An intelligent ventilation system for airborne pollutants and a control method thereof

Abstract

The application discloses an intelligent ventilation system for airborne pollutants and a control method thereof, and belongs to the technical field of industrial ventilation and dust removal; the system realizes real-time collection of dust concentration, wind speed and wind pressure data of each local ventilation and dust removal platform through a sensing and monitoring module, and realizes intelligent control through a remote monitoring module; the innovation lies in the fusion of dust concentration prediction, online parameter self-learning, multi-agent collaborative optimization and digital twin simulation algorithms, dynamic calculation of the required air volume of each branch, and synchronous adjustment of the branch air valve and the main fan speed through the PLC, so that the precise on-demand distribution of the air volume and the global energy saving of the system are realized; the application changes the passive response into the forward-looking collaborative control, and improves the dust removal efficiency, system safety and energy utilization rate.

Description

Technical Field

[0001] This invention relates to the field of industrial ventilation and dust removal technology, specifically to an intelligent ventilation system and control method for airborne pollutants. Background Technology

[0002] In dust-polluted environments such as laboratories and industrial production workshops, local ventilation and dust removal systems are crucial equipment for ensuring air quality and human health. Traditional ventilation and dust removal systems often operate with fixed airflow or manually adjusted branch dampers, which has the following significant drawbacks: 1. High energy consumption: Regardless of the amount of dust generated by the pollution source, the fans often operate at a fixed high speed, resulting in energy waste; 2. Poor control precision: They cannot respond to changes in dust concentration in real time, potentially leading to over-extraction when dust concentration is low and under-extraction when concentration is high, posing health and safety hazards; 3. Low level of automation: They rely on manual experience for adjustment, resulting in delayed response and difficulty in achieving airflow balance among multiple branches.

[0003] Existing technologies include some automatic control systems that incorporate sensors, but most of them only achieve alarms or simple fan start-stop. They fail to accurately and dynamically calculate and allocate the optimal air volume required for each pollution point based on dust concentration, and to perform closed-loop PID control of the fan and valves in coordination to achieve a balance between energy saving and efficient dust removal. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and provide an intelligent ventilation system and control method for airborne pollutants.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] This application provides an intelligent ventilation system for airborne pollutants, comprising:

[0007] The main ventilation module includes an exhaust fan for ventilation and a supply fan for air supply.

[0008] At least two local ventilation and dust removal stations, each of which is connected to the main ventilation module through an independent ventilation branch, and each ventilation branch is equipped with a branch air valve with adjustable opening.

[0009] The sensing and monitoring module includes a dust concentration sensor, a wind speed sensor, and a wind pressure sensor installed on each local ventilation and dust removal station or its corresponding ventilation branch.

[0010] The electrical control module includes a PLC controller, which is electrically connected to the sensor monitoring module, the branch air valve, the exhaust fan, and the air supply fan.

[0011] The remote monitoring module is communicatively connected to the PLC controller;

[0012] The remote monitoring module is configured as follows:

[0013] It receives and displays dust concentration, wind speed, and wind pressure data from the sensor monitoring module in real time;

[0014] Based on the real-time dust concentration at each local ventilation and dust removal station, the target air volume required for that local ventilation and dust removal station is calculated using a preset dust sparseness formula.

[0015] Based on the calculated target air volume and real-time wind speed / pressure data, PID control commands are generated, and the opening degree of the corresponding branch air valves and the speed of the exhaust fan and the supply fan are synchronously adjusted through the PLC controller so that the actual air volume of each ventilation branch dynamically approaches its corresponding target air volume.

[0016] Furthermore, the dust sparseness formula is a formula obtained by fitting experimental data, which describes the functional relationship between dust concentration and the air volume required to dilute it to a safe standard concentration.

[0017] Furthermore, the remote monitoring module is a host computer with configuration software installed.

[0018] Furthermore, the remote monitoring module also includes a dust concentration prediction unit, which uses a long short-term memory network model based on an attention mechanism for prediction. The calculation process is as follows:

[0019] The model input is a dust concentration sequence over T historical time steps:

[0020] ;

[0021] Model output is the future Predicted concentration sequences at each time step:

[0022] ;

[0023] The calculation formula for the model is:

[0024]

[0025] in, The model parameters are obtained through training with historical data; the remote monitoring module is based on the predicted concentration. Calculate the target air volume in advance and generate control commands.

