Partition and time-sharing intelligent regulation and control method and device for heating, ventilation and air conditioning system of transportation hub building fused with multi-source perception, electronic equipment and storage medium

Abstract

The invention relates to a partition and time-sharing intelligent regulation and control method and device for a heating, ventilation and air conditioning system of a transportation hub building fused with multi-source perception, electronic equipment and a storage medium, and the method comprises the steps that real-time people flow data of a target moment in the building are determined through image information collected by a camera network in the building; dividing a plurality of sub-regions of the building at the target moment according to the people flow data; according to the people flow density of each sub-region in a continuous historical time period before the target moment, predicting the people flow density trend of each sub-region in a first preset time period after the target moment; based on the parameters, determining a target dynamic comfortable temperature interval of each sub-region at the target moment; obtaining a target set point sequence of each sub-region according to the target dynamic comfortable temperature interval; and based on the target set point sequence, the heating ventilation and air conditioning system equipment is regulated and controlled. By using the method disclosed by the invention, more accurate and intelligent control on the heating ventilation air conditioning system in the transportation hub building can be realized.

Description

Technical Field

[0001] This disclosure relates to the field of building environment control and intelligent energy-saving technology, and in particular to a method and device for zoned and timed intelligent control of HVAC systems in transportation hub buildings that integrates multi-source sensing, as well as electronic equipment and storage media. Background Technology

[0002] In recent years, with the continuous growth of travel demand, large and super-large transportation hub buildings have emerged, with increasingly complex spatial structures, typically featuring tall spaces, large glass curtain walls, and diverse functional zones. However, due to the influence of climate change and the characteristics of transportation hub buildings, their internal thermal environment exhibits significant spatiotemporal heterogeneity. The enormous spatial volume, high-intensity personnel flow, and intense solar radiation heat gain from the large glass curtain walls within transportation hub buildings make the internal thermal environment exceptionally complex. Areas near the glass curtain walls are significantly affected by changes in outdoor weather and solar radiation, resulting in drastic temperature fluctuations and a huge temperature difference with the internal core area. Traditional centralized temperature control methods are ill-suited to this localized difference, easily leading to thermal discomfort in areas near the curtain walls and energy waste due to excessive cooling in the internal areas.

[0003] Meanwhile, there are significant differences in the activity levels of people in different functional areas within the transportation hub building. For example, passengers in check-in queues and security checkpoints are more active and have higher metabolic rates, while those in waiting areas are mostly seated. This difference in activity intensity directly leads to varying comfort requirements for parameters such as ambient temperature and wind speed, further increasing the difficulty of unified environmental control. In addition, the distribution of people within the terminal is highly dynamic and uneven, with flight takeoffs and landings causing a significant "tidal effect" in the boarding gate area. For example, some boarding gates within the terminal are idle or underutilized during flight intervals; if the air conditioning system continues to operate at full capacity, it will result in huge energy waste.

[0004] Currently, research on energy-saving control of transportation hub buildings in related technologies mainly focuses on improving the overall energy efficiency of the system and optimizing equipment, or on relatively extensive management of fixed zones, failing to finely couple the dynamic distribution of pedestrian flow with external environmental disturbances for on-demand regulation. Existing control strategies often struggle to respond in real time to the differentiated heating and cooling load demands of different areas caused by changes in pedestrian density and solar radiation intensity, easily leading to unreasonable resource allocation problems such as "oversupply" to sparsely populated areas and "insufficient supply" to high-density populations and areas with strong radiation. Summary of the Invention

[0005] In view of this, this disclosure proposes a zoned and time-based intelligent control scheme for the HVAC system of a transportation hub building that integrates multi-source sensing.

[0006] According to one aspect of this disclosure, a method for zoned and timed intelligent control of HVAC systems in transportation hub buildings, integrating multi-source sensing, is provided, comprising: determining real-time pedestrian flow data within the building at a target time based on image information collected by a network of cameras within the building; dividing the building into multiple sub-regions at the target time based on the real-time pedestrian flow data; predicting the pedestrian density trend of each sub-region in a first preset time period after the target time based on the pedestrian density in each sub-region in a continuous historical time period before the target time; determining a target dynamic comfort temperature range for each sub-region at the target time based on the pedestrian density trend, a first set of influencing parameters, and basic comfort temperature ranges of multiple functional areas within the building; obtaining a target setpoint sequence for HVAC system equipment in each sub-region in a second time period after the target time based on the target dynamic comfort temperature range; and controlling the HVAC system equipment based on the target setpoint sequence.

[0007] In one possible implementation, predicting the population density trend of each sub-region in a first preset time period after the target time based on the population density in a continuous historical time period before the target time of each sub-region includes: obtaining a first external disturbance variable at the target time, the external disturbance variable including weather data and flight schedule data; inputting the population density in the continuous historical time period before the target time of each sub-region and the first external disturbance variable into the prediction model to obtain the population density trend of each sub-region in the first preset time period after the target time.

