmental parameter adjustment component automatically adjusts the operating status of the HVAC system based on real-time data. If it is detected that the temperature or humidity in the clean room exceeds the set range, the HVAC system automatically adjusts the air conditioning temperature or humidity level. The environmental parameter adjustment module adjusts the fan speed according to data such as temperature, humidity and air quality to ensure the smoothness and cleanliness of air flow; the HVAC system is optimized through the support vector machine to maintain the environmental conditions required for the clean room, as shown in the following formula:

[0047] f(x)=w T φ(x)+b

[0048] Where: f(x): prediction output of support vector machine (such as predicted values ​​of temperature and humidity); w: weight vector, representing the learning result of the model; φ(x): mapping function of input data, mapped to high-dimensional feature space; b: bias term.

[0049] The neural network model is used to optimize the regulation of environmental parameters such as temperature and humidity to maintain constant indoor air quality. As shown in the following formula:

[0050] y t =σ(W·x t +b)

[0051] Where: y t : the output of the neural network (e.g., the temperature control value of the HVAC system); W: the weight matrix of the neural network; x t: Input data (such as current temperature and humidity, CO2 concentration, etc.); σ: activation function (such as ReLU, Sigmoid, etc.).

[0052] S5: The sensor monitors the pressure difference and flow rate changes of the filter, and the adaptive control unit adjusts the working state of the filtration system based on the air quality data. If the concentration of air pollutants increases, the fan speed is automatically increased; the air quality data adaptive control unit adjusts the working state of the filtration system using an adaptive filtering algorithm, as shown in the following formula:

[0053]

[0054] Where: The filter output at the next moment; The filter output at the current moment; K t : Gain factor, which determines the impact of new observations on the filter output; z t : The observed value of the sensor.

[0055] S6: The energy efficiency optimization module monitors the energy consumption of each device in the clean room in real time, analyzes the relationship between the energy consumption of the equipment and environmental data, automatically adjusts the operating status of the system, optimizes energy use, and intelligently adjusts the operating timing and intensity of the equipment according to the system operating status to ensure that the clean room always maintains the required air quality while reducing energy waste; the operating timing and intensity of the equipment are intelligently adjusted using energy efficiency optimization algorithms, and according to the system's prediction results, the power usage of each device is adjusted to achieve the lowest energy consumption and the best air quality. As shown in the following formula:

[0056]

[0057] Where: E opt : Optimized energy efficiency; P: equipment power; C(P): equipment energy cost; Q(P): air quality, which changes with power; λ: trade-off factor, which controls the trade-off between energy consumption and air quality.

[0058] S7: Monitor the operating status of key equipment through sensors and predict the risk of equipment failure. The prediction model compares historical data with real-time data to identify potential equipment failures in advance, and issues alarms through the remote diagnosis and early warning system to notify operators to conduct inspections and maintenance.

[0059] S8: Through the remote diagnosis and early warning system, operators can view the air quality, equipment operating status, and maintenance records of the clean room in real time through the APP or PC. When equipment abnormalities occur, the system automatically sends alarm information to the operator through the early warning mechanism and provides preliminary fault diagnosis suggestions. Operators can take timely measures according to the system prompts.

[0060] Example:

[0061] A certain electronic product production plant requires the cleanliness level to reach ISO 7. The system of the present invention is now deployed to achieve air purification, energy efficiency optimization and intelligent management.

[0062] Implementation steps:

[0063] System installation: Deploy sensors such as PM2.5, CO2 concentration, temperature and humidity in the factory and connect them to the adaptive control unit.

[0064] Real-time monitoring: The sensor collects air quality data (such as PM2.5 concentration), which is then denoised through Kalman filtering and uploaded to the AI ​​control platform.

[0065] Data analysis and prediction: The AI ​​platform uses DNN to predict the trend of air quality changes, and the RL algorithm dynamically adjusts the fan speed and air volume.

[0066] Air flow control: Based on the fluid mechanics model, optimize air flow distribution to avoid pollutant retention.