[0026] Furthermore, the remote monitoring module also includes an adaptive airflow calculation engine, which uses an online learning algorithm to dynamically update the correction coefficients of the dust dispersion formula. Its update rules are as follows:

[0027]

[0028] in, These are the updated correction factors. This is the current correction factor. For learning rate, To actually measure the air volume, The air volume predicted based on the current formula. This represents the real-time dust concentration. The concentration is within the safe standard range.

[0029] Furthermore, the remote monitoring module employs a game theory-based multi-agent cooperative optimization algorithm for global airflow allocation, treating each ventilation branch as an agent with the following utility function:

[0030]

[0031] in, For the first The utility of an intelligent agent The air volume allocated to this intelligent agent. Based on its basic required air volume, , , These are the weighting coefficients; the system determines the optimal airflow for each branch by iteratively solving for the Nash equilibrium point. .

[0032] Furthermore, the remote monitoring module also includes a digital twin simulation unit, which constructs a virtual system based on a computational fluid dynamics model. Its governing equations include the Navier-Stokes equations and the dust transport equations.

[0033]

[0034]

[0035] in, air density, It is a velocity vector. For pressure, For stress tensor, For volume forces, The dust mass fraction. The diffusion coefficient is... For dust sources; the remote monitoring module pre-simulates control strategies in the digital twin environment and only sends the optimal strategy to the physical system.

[0036] Furthermore, the remote monitoring module also includes a safety boundary assessment module, which calculates in real time the safety margin of dust concentration at each point from the lower explosive limit (LEL). :

[0037]

[0038] when < When this happens, the system automatically switches to safety priority mode, ignores energy-saving targets, uses maximum airflow for ventilation, and issues a warning signal.

[0039] Secondly, this application provides a control method for an intelligent ventilation system for airborne pollutants, the control method comprising the following steps:

[0040] S1. The dust concentration value at each local ventilation and dust removal station is collected in real time by each dust sensor;

[0041] S2. Input the real-time dust concentration value at each local ventilation and dust removal station into the preset dust dilution formula, and calculate the target air volume required for each local ventilation and dust removal station respectively;

[0042] S3. Obtain real-time wind speed and / or wind pressure data for each ventilation branch;

[0043] S4. For each ventilation branch, the deviation between its target air volume and real-time air volume is used as the input of PID control to calculate the adjustment amount of the air valve of the branch and generate the first control command.

[0044] S5. Based on the total target air volume of all ventilation branches and the pressure parameters of the main ventilation duct, calculate the speed adjustment of the exhaust fan and the supply fan using a PID algorithm, and generate a second control command;

[0045] S6. The first control command and the second control command are sent to the PLC controller, which then performs synchronous adjustment of the branch air valves, exhaust fans and supply fans to realize the dynamic on-demand allocation of the total system air volume and the branch air volume.

[0046] Further, in step S2, the dust sparseness formula is:

[0047]

[0048] in, For the first The target air volume required for a local ventilation and dust removal station This represents the real-time dust concentration at the local ventilation and dust removal station. To set a safe standard concentration, This is a correction factor related to the characteristics of the local ventilation dust removal platform.

[0049] Compared with the prior art, this application has the following beneficial effects:

[0050] This invention proposes an intelligent ventilation system and its control method for airborne pollutants. It achieves feedforward and feedback composite control through a predictive model, transforming passive response into active intervention, significantly reducing adjustment lag. While meeting dust removal requirements, it minimizes the system's total energy consumption through a global optimization algorithm. Based on real-time feedback and online learning mechanisms, the airflow calculation model can adapt to changes in operating conditions, achieving more precise airflow allocation. Multi-agent collaborative optimization overcomes the control challenges of strongly coupled multi-variable systems. The entire process, from perception, prediction, optimization to execution, is fully automated and requires no manual intervention. The system possesses lifelong learning capabilities, continuously optimizing its performance. Advanced algorithms enable efficient coordination between fans and valves, maintaining airflow balance and proactively optimizing pipeline pressure distribution, thus improving system stability. Integrating dynamic safety boundary assessment and digital twin pre-simulation, it can proactively identify risks and take preventative measures, or perform safety verification of control commands, greatly enhancing the system's inherent safety level and decision reliability. Attached Figure Description

[0051] Figure 1 This is a schematic diagram of the overall structure of the intelligent ventilation system for airborne pollutants according to the present invention.