[0008] In one possible implementation, the first set of influencing parameters includes the average dwell time of people and the environmental state vector at the target time within the sub-region. The environmental state vector includes indoor temperature, indoor humidity, indoor solar radiation intensity, indoor wind speed, indoor humidity, outdoor temperature, outdoor humidity, outdoor solar radiation intensity, and outdoor wind speed. The step of obtaining the target dynamic comfort temperature range for each sub-region at the target time based on the population density trend, the first set of influencing parameters, and the basic comfort temperature ranges of multiple functional areas within the building includes: inputting the population density trend and parameters from the first set of influencing parameters into their respective corresponding correction functions to obtain multiple correction amounts for the basic comfort temperature range; assigning corresponding weighting coefficients to the multiple correction amounts and performing a weighted summation to obtain a weighted correction amount; and adding the weighted correction amount to the maximum and minimum values ​​of the basic comfort temperature range to obtain the target dynamic comfort temperature range for each sub-region at the target time.

[0009] In one possible implementation, the method further includes: acquiring the outdoor carbon dioxide concentration and the predicted number of people in each sub-area at the target time, wherein the predicted number of people is obtained by performing people identification and shift time series data prediction on the image information; and obtaining a target carbon dioxide concentration setpoint based on the outdoor carbon dioxide concentration and the predicted number of people.

[0010] In one possible implementation, obtaining the target setpoint sequence for the HVAC system equipment in each sub-region during a second time period after the target time, based on the target dynamic comfort temperature range, includes: using model predictive control (MPC) to perform rolling optimization with constraints as the optimization objective, and solving to obtain the target setpoint sequence. The constraints include: ensuring that the actual temperature of each sub-region is within its corresponding target dynamic comfort temperature range; ensuring that the actual carbon dioxide concentration of each sub-region is not higher than its corresponding target carbon dioxide concentration setpoint; and minimizing the total energy consumption of the HVAC system. The target setpoint sequence includes a supply air temperature setpoint sequence, a supply air volume setpoint sequence, a fresh air ratio setpoint sequence, and a system-level chilled water supply temperature setpoint sequence.

[0011] In one possible implementation, regulating the HVAC system equipment based on the target setpoint sequence includes: adjusting the fresh air system of the HVAC system equipment according to the fresh air ratio setpoint sequence; adjusting the water system of the HVAC system equipment according to the system-level chilled water supply temperature setpoint sequence when the carbon dioxide concentration in the sub-region at the target time does not meet the first preset condition; and adjusting the air system of the HVAC system equipment according to the supply air temperature setpoint sequence and the supply air volume setpoint sequence.

[0012] In one possible implementation, the method further includes: monitoring the environmental state vector and the total energy consumption of the HVAC system for each sub-region; if the environmental state vector and the total energy consumption of the HVAC system do not meet a second preset condition, constructing a parameter updater using a machine learning algorithm; and optimizing the target dynamic comfort temperature range based on the parameter updater.

[0013] According to another aspect of this disclosure, a multi-source sensing-integrated intelligent control device for the heating, ventilation, and air conditioning (HVAC) system of a transportation hub building is provided, comprising: a pedestrian flow data acquisition module, used to determine real-time pedestrian flow data within the building at a target time by acquiring image information through a network of cameras within the building; a sub-area division module, used to divide the building into multiple sub-areas at the target time based on the real-time pedestrian flow data; a pedestrian flow density trend acquisition module, used to predict the pedestrian flow density trend of each sub-area in a first preset time period after the target time based on the pedestrian flow density in a continuous historical time period before the target time; a target dynamic comfort temperature range acquisition module, used to determine the target dynamic comfort temperature range of each sub-area at the target time based on the pedestrian flow density trend, a first set of influencing parameters, and the basic comfort temperature ranges of multiple functional areas within the building; a target setpoint sequence acquisition module, used to obtain the target setpoint sequence of the HVAC system equipment in each sub-area in a second time period after the target time based on the target dynamic comfort temperature range; and a control module, used to control the HVAC system equipment based on the target setpoint sequence.

[0014] In one possible implementation, the crowd density trend acquisition module is used to acquire a first external disturbance variable at the target time, the external disturbance variable including weather data and shift time series data; and input the crowd density of each sub-region in the continuous historical time period before the target time and the first external disturbance variable into the prediction model to obtain the crowd density trend of each sub-region in the first preset time period after the target time.

[0015] In one possible implementation, the first set of influencing parameters includes the average dwell time of people and the environmental state vector at the target time within the sub-region. The environmental state vector includes indoor temperature, indoor humidity, indoor solar radiation intensity, indoor wind speed, and outdoor temperature, outdoor humidity, outdoor solar radiation intensity, and outdoor wind speed. The target dynamic comfort temperature range acquisition module is used to input the population density trend and the parameters in the first set of influencing parameters into their respective correction functions to obtain multiple correction values ​​for the basic comfort temperature range. Corresponding weight coefficients are assigned to the multiple correction values, and a weighted sum is performed to obtain a weighted correction value. The weighted correction value is then added to the maximum and minimum values ​​of the basic comfort temperature range to obtain the target dynamic comfort temperature range for each sub-region at the target time.

