Energy-saving optimization control method for fresh air in terminal buildings based on passenger spatiotemporal distribution prediction

Abstract

The present invention discloses an energy-saving optimization control method for fresh air in a terminal building based on the prediction of the spatiotemporal distribution of passengers, which relates to the field of intelligent energy-saving technology and includes the following steps: S100, collecting the real-time position and movement speed of passengers, constructing a model of the spatiotemporal distribution change of passengers, and forming a prediction basis for dynamic adjustment of air pressure; S200, based on the spatiotemporal distribution change model of passengers, calculating the air pressure gradient distribution of each functional area of ​​the terminal building, identifying local positive pressure and adjacent negative pressure areas, establishing a micro-pressure difference graded warning index, and determining ventilation adjustment needs. The present invention dynamically predicts passenger distribution, adjusts the supply air and static pressure in real time, combines pressure difference warning, airflow monitoring and backflow sealing control, ensures orderly airflow, prevents the backflow of polluted gases, and ensures air quality and comfort. Based on the control model and energy-saving strategy library trained with full-process data, intelligent adaptive adjustment of the fresh air system is achieved, ventilation safety and energy efficiency are improved, and the service experience and operational efficiency of the terminal building are optimized.

Description

Technical Field

[0001] The present invention relates to the field of intelligent energy-saving technology, and in particular to an energy-saving optimization control method for fresh air in an airport terminal based on passenger spatiotemporal distribution prediction. Background Art

[0002] Terminal fresh air energy-saving optimization control based on passenger spatiotemporal distribution prediction refers to an energy-saving control method that uses real-time or historical data mining and machine learning technologies to predict the dynamic distribution of passengers in different time periods and areas within the terminal. Based on this prediction, the fresh air system's air volume, air supply area, and operating strategy are dynamically adjusted to achieve on-demand ventilation and precise air supply. This method establishes a spatiotemporal distribution model of passenger flow, predicts crowd density and activity heat load, and intelligently schedules the fresh air system's air volume, fan speed, and ventilation mode based on air quality requirements and energy consumption constraints. This allows the terminal to minimize fresh air processing energy consumption while meeting air quality and comfort standards, avoiding the energy waste and indoor environmental imbalance caused by excessive or insufficient ventilation in traditional systems.

[0003] The existing technology has the following deficiencies:

[0004] Existing optimization control algorithms for terminal fresh air systems, designed for energy conservation, typically predict passenger distribution and shut down air supply in sparsely populated areas, or dynamically adjust duct static pressure to achieve centralized allocation of fresh air resources. However, these optimization strategies generally rely on static or infrequently updated passenger flow forecasts, failing to fully account for regional pressure fluctuations caused by rapid passenger migration or localized concentrations. In particular, when air supply is shut down or pressure is reduced in certain areas, a small but persistent pressure differential may develop between passenger-concentrated areas and adjacent low-pressure zones, triggering an abnormal reversal of air flow direction and resulting in reverse airflow. Once this occurs, polluted gases, odors, or hot and humid air from adjacent spaces can flow back into passenger-concentrated areas, impairing indoor air quality and thermal comfort. This can also cause discomfort, increase the risk of infection, and disrupt ventilation system operation, severely impacting terminal safety and service experience.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not form the prior art that is already known to a person of ordinary skill in the art. Summary of the Invention

[0006] The purpose of this invention is to provide an energy-saving optimization control method for terminal fresh air based on the prediction of passenger spatiotemporal distribution. By dynamically predicting passenger distribution, the air supply and static pressure are adjusted in real time. Combined with pressure differential warning, airflow monitoring and backflow control, this ensures orderly airflow, prevents the backflow of polluted gases, and guarantees air quality and comfort. Based on the control model and energy-saving strategy library trained with full-process data, intelligent adaptive adjustment of the fresh air system is achieved, ventilation safety and energy efficiency are improved, and the terminal service experience and operational efficiency are optimized to solve the problems in the above-mentioned background technology.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a terminal building fresh air energy-saving optimization control method based on passenger spatiotemporal distribution prediction, comprising the following steps:

[0008] S100 collects passengers' real-time locations and movement speeds, constructs a model of passenger temporal and spatial distribution changes, and forms a prediction basis for dynamic air pressure regulation;

[0009] S200, based on a model of passenger spatiotemporal distribution changes, calculates the pressure gradient distribution in each functional area of ​​the terminal, identifies local positive pressure and adjacent negative pressure areas, establishes micro-pressure difference classification warning indicators, and determines ventilation adjustment needs;

[0010] S300, based on micro-pressure differential graded warning indicators, adjusts the air supply and static pressure levels in each functional area. By increasing the air supply and static pressure in negative pressure areas and reducing the air supply in positive pressure areas, a dynamic static pressure adjustment solution is formed to achieve regional pressure differential balance.

[0011] S400, based on a dynamic static pressure regulation solution, monitors airflow direction and velocity in real time, identifies reverse airflow trends, and locates abnormal flow areas and flow field characteristics;

[0012] S500, in response to the reverse propagation trend of airflow, adjusts the supply and exhaust air ratio and the air valve opening, links the anti-backdraft valve and airflow isolation device, closes the reverse airflow path, reconstructs the flow field structure, and ensures consistent airflow direction;

[0013] S600, after completing the closed reverse airflow path and reconstructing the flow field structure, extracts the passenger distribution, pressure difference response, airflow regulation and feedback control data of the entire process, trains the pressure difference regulation and flow direction correction model, establishes an energy-saving control strategy library, and realizes the steady-state operation of the fresh air system and closed-loop air quality assurance.

[0014] Preferably, step S100 includes:

[0015] Collect passengers' real-time location data and movement speed data to build a passenger spatiotemporal distribution change model;

[0016] Based on the passenger spatiotemporal distribution change model, the passenger density, average movement speed, residence time and crowd flow frequency of each functional area are obtained to form the passenger flow behavior factor;

[0017] Establish an air circulation coefficient based on the ventilation efficiency, air supply and exhaust paths, air circulation resistance, and space volume of the functional area;

[0018] Based on passenger flow behavior factors and air circulation coefficients, a mathematical model of air pressure response is constructed to calculate the weight of the impact of passenger flow on the air pressure in each area, forming a predictive basis for dynamic regulation of air pressure.

[0019] Preferably, step S200 includes:

[0020] Based on the passenger spatiotemporal distribution change model, the passenger distribution density, passenger inflow and outflow numbers, average passenger movement speed, stay time and gathering and dispersing behavior patterns of each functional area are extracted;

[0021] Calculate the total heat load increment and flow disturbance intensity per unit time in each functional area to form a basic database of passenger dynamic behavior and air pressure changes;

[0022] Combined with the physical properties of the functional area, the real-time air pressure value and the pressure difference between areas are calculated based on the air flow continuity equation, energy conservation equation and gas state equation;

[0023] Establish micro-pressure difference grading warning indicators based on pressure difference amplitude, change rate and duration to clarify ventilation adjustment needs.

[0024] Preferably, step S300 includes:

[0025] Determine the adjustment targets for air supply volume and static pressure in each functional area based on the micro-pressure difference classification warning indicators;

[0026] Adjust the speed of the air supply fan, the opening of the air supply valve and the frequency conversion drive frequency of the air supply duct to dynamically adjust the air supply volume and static pressure level to achieve pressure difference balance between positive and negative pressure areas;

[0027] Real-time monitoring of air pressure changes and airflow direction, and multiple rounds of progressive adjustments based on the monitoring results to ensure dynamic and stable air pressure;

[0028] The adjustment parameters and pressure difference response effects are recorded as data files, and the static pressure adjustment strategy is optimized through machine learning to achieve adaptive closed-loop control.

[0029] Preferably, step S400 includes:

[0030] Real-time monitoring of the airflow direction and velocity in each functional area to form a three-dimensional airflow monitoring data set;

[0031] Compare real-time monitoring data with the baseline airflow state before static pressure adjustment to identify sensitive areas of flow deviation and flow velocity drop;

[0032] Based on the three-dimensional computational fluid dynamics simulation method of the terminal building's spatial structure, the airflow field in sensitive areas is simulated and dynamically reproduced to confirm the reverse propagation area and flow field characteristics of the airflow;

[0033] The locked airflow reverse propagation area and flow field characteristic parameters are archived to form an airflow dynamic adjustment and abnormal feedback database for subsequent adjustment strategy optimization and intelligent decision-making.