[0052] Figure 2 This is a schematic diagram of the overall connection structure of the intelligent ventilation system for airborne pollutants according to the present invention.

[0053] Figure 3 This is a schematic diagram of a ventilation branch module.

[0054] Figure 4 This invention relates to a remote software system for an intelligent ventilation system targeting airborne pollutants.

[0055] Figure label:

[0056] 1. Exhaust fan; 2. Air supply fan; 3. Local ventilation and dust removal platform; 4. Ventilation branch; 5. Branch air valve; 6. Dust concentration sensor; 7. Wind speed sensor; 8. Wind pressure sensor; 9. PLC controller; 10. Remote monitoring module. Detailed Implementation

[0057] 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.

[0058] Furthermore, in this invention, an element referred to as fixed to or disposed on another element may be directly disposed on the other element, or there may be an intermediate element. When an element is considered to be connected to another element, it may be directly connected to the other element, or there may be an intermediate element present simultaneously. The terms vertical, horizontal, left, right, and similar expressions used herein are for illustrative purposes only and do not represent the only possible implementation.

[0059] Example 1

[0060] See Figures 1-4 This application provides an intelligent ventilation system for airborne pollutants, comprising:

[0061] The main ventilation module includes an exhaust fan 1 for ventilation and an air supply fan 2 for air supply.

[0062] At least two local ventilation and dust removal stations 3, each local ventilation and dust removal station 3 is connected to the main ventilation module through an independent ventilation branch 4, and each ventilation branch 4 is equipped with a branch air valve 5 with an adjustable opening.

[0063] The sensing and monitoring module includes a dust concentration sensor 6, a wind speed sensor 7, and a wind pressure sensor 8 installed on each local ventilation and dust removal table 3 or its corresponding ventilation branch 4;

[0064] The electrical control module includes a PLC controller 9, which is electrically connected to the sensing and monitoring module, the branch air valve 5, the exhaust fan 1, and the air supply fan 2.

[0065] The remote monitoring module 10 is communicatively connected to the PLC controller 9;

[0066] The remote monitoring module 10 is configured as follows:

[0067] It receives and displays dust concentration, wind speed, and wind pressure data from the sensor monitoring module in real time;

[0068] Based on the real-time dust concentration at each of the three local ventilation and dust removal stations, the target air volume required for the local ventilation and dust removal station is calculated based on the preset dust sparseness formula.

[0069] Based on the calculated target air volume and real-time wind speed / pressure data, PID control commands are generated, and the opening degree of the corresponding branch air valve 5 and the speed of the exhaust fan 1 and the supply fan 2 are synchronously adjusted through the PLC controller 9 so that the actual air volume of each ventilation branch 4 dynamically approaches its corresponding target air volume.

[0070] This embodiment defines the basic architecture of an intelligent ventilation system for airborne pollutants. Traditional systems employ fixed airflow or manual adjustment modes, resulting in three major drawbacks: high energy consumption, poor control accuracy, and low automation level. The system hardware comprises: a main ventilation module (exhaust fan 1 and supply fan 2), at least two local ventilation and dust removal platforms 3, ventilation branches 4 and branch dampers 5, a sensor monitoring module (dust 6, wind speed 7, and wind pressure 8 sensors), an electrical control module (PLC controller 9), and a remote monitoring module 10. While these hardware components may be existing technologies, the inventiveness of this invention lies in their organization into a collaborative organic whole through specific intelligent control logic. It is particularly noteworthy that the core feature of this invention lies in the three main functional configurations of the remote monitoring module 10:

[0071] a. Real-time data reception and display;

[0072] b. Calculate the target air volume from the concentration based on the dust sparseness formula;

[0073] c. Based on the target air volume and actual air volume data, generate PID commands to synchronously adjust the branch air valves and fans.

[0074] This technical solution directly addresses the shortcomings of the background technology by changing the fixed air volume to a variable air volume based on real-time concentration calculation to solve the problem of high energy consumption, and changing the manual or simple automatic control to a closed-loop PID control based on a mathematical model to solve the problems of poor control accuracy and low automation level, thus achieving the technical effects of on-demand allocation and dynamic balance.

[0075] Furthermore, the dust sparseness formula is a formula obtained by fitting experimental data, which describes the functional relationship between dust concentration and the air volume required to dilute it to a safe standard concentration.