[0016] In one possible implementation, the device further includes: a target carbon dioxide concentration setpoint acquisition module, used to acquire the outdoor carbon dioxide concentration and the predicted number of people in each sub-area at the target time, wherein the predicted number of people is obtained by performing people recognition and shift time series data prediction on the image information; and to obtain the target carbon dioxide concentration setpoint based on the outdoor carbon dioxide concentration and the predicted number of people.

[0017] In one possible implementation, the target setpoint sequence acquisition module is used to obtain the target setpoint sequence by performing rolling optimization through model predictive control (MPC) with the constraint conditions as the optimization objective. The constraint conditions include: ensuring that the actual temperature of each sub-region is within its corresponding target dynamic comfort temperature range; ensuring that the actual carbon dioxide concentration of each sub-region is not higher than its corresponding target carbon dioxide concentration setpoint; and minimizing the total energy consumption of the HVAC system. The target setpoint sequence includes a supply air temperature setpoint sequence, a supply air volume setpoint sequence, a fresh air ratio setpoint sequence, and a system-level chilled water supply temperature setpoint sequence.

[0018] In one possible implementation, the control module is used to adjust the fresh air system of the HVAC system equipment according to the fresh air ratio setpoint sequence; when the carbon dioxide concentration of the sub-region at the target time does not meet the first preset condition, adjust the water system of the HVAC system equipment according to the system-level chilled water supply temperature setpoint sequence; and adjust the air system of the HVAC system equipment according to the supply air temperature setpoint sequence and the supply air volume setpoint sequence.

[0019] In one possible implementation, the device further includes: a monitoring module for monitoring the environmental state vector and the total energy consumption of the HVAC system for each sub-region; if the environmental state vector and the total energy consumption of the HVAC system do not meet a second preset condition, constructing a parameter updater using a machine learning algorithm; and optimizing the target dynamic comfort temperature range based on the parameter updater.

[0020] According to another aspect of this disclosure, a non-volatile computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, implements the steps of the above-described method.

[0021] According to another aspect of this disclosure, a computer program product is provided, including a computer program or a non-volatile computer-readable storage medium carrying the computer program, wherein the computer program, when executed by a processor, implements the steps of the above-described method.

[0022] In this embodiment, real-time pedestrian flow data at a target time is determined by collecting image information from a network of cameras within the building. Based on the pedestrian flow data, multiple sub-regions of the building at the target time are divided. The pedestrian density trend of each sub-region during a first preset time period after the target time is predicted based on the pedestrian density over a continuous historical period prior to the target time. Based on the pedestrian density trend, a first set of influencing parameters, and the basic comfort temperature ranges of multiple functional areas within the building, a target dynamic comfort temperature range for each sub-region at the target time is determined. Based on the target dynamic comfort temperature range, a target setpoint sequence for the HVAC system equipment in each sub-region during a second time period after the target time is obtained. The HVAC system equipment is then adjusted based on the target setpoint sequence. This enables more precise and intelligent control of the HVAC system within a transportation hub building.

[0023] In this embodiment, by dividing the building into multiple sub-zones based on pedestrian flow data at different times, it can better adapt to dynamically changing pedestrian demand and avoid resource waste caused by fixed zoning. By analyzing pedestrian density data over consecutive historical time periods before the target time, the trend of pedestrian flow changes in the target area can be accurately captured, providing a basis for adjusting the operating parameters of the air conditioning system in advance, making the prediction results more accurate. By determining the target dynamic comfort temperature range based on pedestrian density trends, a first set of influencing parameters, and multiple influencing factors such as the basic comfort temperature range of multiple functional areas within the building, the accuracy of the target dynamic comfort temperature range is improved, and the adaptability of the HVAC system under various influencing factors is enhanced. By generating a sequence of target setpoints for HVAC equipment in future time periods for each sub-zone based on the target dynamic comfort temperature range, the system can use this sequence value to adjust the corresponding equipment in real time, significantly improving the ability to control the temperature of each sub-zone and keeping it more stably within the desired comfort range. Thus, more precise and intelligent control of the HVAC system in the transportation hub building can be achieved.

[0024] Other features and aspects of this disclosure will become clear from the following detailed description of exemplary embodiments with reference to the accompanying drawings. Attached Figure Description

[0025] The accompanying drawings, which are included in and form part of this specification, illustrate exemplary embodiments, features, and aspects of this disclosure together with the specification and serve to explain the principles of this disclosure.

[0026] Figure 1 A flowchart is shown for a method for zoned and time-based intelligent control of a HVAC system for a transportation hub building that integrates multi-source sensing, according to an embodiment of the present disclosure.