[0034] Preferably, step S500 includes:

[0035] Based on the locked abnormal flow area and abnormal pressure boundary characteristics, the air supply and exhaust ratio is dynamically adjusted to increase the air supply volume and reduce the exhaust volume to suppress the formation of backflow;

[0036] Based on real-time airflow direction and velocity data, dynamically optimize the air valve openings in abnormal areas and adjacent areas, expand the air supply valve opening, and reduce the air exhaust valve opening to prevent negative pressure from increasing and countercurrent diffusion;

[0037] Linking the anti-backflow valve and the airflow isolation device to block the backflow path and physically isolate the backflow propagation;

[0038] Continuously monitor airflow direction, velocity, and pressure levels, evaluate the effects of containment and flow field reconstruction, and record adjustment and control data for model training and strategy optimization.

[0039] Preferably, step S600 includes:

[0040] Extract passenger distribution evolution data, pressure difference response data, airflow adjustment data, and feedback control data during the entire process of counterflow control and airflow reconstruction;

[0041] The extracted full data is cleaned, denoised, time-series aligned, and spatially normalized before being fed into a deep learning model to establish a control model linking passenger flow, pressure differentials, and airflow changes.

[0042] Continuously iteratively optimize model parameters based on model training results and adjustment feedback to improve the prediction accuracy of pressure difference control and flow direction correction;

[0043] Based on the optimized control model, an energy-saving control strategy library is constructed to support the adaptive adjustment and steady-state operation of the fresh air system.

[0044] Preferably, the specific steps for implementing adaptive adjustment and steady-state operation of the fresh air system through the energy-saving control strategy library are as follows:

[0045] Based on the energy-saving control strategy library, the system automatically matches the corresponding pressure differential control and airflow correction schemes based on the real-time monitoring results of passenger distribution and pressure differential dynamics;

[0046] According to the matching scheme, the air supply volume, exhaust volume, air valve opening and fan speed of each functional area are automatically adjusted to achieve dynamic regulation of air pressure and airflow direction;

[0047] Real-time monitoring of the air pressure level, airflow direction and velocity changes in each functional area after adjustment to evaluate whether the adjustment effect has achieved the predicted target;

[0048] The adjustment results are compared with the historical case data in the energy-saving control strategy library, and the plans and parameters in the strategy library are updated according to the evaluation results to achieve adaptive optimization of the strategy library.

[0049] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0050] This invention dynamically predicts passenger distribution, adjusts air supply and static pressure levels in real time, and coordinates pressure differential warnings, airflow direction monitoring, and backflow control to ensure that airflow always flows in an orderly manner along the predetermined path. This effectively prevents polluted gases, odors, or hot and humid air from flowing back into passenger-dense areas, thus ensuring air quality and thermal comfort from the source. Furthermore, a control model and energy-saving strategy library based on full-process data training enable intelligent, adaptive adjustment of the fresh air system under varying passenger flow conditions, improving air quality and ventilation safety while optimizing energy utilization, ultimately enhancing the terminal's overall service experience and operational efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction to the drawings required for use in the embodiments will be given below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0052] Figure 1 This is a flow chart of the method for energy-saving optimization control of fresh air in terminal buildings based on passenger spatiotemporal distribution prediction of the present invention. DETAILED DESCRIPTION

[0053] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of this disclosure will be thorough and complete and will fully convey the concepts of the example embodiments to those skilled in the art.

[0054] The present invention provides Figure 1 The terminal building fresh air energy-saving optimization control method based on passenger spatiotemporal distribution prediction includes the following steps:

[0055] S100, through the global passenger flow sensing network, continuously collects real-time passenger location and movement speed data. Based on this data, it dynamically builds a passenger spatiotemporal distribution change model, which forms the prediction basis for dynamic air pressure adjustment.

[0056] The global passenger flow sensing network continuously collects real-time passenger location and movement speed data, and dynamically builds a passenger spatiotemporal distribution change model based on the collected data to form a predictive basis for dynamic air pressure regulation. The specific steps are as follows to achieve comprehensive perception and accurate modeling of the dynamic flow characteristics of passengers within the terminal:

[0057] A global passenger flow sensing network covering all public areas, corridors, waiting areas, boarding gates, commercial areas, and service areas of the terminal is constructed. This global passenger flow sensing network collects information by deploying multi-source heterogeneous sensor devices, including spatial positioning devices based on millimeter-wave radar, temperature sensing devices based on infrared imaging, behavior capture devices based on video image recognition, and mobile device sensing devices based on Bluetooth or Wi-Fi probes. These various sensor devices work together to continuously sense and collect data on passengers' real-time location coordinates, spatial dwell time, movement paths, and travel speeds in different spatial areas of the terminal, generating a multi-dimensional, multi-timeframe data set on passenger behavior and flow. To ensure data integrity and timeliness, all sensor devices are equipped with a high-frequency sampling cycle and precise spatiotemporal synchronization mechanisms, ensuring that the timestamp and spatial positioning accuracy of each collection point is within milliseconds and sub-meter levels, enabling continuous recording of passengers' dynamic movement in three-dimensional space.

[0058] The multi-source passenger behavior and flow information collected through the global passenger flow sensing network is uniformly connected to the data fusion and preprocessing process. The collected data is cleaned, denoised, time-series aligned, and spatial coordinates standardized. Outliers and noise points caused by environmental interference or data loss are eliminated to ensure the continuity and authenticity of the data. On this basis, a time-series-based deep learning model is used to extract features of passenger movement trajectories and speed changes. Graph convolutional networks and attention mechanisms are further used to model the passenger flow relationships between different functional areas in the terminal, dynamically capturing the migration and aggregation trends of passengers in different time periods. Through the training and deduction of this model, a passenger spatiotemporal distribution change model is established that reflects the flow and aggregation patterns of passenger groups in spatial and temporal dimensions, realizing dynamic predictions of the distribution status and change trends of passengers in various regions in future time periods. The model has the ability of continuous self-learning and adaptive updating, and can continuously optimize the prediction accuracy based on actual passenger flow data.

[0059] Time series-based deep learning models use neural networks capable of processing sequential data, such as long short-term memory (LSTM) networks or gated recurrent unit (GRU) networks, to extract features from continuous, time-series dynamic data such as passenger movement trajectories and speed changes, thereby uncovering passenger behavior patterns and trends over time. This model can capture the inherent correlations between passengers' location changes, length of stay, and speed changes at different time points, enabling modeling of the dynamic evolution of individual and group flows. A graph convolutional network (GCN) is a deep learning method for processing graph-structured data that can model spatially correlated node data. Here, the terminal's functional areas are abstracted as nodes in a graph structure. The passenger flow paths and frequencies between different areas form edges and weights between nodes. The graph convolutional network learns the spatial connectivity between areas and the intensity of passenger flow. The attention mechanism, a computational method that assigns different weights to different input features, is used to enhance the model's focus on passenger flow changes in key areas and at key moments, dynamically adjusting the model's sensitivity to different spatiotemporal information, thereby improving the accuracy of predicting migration and aggregation trends. The specific steps are:

[0060] The passenger's trajectory and speed data at each time slice are chronologically constructed into a time series and input into the LSTM or GRU model to extract the flow characteristics in the time dimension;

[0061] The functional areas of the terminal and their passenger flow connections are constructed as a graph data structure. The graph is processed using a graph convolutional network to learn the flow connections and spatial dependencies between areas.

[0062] Combining the output of the graph convolutional network, an attention mechanism is introduced to assign different weights to the regional flow intensity and migration frequency in each time period, enhancing the perception of changes in important areas and time periods.

[0063] By integrating time series features with the results of the graph convolution-attention mechanism, the migration trend and aggregation status of passengers in the entire terminal are dynamically output, providing a dynamic prediction basis for subsequent air pressure regulation and ventilation control.

