Outdoor workstation air quality monitoring and intelligent ventilation system and method

Abstract

The invention discloses an outdoor workstation air quality monitoring and intelligent ventilation system and method, and belongs to the technical field of environment monitoring and intelligent ventilation control. The system is composed of an environment sensing module, an intelligent control unit and a ventilation execution device. And the environment sensing module realizes pollutant concentration field reconstruction through a multi-dimensional data fusion algorithm. The intelligent control unit constructs a dynamic ventilation model based on an improved fuzzy PID algorithm, and generates an optimal ventilation strategy in combination with meteorological prediction data and workstation structure parameters. And the ventilation execution device adopts a modular design to realize directional air supply and pollutant blocking. The method comprises four steps of real-time environment data acquisition, dynamic risk assessment, ventilation parameter optimization and execution mechanism linkage control, and remote intelligent regulation and control are realized through cooperation of edge computing and a cloud platform. According to the scheme, a passive response mode of a traditional ventilation system is broken through, and the technical advantages of accurate pollutant tracing, dynamic optimization of ventilation efficiency and intelligent adjustment of equipment energy consumption are achieved.

Description

Technical Field

[0001] This invention relates to the field of environmental monitoring and intelligent ventilation control technology, and in particular to an outdoor workstation air quality monitoring and intelligent ventilation system and method. Background Technology

[0002] Outdoor workstations (such as construction sites, mining camps, and scientific research bases) often face air quality issues such as dust, harmful gases, and abnormal temperature and humidity during operations, directly impacting the health of workers and the safety of equipment operation. Current mainstream technologies employ a combination of fixed sensors and mechanical ventilation, for example, installing a single gas detector on the top of the workstation and using an axial flow fan for air exchange. However, such systems suffer from insufficient sensor coverage and rigid ventilation strategies. Especially in open or semi-enclosed outdoor environments, it is difficult to track the migration path of pollution sources in real time, and ventilation parameters cannot be dynamically adjusted according to weather changes, leading to energy waste and localized air quality control failures.

[0003] In the prior art, a Chinese patent (application number CN201320323587.0) discloses an "Indoor Air Purification and Filtration System Based on the Internet of Things," which monitors indoor PM2.5 concentration through a distributed sensor network and controls the start and stop of the fresh air unit based on a preset threshold. While this solution increases the sensor deployment density, its ventilation strategy relies on a fixed threshold and does not consider the coupling effect between the dynamic characteristics of pollutant diffusion and the outdoor environment. Furthermore, its embodiment uses a two-dimensional planar sensor layout, which cannot adapt to the heterogeneous three-dimensional environment of outdoor workstations, and it lacks an integrated mechanism for correcting ventilation efficiency based on meteorological data, resulting in a significant decrease in control accuracy under multi-variable interference scenarios outdoors.

[0004] Based on the aforementioned existing technologies, the following technical bottlenecks still exist in the field of air quality control for outdoor workstations: 1) Traditional sensor network layouts struggle to achieve three-dimensional spatial source tracing and concentration field reconstruction of pollutants; 2) Ventilation control models lack the ability to perform multi-factor coupled analysis of meteorological conditions, workstation structural parameters, and pollutant diffusion dynamics; 3) Actuators employ a single response mode, failing to coordinate directional air supply, filtration and purification, and energy consumption optimization objectives. Therefore, a multi-dimensional collaborative control system capable of dynamically sensing environmental parameters, intelligently generating ventilation strategies, and precisely executing them is urgently needed to solve the technical challenges of real-time air quality control in special outdoor scenarios. Summary of the Invention

[0005] In view of the aforementioned existing problems, the present invention is proposed.

[0006] Therefore, the present invention provides an outdoor workstation air quality monitoring and intelligent ventilation system and method that solves the problem of real-time air quality control in special outdoor scenarios.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions: In a first aspect, the present invention provides an outdoor workstation air quality monitoring and intelligent ventilation system, including an environmental sensing module, an intelligent control unit, and a ventilation actuator; The environmental sensing module consists of a sensor array composed of distributed PM2.5 sensors, VOCs sensors, temperature and humidity sensors, and air pressure sensors. It is used to collect air quality parameters and environmental parameters inside and outside the workstation in real time, and generate a heat map of pollution source distribution through a multi-dimensional data fusion algorithm. The intelligent control unit has a built-in dynamic ventilation model, which generates an adaptive ventilation strategy based on the output data of the environmental sensing module, workstation structural parameters, and external meteorological forecast data, combined with a preset air quality threshold. The dynamic ventilation model optimizes the fan speed, duct turning angle, and filter component working mode based on an improved fuzzy PID algorithm. The ventilation actuator includes a steerable duct module, a high-efficiency filter component, and a variable frequency fan array. It is used to execute the ventilation strategy generated by the intelligent control unit and achieve the directional dispersion of pollutants and the supply of clean air by adjusting the duct direction, filtration level, and air volume intensity.

[0008] As a preferred embodiment of the outdoor workstation air quality monitoring and intelligent ventilation system described in this invention, the redundant deployment of the sensor array is specifically as follows: sensor nodes are arranged in a three-dimensional grid topology inside the workstation, with the distance between adjacent nodes not exceeding 50% of the maximum detection radius of the sensor, and at least one backup node is set in each grid plane; meteorological monitoring nodes are arranged every 10 meters along the perimeter of the workstation, and the meteorological monitoring nodes integrate anemometers, rain gauges, and ultraviolet intensity detection modules, and communicate with edge computing nodes through a wireless self-organizing network to form an internal and external interconnected environmental perception network.

[0009] As a preferred embodiment of the outdoor workstation air quality monitoring and intelligent ventilation system described in this invention, the dynamic ventilation model generates and corrects the ventilation strategy in the following manner: based on the time-series data of pollutant concentration and historical meteorological data of the workstation over the past 30 days, a long short-term memory neural network (LSTM) is used to train a spatiotemporal prediction model for pollutant diffusion; during real-time operation, the wind speed and air pressure data of the current meteorological monitoring node are timestamped with the output of the prediction model, and the fan start-up threshold and duct turning priority in the ventilation strategy are corrected by a weighted moving average algorithm.

[0010] The pollutant diffusion prediction curve of the dynamic ventilation model is generated by the following formula: in, For position In time The predicted values ​​of pollutant concentrations This is the initial concentration. β is the natural attenuation coefficient, and β is the meteorological influence factor. For position In time The wind speed component, The diffusion kernel function characterizes the diffusion characteristics of pollutants in time and space.