[0067] Filter optimization: Adaptive filtering algorithm dynamically adjusts the filter working status and promptly reminds operators to clean or replace it.

[0068] Temperature and humidity control: The temperature and humidity parameters are adjusted in real time through SVM and neural network, and the HVAC system is linked to ensure a constant environment.

[0069] Energy efficiency optimization: The energy efficiency optimization module analyzes energy consumption data, adjusts equipment operation intensity, and achieves energy conservation and emission reduction.

[0070] Remote maintenance: Operators can view the equipment status in real time through the APP and perform maintenance in a timely manner after receiving fault prediction reminders.

[0071] When the system is running, the air cleanliness can be maintained at ISO 7 standards, energy consumption is reduced by 15%, and equipment failure rate is reduced by 30%, significantly improving production efficiency.

[0072] The technical solution of the present invention is not limited to the above-mentioned specific embodiments. All technical variations made according to the technical solution of the present invention fall within the protection scope of the present invention.

Claims

1. A clean workshop air purification system, characterized by: The system comprises an adaptive control unit, an air quality monitoring module, an air flow control component, an environmental parameter adjustment component, a self-optimizing filter component, a human-machine interaction interface and a predictive maintenance system, an energy efficiency optimization module and a remote diagnosis and early warning system. The adaptive control unit predicts and adjusts various parameters of the clean room environment based on real-time data and historical data through a machine learning algorithm. The air quality monitoring module collects air quality data in real time by arranging sensors in the clean room, and transmits the data to the adaptive control unit to optimize the operation of the system in real time. The self-optimizing filter component dynamically adjusts the operating state of the filter system in combination with the adaptive control unit algorithm and the air quality data. The environmental parameter adjustment component dynamically adjusts the operating state of the HVAC system by analyzing the environmental parameters in real time. The air flow control component optimizes and controls the air flow of the air duct. The human-machine interaction interface and the predictive maintenance system provide an intuitive monitoring interface for operators, and diagnose potential equipment problems in advance based on the adaptive control unit and big data. The energy efficiency optimization module optimizes and manages energy consumption in combination with the data of the adaptive control unit and big data analysis. The remote diagnosis and early warning system understands the operating state of the air purification system in real time through intelligent remote monitoring, and performs remote fault diagnosis and early warning.

2. The clean room air purification system according to claim 1 is characterized in that: The machine learning algorithm of the adaptive control unit is based on reinforcement learning algorithm, deep neural network algorithm and data fusion algorithm. The sensors of the air quality monitoring module include M2.5 sensor, CO2 sensor, temperature and humidity sensor, air flow velocity sensor, particulate matter concentration sensor, and microbial detection sensor. Based on the data, the air pollution trend is predicted, the air treatment strategy is automatically adjusted, and the changes in the pollution source are responded to in time. The self-optimizing filter component monitors the load of the filter and adjusts the fan speed according to the filter status and air quality. The environmental parameter adjustment component automatically adjusts the HVAC system according to the current environmental changes in the clean room. The air flow control component analyzes the air flow path to avoid local air stagnation or pollution diffusion, and adjusts the air supply and exhaust volume according to the real-time monitoring of air quality. The remote diagnosis and early warning system checks the air quality, equipment status, and maintenance status of the clean room in real time through the APP or PC terminal, predicts potential equipment failures or air quality problems, and sends alarm information in advance. The energy efficiency optimization module automatically adjusts the system operation status to reduce energy consumption while maintaining the air quality required for the clean room; the human-computer interaction interface and predictive maintenance system predict equipment failure or performance degradation based on historical data and real-time data, and perform maintenance in advance.