[0064] Based on the constructed passenger spatiotemporal distribution change model, the predicted density and movement speed of passengers in each functional area within each time segment are further mapped into the overall and local air pressure change trends of the terminal. By establishing a mapping relationship model between passenger flow and spatial air pressure response, combined with parameters such as air flow within the terminal, ventilation efficiency, and regional space volume, the impact weight of passenger flow behavior on the air pressure in each area is dynamically calculated. This calculation model not only takes into account changes in the number of passengers, but also introduces the superimposed effects of passenger average speed, residence time, and crowd density on local heat load and air pressure changes. It accurately predicts the dynamic changes of positive pressure areas, negative pressure areas, and air pressure gradients that may appear in each functional area in the future period, which serves as a direct basis for subsequent air pressure regulation and fresh air control strategies.

[0065] Establishing a mapping relationship model between passenger flow and spatial air pressure response includes the following steps:

[0066] Based on the passenger spatiotemporal distribution change model, dynamic behavioral data such as passenger density, average movement speed, dwell time, and crowd flow frequency in each functional area at different time periods are obtained. These data are parameterized into behavioral factors that describe the local crowd flow energy and activity intensity.

[0067] For each functional area within the terminal, considering its structural characteristics, we collect and quantify air flow characteristic parameters, including ventilation efficiency, supply and exhaust air paths, air flow resistance, degree of openness or enclosure, and regional spatial volume. We also establish air flow coefficients within a region and between adjacent areas to reflect the ease of air flow and diffusion between different areas.

[0068] Based on the principles of mass conservation and energy balance, a mathematical model of air pressure response was constructed, coupling passenger flow behavior factors with air circulation characteristic parameters. By considering the contribution of passenger flow to local heat load, the intensity of airflow disturbances, and the cumulative effect on spatial pressure, the gain or reduction effect of passenger flow on local and adjacent area air pressure changes was calculated, and the corresponding air pressure impact weight for each area was quantified.

[0069] The model is fitted and corrected using historical measured data, and the accuracy of air pressure response prediction is continuously optimized using multivariate regression or machine learning methods. This enables dynamic calculation and prediction of air pressure change trends in various functional areas under different passenger flow scenarios, providing a quantitative basis and real-time reference for subsequent static pressure regulation and ventilation control.

[0070] The predicted pressure change trend is combined with the terminal's existing fresh air adjustment strategy and parameters to form a predictive basis for dynamic pressure adjustment, and this predictive basis is used as the core input parameter for the subsequent dynamic adjustment of the ventilation system. By linking the passenger spatiotemporal distribution change model with the pressure prediction model, a pre-adjustment plan for fresh air supply volume, static pressure setting, and air supply path is formulated to ensure that the terminal can proactively optimize air pressure and airflow organization in advance when passenger flow changes intensively, preventing pressure imbalance and airflow anomalies caused by sudden passenger gatherings or rapid migration. At the same time, the predicted and actual monitored pressure data are dynamically compared, and the model parameters are continuously corrected to improve the model's prediction accuracy and the real-time adjustment response, thereby achieving a high degree of coordination and closed-loop control between passenger dynamic flow, air pressure response, and fresh air adjustment.

[0071] This step aims to provide precise basic sensing and dynamic prediction capabilities for the terminal's energy-efficient fresh air optimization control, resolving issues inherent in traditional ventilation control, such as delayed response to passenger flow, blind adjustments, and inaccurate regional air quality control. Through a global passenger flow sensing network covering all functional areas and routes within the terminal, the terminal can capture every passenger's location, speed, and trajectory in real time, generating highly timely and spatially-resolution passenger flow data. This global, real-time sensing approach enables the terminal to dynamically understand the spatial distribution density, movement trends, and potential gathering areas of people, transcending the limitations of traditional methods that rely solely on static passenger flow statistics or low-frequency monitoring. A passenger spatiotemporal distribution model, constructed based on this collected data, not only reflects the current passenger distribution but also, through deep learning and time series analysis, predicts future passenger flow trends and changes in densely populated areas. This model provides an accurate decision-making basis for subsequent dynamic adjustment of air pressure, enabling the fresh air system to adjust the ventilation strategy in advance according to the predicted changes in pedestrian flow, achieve on-demand air supply, dynamically balance regional air pressure, and avoid local pressure difference imbalance and air flow backflow caused by rapid changes in pedestrian flow, thereby improving air quality and energy utilization efficiency.

[0072] S200, based on a model of passenger spatiotemporal distribution changes, calculates the pressure gradient distribution of each functional area within the terminal in real time, identifies local positive pressure areas formed by passenger concentrations and their adjacent relatively negative pressure areas, establishes corresponding micro-pressure difference classification warning indicators, and clarifies the dynamic adjustment needs of ventilation control;

[0073] Based on the passenger spatiotemporal distribution change model, the air pressure gradient distribution of each functional area in the terminal is calculated in real time, the local positive pressure areas formed by passenger gathering and their adjacent relatively negative pressure areas are identified, the corresponding micro-pressure difference classification warning indicators are established, and the dynamic adjustment needs of ventilation adjustment are clarified. The specific steps are as follows to achieve dynamic control of the air pressure situation inside the terminal and scientific guidance for ventilation control.

[0074] Based on the passenger spatiotemporal distribution change model, the passenger distribution density, the number of passenger inflows and outflows per unit time, the average passenger movement speed, the length of stay, and the passenger gathering and dispersion behavior patterns in each functional area within the entire terminal area are extracted for the current and predicted time periods. For each functional area, the total heat load increment and flow disturbance intensity generated in the area per unit time are calculated based on the dynamic activities of passengers in the local space. In this calculation process, the heat dissipation from the passenger surface, the heat and moisture exchange caused by breathing, the local heat accumulation caused by the agglomeration density, and the stirring effect of passenger flow on the surrounding air flow field are explicitly taken as input factors to establish a multivariate coupling model of the effect of passenger flow on the thermal dynamic changes of the local air space, thereby forming a basic database that reflects the quantitative relationship between passenger dynamic behavior and spatial air pressure changes.

[0075] Calculating the total heat load increment and flow disturbance intensity generated per unit time in a functional area of ​​the terminal includes the following steps:

[0076] Statistics are collected on the changes in the number of passengers in the area per unit time, and the number of passengers entering and leaving the area, the average length of stay in the area, and the average movement speed are recorded. At the same time, the passengers' body heat dissipation parameters are collected, including basal metabolic heat release, surface convective heat transfer, respiratory moist heat release, and metabolic heat differences related to activity intensity. These parameters are multiplied by the corresponding number of passengers and accumulated to form the total body heat load per unit time.

[0077] Taking into account the spatial volume and degree of enclosure of the area, the temperature rise per unit volume of air due to the gathering of passengers is calculated. Based on the energy conservation equation, the heat gain is converted into quantitative results of air temperature rise and pressure change.

[0078] Calculate the flow disturbance intensity based on the average passenger movement speed, flow direction change frequency and regional spatial layout (such as channel width and obstacle density). Use the flow field disturbance model to infer the local wind speed fluctuations and air streamline disturbance amplitude caused by the movement of passenger groups, and quantify the disturbance energy intensity value.

[0079] The total heat load increment is combined with the flow disturbance intensity as a comprehensive influencing factor of the area's changes in air pressure and air flow field, which is then used by subsequent pressure gradient and flow direction prediction models to achieve a dynamic linkage assessment of passenger activities and environmental changes.

[0080] Based on this basic database, dynamic pressure simulations were conducted by integrating the heat load variations and disturbance intensity of each functional area with the terminal's spatial physical properties. These physical properties include the functional area's actual volume, floor height dimensions, openness, window and door permeability, the number and length of connecting passages, the location and diameter of ventilation risers and ducts, the configuration of supply and exhaust vents, and ventilation efficiency. By comprehensively integrating these static and dynamic parameters, the air flow continuity equation was applied to ensure the conservation of inflow and outflow airflow. The energy conservation equation was combined to calculate the impact of local heat accumulation on air temperature and flow velocity. Furthermore, the gas equation of state was used to derive internal pressure changes. Based on this, real-time pressure values ​​were dynamically generated for each functional area, and the pressure differences between adjacent areas were calculated one by one. Ultimately, a dynamic pressure gradient distribution map was created for the entire terminal. This pressure gradient map clearly identifies localized positive pressure areas caused by passenger congestion, delays, and frequent movement, while also identifying adjacent areas of relatively negative pressure, accurately reflecting the spatial pressure dynamics and gradient relationships.