[0011] As a preferred embodiment of the outdoor workstation air quality monitoring and intelligent ventilation system described in this invention, the steerable duct module comprises the following: a multi-degree-of-freedom rotation mechanism using a combination of a spherical universal joint and an electric push rod to achieve 360° horizontal rotation and ±45° pitch adjustment; guide vanes consisting of 12 sets of aviation aluminum arc-shaped blades, with the blade spacing dynamically adjusted by a micro stepper motor with a precision of 0.5°; a high-efficiency filter assembly with a built-in parallel duct, switching between HEPA filtration mode and activated carbon adsorption mode via an electric valve; and each fan in the variable frequency fan array having an independently configured PID controller for its drive circuit, and achieving coordinated airflow distribution via a CAN bus protocol.

[0012] As a preferred embodiment of the outdoor workstation air quality monitoring and intelligent ventilation system described in this invention, the redundant deployment of the sensor array satisfies the spatial coverage formula: in, This is the coverage coefficient. The total number of sensors, For the effective monitoring area of ​​a single sensor, The workstation volume is specified; sensors within the workstation are arranged according to a three-dimensional grid spacing d, satisfying the following requirements. , This represents the sensor's maximum detection radius.

[0013] As a preferred embodiment of the outdoor workstation air quality monitoring and intelligent ventilation system described in this invention, the deflection angle of the guide vanes of the steerable duct module is determined by both the target air supply direction and the pollutant concentration gradient, and its calculation formula is as follows: in, Coordinates of the target area requiring enhanced ventilation. This refers to the real-time coordinates of the air duct outlet. This is the proportional adjustment coefficient, with a value range of 0.1. 0.5, The spatial gradient of pollutant concentration at the current location is calculated using differential data from the sensor array. This spatial gradient serves as a control factor for the deflection angle of the guide vanes, adjusting the target airflow direction. The deflection angle of the guide vanes in the steerable duct module is jointly determined by the target airflow direction and the pollutant concentration gradient; therefore, it is added to the angle of the target airflow direction. The portion that is jointly adjusted by the proportional adjustment coefficient and the spatial gradient of pollutant concentration on the angle gradient of the target air supply direction at the current location.

[0014] As a preferred embodiment of the outdoor workstation air quality monitoring and intelligent ventilation method described in this invention The solution includes the following steps: Step S1: Real-time collection of PM2.5 concentration, VOCs concentration, temperature, humidity and air pressure data inside and outside the workstation through a distributed sensor array, and transmission to the edge computing node; Step S2: Reconstruct the pollutant concentration field distribution within the workstation based on a multi-dimensional data fusion algorithm, and calculate the pollutant diffusion trend by combining meteorological forecast data; Step S3: Generate a pollution level index through a dynamic risk assessment model based on the concentration field distribution, diffusion trend, and preset air quality safety threshold; Step S4: Dynamically optimize ventilation parameters, including fan speed, duct turning angle, and filtration mode, based on the pollution level index and workstation structural parameters using an improved fuzzy PID algorithm; Step S5: Send control commands to the ventilation actuators to synchronously drive the steerable duct, variable frequency fan, and filter components to achieve rapid dilution of pollutants and directional replenishment of clean air.

[0015] As a preferred embodiment of the outdoor workstation air quality monitoring and intelligent ventilation method of the present invention, wherein the multi-dimensional data fusion algorithm in step S2 reconstructs the concentration field using the following formula: in, For position The fusion concentration value, For spatial interpolation weights, These are observations from nearby sensors. For Kalman filter gain, To predict concentration, This represents the actual observed concentration.

[0016] As a preferred embodiment of the outdoor workstation air quality monitoring and intelligent ventilation method of the present invention, wherein the pollution level index R in step S3 is calculated using the following formula: in, For the first Type of pollutants in time concentration, For the corresponding safety threshold, As a toxicity weighting factor, Given the current population density in the area, To ensure the upper limit of safe density, To calculate the cumulative exposure time, To allow for safe exposure time, This is the dynamic adjustment coefficient.

[0017] As a preferred embodiment of the outdoor workstation air quality monitoring and intelligent ventilation method of the present invention, wherein: the model parameter update formula for ventilation efficiency optimization in step S5 is: in, The adaptive learning rate is set to 0.01. 0.1, The regional safe concentration threshold, For time The average actual monitored concentration, For a sign function, when the partial derivatives If positive, add 1; otherwise, take 1. .

[0018] In a second aspect, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the outdoor workstation air quality monitoring and intelligent ventilation method described in the first aspect of the present invention.

[0019] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein, when the computer program is executed by a processor, it implements any step of the method for monitoring air quality and intelligent ventilation of an outdoor workstation as described in the first aspect of the present invention.

[0020] The beneficial effects of this invention are: Enhanced 3D pollution monitoring capabilities: Through redundant deployment of a 3D gridded sensor array within the workstation and perimeter meteorological nodes (coverage coefficient ≥ 1.2), combined with multi-dimensional data fusion algorithms, the spatial reconstruction accuracy of pollutant concentration fields is improved by more than 40%, effectively solving the problem of pollution source location deviation caused by traditional 2D layout.

[0021] Dynamic ventilation intelligent decision-making: Based on the diffusion prediction model trained by LSTM and the real-time meteorological data correction mechanism, the response speed of ventilation strategy is improved by 50%. In the event of a sudden pollution incident, the peak concentration of pollutants can be suppressed for less than 30 minutes, while avoiding energy waste caused by excessive ventilation.

[0022] Optimized directional purification efficiency: The steerable air duct module achieves a clean air delivery direction error of less than 5° through multi-degree-of-freedom adjustment (360° horizontal rotation and ±45° pitch) and dynamic spacing control of the guide vanes. Combined with the HEPA / activated carbon dual-mode filter components, the removal efficiency of PM2.5 and VOCs in the target area reaches more than 98% and 85%, respectively.

[0023] System energy consumption coordinated control: The distributed PID drive of the variable frequency fan array and the dynamic adaptation mechanism of the power grid load reduce energy consumption by 25% compared with the traditional fixed frequency system while ensuring ventilation requirements are met. 40%, and through collaborative optimization of edge computing and cloud platforms, daily carbon emissions are reduced by 15%.