[0076] In this embodiment, the dust dispersion formula is further defined; this formula is based on a concentration-airflow conversion model established by the law of conservation of mass, for example: and This limitation clarifies the scientific basis for the conversion from concentration to air volume, elevating control decisions from experience-based judgment to model-based calculation. This formula is the core algorithm for the system to achieve precise control and intelligent calculation, which differs from existing technologies that only determine concentration thresholds and trigger alarms or start / stop functions.

[0077] Furthermore, the remote monitoring module 10 is a host computer with configuration software installed.

[0078] In this embodiment, the remote monitoring module 10 is specifically defined as a host computer with configuration software installed. The remote monitoring module 10 is an industrial computer with configuration software installed; this definition indicates the preferred and typical implementation carrier of the present invention, proving the practicality and feasibility of the technical solution; the configuration software is a mature platform in the field of industrial control, and is used here to carry and run the core control logic and algorithm of the present invention.

[0079] Furthermore, the remote monitoring module 10 also includes a dust concentration prediction unit, which uses a long short-term memory network model based on an attention mechanism for prediction. The calculation process is as follows:

[0080] The model input is a dust concentration sequence over T historical time steps:

[0081] ;

[0082] Model output is the future Predicted concentration sequences at each time step:

[0083] ;

[0084] The calculation formula for the model is:

[0085] ;

[0086] in, The model parameters are obtained through training with historical data; the remote monitoring module 10 calculates the predicted concentration. Calculate the target air volume in advance and generate control commands.

[0087] In this embodiment, a dust concentration prediction unit and a Long Short-Term Memory (LSTM) network model based on an attention mechanism are introduced. Existing traditional systems suffer from response lag; this embodiment provides a solution: using an LSTM-Attention model, with historical concentration sequences as input. Output future prediction sequence This technology enables the system to anticipate the trend of dust concentration changes, thereby calculating the target air volume in advance and generating control commands. It achieves a qualitative change from "passive feedback" to active feedforward and feedback composite control, effectively overcoming control lag and improving system response speed and control quality.

[0088] Furthermore, the remote monitoring module 10 also includes an adaptive airflow calculation engine, which uses an online learning algorithm to dynamically update the correction coefficients of the dust dispersion formula. Its update rules are as follows:

[0089] ;

[0090] in, These are the updated correction factors. This is the current correction factor. For learning rate, To actually measure the air volume, The air volume predicted based on the current formula. This represents the real-time dust concentration. The concentration is within the safe standard range.

[0091] In this embodiment, an adaptive airflow calculation engine and an online learning algorithm are introduced. This embodiment addresses the issue that parameters (such as the correction coefficient K) in the dust sparseness formula of the basic scheme may become inaccurate due to changes in operating conditions, and proposes an online learning mechanism. Its core formula is: ;

[0092] Essentially, it's a gradient descent parameter update strategy; it updates parameters by comparing actual airflow in real time. With model-predicted air volume To correct the deviation, adjust the model parameters in the opposite direction. This enables the air volume calculation model to continuously self-calibrate and adapt to slow time-varying factors such as changes in pipeline resistance and filter clogging, thereby ensuring the control accuracy and robustness of the system under long-term operation and demonstrating the system's self-learning capability.

[0093] Furthermore, the remote monitoring module 10 employs a game theory-based multi-agent cooperative optimization algorithm for global airflow allocation, treating each ventilation branch 4 as an agent with the following utility function:

[0094] ;

[0095] in, For the first The utility of an intelligent agent The air volume allocated to this intelligent agent. Based on its basic required air volume, , , These are the weighting coefficients; the system determines the optimal airflow for each branch by iteratively solving for the Nash equilibrium point. .

[0096] This embodiment introduces a multi-agent cooperative optimization algorithm and a game theory-based utility function. In a multi-branch ventilation system, the airflow regulation of each branch is coupled, making it a complex multivariate optimization problem. This embodiment models each ventilation branch as an agent and designs a comprehensive utility function for it:

[0097] ;

[0098] This function simultaneously considers control effectiveness (concentration deviation penalty), operational stability (airflow fluctuation penalty), and the collaborative relationship between branches. By solving the Nash equilibrium that maximizes the utility of each agent in a distributed manner, the system can find the globally optimal airflow allocation scheme with lower total energy consumption and more coordinated response while meeting the needs of all dust removal points. This effectively solves the problem of collaborative optimization of dynamic balance of multiple branches and global energy saving.