[0027] Figure 2A schematic diagram is shown of a process for generating a target dynamic comfort temperature range according to an embodiment of the present disclosure.

[0028] Figure 3 A schematic diagram of a model predictive control rolling optimization according to an embodiment of the present disclosure is shown.

[0029] Figure 4 A schematic diagram of a process for controlling a heating, ventilation, and air conditioning system device according to an embodiment of the present disclosure is shown.

[0030] Figure 5 This diagram illustrates the flow of an architecture for a zoned and time-based intelligent control method for a HVAC system in a transportation hub building that integrates multi-source sensing, according to an embodiment of the present disclosure.

[0031] Figure 6 A block diagram of a zone-based, time-based intelligent control device for a HVAC system in a transportation hub building, which integrates multi-source sensing according to an embodiment of the present disclosure, is shown.

[0032] Figure 7 A block diagram of an electronic device according to an embodiment of the present disclosure is shown. Detailed Implementation

[0033] Various exemplary embodiments, features, and aspects of this disclosure will now be described in detail with reference to the accompanying drawings. The same reference numerals in the drawings denote elements that have the same or similar functions. Although various aspects of the embodiments are shown in the drawings, they are not necessarily drawn to scale unless specifically indicated otherwise.

[0034] As used herein, the terms “comprising,” “including,” “having,” or variations thereof are open-ended and include one or more of the stated features, integrals, elements, steps, components, or functions, but do not exclude the presence or addition of one or more other features, integrals, elements, steps, components, functions, or groups thereof.

[0035] When an element is referred to as “connected,” “coupled,” “responding,” or a variation thereof relative to another element, it may be directly connected, coupled, or responding to another element, or there may be an intermediate element present.

[0036] Although the terms first, second, third, etc., may be used herein to describe various elements / operations, these elements / operations should not be limited by these terms. These terms are only used to distinguish one element / operation from another. Therefore, without departing from the teachings of the inventive concept, a first element / operation in some embodiments may be referred to as a second element / operation in other embodiments.

[0037] The term “exemplary” as used herein means “serving as an example, embodiment, or illustration.” Any embodiment illustrated herein as “exemplary” is not necessarily to be construed as superior to or better than other embodiments.

[0038] Furthermore, to better illustrate this disclosure, numerous specific details are set forth in the following detailed description. Those skilled in the art will understand that this disclosure can be practiced without certain specific details. In some instances, methods, means, components, and circuits well known to those skilled in the art have not been described in detail in order to highlight the main points of this disclosure.

[0039] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, data stored, data displayed, etc.) and signals involved in this application are all authorized by the user or fully authorized by all parties, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant regions.

[0040] In this embodiment, real-time pedestrian flow data at a target time is determined by collecting image information from a network of cameras within the building. Based on the pedestrian flow data, multiple sub-regions of the building at the target time are divided. The pedestrian density trend of each sub-region during a first preset time period after the target time is predicted based on the pedestrian density over a continuous historical period prior to the target time. Based on the pedestrian density trend, a first set of influencing parameters, and the basic comfort temperature ranges of multiple functional areas within the building, a target dynamic comfort temperature range for each sub-region at the target time is determined. Based on the target dynamic comfort temperature range, a target setpoint sequence for the HVAC system equipment in each sub-region during a second time period after the target time is obtained. The HVAC system equipment is then adjusted based on the target setpoint sequence. This enables more precise and intelligent control of the HVAC system within a transportation hub building.

[0041] Figure 1 A flowchart illustrating a method for zoned and time-based intelligent control of a HVAC system in a transportation hub building, incorporating multi-source sensing according to an embodiment of this disclosure, is shown. Figure 1 As shown, the method includes:

[0042] S11. Determine the real-time pedestrian flow data inside the building at the target time by collecting image information through the network of cameras inside the building;

[0043] The transportation hub buildings in this disclosure can include public buildings whose core function is to realize large-scale passenger distribution and transfer, which can gather passengers from different directions and using different transportation routes or modes, and redistribute passengers to new directions and routes. The transportation hub buildings can be airport terminals, railway passenger stations, railway stations, long-distance bus stations, etc., and this disclosure does not limit them.

[0044] The target time can be a point in time used for a specific operation, calculation, or state assessment; the target time can also be the current time when step S11 is executed. The camera network within the transportation hub building can be a comprehensive monitoring network formed by cameras deployed in various areas within the transportation hub building.

[0045] In this embodiment, real-time pedestrian flow data at a target time within a building can be obtained through a network of cameras deployed within the building. Specifically, a real-time video stream can be acquired through a network of cameras deployed within the building, and then processed using computer vision technology. This process can begin by using an object detection model (such as YOLO) to identify the positions of all people in the image frame by frame, generating bounding boxes. Then, a multi-object tracking algorithm (such as DeepSORT) is used to correlate these detection results across frames, assigning an ID to each person and forming a continuous motion trajectory. Finally, this trajectory information (including person IDs, person coordinates, timestamps, etc.) can be integrated and calibrated to generate a pedestrian flow dataset corresponding to the target time.