[0024] Scene adaptability and scalability: The modular design supports rapid replacement and expansion of sensor nodes, duct components, and filter units, and can adapt to different volumes (50). An outdoor workstation of 2000m³, operating under extreme temperature and humidity conditions. 30℃ to 60℃) and dust concentration (0 It still maintains stable operation under conditions of 1000 μg / m³. Attached Figure Description

[0025] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a schematic diagram of an outdoor workstation air quality monitoring and intelligent ventilation system in Example 1; Figure 2 This is a flowchart of an outdoor workstation air quality monitoring and intelligent ventilation method in Example 2. Detailed Implementation

[0027] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0028] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0029] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0030] Example 1, referring to Figure 1 This is the first embodiment of the present invention, which provides an outdoor workstation air quality monitoring and intelligent ventilation system, including an environmental sensing module, an intelligent control unit, and a ventilation actuator; The environmental sensing module consists of a sensor array composed of distributed PM2.5 sensors, VOCs sensors, temperature and humidity sensors, and air pressure sensors. It is used to collect air quality parameters and environmental parameters inside and outside the workstation in real time, and generate a heat map of pollution source distribution through a multi-dimensional data fusion algorithm. The intelligent control unit has a built-in dynamic ventilation model, which generates an adaptive ventilation strategy based on the output data of the environmental sensing module, workstation structural parameters, and external meteorological forecast data, combined with a preset air quality threshold. The dynamic ventilation model optimizes the fan speed, duct turning angle, and filter component working mode based on an improved fuzzy PID algorithm. The ventilation actuator includes a steerable duct module, a high-efficiency filter component, and a variable frequency fan array. It is used to execute the ventilation strategy generated by the intelligent control unit and achieve the directional dispersion of pollutants and the supply of clean air by adjusting the duct direction, filtration level, and air volume intensity.

[0031] The redundant deployment of the sensor array is as follows: sensor nodes are arranged in a three-dimensional mesh topology inside the workstation, with the distance between adjacent nodes not exceeding 50% of the maximum detection radius of the sensor, and at least one backup node is set in each mesh plane; meteorological monitoring nodes are arranged at 10-meter intervals along the perimeter of the workstation. The meteorological monitoring nodes integrate wind speed and direction instruments, rainfall sensors and ultraviolet intensity detection modules, and communicate with edge computing nodes through a wireless self-organizing network to form an internal and external linkage environmental perception network.

[0032] The dynamic ventilation model generates and corrects the ventilation strategy in the following way: Based on the time series data of pollutant concentration and historical meteorological data of the workstation over the past 30 days, a long short-term memory neural network (LSTM) is used to train the spatiotemporal prediction model of pollutant diffusion; during real-time operation, the wind speed and air pressure data of the current meteorological monitoring node are timestamped with the output of the prediction model, and the fan start threshold and duct turning priority in the ventilation strategy are corrected by a weighted moving average algorithm.

[0033] The pollutant diffusion prediction curve of the dynamic ventilation model is generated by the following formula: in, For position In time The predicted values ​​of pollutant concentrations This is the initial concentration. β is the natural attenuation coefficient, and β is the meteorological influence factor. For position In time The wind speed component, The diffusion kernel function characterizes the diffusion characteristics of pollutants in time and space.

[0034] The specific structure of the steerable air duct module includes: a multi-degree-of-freedom rotation mechanism that uses a combination of a spherical universal joint and an electric push rod to achieve 360° horizontal rotation and ±45° pitch adjustment; guide vanes consisting of 12 sets of aviation aluminum arc-shaped blades, with the blade spacing dynamically adjusted by a micro stepper motor with a precision of 0.5°; the high-efficiency filtration component with a built-in parallel air duct, which switches between HEPA filtration mode and activated carbon adsorption mode via an electric valve; and the drive circuit of each fan in the variable frequency fan array is independently configured with a PID controller, and the air volume is coordinated and distributed through the CAN bus protocol.

[0035] The redundant deployment of sensor arrays satisfies the spatial coverage formula: in, This is the coverage coefficient. The total number of sensors, For the effective monitoring area of ​​a single sensor, The workstation volume is specified; sensors within the workstation are arranged according to a three-dimensional grid spacing d, satisfying the following requirements. , This represents the sensor's maximum detection radius.

[0036] The deflection angle of the guide vanes in the steerable duct module is determined by both the target airflow direction and the pollutant concentration gradient, and its calculation formula is as follows: in, Coordinates of the target area requiring enhanced ventilation. This refers to the real-time coordinates of the air duct outlet. This is the proportional adjustment coefficient, with a value range of 0.1. 0.5, The spatial gradient of pollutant concentration at the current location is calculated using differential data from the sensor array. This spatial gradient serves as a control factor for the deflection angle of the guide vanes, adjusting the target airflow direction. The deflection angle of the guide vanes in the steerable duct module is jointly determined by the target airflow direction and the pollutant concentration gradient; therefore, it is added to the angle of the target airflow direction. The portion that is jointly adjusted by the proportional adjustment coefficient and the spatial gradient of pollutant concentration on the angle gradient of the target air supply direction at the current location.

[0037] After the environmental sensing module is activated, PM2.5, VOCs, temperature, humidity, and air pressure sensors distributed in a three-dimensional grid inside the workstation synchronously collect data at 5-second intervals. Meteorological monitoring nodes deployed along the perimeter of the workstation capture wind speed, wind direction, and rainfall information in real time. All sensor data is uploaded to the edge computing node via the LoRa wireless communication protocol. A multi-dimensional data fusion algorithm filters outliers and generates a three-dimensional pollutant concentration distribution heatmap within the workstation based on spatial interpolation. High-concentration areas in the heatmap are marked with a red gradient, while low-concentration areas are displayed in green. Simultaneously, external meteorological data is overlaid to form a dynamic diffusion trend prediction layer.

[0038] After receiving the heat map and diffusion trend data output by the environmental sensing module, the intelligent control unit first loads the pre-stored workstation structural parameters (including space volume, door and window locations, and equipment layout). Combined with weather change information for the next two hours provided by the meteorological forecasting platform, it inputs this information into the dynamic ventilation model for strategy calculation. The dynamic ventilation model analyzes the deviation between the current pollution level and the target threshold using an improved fuzzy PID algorithm, dynamically adjusting control parameters. When PM2.5 concentration exceeds the limit, it prioritizes increasing fan speed and activating HEPA filtration mode; when VOCs concentration exceeds the limit, it switches to activated carbon adsorption mode and optimizes the duct turning angle. The strategy is rolled over every 5 minutes based on the latest sensor data to ensure that ventilation commands match real-time environmental changes.