[0099] Furthermore, the remote monitoring module 10 also includes a digital twin simulation unit, which constructs a virtual system based on a computational fluid dynamics model. Its governing equations include the Navier-Stokes equations and the dust transport equations.

[0100] ;

[0101] ;

[0102] in, air density, It is a velocity vector. For pressure, For stress tensor, For volume forces, The dust mass fraction. The diffusion coefficient is... For dust sources; the remote monitoring module 10 pre-simulates the control strategy in the digital twin environment and only sends the optimal strategy to the physical system.

[0103] In this embodiment, a digital twin simulation unit and a model based on computational fluid dynamics (CFD) are introduced. This embodiment constructs a virtual simulation environment corresponding to the physical system, the core of which is the physical equations describing airflow and particulate transport (Navier-Stokes equations and dust transport equations). Any control strategy generated by optimization algorithms can be simulated in this virtual environment at the millisecond level, previewing the wind field, pressure field, and concentration field after its execution. This is equivalent to setting up a strategy sandbox for the control system. Only strategies that are verified to be safe, effective, and efficient through simulation will be issued to the actual equipment for execution. This feature greatly improves the safety and reliability of control decisions and avoids the risks and costs that may be caused by trial and error directly on the physical system.

[0104] Furthermore, the remote monitoring module 10 also includes a safety boundary assessment module, which calculates in real time the safety margin of dust concentration at each point from the lower explosive limit (LEL). :

[0105] ;

[0106] when < When this happens, the system automatically switches to safety priority mode, ignores energy-saving targets, uses maximum airflow for ventilation, and issues a warning signal.

[0107] In this embodiment, a safety boundary assessment module and safety margin calculation are introduced; this invention not only focuses on controlling dust concentration within occupational health standards (… Furthermore, a higher standard for industrial safety is introduced: the lower explosive limit (LEL) of dust; expressed by the formula:

[0108] ;

[0109] The system dynamically calculates the safety margin in real time. When the margin falls below a set threshold, the system will automatically trigger the highest priority safety override control. This feature elevates the system's function from occupational health protection to process safety control, thereby enhancing inherent safety.

[0110] Secondly, this application provides a control method for an intelligent ventilation system targeting airborne pollutants, which can be used in the aforementioned intelligent ventilation system targeting airborne pollutants. The control method includes the following steps:

[0111] S1. The dust concentration value at each local ventilation and dust removal platform 3 is collected in real time by each dust sensor 6;

[0112] S2. Input the real-time dust concentration values ​​at each of the three local ventilation and dust removal stations into the preset dust sparseness formula, and calculate the target air volume required for each local ventilation and dust removal station respectively;

[0113] S3. Obtain real-time wind speed and / or wind pressure data for each ventilation branch 4;

[0114] S4. For each ventilation branch 4, the deviation between its target air volume and real-time air volume is used as the input of PID control to calculate the adjustment amount of the damper 5 of that branch and generate the first control command.

[0115] S5. Based on the total target air volume of all ventilation branches 4 and the pressure parameters of the main ventilation duct, the speed adjustment of the exhaust fan 1 and the supply fan 2 is calculated by the PID algorithm, and a second control command is generated.

[0116] S6. The first control command and the second control command are sent to the PLC controller 9, and the PLC controller 9 performs synchronous adjustment of each branch air valve 5, exhaust fan 1 and supply fan 2 to realize dynamic on-demand allocation of the total system air volume and branch air volume.

[0117] In this embodiment, steps S1 to S6 of the method have clear logical correspondences in the content of this application and in the specific implementation, constituting a complete closed-loop control process of data acquisition, target air volume calculation, PID adjustment calculation, and coordinated execution; the method is the specific unfolding and implementation steps of the system function in terms of timing.

[0118] Further, in step S2, the dust sparseness formula is:

[0119] ;

[0120] in, For the first The target air volume required for a local ventilation and dust removal station This represents the real-time dust concentration at the local ventilation and dust removal station. To set a safe standard concentration, This is a correction factor related to the characteristics of the local ventilation dust removal platform.

[0121] In this embodiment, the following is adopted: This linearized formula is simple in form, with clear physical meaning of the parameters (K is the air volume correction coefficient), and is easy to calibrate and apply in engineering practice, reflecting the practicality and operability of the invention.