[0046] In this embodiment of the disclosure, by continuously collecting and analyzing the image information of the camera, it is possible to achieve dynamic capture and statistics of real-time crowd flow. Compared with the related technologies that rely solely on environmental sensors such as carbon dioxide concentration to estimate the number of people, this embodiment of the disclosure can achieve comprehensive and real-time accurate perception of the number of people and the distribution of crowd density by directly capturing and analyzing visual images through the camera network.

[0047] While acquiring pedestrian flow data, environmental parameters and operational data can also be collected. Specifically, a sensor network located throughout the building (such as temperature and humidity) can be utilized. Environmental data is collected via sensors such as solar radiation sensors. Operational data can be obtained in real time through access to the business systems of transportation hubs, such as airport flight information databases, railway train dispatching systems, and bus station ticketing platforms. After collecting passenger flow data, environmental data, and operational data, the above-mentioned multi-source heterogeneous data can be synchronized in time, unifying the data to the same time base, and then filtered and denoised to eliminate measurement noise and outliers; among them, operational data can include time-series data of train schedules.

[0048] S12. Based on the real-time pedestrian flow data, divide the building into multiple sub-areas at the target time.

[0049] After preprocessing the data, such as time synchronization and filtering / denoising, multiple sub-regions of the building at the target time can be divided based on the pedestrian flow data at that time. Specifically, within the physical boundaries of functional areas (such as check-in halls and waiting areas), each functional area can be divided into smaller sub-regions according to the pedestrian flow density within that functional area; that is, each sub-region must be within a functional area of ​​a certain building. In this embodiment, the building's sub-regions can also be dynamically divided based on pedestrian flow, without being limited to fixed functional boundaries. In this embodiment, algorithms (such as adaptive density clustering algorithms) can be used to analyze the pedestrian flow data at the target time, automatically identify and divide dynamic areas where people gather, and calculate the number of people and pedestrian density in each area.

[0050] In this embodiment of the disclosure, by dynamically dividing sub-regions based on real-time pedestrian flow data, the control of HVAC equipment can be matched with the actual distribution of pedestrian flow, and cooling capacity and air volume can be directly delivered to areas where there is real demand, thereby effectively avoiding energy waste and uneven comfort.

[0051] S13. Based on the pedestrian density in the continuous historical time period before the target time of each sub-region, predict the pedestrian density trend of each sub-region in the first preset time period after the target time.

[0052] Specifically, for each sub-region, the pedestrian density values ​​within a continuous historical window prior to the target time can be extracted first, forming an input sequence arranged by time. This sequence can then be input into a time-series prediction model, which, based on pedestrian density learned from historical data, predicts the pedestrian density change sequence for that sub-region over a preset period starting from the target time. The time-series prediction model can be a Long Short-Term Memory (LSTM) network, a Gated Recurrent Unit (GRU) network, or similar models; this embodiment does not limit the specific model used.

[0053] The continuous historical time period preceding the target time can be a past time window that is immediately adjacent to the target time and uninterrupted. It can be determined based on the passenger flow cycle of the transportation hub. For example, in an airport terminal, the historical window can cover the passenger queuing cycle of the previous 1-2 flights and can be set between 30 minutes and 2 hours. The first preset time period after the target time can be set according to specific control requirements. It can cover the evolution process from the current time to the next passenger flow peak or trough and can be set between 30 minutes and 2 hours. This embodiment of the present disclosure does not impose any restrictions on this.

[0054] In this embodiment of the disclosure, by using the pedestrian density of each sub-region in a continuous historical time period before the target time, the pedestrian density trend of each sub-region in the first preset time period after the target time can be predicted, thereby achieving accurate prediction of the pedestrian density trend of each sub-region.

[0055] In one possible implementation, predicting the population density trend of each sub-region in a first preset time period after the target time based on the population density in a continuous historical time period before the target time of each sub-region includes: obtaining a first external disturbance variable at the target time, the external disturbance variable including weather data and flight schedule data; inputting the population density in the continuous historical time period before the target time of each sub-region and the first external disturbance variable into the prediction model to obtain the population density trend of each sub-region in the first preset time period after the target time.

[0056] Based on the pedestrian density sequence of each sub-region over consecutive historical periods prior to the target time, relevant external disturbance variables (such as weather conditions) within the same time period can be simultaneously fused to form a multivariate time series input. This fused sequence can then be modeled using a time series model (such as LSTM or GRU). The model simultaneously captures the correlation between internal pedestrian flow evolution patterns and the influence of external disturbances, thereby inferring the future pedestrian density change trend ρ of each sub-region. forecast (i, t+Δt).

[0057] Specifically, when the transportation hub building is an airport terminal, the time sequence data can include flight information, etc.; when the transportation hub building is a train station, the time sequence data can include real-time train arrival and departure times, delay status, and carriage capacity information, etc.; when the building is a long-distance bus station, the time sequence data can include long-distance bus timetables, road condition information of major roads, etc.