[0081] Based on real-time pressure gradient distribution maps, a comprehensive micro-pressure differential grading warning indicator system has been established. This system assigns warning levels based on three dimensions: pressure differential amplitude, pressure change rate, and pressure differential duration. These levels are categorized into mild, moderate, and severe warnings. A mild warning corresponds to a pressure differential of less than 100 Pa and lasting no longer than 10 minutes, prompting attention but no immediate adjustments. A moderate warning corresponds to a pressure differential between 100 and 300 Pa, or a pressure differential change rate exceeding 20 Pa per minute, prompting appropriate adjustments to the supply or exhaust system. A severe warning corresponds to a pressure differential exceeding 300 Pa, persisting for more than 30 minutes, or accompanied by an abnormal reversal of airflow direction. In these cases, targeted ventilation adjustments and damper control are necessary. The warning indicators also prioritize the sensitivity of different functional areas to pressure differentials. For example, waiting areas and VIP lounges are more sensitive to pressure differentials and are assigned lower warning thresholds, while baggage claim areas and equipment rooms can have higher warning tolerances.

[0082] The established micro-pressure difference graded warning indicators are directly linked to the adjustment logic of the terminal's fresh air control to generate a dynamic ventilation adjustment demand list. The list clearly stipulates the ventilation adjustment actions corresponding to each warning indicator, including increasing the air supply volume and static pressure setting value of the air supply fan in the functional area in the negative pressure zone, increasing the opening of the air supply port, and synchronously adjusting the air supply load of the adjacent positive pressure area, reducing the air volume and adjusting the exhaust rate accordingly to ensure that the local pressure difference is gradually balanced. In addition, the demand list also clearly defines the response priority, target pressure difference value and adjustment duration of the adjustment, and dynamically compares the adjustment effect with the real-time air pressure monitoring results to form a closed-loop feedback of the pressure difference adjustment, ensure the accuracy and timeliness of ventilation control, and achieve dynamic balance of air pressure and stability of air flow direction in the terminal's internal space.

[0083] This step provides precise pressure-based guidance for the dynamic adjustment of the terminal's fresh air system, addressing regional pressure imbalances, airflow disturbances, and potential air quality deterioration caused by changes in passenger flow and concentration. Through real-time calculations based on a model of passenger spatiotemporal distribution, the system dynamically senses and predicts pressure fluctuations in various functional areas within the terminal as passenger numbers, density, and movement change, providing a comprehensive understanding of the spatial pressure gradient distribution. This step not only accurately identifies localized positive pressure areas caused by high passenger density, but also simultaneously identifies adjacent relatively negative pressure areas. This identification is crucial for preventing reverse airflow propagation caused by localized pressure differentials. By establishing a micro-pressure differential early warning indicator that comprehensively quantifies factors such as the magnitude, rate of change, and duration of pressure differentials, it is possible to set clear warning levels for different pressure differential levels, guiding the fresh air system to make timely and graded adjustments to air supply volume, pressure, and exhaust strategies. This early warning mechanism avoids problems such as air backflow, deterioration in air quality, and impairment of passenger comfort caused by delayed response or improper adjustment of the ventilation system. It achieves precise linkage between air pressure changes and ventilation adjustment, laying a dynamic response foundation for the safe, healthy and energy-saving operation of the terminal's internal environment.

[0084] The S300 system uses micro-pressure differential graded warning indicators to control the zoned air duct controllers of the fresh air system to dynamically adjust the air supply volume and static pressure levels in each functional area. By increasing the air supply volume and static pressure levels in negative pressure areas and simultaneously reducing the air supply load in positive pressure areas, a dynamic static pressure adjustment scheme is formed to achieve dynamic balance of regional pressure differentials.

[0085] Based on the micro-pressure differential graded warning indicators, the zoned air duct controller that controls the fresh air system dynamically adjusts the air supply volume and static pressure level of each functional area. By increasing the air supply volume and static pressure level in the negative pressure area and simultaneously reducing the air supply load in the positive pressure area, a dynamic static pressure adjustment scheme is formed to achieve dynamic balance of regional pressure differences. The specific steps are as follows to achieve precise control of the dynamic balance of global and local air pressure inside the terminal.

[0086] Based on real-time calculated pressure gradient distribution and micro-pressure differential grading warning indicators, the current air pressure status of each functional area of ​​the terminal is comprehensively identified, distinguishing between areas experiencing negative pressure, positive pressure, and near-normal pressure. Dynamic air flow adjustment requirements are determined for each area based on the magnitude, rate of change, and duration of the pressure differential in each area. For negative pressure areas, air flow adjustment targets are set at three levels: basic, intermediate, and advanced, depending on the warning level. Basic corresponds to mild negative pressure, with an air flow increase of 10% to 15% and a static pressure increase target of 8% to 12%. Intermediate corresponds to moderate negative pressure, with an air flow increase of 20% to 30% and a static pressure increase of 15% to 20%. Advanced corresponds to severe negative pressure, with an air flow increase of more than 30% and a static pressure increase of more than 20%. Conversely, for positive pressure areas, air flow reductions are set based on the magnitude and duration of the positive pressure: 10% for mild positive pressure, 20% for moderate pressure, and more than 30% for severe pressure. Static pressure is also reduced proportionally, ensuring that pressure in positive pressure areas gradually returns to a balanced range. Each type of adjustment parameter is set based on the regional functional positioning, personnel density sensitivity and airflow organization requirements within the terminal to ensure that the adjustment standards are scientific and accurate.

[0087] Based on the aforementioned regulation targets, air volume and static pressure are dynamically controlled through fine-tuning of the fresh air system's fan speed, air outlet damper opening, and the variable frequency drive frequency within the air supply duct. For negative pressure areas, the supply fan speed is prioritized. This increases air volume output by increasing the fan's rated speed, while simultaneously widening the mechanical opening of the air outlet to free up more air supply cross-section and improve air supply coverage in the local space. Simultaneously, the frequency of the variable frequency drive within the duct is increased to increase the fan's output pressure, thereby simultaneously raising the static pressure level. For positive pressure areas, the supply fan speed is reduced, the air outlet opening is narrowed, and the variable frequency drive frequency is lowered to reduce air volume and moderately lower static pressure to prevent further pressure increases. All regulation actions are implemented in progressive increments of varying magnitudes based on the warning level. Each adjustment action is verified in real time against the feedback value of the pressure differential change to avoid pressure fluctuations or air flow disturbances caused by insufficient or excessive regulation.

[0088] After completing the initial adjustment, the air pressure values, air pressure change rates and pressure gradients between areas in all functional areas are continuously monitored in real time. The pressure difference status before and after the adjustment is dynamically compared to assess whether the target threshold for pressure difference balance has been reached. If the monitoring results show that the pressure difference in the local area is still higher or lower than the set balance range, or the pressure difference change rate is lower than the adjustment response rate of 5 Pa per minute, the adjustment requirements for the air supply volume and static pressure are recalculated based on the real-time data, and secondary or even multiple adjustments are performed to ensure that the air pressure in all functional areas is in a dynamically stable equilibrium state. In addition, the flow direction changes, flow velocity distribution and streamline continuity of the airflow during the adjustment process are monitored in real time to ensure that the airflow continues to flow towards the established supply and exhaust air paths after adjustment to avoid air backflow, turbulence or accumulation in dead corners, and to ensure that air quality and thermal comfort are not affected.

[0089] The specific parameters of each air volume and static pressure adjustment involved in the dynamic adjustment process, the pressure difference response effect, the adjustment time, the regional feedback data, and the corresponding warning level are recorded item by item to form a standardized data file. This data file is multi-dimensionally associated with the dynamic data of passenger distribution, the historical curve of air pressure changes, and the effect data of the adjustment response. The relationship between the static pressure adjustment strategy and the pressure difference response is modeled and optimized through machine learning methods to form an experience library and strategy set based on actual scenarios. In the future, when faced with similar abnormal situations of passenger flow and pressure difference, the most targeted and fastest-responding static pressure adjustment plan can be quickly generated by calling the experience library, realizing closed-loop control of the adaptive adjustment of the fresh air system and the dynamic balance of the pressure difference, thereby effectively ensuring the safety, comfort and energy efficiency of the terminal's internal environment.