[0122] Example 2

[0123] As attached Figure 1 As shown, this system is equipped with two local ventilation and dust removal stations (3a, 3b). The local ventilation and dust removal station 3a is connected to the branch (4a), and a branch air valve (5a) is installed on the branch (4a). The PLC controller 9 is a Siemens Smart-200 series. The remote monitoring module 10 is an industrial computer with configuration software installed.

[0124] The dust sparse formula uses a simplified model: ,in The volume of the operating space for the local ventilation and dust removal station is given by K, which is an empirical coefficient, and C0 is set to 8 mg / m³ (national indoor standard). This formula is pre-installed in the software of the remote monitoring module 10.

[0125] After the system starts, data from each sensor is uploaded in real time. Assuming that at a certain moment, the dust concentration C1 of the local ventilation dust removal station (3a) rises to 50mg / m³, the remote monitoring module 10 immediately calculates its target air volume Q1 as 800m³ / h; at the same time, the concentration C2 of the local ventilation dust removal station (3b) is 10mg / m³, and Q2 is calculated to be 200m³ / h; the total target air volume of the system is 1000m³ / h.

[0126] The PID controller in the remote monitoring module 10 starts working: For the ventilation branch (4a), if its current air volume is 600m³ / h, which deviates from the target value of 800m³ / h, the PID algorithm outputs a command to increase the opening of the branch air valve (5a) by a certain proportion through the PLC controller 9; at the same time, in order to ensure that the total air volume reaches 1000m³ / h and the main pipeline pressure is stable, the PID algorithm synchronously calculates that the speed of the exhaust fan 1 and the supply fan 2 needs to be increased to the corresponding frequency; each actuator operates synchronously under the coordination of the PLC.

[0127] A few seconds later, the system reached a new equilibrium: the air volume of each branch stabilized near the target value, the dust concentration C1 began to decrease and approached C0, the whole process was completed automatically and displayed in real time on the remote monitoring interface.

[0128] Example 3

[0129] Based on the aforementioned Embodiments 1 and 2, this embodiment further clarifies how the present invention can be applied in actual industrial scenarios.

[0130] A metal grinding workshop has two independent grinding stations, corresponding to local ventilation and dust removal stations 3a and 3b respectively. The system hardware is configured as follows: the main ventilation module uses two variable frequency fans (exhaust fan 1 and supply fan 2), each with a rated air volume of 2000 m³ / h; each station is equipped with a dust collection hood, which is connected to the main duct via galvanized steel ducts (ventilation branches 4a and 4b); electrically adjustable branch dampers (5a, 5b) are installed on each branch duct to precisely control the air volume; the sensing and monitoring module includes a laser dust concentration sensor 6 installed in the dust collection hood, and a Testo thermal anemometer 7 and a micro-differential pressure sensor 8 installed on the branch ducts; the core of the electrical control module is a Siemens PLC controller 9, which achieves high-speed communication with all sensors, actuators, and remote monitoring modules through the Modbus communication protocol; the remote monitoring module 10 is a computer running a dedicated intelligent ventilation monitoring software developed independently based on C#, which integrates all the advanced algorithm modules described in this patent.

[0131] During the system initialization phase, key parameters are preset in the monitoring software: safety standard concentration. =8mg / m³, the lower explosive limit (LEL) of the dust involved in this grinding process is 60g / m³; the basic air volume calculation uses the formula And set initial correction coefficients for the two workstations. The dust concentration prediction module loads a pre-trained LSTM neural network model with an online learning rate of [missing information]. Set to 0.01. The alarm threshold for the safety boundary assessment module is set to... =30%.

[0132] The system starts up and enters normal operation. At this time, station 3b is not operating, and its dust concentration C2 is stable at 2 mg / m³. According to the formula, the required air volume Q2 is close to 0 m³ / h. Therefore, branch damper 5b is kept at its minimum opening to maintain a slight negative pressure on the dust collection hood. Station 3a is in a light polishing state, with a concentration C1 of 12 mg / m³. The target air volume is calculated. At this time, the total air volume required by the system is low, and the main fan maintains a low-speed operation mode under the control of the PLC, so the overall energy consumption is low.

[0133] Subsequently, workstation 3a was switched to heavy grinding operation, and the amount of dust generated increased sharply; the dust concentration sensor data showed that the C1 value rapidly climbed from 12mg / m³ to 50mg / m³ within 10 seconds; this change immediately triggered the system's intelligent response chain.