[0058] In this embodiment of the disclosure, by using the population density sequence of each sub-region in the continuous historical period before the target time and incorporating relevant external disturbance variables in the same time period, the model can still maintain good predictive performance when facing emergencies.

[0059] For example, when the prediction model is LSTM, the core prediction model can be expressed as formula (1):

[0060] ρ forecast (i, t+Δt) = f LSTM ( [ρ occupancy [i, tn:t)], X exogenous (t), θ LSTM (1)

[0061] Where, θ LSTMFor model parameters, tn:t represents a historical time series window of length n, ρ occupancy (i, tn:t) represents the population density sequence at each time point within the historical time series window, X exogenous (t) represents the external disturbance variable at time t, f LSTM This is an LSTM model used to process time series data and generate prediction results.

[0062] S14. Based on the population density trend, the first set of influencing parameters and the basic comfort temperature range of multiple functional areas in the building, determine the target dynamic comfort temperature range of each sub-area at the target time.

[0063] After obtaining the population density trend ρ forecast After (i, t+Δt), the basic temperature comfort range of each sub-region can be corrected based on the trend of pedestrian density, the first set of influencing parameters, and the basic temperature comfort ranges of multiple functional areas, to obtain the target dynamic comfort temperature range of each sub-region at the target time. In one possible implementation, the first set of influencing parameters includes the average dwell time of people and the environmental state vector within the sub-region at the target time. The environmental state vector includes indoor temperature, indoor humidity, indoor solar radiation intensity, indoor wind speed, indoor humidity, outdoor temperature, outdoor humidity, outdoor solar radiation intensity, outdoor wind speed, and outdoor humidity.

[0064] In one possible implementation, after dividing the building into multiple sub-regions at the target time, these dynamically divided regions can be used as spatial units to integrate pedestrian flow data, environmental data, and operational data. This involves combining environmental and operational data with multi-dimensional features from the pedestrian flow data, such as real-time numbers and pedestrian density, to construct a unified vector for each sub-region. For example, the pedestrian density ρ in each sub-region at the target time t... occupancy (i,t), average time of stay for people t dwell (i,t) and the environmental state vector E(i,t) = [T in (i,t), RH in (i,t), Solar in (i,t), T out (t), RH out (t), Solar out (t), v in , v out ...]. Where i is a sub-region, T in For indoor temperature, Solar in Indoor solar radiation intensity, RH in For indoor humidity, T out For outdoor temperature, Solarout Outdoor solar radiation intensity, RH out outdoor humidity, v in For indoor wind speed, v out This refers to the outdoor wind speed.

[0065] In this embodiment of the disclosure, by integrating multi-source heterogeneous data such as pedestrian flow data, environmental data, and operational data, the supply of key resources such as cooling capacity and fresh air volume in the corresponding sub-area can be adjusted in advance before the actual arrival of peak passenger flow, thereby improving the system's operational energy efficiency while ensuring environmental comfort.

[0066] In this embodiment of the disclosure, the basic comfort temperature range of each sub-area can be determined according to its division method. When divided according to the physical boundaries of functional areas and the dynamic flow of people, each sub-area can inherit the basic comfort temperature range of its functional area. When the sub-areas of a building are not limited to fixed functional boundaries and are divided according to the dynamic flow of people, the basic comfort temperature range of each such dynamic sub-area can be determined according to the attributes of the main functional areas within its coverage area (for example, if it mostly falls within the waiting area, then the benchmark of the waiting area is adopted).

[0067] The basic comfortable temperature ranges for multiple functional areas within the building can be obtained first. These functional areas can be divided according to the functions within the building. For example, the functional areas of an airport terminal may include check-in areas, security check areas, waiting areas, baggage claim halls, etc., while the functional areas of a train station may include waiting halls, ticket gates, exit passages, etc. This disclosure does not impose any limitations on these aspects.

[0068] The basic comfort temperature range can be set according to the function of different areas. For example, in high-activity areas where people are constantly walking and carrying luggage (such as security checkpoints and transfer channels), the human body generates a lot of heat through metabolism, so a relatively lower temperature range (such as 23-25℃) can be set to quickly offset the heat load. In contrast, in waiting areas where people sit for long periods of time, the metabolic rate is low and they are sensitive to temperature fluctuations, so a more stable standard range (such as 24-26℃) can be set to ensure static comfort.