[0090] This step aims to achieve dynamic pressure balance between functional areas within the terminal through precise regulation based on micro-pressure differential graded warning indicators, thereby ensuring indoor air quality, passenger comfort, and ventilation system efficiency. Rapid changes in passenger distribution within the terminal can easily lead to uneven air pressure in local areas. This means that some areas experience positive pressure due to dense passenger flow, while adjacent areas experience negative pressure due to sparse passenger flow or insufficient air supply. If this pressure differential persists, it can cause abnormal reversals in airflow direction, leading to the reverse diffusion of polluted air, odors, or highly humid and hot air, seriously impacting air quality and passenger health and safety. This step controls the zoned air duct controllers of the fresh air system and dynamically adjusts the air supply volume and static pressure in positive and negative pressure areas based on micro-pressure differential graded warning indicators. This increases the air supply volume and static pressure in negative pressure areas, enhancing fresh air replenishment and airflow kinetic energy, and rapidly restoring local air pressure. Simultaneously, it reduces the air supply load in positive pressure areas, lowering static pressure and preventing pressure buildup that can lead to airflow disturbances. Through this differentiated and dynamic static pressure regulation mechanism, the air pressure gradient of each functional area can be continuously maintained in a safe and stable range in the global and local scope, preventing air flow reversal and air quality deterioration. At the same time, the ventilation system can achieve the goals of on-demand precise air supply and energy saving and consumption reduction, providing intelligent and automated control support for the ventilation environment and operational efficiency in the terminal.

[0091] The S400, based on a dynamic static pressure control solution, monitors the airflow direction and velocity in each functional area of ​​the terminal in real time. It compares airflow changes before and after static pressure control, identifies the reverse propagation trend of airflow after static pressure control, and locates specific areas with abnormal flow direction and the corresponding flow field characteristics.

[0092] Real-time monitoring of the airflow direction and velocity in each functional area of ​​the terminal building is carried out. The airflow changes before and after static pressure adjustment are compared to identify the reverse propagation trend of airflow after static pressure adjustment. The specific areas with abnormal flow direction and the corresponding flow field characteristics are identified. The specific steps are as follows to achieve continuous monitoring of dynamic changes in airflow and comprehensive identification of abnormal trends.

[0093] After implementing the dynamic static pressure adjustment plan, multiple types of high-precision airflow monitoring sensors were deployed in various functional areas within the terminal to continuously collect data on the direction and velocity of airflow in each area. Specifically, ultrasonic anemometers, hot wire anemometers, and laser Doppler flow sensors were installed in the terminal's main air supply ducts, exhaust ducts, ventilation risers, air duct branches, space entrances and exits of key functional areas, passenger-intensive waiting areas, boarding gates, security checkpoints, rest areas, and commercial service areas. These sensors are used to accurately measure the velocity and direction of airflow in different directions. These sensors continuously record changes in airflow velocity and direction at a sampling frequency of no less than 10 times per second. At the same time, all sensor devices are uniformly equipped with a nanosecond time synchronization mechanism to ensure precise alignment of data in time and space, thereby forming a complete three-dimensional airflow monitoring network and real-time airflow dataset.

[0094] The real-time airflow direction and velocity data obtained from monitoring are compared with the baseline airflow status data before the implementation of the dynamic static pressure adjustment scheme. Through comparison, the airflow changes in each functional area after the air supply volume and static pressure are adjusted are analyzed one by one, and the flow velocity change rate and flow direction deviation amplitude of each area are calculated. Specifically, whether the deviation angle of the airflow direction before and after adjustment exceeds 15 degrees is used as the judgment standard, and whether the flow velocity drops by more than 20% per unit time is detected at the same time. For areas where the flow velocity is detected to have dropped significantly or the flow direction is significantly deflected, they are further marked as airflow change sensitive areas as early warning areas where there may be a risk of reverse propagation of airflow. In addition, by comparing the continuity of the flow direction and the stability of the flow velocity, it is screened whether there is a trend of the airflow reversing from the supply direction to the exhaust direction, forming a preliminary identification result of reverse propagation.

[0095] For the sensitive areas screened out, a three-dimensional computational fluid dynamics simulation method based on the spatial structure of the terminal building is further used, combined with the real-time collected flow velocity and flow direction data, to conduct detailed simulation and dynamic reproduction of the airflow state in the target area. The local velocity field, pressure field and streamline distribution obtained through simulation are cross-validated with the real-time monitoring data to accurately identify whether pressure reversal, airflow backflow or circulation is formed in the local space. In particular, attention is paid to the judgment criteria of whether there are continuous countercurrent streamlines with a length greater than 3 meters in the airflow direction, whether the countercurrent flow velocity exceeds 20% of the original flow velocity, and whether the countercurrent duration exceeds 300 seconds. Combining these quantitative indicators, it is confirmed that there is a stable existence of reverse airflow propagation. For areas that meet the above judgment conditions, the path of the countercurrent, the flow deviation angle, the peak value of the countercurrent velocity and the duration are further extracted to construct a complete description of the flow abnormality characteristics.

[0096] The three-dimensional computational fluid dynamics simulation method based on the terminal building's spatial structure refers to a method that constructs a complete three-dimensional digital model based on the terminal's actual physical structure, spatial layout, functional area division, and architectural design parameters for ventilation and supply and exhaust air paths, and uses computational fluid dynamics (CFD) numerical simulation technology to perform high-precision simulation of the velocity field, pressure field, and flow streamlines during the air flow process within the terminal. In the present invention, the role of this simulation method is to combine the terminal's architectural physical characteristics with real-time monitored dynamic data such as airflow velocity and direction to dynamically reproduce the flow pattern, pressure changes, and flow direction evolution process of the airflow at a specific time and in a specific spatial area, accurately restore the actual behavior of the airflow before and after adjustment, and then identify whether there are abnormal phenomena such as reverse propagation, backflow, vortexes, or local turbulence in the airflow. The specific steps are as follows: First, based on the terminal building's architectural design blueprint, spatial three-dimensional surveying data and structural parameters, a complete three-dimensional building structure model is established to accurately depict the spatial scale, floor height, partitions, door and window permeability, air duct layout and supply and exhaust vent locations of each functional area; second, the real-time collected air flow velocity, flow direction, static pressure and other monitoring data are used as the boundary conditions and initial conditions of the simulation model to ensure that the simulation is consistent with the actual working conditions; third, CFD simulation software is used to simulate the air flow process in the building space based on the NS equation, continuity equation and energy conservation equation, and the velocity field, pressure field and flow direction streamline diagram at each moment are generated to dynamically reflect the changing trend and stability of the air flow; finally, by comparing and verifying the simulation results with the real-time monitoring data, it is analyzed whether a countercurrent path, pressure difference reversal or flow turbulence is formed in a specific area. At the same time, the flow velocity peak, flow direction deviation angle and flow field pressure distribution in the abnormal area are extracted as the basis for judging the reverse propagation of air flow and formulating subsequent supply and exhaust air adjustment strategies.

[0097] The locked airflow reverse propagation areas and their flow field characteristic parameters are uniformly archived to form a standardized data structure, which includes the geographical location of each abnormal area, the length of the reverse flow path, the reverse flow velocity change curve, the degree of flow direction deviation, the flow field pressure distribution, the timestamp of the abnormality, and the real-time relationship with the passenger distribution. At the same time, the data is associated with the specific parameters of dynamic static pressure regulation, supply and exhaust air settings, fan speed, and air valve opening adjustment records to form an integrated airflow dynamic regulation and abnormality feedback database. This database is not only used to support the dynamic response and regulation strategy optimization of the fresh air system, but also serves as a training data source for machine learning models to continuously improve the accuracy of complex flow field evolution and airflow anomaly prediction, and realize the adaptive and intelligent decision-making of airflow regulation in the terminal under passenger dynamics and environmental changes.