[0134] First, the dust concentration prediction unit, based on the current concentration rise rate and historical time-series data, uses an LSTM model to quickly extrapolate and predicts that the concentration at this point may reach 80 mg / m³ in about 30 seconds. Next, the multi-agent collaborative optimization unit is activated. This unit uses the predicted concentration as a forward-looking input to recalculate the airflow requirements of each branch. In the optimization calculation, the algorithm not only considers meeting the urgent needs of workstation 3a but also comprehensively evaluates the overall system status. It identifies that the current demand for branch 4b is extremely low and can be used as an adjustment margin. After rapid optimization, the algorithm decides on a global collaborative strategy: significantly increasing the opening of branch damper 5a from 30% to 85% to quickly increase suction capacity; simultaneously, to balance the main pipe pressure and avoid system fluctuations due to a surge in airflow, the strategy moderately reduces the opening of branch damper 5b from 10% to 5%; to match the increased demand for total airflow, the operating frequency of the main fan needs to be increased from 35Hz to 48Hz.

[0135] Before distributing this optimization strategy to the physical system, the digital twin simulation unit performed a real-time simulation verification at the second level. Based on the CFD model, the unit simulated the flow state of the pipeline network after the adjustment of the valves and fans, confirming that the wind speed of branch 4a after adjustment could meet the expected requirements, and that the pressure distribution of the entire pipeline network was stable, without any violent fluctuations or resonance risks. At the same time, the simulation showed that the dust concentration safety margin M1 at the key points was always higher than 50%, meeting the safety requirements. After the simulation passed, a set of synchronous control instructions was sent to the PLC through the Profibus-DP bus. The PLC controller then coordinated the actions of each actuator, and after about 5 seconds, the system reached a new dynamic equilibrium. At this time, the measured air volume at station 3a stabilized at about 820 m³ / h. The strong suction capacity effectively curbed the upward trend of dust concentration and began to decline steadily.

[0136] During the control process, the adaptive airflow calculation engine continuously operates; it compares the actual stabilized airflow (820 m³ / h) with the theoretical airflow calculated from the basic formula based on the final stable concentration, utilizing online learning rules ( Correction factor for workstation 3a Fine-tuning was performed to make the calculation model more closely match the actual dust generation and collection characteristics of the workstation, demonstrating the system's self-learning capability.

[0137] Throughout the entire incident, the safety boundary assessment module maintained real-time monitoring; assuming an extreme situation, such as a grinding spark igniting localized dust, causing the concentration to spike abnormally to 45 g / m³, the module immediately calculated the current safety margin. The value has fallen below the preset threshold of 30%; the system then triggers the highest priority override control: immediately issues an audible and visual alarm, and forcibly switches the branch air valves 5a and 5b to the fully open state, while driving the main fan to the maximum speed of the power frequency, so as to carry out emergency dilution and discharge with the maximum ventilation capacity, and do everything possible to ensure safety.

[0138] As can be seen from this embodiment, the system of the present invention upgrades traditional ventilation into an overall intelligent solution with sensing, prediction, optimization, simulation and learning capabilities, and achieves safe, accurate, efficient and energy-saving ventilation and dust removal control in real industrial scenarios.

[0139] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0140] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. An intelligent ventilation system for airborne pollutants, characterized in that, include: The main ventilation module includes an exhaust fan (1) for exhausting air and an air supply fan (2) for supplying air. At least two local ventilation dust removal stations (3), each local ventilation dust removal station (3) is connected to the main ventilation module through an independent ventilation branch (4), and each ventilation branch (4) is equipped with a branch air valve (5) with adjustable opening; The sensing and monitoring module includes a dust concentration sensor (6), a wind speed sensor (7), and a wind pressure sensor (8) installed on each local ventilation dust removal station (3) or its corresponding ventilation branch (4); The electrical control module includes a PLC controller (9), which is electrically connected to the sensor monitoring module, the branch air valve (5), the exhaust fan (1), and the air supply fan (2); and a remote monitoring module (10), which is communicatively connected to the PLC controller (9). The remote monitoring module (10) is configured as follows: It receives and displays dust concentration, wind speed, and wind pressure data from the sensor monitoring module in real time; Based on the real-time dust concentration at each local ventilation dust removal station (3), the target air volume required for the local ventilation dust removal station is calculated based on the preset dust sparseness formula. Based on the calculated target air volume and real-time wind speed / pressure data, PID control instructions are generated, and the opening degree of the corresponding branch air valve (5) and the speed of the exhaust fan (1) and the supply fan (2) are synchronously adjusted through the PLC controller (9) so that the actual air volume of each ventilation branch (4) dynamically approaches its corresponding target air volume. The remote monitoring module (10) also includes a dust concentration prediction unit, which uses a long short-term memory network model based on an attention mechanism for prediction. The calculation process is as follows: The model input is a dust concentration sequence over T historical time steps: ； The model outputs a predicted concentration sequence for future time steps: ； The calculation formula for the model is: ； in, The model parameters are obtained through training with historical data; the remote monitoring module (10) calculates the concentration based on the predicted concentration. Calculate the target air volume in advance and generate control commands.