[0069] In one possible implementation, preset comfort temperature ranges for different functional areas within the terminal can be determined based on current seasonal information to establish a comfort temperature setting benchmark that matches the season and functional areas. Specifically, the basic comfort temperature ranges can also be set according to the functions of the functional areas within the building in different seasons. Seasons can be divided into heating season, cooling season, transitional season, etc. When the transportation hub building is an airport terminal, the basic comfort temperature ranges for different functional areas can be as follows:

[0070]

[0071] In one possible implementation, obtaining the target dynamic comfort temperature range for each sub-region at a target time based on the pedestrian density trend, the first set of influencing parameters, and the basic comfort temperature ranges of multiple functional areas within the building includes: inputting the pedestrian density trend and the parameters from the first set of influencing parameters into their respective corresponding correction functions to obtain multiple correction amounts for the basic comfort temperature range; assigning corresponding weight coefficients to the multiple correction amounts and performing a weighted summation to obtain a weighted correction amount; and adding the weighted correction amount to the maximum and minimum values ​​of the basic comfort temperature range to obtain the target dynamic comfort temperature range for each sub-region at the target time.

[0072] After obtaining the basic comfort range for each sub-region, the basic comfort temperature range can be corrected based on the population density trend and the first set of influencing parameters.

[0073] The pedestrian density trend and the parameters in the first set of influencing parameters can be input into their respective correction functions. That is, the pedestrian density trend, average dwell time, indoor temperature, indoor humidity, indoor solar radiation intensity, indoor wind speed, indoor humidity, outdoor temperature, outdoor humidity, outdoor solar radiation intensity, outdoor wind speed, and outdoor humidity can all have their own corresponding correction functions. Those skilled in the art will understand that the specific form of each correction function can be selected and designed according to actual needs. It can be a linear function, a piecewise linear function, a polynomial function, an exponential function, a discrete mapping relationship based on a lookup table, or a nonlinear function obtained by fitting a machine learning model, etc. This disclosure does not limit this.

[0074] After inputting the above parameters into their respective correction functions, multiple correction values ​​for the basic comfort temperature range can be obtained. Then, weighting coefficients can be assigned to these correction values, and a weighted sum can be performed to obtain a weighted correction value. This weighted correction value is then added to the maximum and minimum values ​​of the basic comfort temperature range to obtain the target dynamic comfort temperature range for each sub-region at the target time. The values ​​of the weighting coefficients can be determined based on the specific needs of those skilled in the art, or based on the functional characteristics of the region or historical data. For example, in a waiting area, the cumulative heat load from prolonged personnel stays is critical, so the weight of personnel stay time can be increased accordingly.

[0075] In one possible implementation, the target dynamic comfort temperature range can be found in Equations (2) and (3).

[0076] T set min (i, t) = T base min (j) + α1×F(ρ) + α2×G(tdwell ) + α3×H(Solar in ) + α4×I(v in (2)

[0077] T set max (i, t) = T base max (j) + β1×F(ρ) + β2×G(t dwell ) + β3×H(Solar in ) + β4×I(v in (3)

[0078] Where j is a sub-region, T base min (j) is the minimum value of the basic comfort temperature range, T base max (j) is the maximum value of the basic comfort temperature range, ρ is the human flow density trend, and t dwell For average staff stay time, Solar in v represents the indoor solar radiation intensity. in Let F be the indoor wind speed, G be the correction function for the trend of population density, H be the correction function for the average dwell time of people, I be the correction function for indoor solar radiation intensity, and α be the correction function for indoor wind speed. 1、 α 2、 α 3、 α 4、 β 1、 β 2、 β 3、 β4 is the weighting coefficient. T set min (i, t) represents the minimum set temperature of the i-th region at time t, where T set max (i, t) represents the maximum set temperature of the i-th region at time t.

[0079] For example, in the above formula, taking the correction function F of the pedestrian density trend as an example, the correction function F can be obtained by using a discrete mapping relationship based on a lookup table. According to the experience of those skilled in the art or data analysis, the pedestrian density can be divided into several typical intervals, and each interval can be directly assigned an optimized fixed temperature offset value. Thus, a lookup table from density intervals to correction values ​​can be established, and the correction value can be quickly obtained by looking up the table. Other correction functions can be obtained in the same way.

[0080] In this embodiment of the disclosure, the weighted correction amount obtained by weighted summation can effectively balance the influence of each influencing factor on the comfort temperature, thereby controlling the temperature of each sub-region within the optimal range and improving the indoor environmental comfort of each sub-region.

[0081] In one possible implementation, the method further includes: acquiring the outdoor carbon dioxide concentration and the predicted number of people in each sub-area at the target time, wherein the predicted number of people is obtained by performing people identification and shift time series data prediction on the image information; and obtaining a target carbon dioxide concentration setpoint based on the outdoor carbon dioxide concentration and the predicted number of people.

[0082] Among them, the trend of pedestrian density ρ was obtained in the previous text. forecast After (i, t+Δt), the predicted number of people can be obtained by multiplying the population density trend by the area of ​​the sub-region, which gives the predicted number of people in the sub-region. Alternatively, it can be obtained by performing population identification on the image information at the target time to get the number of people at the target time, and then predicting the number of people at the target time (within the first preset time period afterward) based on the shift time series data. The specific prediction implementation method is the same as that used in the previous text to obtain the population density trend ρ. forecast Similarly, the embodiments disclosed herein will not be elaborated upon in detail.