[0039] After receiving instructions from the intelligent control unit, the ventilation actuator first activates the multi-degree-of-freedom rotation mechanism of the directional duct module. This mechanism drives the spherical universal joint to rotate to the target angle via an electric push rod, while the guide vanes simultaneously adjust their spacing to focus the airflow. The high-efficiency filter component switches electric valves according to instructions, directing airflow towards the HEPA filter or activated carbon adsorption layer; the filtration mode switching time is no more than 10 seconds. Each fan in the variable frequency fan array receives independent speed commands via a CAN bus, employing a ramp-up strategy to avoid current surges, and enters a steady-state PID control mode after reaching the target airflow. During execution, the environmental sensing module continuously monitors the rate of decrease in pollutant concentration in the target area. If the expected effect is not achieved, the edge computing node triggers a strategy replanning process, forming a closed-loop control of "perception-decision-execution-feedback". The entire system automatically activates an anti-interference mode under extreme weather conditions, maintaining ventilation stability through an air pressure compensation algorithm, and maintaining a minimum safe operating state by relying on locally cached data during network interruptions.

[0040] Example 2, refer to Figure 2 This is the second embodiment of the present invention, which provides a method for monitoring air quality and intelligent ventilation in an outdoor workstation, including the following steps: Step S1: Collect PM2.5 concentration, VOCs concentration, temperature, humidity and air pressure data inside and outside the workstation in real time through a distributed sensor array, and transmit them to the edge computing node; During the system startup and initialization phase, the 3D gridded sensor array inside the workstation is activated synchronously with the perimeter meteorological monitoring nodes. The sensor nodes inside the workstation are deployed according to preset 3D spatial coordinates, with the spacing between nodes in each grid layer strictly controlled within 50% of the sensor's maximum detection radius (e.g., when the PM2.5 sensor's detection radius is 5 meters, the node spacing is set to 2.5 meters). Redundant backup nodes are added at key intersections in each horizontal layer (such as passageway intersections and areas with dense equipment), forming a dual data acquisition link. The PM2.5 sensor in the sensor array uses the laser scattering principle, completing a particulate matter concentration sample every 2 seconds; the VOCs sensor is based on electrochemical detection technology, outputting volatile organic compound concentration values ​​every 5 seconds; the temperature and humidity sensor and the barometric pressure sensor are integrated into the same probe, updating environmental parameters every 1 second. The perimeter meteorological monitoring nodes outside the workstation are fixed to the fence or bracket at 10-meter intervals. The internal components include an ultrasonic anemometer (measurement range 0-30m / s, accuracy ±0.3m / s), a tipping bucket rain gauge (resolution 0.2mm), and an ultraviolet intensity sensor (wavelength range 280-400nm). All meteorological data are collected in 10-second cycles and transmitted via the LoRaWAN protocol.

[0041] During data acquisition, edge computing nodes ensure time alignment between internal and external sensors through a timestamp synchronization mechanism. Each sensor node has a built-in high-precision clock chip, which synchronizes with the edge node via the NTP protocol at 00:00 every day, with clock deviation controlled within ±10 milliseconds. The collected raw data undergoes preliminary preprocessing at the sensor end: PM2.5 data is filtered using a sliding window averaging filter (window length of 5 sampling points), VOCs data is filtered using median filtering to remove transient interference, and temperature and humidity data are verified for reasonableness based on historical 24-hour curves (an anomaly flag is triggered if a sudden temperature change exceeds 10℃ / minute). The preprocessed data packet contains the sensor ID, timestamp, calibrated values, and data quality identifier, and is transmitted to the edge node via a LoRa wireless module on the 470MHz band. In areas with severe signal interference, it automatically switches to the 868MHz backup band to ensure a transmission success rate of ≥99.5%.

[0042] After receiving data, the edge computing node first performs data integrity verification: checking the data packet header checksum, timestamp continuity, and the validity of the numerical range (e.g., if the PM2.5 concentration exceeds 1000 μg / m³, it is determined to be a sensor malfunction). Verified data enters the cache queue and is reconstructed according to spatial topology to build a spatiotemporal matrix of environmental parameters inside and outside the workstation. For missing data due to transmission packet loss, a triple redundancy strategy is used for recovery: replacement data is preferentially obtained from backup nodes in the same layer of the grid; if no backup is available, linear interpolation is performed based on data from adjacent nodes; finally, historical data from the same period (average of the same time over the past 7 days) is used as supplementary data. After data reconstruction, the edge node marks the complete spatiotemporal dataset as "ready" and pushes it to the data preprocessing interface of the intelligent control unit, while simultaneously retaining a rolling cache of the most recent hour locally for emergency backtracking. The entire acquisition and transmission link activates an anti-interference mode under extreme environments (such as heavy rain and sandstorms), reducing the sensor sampling frequency by 50% to improve signal stability and reducing the size of a single data packet through an incremental transmission mechanism to ensure continuous monitoring of core parameters.

[0043] Step S2: Reconstruct the pollutant concentration field distribution within the workstation based on a multi-dimensional data fusion algorithm, and calculate the pollutant diffusion trend by combining meteorological forecast data; The multi-dimensional data fusion algorithm described in step S2 reconstructs the concentration field using the following formula: in, For position The fusion concentration value, For spatial interpolation weights, These are observations from nearby sensors. For Kalman filter gain, To predict concentration, This represents the actual observed concentration.

[0044] After receiving the complete spatiotemporal dataset transmitted in step S1, the edge computing node starts the multi-dimensional data fusion engine. First, spatial interpolation is performed on the 3D mesh sensor data within the workstation: for areas not directly covered by sensors (such as equipment gaps or corners), the concentration contribution weights of the eight adjacent nodes are calculated based on the inverse distance weighted algorithm (IDW). Combined with the attenuation characteristics of the sensor detection radius, a 3D voxel model of pollutant concentrations covering the entire workstation is generated. For PM2.5 and VOCs concentration data, independent interpolation layers are established, and the interpolation results are smoothed using convolutional kernels to eliminate local abrupt changes caused by sensor noise. After interpolation, the system projects the 3D voxel model onto horizontal and vertical cross-sections to generate a visual heatmap, where the color gradient transitions from dark red (high concentration) to green (low concentration), and the geographical coordinates of areas exceeding concentration limits are marked.

[0045] Meanwhile, the meteorological forecast data access module obtains refined forecast information for the next two hours from an external meteorological service API, including wind speed and direction matrices updated every 10 minutes, rainfall intensity, and atmospheric stability levels. The system spatiotemporally aligns the forecast data with real-time observations from perimeter meteorological monitoring nodes, decomposes the forecast wind speed into U / V components through timestamp matching, and calculates the difference between the forecast and real-time anemometer data in the same coordinate system (such as the northeast-sky coordinate system) to generate a wind speed correction coefficient field. This correction coefficient field is superimposed onto the workstation spatial model in a grid format to correct the predicted results of pollutant diffusion direction and rate.