[0098] The purpose of this step is to monitor and accurately analyze the dynamic changes in the airflow organization within the terminal after static pressure adjustment throughout the entire process, ensuring that ventilation adjustment will not cause new airflow anomalies due to the redistribution of regional pressure differences, especially the risk of reverse airflow propagation. The internal spatial structure of the terminal is complex, with numerous functional areas that are interconnected. During the static pressure adjustment process, an increase or decrease in the air supply volume and static pressure in a local area can easily cause the pressure balance in adjacent areas to be disrupted, thereby causing the airflow direction to deflect or even reverse. Once reverse airflow propagation occurs, polluted air, odorous or humid air that should have been discharged to the outside may flow back to areas with dense passenger traffic, causing air quality to deteriorate, comfort to worsen, and even health risks such as disease transmission. Therefore, this step can dynamically grasp the air flow status of each functional area through real-time monitoring of airflow direction and flow rate, and by comparing it with the airflow baseline state before static pressure adjustment, promptly identify whether the airflow has abnormal flow direction, abnormal flow rate attenuation, or a trend of flow reversal. Once an abnormal flow direction is detected, by identifying the specific abnormal area and its corresponding flow field characteristics, such as the backflow path, flow deviation angle, and pressure distribution, a clear basis for intervention can be provided for subsequent supply and exhaust air adjustment, damper control, and backflow prevention measures. This prevents the continued deterioration of airflow disorder and ensures the air quality, comfort, and stability of the terminal's internal environment. Furthermore, the data accumulated in this step provides accurate measurement support for the optimization of ventilation control models and the prediction of airflow dynamic behavior, enhancing the intelligence and adaptability of the entire fresh air energy-saving optimization control system.

[0099] After identifying the reverse propagation trend of airflow, S500 adjusts the supply and exhaust ratio and optimizes the air valve opening according to the locked abnormal flow direction area and abnormal pressure boundary characteristics. It also links the anti-backflow valve and airflow isolation device to complete the real-time closure of the reverse airflow path and reconstruct the flow field structure to ensure the target consistency of airflow direction;

[0100] After identifying the reverse propagation trend of airflow, the supply and exhaust ratio is adjusted and the air valve opening is optimized according to the locked flow abnormality area and abnormal air pressure boundary characteristics. The anti-backdraft valve and airflow isolation device are linked to complete the real-time closure of the reverse airflow path and reconstruction of the flow field structure. The specific steps to ensure the target consistency of airflow direction are as follows, so as to achieve rapid response and dynamic correction to reverse airflow, and ensure the continuous stability and precise control of air flow direction inside the terminal.

[0101] After identifying a trend of reverse airflow propagation, the supply and exhaust air volume ratio for the abnormal area and its adjacent functional areas is immediately adjusted based on the previously identified flow anomaly region and its corresponding pressure boundary characteristics. The optimization target for the supply and exhaust air ratio is dynamically determined based on the pressure differential amplitude, reverse flow velocity, and flow deviation angle in the abnormal area. For example, when the pressure differential between the abnormal area and its upstream positive pressure zone is less than 50 Pa and the reverse flow velocity is greater than 15% of the original flow velocity, the supply air volume needs to be increased by 20% to 30%, while the exhaust air volume is simultaneously reduced by 10% to 15%. If the pressure differential exceeds 100 Pa and the reverse flow velocity is greater than 25% of the original flow velocity, the supply air volume needs to be increased by more than 30%, while the exhaust air volume is simultaneously reduced by more than 20%. This ratio adjustment is achieved by directly adjusting the speed and outlet pressure of the supply and exhaust fans, ensuring that the formation and spread of reverse airflow is quickly suppressed by the coordinated approach of increasing the positive pressure of the supply air and reducing the negative pressure of the exhaust air.

[0102] While adjusting the supply and exhaust air ratio, the opening of the air valves in the supply and exhaust ducts of the abnormal area and adjacent areas is dynamically optimized. Through the intelligent air valve control device, based on the real-time monitored air flow direction and flow rate data, the opening of the supply air valves in the abnormal area is amplified to increase the supply air volume and flow rate. At the same time, the opening of the exhaust air valves in the area is reduced to moderately suppress the exhaust intensity and prevent the negative pressure from further increasing. For adjacent non-abnormal areas, in order to prevent the pressure difference from being transmitted and causing backflow and diffusion, the air valve openings in these areas are fine-tuned synchronously to maintain the relative stability of their air pressure and air flow direction. The adjustment range of the air valve opening is dynamically set according to the actual pressure difference and flow rate feedback, usually controlled within an adjustment range of ±20%. After each adjustment, the adjustment effect is continuously monitored at a frequency of seconds to ensure that the adjustment and feedback form an efficient closed loop.

[0103] The anti-backdraft valve and the airflow isolation device are activated in conjunction to implement physical sealing of the identified backflow path. The anti-backdraft valve uses a structure with one-way opening and automatic closing functions, and is deployed at the node position of the exhaust duct and the main air supply branch channel in the abnormal area. When the airflow detects a reversal of flow direction, the anti-backdraft valve is immediately closed to block the reverse flow of the airflow. At the same time, the airflow isolation device is activated for the channel or space intersection position where the backflow path passes, including controllable automatic partition doors, adjustable louvers and dynamically controlled airflow curtain walls. It automatically opens according to the spatial distribution of the abnormal flow direction to form an aerodynamic isolation barrier to physically block the continued spread of the backflow. The opening and closing logic of the anti-backdraft valve and the airflow isolation device is dynamically controlled based on real-time monitoring data and simulated flow field characteristics to ensure that it can automatically restore to the normal ventilation mode after the backflow hidden danger is eliminated.

[0104] After completing the above-mentioned supply and exhaust air ratio adjustment, air valve opening optimization, and linkage control of the anti-backdraft valve and airflow isolation device, the airflow direction, flow rate and air pressure level of the target area and its adjacent areas are continuously monitored and dynamically evaluated. By comparing the flow direction stability, flow rate uniformity and pressure difference change curve before and after the adjustment, the effect of backflow control and flow field reconstruction is verified to ensure that the airflow direction is restored to the original design of the supply and exhaust air path, and the target consistency of the airflow direction is achieved at all monitoring points. All operating parameters, timestamps, monitoring feedback and control effects of the adjustment and linkage control are recorded in the database, and associated with the dynamic distribution of passengers and ventilation adjustment logs for subsequent model training and strategy optimization, so as to achieve rapid response and precise control of similar scenarios, thereby ensuring the long-term stability of the airflow organization inside the terminal, the continuous compliance of air quality standards and the optimization of energy utilization.

[0105] The purpose of this step is to quickly and efficiently intervene and correct the backflow risk areas through ventilation and physical isolation after the reverse propagation trend of airflow is accurately identified, ensuring that the airflow direction inside the terminal always flows in an orderly manner according to the designed supply and exhaust paths, preventing the backflow from damaging air quality and passenger comfort. When the airflow direction in some areas of the terminal reverses due to changes in the dynamic distribution of passengers, imbalances in static pressure regulation, or sudden changes in local pressure differences, that is, the backflow phenomenon, polluted air, hot and humid air, or air mixed with potential pathogens may flow back from the exhaust path into passenger-dense areas or functional areas with high air quality requirements, causing air pollution, the spread of odors, uncontrolled humidity, and even public health risks. This step, based on the identification of abnormal flow direction areas and pressure anomaly boundaries, first dynamically adjusts the supply and exhaust air ratio, increasing the supply volume in positive pressure areas and reducing the exhaust intensity, thereby suppressing the conditions for backflow through aerodynamic forces. Secondly, through fine-tuning the damper opening, increasing the damper opening on the supply side to increase the kinetic energy of the air supply and decreasing the damper opening on the exhaust side to reduce the suction force generated by negative pressure, thereby bidirectionally suppressing the formation and spread of backflow paths. Furthermore, by linking the backflow prevention valve and airflow isolation device in real time, key nodes and channels along the backflow path are physically sealed, completely blocking the propagation path of the backflow and preventing backflow gas from invading sensitive areas. Finally, through multi-dimensional and multi-strategy combined control, the flow field structure is dynamically reconfigured, restoring the airflow direction to the intended direction, ensuring the orderly operation of the ventilation system and the safety, cleanliness, and comfort of the terminal air environment. This step not only plays a direct role in eliminating the risk of backflow, but its control parameters and feedback mechanism also enable continuous optimization of the adjustment strategy, enhancing the intelligent level of ventilation control and emergency response capabilities.