3. A purification method for a clean room air purification system as claimed in claim 2, characterized in that: The following steps are involved: S1: Install a variety of air quality sensors in the clean room to monitor and collect air quality data in real time, and transmit the collected sensor data to the adaptive control unit in real time through the Internet of Things technology; S2: The adaptive control unit uses historical data and real-time data, adopts data fusion algorithm, integrates data from different sources into a unified analysis platform, and predicts the trend of clean room air quality based on machine learning algorithm, determines the future air pollution trend, and automatically adjusts the operation strategy of the air purification system; S3: Use sensors to monitor the air flow status in real time, adjust the air supply and exhaust volume according to the changes in air quality, and dynamically adjust the air flow direction in the air duct according to the analysis of the air flow control component to ensure that the air flows evenly in the clean room; S4: The environmental parameter adjustment component automatically adjusts the operating status of the HVAC system based on real-time data. If it detects that the temperature or humidity in the clean room exceeds the set range, the HVAC system automatically adjusts the air conditioning temperature or humidity level. The environmental parameter adjustment module adjusts the fan speed according to data such as temperature, humidity and air quality to ensure the smoothness and cleanliness of air flow; S5: The sensor monitors the pressure difference and flow rate changes of the filter, and the adaptive control unit adjusts the working state of the filtration system based on the air quality data. If the concentration of air pollutants increases, the fan speed is automatically increased; S6: Energy efficiency optimization module monitors the energy consumption of each device in the clean room in real time, analyzes the relationship between the energy consumption of the device and environmental data, automatically adjusts the operating status of the system, optimizes energy use, and intelligently adjusts the operating timing and intensity of the equipment according to the system operating status to ensure that the clean room always maintains the required air quality while reducing energy waste; S7: Monitor the operating status of key equipment through sensors and predict the risk of equipment failure. The prediction model compares historical data with real-time data to identify potential equipment failures in advance, and issues alarms through the remote diagnosis and early warning system to notify operators to conduct inspections and maintenance. S8: Through the remote diagnosis and early warning system, operators can view the air quality, equipment operating status, and maintenance records of the clean room in real time through the APP or PC. When equipment abnormalities occur, the system automatically sends alarm information to the operator through the early warning mechanism and provides preliminary fault diagnosis suggestions. Operators can take timely measures according to the system prompts.

4. The purification method of the clean room air purification system according to claim 3, characterized in that: The data fusion algorithm in step S2 is as follows: Where: The current status of the estimate; The estimated state at the previous moment, K k : gain; z k : Observation value of the sensor; H k : observation matrix; The machine learning algorithm is: Where: Predicted air quality output; W1, W2: weight matrix of the neural network, x t : Input data at the current moment; h t-1 : The hidden layer output at the previous moment; b1, b2: bias terms.

5. The purification method of the clean room air purification system according to claim 4, characterized in that: In step S3, the airflow direction in the air duct is dynamically adjusted using a fluid mechanics model and a reinforcement learning algorithm to analyze the air flow. The fluid mechanics model is as follows: (Conservation of mass) (Navier-Stokes equations) Where: v: air velocity; ρ: air density; t: time; p: pressure; μ: dynamic viscosity coefficient of air; f: volume force; The reinforcement learning algorithm is as follows: Where: Q(s t , a t ): state-action value function, indicating that in state s t Take action a t The value of; α: learning rate; r t : Rewards for the current step, related to air quality improvement or energy efficiency optimization; γ: discount factor, controlling the impact of future rewards, max a′ Q(s t+1 , a′): in the next state s t+1 Next, choose the value of the optimal action.

6. The purification method of the clean room air purification system according to claim 5, characterized in that: In step S5, the air quality data adaptive control unit adjusts the working state of the filtering system using an adaptive filtering algorithm, as shown in the following formula: Where: The filter output at the next moment; The filter output at the current moment; K t : Gain factor, which determines the impact of new observations on the filter output; z t : The observed value of the sensor.

7. The purification method of the clean room air purification system according to claim 6, characterized in that: In step S6, the operation timing and intensity of the intelligent adjustment device are adjusted using an energy efficiency optimization algorithm, as shown in the following formula: Where: E opt : Energy efficiency after optimization; P: Equipment power; C(P): Energy consumption cost of equipment; Q(P): Air quality, which changes with power; λ: trade-off factor, controlling the trade-off between energy consumption and air quality.

Images