[0046] The calculation of pollutant diffusion trends is achieved by coupling a physical model with a data-driven model: 1. Physical Model Layer: Based on the principles of Computational Fluid Dynamics (CFD), using the workstation's 3D structural model as boundary conditions and combined with real-time wind speed and direction data, the basic flow field distribution under different meteorological scenarios is pre-calculated. The flow field data is updated every 30 minutes and stored as a velocity vector matrix for real-time retrieval.

[0047] 2. Data-driven layer: Using time series prediction models (such as LSTM networks) trained on historical pollution event datasets, analyze the correlation between the current concentration field and historical similar patterns, and predict the natural decay trend of pollutant concentrations within the next 15 minutes.

[0048] 3. Dynamic Correction: The diffusion direction output by the physical model is fused with the decay rate predicted by the data-driven model, and the prediction results are continuously corrected using a Kalman filter algorithm. Specifically, every 5 minutes, the latest sensor-observed concentration is compared with the predicted value of the previous period, the residual is calculated, and the weight parameters of the diffusion kernel function are adjusted so that the predicted curve gradually approximates the actual diffusion path.

[0049] The final generated diffusion trend is presented in the form of a dynamic vector field. The system automatically marks high-risk diffusion areas (such as concentration accumulation areas near ventilation openings) and potential cross-contamination paths. The system then packages the fused concentration field and diffusion trend data into a structured message through edge nodes and pushes it to the strategy generation module of the intelligent control unit.

[0050] During the data verification phase, the system cross-validates the interpolation results using independent observations from redundant sensor nodes. If the concentration difference reported by different nodes at the same location exceeds 20%, an anomaly detection process is triggered, data is re-collected, and the interpolation model is locally updated. Simultaneously, the diffusion trend prediction results are compared offline with the measured concentration change rate every hour. When the prediction error exceeds 15% for three consecutive times, the model retraining mechanism is automatically triggered to ensure algorithm adaptability.

[0051] Step S3: Based on the concentration field distribution, diffusion trend, and preset air quality safety threshold, generate a pollution level index through a dynamic risk assessment model; The formula for calculating the pollution level index R in step S3 is as follows: in, For the first Type of pollutants in time concentration, For the corresponding safety threshold, As a toxicity weighting factor, Given the current population density in the area, To ensure the upper limit of safe density, To calculate the cumulative exposure time, To allow for safe exposure time, This is the dynamic adjustment coefficient.

[0052] After accurately modeling the distribution and diffusion trends of pollutant concentration fields, the system enters the dynamic risk assessment phase. The intelligent control unit first retrieves preset air quality safety threshold parameters from the local database. These thresholds not only cover the concentration limits for pollutants such as PM2.5 and VOCs stipulated by national standards (e.g., the upper limit for the 24-hour average concentration of PM2.5 is 35 micrograms per cubic meter), but also load customized restrictions based on the specific operation type of the workstation. For example, in areas involving welding operations, an additional industry standard is set that the concentration of welding fumes must not exceed 5 milligrams per cubic meter; in chemical storage areas, sub-thresholds are set for specific VOCs components such as benzene and formaldehyde. The threshold data is bound to the workstation's 3D model in the form of a spatial mapping table to ensure that each area matches the corresponding pollution limit requirements.

[0053] The core operation of the dynamic risk assessment model consists of two stages: multi-level data fusion and risk quantification. First, the model performs spatial cluster analysis on the real-time concentration field, identifying areas continuously exceeding limits and marking them as "core pollution sources." For example, if more than 80% of the sensor nodes within a cubic area report PM2.5 concentrations above 50 micrograms per cubic meter for 10 consecutive minutes, the area is determined to be a heavily polluted core area. Subsequently, combining diffusion trend prediction data, the model analyzes the migration paths of pollutants under airflow, predicting the area that may be affected within the next 10 minutes, especially areas with high population density or sensitive equipment (such as power distribution rooms and control consoles).

[0054] The risk quantification process incorporates a dynamic weighting mechanism, comprehensively calculating the pollution level index from three dimensions: toxicity hazard, exposure duration, and personnel density. Toxicity hazard weights are dynamically adjusted based on pollutant type: PM2.5 is assigned a weight of 1.0 as the baseline unit; benzene-based VOCs are assigned a weight of 2.0 due to their carcinogenicity; formaldehyde-based VOCs are assigned a weight of 1.5; and abnormal temperature and humidity are assigned auxiliary weights ranging from 0.3 to 0.8 based on the percentage deviation from the threshold. The exposure duration correction module tracks the cumulative exceedance time for each area in real time. If the PM2.5 concentration in a certain area exceeds the limit for more than 30 minutes, the risk value will increase non-linearly at a rate of 20% per hour. Personnel density data is collected in real time via UWB positioning tags deployed on safety helmets or work badges. If the personnel density in a high-risk area reaches 0.5 people per square meter or more, the risk level automatically increases by one level and triggers an audible and visual alarm.

[0055] To enhance the foresight of risk prediction, the model includes a "potential high-risk area" prediction mechanism. When the diffusion trend indicates that a pollution plume is moving towards a populated area at a concentration increase of more than 5% per minute, even if the current concentration in that area is not exceeded, the system will still mark it as an orange alert area in advance and activate preventative ventilation strategies. The risk level of all areas is quantified using an index from 0 to 100, divided into four levels: safe (green, 0-30), low risk (yellow, 31-60), moderate risk (orange, 61-80), and high risk (red, 81-100). Each level corresponds to a different response strategy. For example, orange-level areas will activate a directional pressurized air supply mode, while red-level areas will forcibly activate maximum airflow ventilation and push evacuation recommendations to management terminals.

[0056] To ensure the reliability of the assessment results, the system incorporates multiple verification mechanisms. The spatial redundancy verification module cross-compares data from multiple sensors within the same area. If more than 30% of the node data deviates by more than 20%, abnormal data is automatically removed, and backup nodes are invoked for recalculation. The temporal continuity verification module monitors abrupt changes in risk levels. If the risk level of an area jumps by more than two levels within one minute (e.g., from green to orange), the assessment results for that area are frozen, triggering a manual review process, and backup sensors are activated for secondary verification. The prediction backtracking verification module performs offline analysis every hour, comparing the risk prediction results from the past hour with the actual concentration change curve. If the prediction deviation rate exceeds 15%, historical datasets are automatically used to incrementally train the LSTM model and optimize the prediction parameters.