[0106] After sealing off the backflow and reconstructing the airflow, S600 extracts passenger distribution evolution data, pressure differential response data, airflow adjustment data, and feedback control data throughout the entire process. Based on this extracted data, it iteratively trains control models for pressure differential regulation and flow direction correction, establishing an adaptively updated energy-saving control strategy library. This ensures steady-state operation of the fresh air system and closed-loop air quality assurance in a dynamic passenger flow environment.

[0107] After completing the containment of the backflow and airflow reconstruction, the passenger distribution evolution data, pressure difference response data, airflow adjustment data, and feedback control data throughout the entire process are extracted. Based on the extracted full data, the control model for pressure difference regulation and flow direction correction is iteratively trained, and an adaptively updated energy-saving control strategy library is established to achieve steady-state operation of the fresh air system and closed-loop assurance of air quality in a dynamic passenger flow environment. The specific steps are as follows to achieve continuous optimization of the control model and intelligent evolution of the ventilation strategy.

[0108] During the entire process of completing the counter-flow control and airflow organization reconstruction, multi-dimensional dynamic data involved in the entire process are recorded and collected in real time. Specifically, it includes passenger distribution evolution data, covering the changes in the number of passengers in each functional area, movement speed, residence time and flow path; pressure difference response data, recording the pressure difference value, change rate and time node of pressure difference adjustment between each functional area and the adjacent area; airflow adjustment data, covering the adjustment range of supply air volume, exhaust air volume change, supply and exhaust air ratio, air valve opening change and fan speed adjustment; feedback control data, recording the real-time changes in airflow direction and flow rate after each round of adjustment, as well as the time and dynamic stability indicators used for the airflow direction to return to the target orientation. These data are stored in real time in a high-frequency, multi-point synchronous manner to ensure the integrity of the time series and the accuracy of the spatial distribution, laying a sufficient data foundation for subsequent modeling and training.

[0109] The entire collected data set is pre-processed for cleaning and standardization. Passenger distribution data needs to be standardized through coordinates and path classification to form a structured representation of different passenger flow patterns; pressure difference response data and airflow regulation data need to be aligned through a unified timestamp to ensure that the causal relationship between passenger dynamics and ventilation response can be accurately mapped in the same time slice. During the preprocessing process, outliers, missing values, and data drift are corrected to ensure the continuity and validity of the input data. The processed data is input into the time series modeling and spatial correlation modeling links under the deep learning framework to establish a control model that can simultaneously capture the relationship between passenger flow and pressure difference and airflow changes. The model integrates a recurrent neural network to learn the time series characteristics of passenger flow and pressure difference response, and combines it with a graph neural network to characterize the dynamic relationship between airflow between functional areas, forming a multi-dimensional prediction and control capability for pressure difference regulation and flow direction correction.

[0110] To establish a control model that can simultaneously capture the correlation between passenger flow and pressure differences and airflow changes, it is necessary to design a composite deep learning architecture that combines time series modeling and spatial correlation modeling based on the processed full dynamic data. The specific method is as follows:

[0111] Passenger flow data (such as the number of passengers in the area, movement speed, and flow path) are aligned with pressure difference change data, airflow direction and velocity data in a unified time series to construct a multivariate time series input. This input is then fed into a long short-term memory network (LSTM) or a gated recurrent unit network (GRU) to extract the characteristic patterns of passenger flow behavior and the temporal evolution of pressure difference and airflow dynamics.

[0112] The functional zoning and ventilation paths of the terminal are converted into a graph structure, with regions as nodes and the airflow channels and pressure difference relationships between regions as edges. A graph convolutional network (GCN) is used to learn features of spatial correlation and capture the mutual influence of airflow and pressure difference between different regions.

[0113] By integrating the outputs of the time series model and the graph convolution model, and dynamically assigning the importance of different time periods and spatial nodes through the attention mechanism, a unified control model is formed. This model can simultaneously perceive the spatiotemporal coupling relationship between passenger flow trends, pressure difference fluctuations, and airflow distribution, and achieve accurate prediction and adjustment decisions for pressure difference control and airflow correction.

[0114] Based on the trained control model, combined with the feedback of the actual control process, the model parameters are continuously iterated and optimized. Specifically, after each counterflow control, airflow reconstruction and ventilation adjustment, the deviation analysis of the recorded actual pressure difference changes, flow direction recovery effects and the control instructions predicted by the model is carried out. In response to the error between the prediction and the actual effect, the back propagation and adaptive learning rate adjustment mechanism are used to dynamically correct the model weights and parameter configurations. Through this continuous combination of online learning and offline training, the model gradually enhances its control decision-making capabilities under different passenger flow scenarios, different spatial structures, and different pressure difference dynamics, thereby improving the adaptability and robustness of the model in complex dynamic environments.

[0115] Based on a continuously iteratively optimized control model, an energy-saving control strategy library is constructed that covers passenger flow patterns, pressure differential dynamics, airflow regulation responses, and feedback effects. This strategy library is indexed by various passenger distribution and flow combination scenarios, and stores the corresponding optimal pressure differential control and airflow correction schemes, as well as the corresponding supply and exhaust air ratios, air valve opening adjustment strategies, and linkage rules for anti-backflow valves and airflow isolation devices. In the future passenger dynamic environment, the fresh air system can quickly match and execute corresponding adjustment measures based on the optimal solution in the strategy library, realizing intelligent adaptive adjustment without manual intervention, ensuring that the air quality in the terminal continues to meet standards and the airflow organization is stable, while achieving the optimization of energy consumption of the fresh air system and the long-term guarantee of steady operation.

[0116] The purpose of this step is to extract and model the full amount of dynamic data accumulated throughout the entire process of backflow control and airflow reconstruction, thereby achieving continuous iteration and optimization of the control model for pressure differential control and airflow direction correction, and then establishing an energy-saving control strategy library that can be adaptively updated, enabling the fresh air system to maintain steady-state operation and closed-loop air quality assurance in a dynamic and complex passenger flow environment. In the complex environment of a terminal building with multiple areas, multiple channels, and multiple people, the spatial distribution and temporal changes of passenger flow are extremely uneven, resulting in dynamic instability of pressure differentials, airflow directions, and flow rates in different areas. Although a single backflow control and airflow reconstruction can temporarily restore flow field balance, if the evolution of passenger distribution, the response process of pressure differentials, the operating parameters of airflow regulation, and the effects of feedback control are not collected and analyzed during the process, the ventilation system will not be able to adapt to future dynamic changes in passenger flow patterns or abnormal environments. Therefore, by comprehensively extracting the above data and forming structured samples, combined with deep learning and machine learning models, the pressure differential control and flow direction correction strategies are continuously iteratively trained, enabling the model to learn from historical adjustment experience, predict, and optimize the adjustment path. The energy-saving control strategy library established based on this can not only cover a variety of scenarios with abnormal passenger flows and pressure differences, but also quickly call the optimal ventilation adjustment plan based on the real-time perception of the dynamic status, so as to achieve early prevention, precise response and automatic correction, and ultimately realize the adaptive control of the fresh air system to the complex environment of the terminal, while achieving the dual goals of energy saving and safety while ensuring air quality and airflow stability.

[0117] The above-mentioned terminal fresh air energy-saving optimization control method based on the prediction of passenger spatiotemporal distribution can achieve accurate perception and coordinated control of the complex passenger flow and air pressure dynamics inside the terminal, avoiding problems such as local pressure difference imbalance and airflow backpropagation caused by the rapid gathering or flow of passengers. This method dynamically predicts the passenger distribution and adjusts the supply air and static pressure levels in real time. In conjunction with pressure difference warning, airflow direction monitoring and backflow control, it ensures that the airflow always flows in an orderly manner along the predetermined path, effectively preventing polluted gases, odors or hot and humid air from flowing back into passenger-dense areas, and ensuring air quality and thermal comfort from the source. At the same time, the control model and energy-saving strategy library based on full-process data training realize the intelligent adaptive adjustment of the fresh air system under different passenger flow situations, which not only improves air quality and ventilation safety, but also optimizes energy utilization, and overall improves the service experience and operational efficiency of the terminal.