2. The intelligent ventilation system for airborne pollutants according to claim 1, characterized in that, The dust dispersion formula is a formula obtained by fitting experimental data, which describes the functional relationship between dust concentration and the air volume required to dilute it to a safe standard concentration.

3. The intelligent ventilation system for airborne pollutants according to claim 1 or 2, characterized in that, The remote monitoring module (10) is a host computer with configuration software installed.

4. The intelligent ventilation system for airborne pollutants according to claim 1, characterized in that, The remote monitoring module (10) also includes an adaptive airflow calculation engine, which uses an online learning algorithm to dynamically update the correction coefficients of the dust sparseness formula. Its update rules are as follows: ； in, These are the updated correction factors. This is the current correction factor. For learning rate, To actually measure the air volume, The air volume predicted based on the current formula. This represents the real-time dust concentration. The concentration is within the safe standard range.

5. The intelligent ventilation system for airborne pollutants according to claim 1, characterized in that, The remote monitoring module (10) uses a game theory-based multi-agent collaborative optimization algorithm for global air volume allocation, treating each ventilation branch (4) as an agent with the following utility function: ； in, For the first The utility of an intelligent agent The air volume allocated to this intelligent agent. Based on its basic required air volume, , , These are the weighting coefficients; the system determines the optimal airflow for each branch by iteratively solving for the Nash equilibrium point. .

6. The intelligent ventilation system for airborne pollutants according to claim 1, characterized in that, The remote monitoring module (10) also includes a digital twin simulation unit, which constructs a virtual system based on a computational fluid dynamics model. Its governing equations include the Navier-Stokes equations and the dust transport equations. ； ； in, air density, It is a velocity vector. For pressure, For stress tensor, For volume forces, The dust mass fraction. Where is the diffusion coefficient. For dust source items; the remote monitoring module (10) pre-simulates the control strategy in the digital twin environment and only sends the optimal strategy to the physical system.

7. The intelligent ventilation system for airborne pollutants according to claim 1, characterized in that, The remote monitoring module (10) also includes a safety boundary assessment module, which calculates in real time the safety margin of dust concentration at each point from the lower explosive limit (LEL). : ； when < When this happens, the system automatically switches to safety priority mode, ignores energy-saving targets, uses maximum airflow for ventilation, and issues a warning signal.

8. A control method for an intelligent ventilation system for airborne pollutants according to any one of claims 1-7, characterized in that: The control method includes the following steps: S1. The dust concentration value at each local ventilation dust removal station (3) is collected in real time by each dust sensor (6); S2. Input the real-time dust concentration value at each local ventilation dust removal station (3) into the preset dust sparse formula, and calculate the target air volume required for each local ventilation dust removal station respectively; S3. Obtain real-time wind speed and / or wind pressure data for each ventilation branch (4); S4. For each ventilation branch (4), the deviation between its target air volume and real-time air volume is used as the input of PID control to calculate the adjustment amount of the branch air valve (5) and generate the first control command. S5. Based on the total target air volume of all ventilation branches (4) and the pressure parameters of the main ventilation duct, the speed adjustment of the exhaust fan (1) and the supply fan (2) is calculated by PID algorithm, and a second control command is generated. S6. The first control command and the second control command are sent to the PLC controller (9), and the PLC controller (9) performs synchronous adjustment of each branch air valve (5), exhaust fan (1) and air supply fan (2) to realize the dynamic on-demand allocation of the total air volume of the system and the air volume of the branch.

9. The control method according to claim 8, characterized in that, In step S2, the dust sparseness formula is: ； in, For the first The target air volume required for a local ventilation and dust removal station. This represents the real-time dust concentration at the local ventilation and dust removal station. To set the safety standard concentration, This is a correction factor related to the characteristics of the local ventilation dust removal platform.

Images