[0057] The final pollution level index is integrated into the workstation's 3D management platform in a visualized form. Operators can view risk distribution heatmaps of any cross-section through the interactive interface. Clicking on a specific area allows for detailed analysis of risk factor composition, such as displaying decomposed data like "PM2.5 contribution 65% + benzene concentration contribution 25% + personnel density bonus 10%". The assessment results are synchronized in real time to the ventilation strategy engine, assigning dynamic priority tags to each variable frequency fan and duct module to ensure that high-risk areas receive priority ventilation resources. Simultaneously, the system encrypts and uploads the complete risk assessment log (including timestamps, geographic coordinates, risk values, and decision logic chains) to the cloud knowledge base for later use in generating compliance reports, accident tracing, and model iteration optimization. In extreme cases (such as network outages), edge nodes can maintain emergency decision-making capabilities for at least 24 hours based on locally cached historical risk assessment models.

[0058] Step S4: Based on the pollution level index and workstation structural parameters, dynamically optimize ventilation parameters, including fan speed, duct turning angle and filtration mode, using an improved fuzzy PID algorithm. After the pollution level index is generated, the intelligent control unit activates an improved fuzzy PID control engine to dynamically optimize ventilation parameters based on workstation structural parameters (such as space volume, door and window layout, and equipment distribution). First, the system maps the pollution level index to fuzzy control input variables: risk level (low, medium, high), pollution diffusion rate (slow, medium, rapid), and target area priority (ordinary, critical, core). Each input variable is converted into a fuzzy linguistic variable through a membership function. For example, when the risk level is "high" and the diffusion rate is "rapid," rules with a membership degree exceeding 0.8 will trigger a strong intervention mode.

[0059] The control engine loads the 3D model data of the workstation structure and sets differentiated control strategies for different areas. For high-ceilinged areas (such as warehouse areas), priority is given to increasing the vertical air supply intensity; for narrow passages or densely populated equipment areas, the focus is on the accuracy of air duct turning and the suppression of local turbulence. The model dynamically updates boundary conditions based on real-time structural changes (such as the construction of temporary partitions or equipment relocation) to ensure that the ventilation strategy matches the actual spatial form.

[0060] The core improvement of the fuzzy PID algorithm lies in the introduction of an adaptive rule base: 1. Rule self-learning mechanism: Based on historical control effect data (such as the pollution removal efficiency after a 20% increase in fan speed under an orange risk level), the system automatically optimizes the weights of fuzzy rules. For example, if the traditional PID integral term causes duct oscillation in a certain scenario, the system will reduce the integral weight and increase the proportion of the derivative term.

[0061] 2. Environmental Coupling Compensation: Meteorological forecast data (such as wind speed changes in the next 10 minutes) is injected into the control loop in advance to pre-adjust the fan speed before a sudden increase in external wind speed, thereby offsetting the interference of natural wind on directional air delivery.

[0062] 3. Energy balancing module: Dynamically adjusts the priority of control targets based on the real-time load status of the power grid, prioritizing ventilation in the core area during peak electricity consumption periods, while using intermittent air supply in non-core areas to reduce overall power consumption.

[0063] The specific parameter optimization process is as follows: Fan speed regulation: The target air volume is set based on the pollution level index, and the speed adjustment is output through fuzzy PID. For example, when the fan associated with the red risk area needs to increase its speed, a ramp acceleration strategy (increasing the rated speed by 15% per minute) is adopted to avoid current surges; at the same time, according to the adjacent fan coordination rules, airflow collisions caused by multiple fans increasing speed at the same time are prevented.

[0064] Airflow steering control: Combining the coordinates of the pollution source with the diffusion trend vector field, the optimal airflow direction is calculated. After receiving the steering command, the multi-degree-of-freedom rotating mechanism drives the spherical universal joint via an electric actuator to gradually adjust the angle of the guide vanes with an accuracy of 0.5°. The spacing between the guide vanes is simultaneously reduced to focus the airflow, and when it is necessary to cover a large area, the spacing is increased to the maximum allowable value.

[0065] Filtration mode switching: The operating mode of the filter components is determined based on the proportion of pollutant types. When the VOCs concentration exceeds 60%, the electric valve directs the airflow to the activated carbon adsorption layer and performs a 3-second purge process before switching to remove residual particulate matter; when PM2.5 is the main pollution source, the HEPA filter activates the pulse self-cleaning function, which reverses the airflow every 30 minutes to remove accumulated dust and maintains a filtration efficiency of ≥95%.

[0066] During the control process, the system monitors the execution effect in real time, collecting the pollutant concentration reduction rate in the target area every 2 minutes. If the actual rate is lower than 80% of the expected value, a control parameter recalibration process is triggered. For example, when the actual air volume of a certain fan decreases due to filter blockage, the system automatically compensates by increasing the fan speed and marks the fan as needing maintenance and inspection. Simultaneously, edge computing nodes record the execution effect data of each control decision (such as energy consumption increment, pollution removal efficiency, and equipment wear index) for weekly global rule base optimization, gradually improving control accuracy and equipment lifespan.

[0067] To ensure system robustness, multiple fault-tolerance mechanisms are implemented. When a partial failure occurs in the sensor network, the system automatically switches to a backpropagation mode based on a diffusion model. When communication latency exceeds 500 milliseconds, the latest valid control parameters cached locally are used to maintain operation. For critical actuators (such as variable frequency fan drive boards), a dual CAN bus redundancy design is adopted, with millisecond-level switching between primary and backup channels to ensure continuous command execution. The final optimized parameter set is distributed to each actuator via industrial Ethernet and synchronously updated to the cloud platform control log, forming a complete and traceable decision chain.

[0068] Step S5: Send control commands to the ventilation actuator to synchronously drive the steerable air duct, variable frequency fan and filter components to achieve rapid dilution of pollutants and directional supply of clean air.

[0069] The formula for updating the model parameters for optimizing ventilation efficiency in step S5 is as follows: in, The adaptive learning rate is set to 0.01. 0.1, The regional safe concentration threshold, For time The average actual monitored concentration, For a sign function, when the partial derivatives If positive, add 1; otherwise, take 1. .

[0070] After generating the final ventilation parameters, the intelligent control unit sends control commands to the ventilation actuators via a hybrid communication architecture of industrial Ethernet and CAN bus. The command protocol adopts a layered encoding design, with the first byte identifying the target device type (0x01 for steerable air ducts, 0x02 for variable frequency fans, and 0x03 for filter components), and subsequent bytes containing the specific opcode and parameter values. The system uses a timestamp synchronization mechanism to ensure that all actuators complete command reception and parsing within 50 milliseconds, enabling multi-device coordinated operation.