[0118] The above description is merely illustrative of certain exemplary embodiments of the present invention. It goes without saying that those skilled in the art will be able to modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the scope of protection of the claims.

[0119] It should be noted that, in this document, if there are relational terms such as first and second, etc., they are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprises", "comprising" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or device that includes the element.

[0120] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0121] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0122] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0123] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0124] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0125] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0126] The above description is merely illustrative of certain exemplary embodiments of the present invention. It goes without saying that those skilled in the art will be able to modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the scope of protection of the claims.

Claims

1. The terminal building fresh air energy-saving optimization control method based on the prediction of passenger spatiotemporal distribution is characterized by: The following steps are involved: S100 collects passengers' real-time locations and movement speeds, constructs a model of passenger temporal and spatial distribution changes, and forms a prediction basis for dynamic air pressure regulation; S200, based on a model of passenger spatiotemporal distribution changes, calculates the pressure gradient distribution in each functional area of ​​the terminal, identifies local positive pressure and adjacent negative pressure areas, establishes micro-pressure difference classification warning indicators, and determines ventilation adjustment needs; S300, based on micro-pressure differential graded warning indicators, adjusts the air supply and static pressure levels in each functional area. By increasing the air supply and static pressure in negative pressure areas and reducing the air supply in positive pressure areas, a dynamic static pressure adjustment solution is formed to achieve regional pressure differential balance. S400, based on a dynamic static pressure regulation solution, monitors airflow direction and velocity in real time, identifies reverse airflow trends, and locates abnormal flow areas and flow field characteristics; S500, in response to the reverse propagation trend of airflow, adjusts the supply and exhaust air ratio and the air valve opening, links the anti-backdraft valve and airflow isolation device, closes the reverse airflow path, reconstructs the flow field structure, and ensures consistent airflow direction; S600, after completing the closed reverse airflow path and reconstructing the flow field structure, extracts the passenger distribution, pressure difference response, airflow regulation and feedback control data of the entire process, trains the pressure difference regulation and flow direction correction model, establishes an energy-saving control strategy library, and realizes the steady-state operation of the fresh air system and closed-loop air quality assurance.

2. The terminal building fresh air energy-saving optimization control method based on passenger spatiotemporal distribution prediction according to claim 1 is characterized in that: Step S100 includes: Collect passengers' real-time location data and movement speed data to build a passenger spatiotemporal distribution change model; Based on the passenger spatiotemporal distribution change model, the passenger density, average movement speed, residence time and crowd flow frequency of each functional area are obtained to form the passenger flow behavior factor; Establish an air circulation coefficient based on the ventilation efficiency, air supply and exhaust paths, air circulation resistance, and space volume of the functional area; Based on passenger flow behavior factors and air circulation coefficients, a mathematical model of air pressure response is constructed to calculate the weight of the impact of passenger flow on the air pressure in each area, forming a predictive basis for dynamic regulation of air pressure.

3. The terminal building fresh air energy-saving optimization control method based on passenger spatiotemporal distribution prediction according to claim 1 is characterized in that: Step S200 includes: Based on the passenger spatiotemporal distribution change model, the passenger distribution density, passenger inflow and outflow numbers, average passenger movement speed, stay time and gathering and dispersing behavior patterns of each functional area are extracted; Calculate the total heat load increment and flow disturbance intensity per unit time in each functional area to form a basic database of passenger dynamic behavior and air pressure changes; Combined with the physical properties of the functional area, the real-time air pressure value and the pressure difference between areas are calculated based on the air flow continuity equation, energy conservation equation and gas state equation; Establish micro-pressure difference grading warning indicators based on pressure difference amplitude, change rate and duration to clarify ventilation adjustment needs.

4. The terminal building fresh air energy-saving optimization control method based on passenger spatiotemporal distribution prediction according to claim 1 is characterized in that: Step S300 includes: Determine the adjustment targets for air supply volume and static pressure in each functional area based on the micro-pressure difference classification warning indicators; Adjust the speed of the air supply fan, the opening of the air supply valve and the frequency conversion drive frequency of the air supply duct to dynamically adjust the air supply volume and static pressure level to achieve pressure difference balance between positive and negative pressure areas; Real-time monitoring of air pressure changes and airflow direction, and multiple rounds of progressive adjustments based on the monitoring results to ensure dynamic and stable air pressure; The adjustment parameters and pressure difference response effects are recorded as data files, and the static pressure adjustment strategy is optimized through machine learning to achieve adaptive closed-loop control.

5. The terminal building fresh air energy-saving optimization control method based on passenger spatiotemporal distribution prediction according to claim 1 is characterized in that: Step S400 includes: Real-time monitoring of the airflow direction and velocity in each functional area to form a three-dimensional airflow monitoring data set; Compare real-time monitoring data with the baseline airflow state before static pressure adjustment to identify sensitive areas of flow deviation and flow velocity drop; Based on the three-dimensional computational fluid dynamics simulation method of the terminal building's spatial structure, the airflow field in sensitive areas is simulated and dynamically reproduced to confirm the reverse propagation area and flow field characteristics of the airflow; The locked airflow reverse propagation area and flow field characteristic parameters are archived to form an airflow dynamic adjustment and abnormal feedback database for subsequent adjustment strategy optimization and intelligent decision-making.

6. The terminal building fresh air energy-saving optimization control method based on passenger spatiotemporal distribution prediction according to claim 1 is characterized in that: Step S500 includes: Based on the locked abnormal flow area and abnormal pressure boundary characteristics, the air supply and exhaust ratio is dynamically adjusted to increase the air supply volume and reduce the exhaust volume to suppress the formation of backflow; Based on real-time airflow direction and velocity data, dynamically optimize the air valve openings in abnormal areas and adjacent areas, expand the air supply valve opening, and reduce the air exhaust valve opening to prevent negative pressure from increasing and countercurrent diffusion; Linking the anti-backflow valve and the airflow isolation device to block the backflow path and physically isolate the backflow propagation; Continuously monitor airflow direction, velocity, and pressure levels, evaluate the effects of containment and flow field reconstruction, and record adjustment and control data for model training and strategy optimization.

7. The terminal building fresh air energy-saving optimization control method based on passenger spatiotemporal distribution prediction according to claim 1 is characterized in that: Step S600 includes: Extract passenger distribution evolution data, pressure difference response data, airflow adjustment data, and feedback control data during the entire process of counterflow control and airflow reconstruction; The extracted full data is cleaned, denoised, time-series aligned, and spatially normalized before being fed into a deep learning model to establish a control model linking passenger flow, pressure differentials, and airflow changes. Continuously iteratively optimize model parameters based on model training results and adjustment feedback to improve the prediction accuracy of pressure difference control and flow direction correction; Based on the optimized control model, an energy-saving control strategy library is constructed to support the adaptive adjustment and steady-state operation of the fresh air system.

8. The terminal building fresh air energy-saving optimization control method based on passenger spatiotemporal distribution prediction according to claim 7 is characterized in that: The specific steps to achieve adaptive adjustment and steady-state operation of the fresh air system through the energy-saving control strategy library are as follows: Based on the energy-saving control strategy library, the system automatically matches the corresponding pressure differential control and airflow correction schemes based on the real-time monitoring results of passenger distribution and pressure differential dynamics; According to the matching scheme, the air supply volume, exhaust volume, air valve opening and fan speed of each functional area are automatically adjusted to achieve dynamic regulation of air pressure and airflow direction; Real-time monitoring of the air pressure level, airflow direction and velocity changes in each functional area after adjustment to evaluate whether the adjustment effect has achieved the predicted target; The adjustment results are compared with the historical case data in the energy-saving control strategy library, and the plans and parameters in the strategy library are updated according to the evaluation results to achieve adaptive optimization of the strategy library.

Images