[0071] Upon receiving a steering command, the steerable duct module initiates a multi-degree-of-freedom rotation mechanism that activates a coordinated control process. An electric actuator drives a spherical universal joint based on the target angle (e.g., 120° horizontal azimuth and 15° pitch), while a stepper motor fine-tunes the guide vane spacing with a resolution of 0.1°. The curved surface design of the guide vanes increases the airflow outlet velocity by 20%, and a laser positioning sensor provides real-time feedback on the actual angle; if the deviation exceeds 0.5°, three recalibration cycles are triggered. For applications requiring coverage of multiple pollution sources, the duct employs a "fan-shaped sweeping mode," periodically oscillating at 2° per second to form a clean air curtain 5 meters wide and 15 meters in range.

[0072] After receiving the speed command, the variable frequency fan array uses an independent PID controller to perform ramp speed regulation on each fan drive board. A gradual increase curve (0) is used during the start-up phase. The rated speed is maintained for 30 seconds to avoid grid impact. During steady-state operation, the speed is dynamically fine-tuned based on real-time wind pressure sensor data to maintain the target airflow error ≤3%. When multiple fans need to work together (such as in parallel air supply scenarios), the master control node allocates speed weights through a load balancing algorithm. The fan closest to the pollution source is increased to 90% of its rated speed, the secondary fans are maintained at 70%, and the terminal fans are kept on standby at 50%. The fan blades are coated with a dust-repellent coating and automatically perform a 10-second reverse pulse purging every 8 hours of operation to prevent dust accumulation from causing dynamic imbalance.

[0073] The high-efficiency filter module switches its operating mode according to the type of pollution. In HEPA mode, the electric valve closes the activated carbon bypass, and the airflow passes through the pre-filter to intercept large particles before entering the H13 grade pleated filter paper, achieving a filtration efficiency of ≥99.97%. In activated carbon mode, the valve switches to the parallel adsorption chamber, and the airflow simultaneously passes through the impregnated activated carbon layer and the zeolite molecular sieve layer, achieving adsorption capacities of 35 mg / g for benzene compounds and 28 mg / g for formaldehyde. The mode switching process includes a 3-second pre-purging phase: closing the inlet valve and activating a reverse pulse airflow to remove residual pollutants, ensuring no risk of cross-contamination. The real-time filter resistance monitoring module triggers a filter media replacement reminder when the pressure difference exceeds 500 Pa and automatically downgrades to a low-flow mode to maintain basic filtration functionality.

[0074] During execution, edge computing nodes capture real-time changes in pollutant concentrations through a distributed sensor network, generating an effectiveness assessment report every 15 seconds. If the PM2.5 concentration reduction rate in the target area is less than 80% of the expected value, the system automatically upgrades the intervention level: increasing the fan speed by 10% and correcting the duct turning angle by 2°. 5°C, and activate the backup filter unit. When a sudden pollution event is detected (such as a 200% surge in VOC concentration within 5 minutes), the emergency protocol takes effect immediately. Ventilation in all non-critical areas is downgraded to the lowest power consumption mode, 90% of the air volume resources are concentrated to treat the core pollution area, and the backup power supply is activated to ensure the continuous operation of critical equipment. The system is equipped with multiple security interlock mechanisms: 1. Mechanical limit protection: When the duct turning angle exceeds the physical limit, the Hall sensor triggers an emergency stop signal and activates the hydraulic damping buffer; 2. Current overload protection: The fan drive board monitors the winding temperature and current harmonics in real time. When the temperature exceeds 85℃ or the harmonic distortion rate is >15%, it will automatically derate. 3. Emergency response to filter media failure: When the HEPA filter is damaged and the difference in particle concentration between upstream and downstream is less than 10%, shut down the filtration channel and activate the adjacent redundant unit.

[0075] All execution data (including instruction content, equipment status, and environmental feedback) is recorded in the black box storage at millisecond-level time resolution and simultaneously uploaded to the cloud platform to generate a visualized operation log. Maintenance personnel can view the overlaid ventilation vector field using AR glasses and manually adjust local parameters. After completing each ventilation task, the system automatically enters the energy efficiency assessment phase, calculating the comprehensive energy consumption per unit of pollutant removal (kWh / mg), comparing it with historical best values, and generating optimization suggestions for pre-learning and fine-tuning of the control strategy in the next cycle.

[0076] This embodiment also provides a computer device applicable to an outdoor workstation air quality monitoring and intelligent ventilation method, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the outdoor workstation air quality monitoring and intelligent ventilation method proposed in the above embodiment.

[0077] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0078] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the method for monitoring air quality and providing intelligent ventilation for an outdoor workstation as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0079] In summary, this invention, through redundant deployment of a 3D meshed sensor array and perimeter meteorological nodes within the workstation (coverage coefficient ≥ 1.2), combined with a multi-dimensional data fusion algorithm, achieves a spatial reconstruction accuracy of pollutant concentration fields exceeding 40%, effectively solving the pollution source location deviation problem caused by traditional 2D layouts. The LSTM-trained diffusion prediction model and real-time meteorological data correction mechanism improve the ventilation strategy response speed by 50%, shortening the peak pollutant concentration suppression time to within 30 minutes during sudden pollution events, while avoiding energy waste caused by excessive ventilation. The steerable duct module, through multi-degree-of-freedom adjustment (360° horizontal rotation and ±45° pitch) and dynamic spacing control of the guide vanes, achieves a clean air delivery direction error of less than 5°. Combined with HEPA / activated carbon dual-mode filtration components, it achieves PM2.5 and VOCs removal efficiencies of over 98% and 85% respectively in the target area. The distributed PID drive and dynamic grid load adaptation mechanism of the variable frequency fan array reduce energy consumption by 25% compared to traditional fixed-frequency systems while ensuring ventilation requirements are met. It reduces carbon emissions by 40%, and through collaborative optimization with edge computing and cloud platforms, it achieves a 15% reduction in daily carbon emissions. The modular design supports rapid replacement and expansion of sensor nodes, air duct components, and filter units, and can be adapted to outdoor workstations of different volumes (50-2000m³). It also maintains stable operation in extreme temperature and humidity environments (30℃ to 60℃) and dust concentrations (0-1000μg / m³).

[0080] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An outdoor workstation air quality monitoring and intelligent ventilation system, characterized in that, Includes an environmental sensing module, an intelligent control unit, and a ventilation actuator; The environmental sensing module consists of a sensor array composed of distributed PM2.5 sensors, VOCs sensors, temperature and humidity sensors, and air pressure sensors. It is used to collect air quality parameters and environmental parameters inside and outside the workstation in real time, and generate a heat map of pollution source distribution through a multi-dimensional data fusion algorithm. The intelligent control unit has a built-in dynamic ventilation model, which generates an adaptive ventilation strategy based on the output data of the environmental sensing module, workstation structural parameters, and external meteorological forecast data, combined with a preset air quality threshold. The dynamic ventilation model optimizes the fan speed, duct turning angle, and filter component working mode based on an improved fuzzy PID algorithm. The ventilation actuator includes a steerable duct module, a high-efficiency filter component, and a variable frequency fan array. It is used to execute the ventilation strategy generated by the intelligent control unit and achieve the directional dispersion of pollutants and the supply of clean air by adjusting the duct direction, filtration level, and air volume intensity. The dynamic ventilation model generates and corrects the ventilation strategy in the following way: Based on the time series data of pollutant concentration and historical meteorological data of the workstation over the past 30 days, a long short-term memory neural network (LSTM) is used to train a spatiotemporal prediction model for pollutant diffusion; during real-time operation, the wind speed and air pressure data of the current meteorological monitoring node are timestamped with the output of the prediction model, and the fan start threshold and duct turning priority in the ventilation strategy are corrected by a weighted moving average algorithm. The pollutant diffusion prediction curve of the dynamic ventilation model is generated by the following formula: in, For position In time The predicted values ​​of pollutant concentrations This is the initial concentration. β is the natural attenuation coefficient, and β is the meteorological influence factor. For position In time The wind speed component, The diffusion kernel function characterizes the diffusion characteristics of pollutants in time and space.

2. The outdoor workstation air quality monitoring and intelligent ventilation system as described in claim 1, characterized in that, The redundant deployment method of the sensor array is as follows: sensor nodes are arranged in a three-dimensional mesh topology inside the workstation, with the distance between adjacent nodes not exceeding 50% of the maximum detection radius of the sensor, and at least one backup node is set in each mesh plane; meteorological monitoring nodes are arranged at 10-meter intervals along the perimeter of the workstation. The meteorological monitoring nodes integrate wind speed and direction instruments, rainfall sensors and ultraviolet intensity detection modules, and communicate with edge computing nodes through a wireless self-organizing network to form an internal and external linkage environmental perception network.

3. The outdoor workstation air quality monitoring and intelligent ventilation system as described in claim 1, characterized in that, The specific structure of the steerable air duct module includes: a multi-degree-of-freedom rotation mechanism that uses a combination of a spherical universal joint and an electric push rod to achieve 360° horizontal rotation and ±45° pitch adjustment; guide vanes composed of 12 sets of aviation aluminum arc-shaped blades, with the blade spacing dynamically adjusted by a micro stepper motor with a precision of 0.5°; the high-efficiency filtration component has a built-in parallel air duct, which switches between HEPA filtration mode and activated carbon adsorption mode via an electric valve; and the drive circuit of each fan in the variable frequency fan array is independently configured with a PID controller, and the air volume is coordinated and distributed through the CAN bus protocol.

4. The outdoor workstation air quality monitoring and intelligent ventilation system as described in claim 1, characterized in that, The redundant deployment of the sensor array satisfies the spatial coverage formula: in, This is the coverage coefficient. The total number of sensors, For the effective monitoring area of ​​a single sensor, The workstation volume is specified; sensors within the workstation are arranged according to a three-dimensional grid spacing d, satisfying the following requirements. , This represents the sensor's maximum detection radius.

5. The outdoor workstation air quality monitoring and intelligent ventilation system as described in claim 1, characterized in that, The deflection angle of the guide vanes of the steerable air duct module Determined by both the target air supply direction and the pollutant concentration gradient, its calculation formula is as follows: in, Coordinates of the target area requiring enhanced ventilation. This refers to the real-time coordinates of the air duct outlet. This is the proportional adjustment coefficient, with a value range of 0.

1. 0.5, The spatial gradient of pollutant concentration at the current location is calculated using differential data from the sensor array. This spatial gradient serves as a control factor for the deflection angle of the guide vanes, adjusting the target airflow direction. The deflection angle of the guide vanes in the steerable duct module is jointly determined by the target airflow direction and the pollutant concentration gradient; therefore, it is added to the angle of the target airflow direction. The portion that is jointly adjusted by the proportional adjustment coefficient and the spatial gradient of pollutant concentration on the angle gradient of the target air supply direction at the current location.

6. A method for air quality monitoring and intelligent ventilation of an outdoor workstation, implemented based on an outdoor workstation air quality monitoring and intelligent ventilation system as described in any one of claims 1 to 5, characterized in that, The following steps are involved: Step S1: Collect PM2.5 concentration, VOCs concentration, temperature, humidity and air pressure data inside and outside the workstation in real time through a distributed sensor array, and transmit them to the edge computing node; Step S2: Reconstruct the pollutant concentration field distribution within the workstation based on a multi-dimensional data fusion algorithm, and calculate the pollutant diffusion trend by combining meteorological forecast data; Step S3: Based on the concentration field distribution, diffusion trend, and preset air quality safety threshold, generate a pollution level index through a dynamic risk assessment model; Step S4: Based on the pollution level index and workstation structural parameters, dynamically optimize ventilation parameters, including fan speed, duct turning angle and filtration mode, using an improved fuzzy PID algorithm. Step S5: Send control commands to the ventilation actuator to synchronously drive the steerable air duct, variable frequency fan and filter components to achieve rapid dilution of pollutants and directional supply of clean air.

7. The method for air quality monitoring and intelligent ventilation of an outdoor workstation as described in claim 6, characterized in that, The multi-dimensional data fusion algorithm in step S2 reconstructs the concentration field using the following formula: in, For position The fusion concentration value, For spatial interpolation weights, These are observations from nearby sensors. For Kalman filter gain, To predict concentration, This represents the actual observed concentration.

8. The method for air quality monitoring and intelligent ventilation of an outdoor workstation as described in claim 7, characterized in that, The formula for calculating the pollution level index R in step S3 is as follows: in, For the first Type of pollutants in time concentration, For the corresponding safety threshold, As a toxicity weighting factor, Given the current population density in the area, To ensure a safe density limit, To calculate the cumulative exposure time, To allow for safe exposure time, This is the dynamic adjustment coefficient.

9. The method for air quality monitoring and intelligent ventilation of an outdoor workstation as described in claim 8, characterized in that, The model parameter update formula for optimizing ventilation efficiency in step S5 is as follows: in, The adaptive learning rate is set to 0.

01. 0.1, The regional safe concentration threshold, For time The average actual monitored concentration, For a sign function, when the partial derivatives If positive, add 1; otherwise, take 1. .

Images