[0083] Simultaneously, predicted pedestrian flow can be used to determine indoor air quality (IAQ) control targets, such as carbon dioxide concentration setpoints. Specifically, the outdoor carbon dioxide concentration and the predicted number of people in each sub-area at the target time can be obtained first, and then the target carbon dioxide concentration setting value can be obtained based on the outdoor carbon dioxide concentration and the predicted number of people.

[0084] In one possible implementation, the carbon dioxide concentration setpoint For the specific determination method, please refer to formula (4).

[0085] (4)

[0086] Where j is a sub-region, CO 2out Outdoor carbon dioxide concentration, Let j be the predicted number of people in sub-region j at target time t. Let H be the area of ​​zone j (m²), and H be the average ceiling height of zone j (m). The adjustment coefficient (usually 0.5-1.0) is used to control the nonlinear sensitivity of the set value to changes in population density. The specific value can be determined according to the actual needs of those skilled in the art. For example, if a higher quality of air safety and comfort is desired, a higher value (such as 0.8-1.0) can be used; if the stability and energy saving of the HVAC system are emphasized, a lower value (such as 0.5-0.7) can be used. This disclosure does not limit this.

[0087] In this embodiment of the disclosure, by simultaneously determining the target carbon dioxide concentration setpoint, the HVAC system can achieve precise control of air quality, ensuring indoor air quality and human comfort.

[0088] For example, Figure 2 A schematic diagram is shown illustrating a process for generating a target dynamic comfort temperature range according to an embodiment of the present disclosure. See also... Figure 2 The basic comfort (temperature) range of key functional areas during seasonality and correction factors, including the trend of pedestrian density ρ and the average dwell time t, can be used. dwell Indoor solar radiation intensity in Indoor wind speed v in This allows us to obtain the temperature setpoint range and the carbon dioxide setpoint through a correction factor. By setting the temperature range and carbon dioxide value, the goal of creating a healthy and comfortable environment in different zones can be achieved.

[0089] In this embodiment of the disclosure, by adjusting the set temperature and carbon dioxide concentration of the HVAC system in each zone, the goal of creating a personalized, healthy, and comfortable indoor environment for different areas is achieved.

[0090] S15. Based on the target dynamic comfort temperature range, obtain the target setpoint sequence of the HVAC system equipment in the second time period after the target time for each sub-region;

[0091] After obtaining the target dynamic comfort temperature range, the target setpoint sequence of the HVAC system equipment in the second time period after the target time can be obtained based on the target dynamic comfort temperature range.

[0092] Specifically, based on the dynamic comfortable temperature range of each sub-region after the target time, the control algorithm (such as model predictive control) can be used to transform it into a sequence of specific control commands for the HVAC system equipment in the future time period, so that the HVAC system equipment can accurately maintain the temperature of the region within the comfortable range.

[0093] In one possible implementation, obtaining the target setpoint sequence of HVAC system equipment for each sub-region during a second time period after the target time, based on the target dynamic comfort temperature range, includes: using model predictive control (MPC) as the optimization objective to perform rolling optimization and solve for the target setpoint sequence. The constraints include: ensuring that the actual temperature of each sub-region is within its corresponding target dynamic comfort temperature range; ensuring that the actual carbon dioxide concentration of each sub-region is not higher than its corresponding target carbon dioxide concentration setpoint; and minimizing the total energy consumption of the HVAC system. The target setpoint sequence includes a supply air temperature setpoint sequence, a supply air volume setpoint sequence, a fresh air ratio setpoint sequence, and a system-level chilled water supply temperature setpoint sequence.

[0094] Model Predictive Control (MPC) is an advanced process control strategy based on dynamic mathematical models. Its characteristics include model-based look-ahead prediction, online optimization in the finite time domain, and rolling feedback correction.

[0095] In this embodiment, the optimization objective can be to satisfy constraints, and multi-objective setpoint rolling optimization can be performed using Model Predictive Control (MPC). Specifically, a Model Predictive Control (MPC) rolling optimization model can be established with the prediction time domain being the second time period after the target time and the control time domain being the third time period after the target time, thereby obtaining a sequence of target setpoints. This model can use the population density trend and carbon dioxide concentration setpoint as feedforward inputs to determine the optimal operating setpoint of the system by solving a constrained multi-objective optimization problem. The third time period after the target time is shorter than the second time period after the target time. Constraints can include ensuring that the actual temperature of each sub-region is within its corresponding target dynamic comfort temperature range; ensuring that the actual carbon dioxide concentration of each sub-region is not higher than its corresponding target carbon dioxide concentration setpoint; and minimizing the total energy consumption of the HVAC system. The sequence of target setpoints can include a supply air temperature setpoint sequence, a supply air volume setpoint sequence, a fresh air ratio setpoint sequence, and a system-level chilled water supply temperature setpoint sequence.

[0096] For example, the prediction time domain Hp of the MPC model can be 1 hour, and the control time domain Hc can be 15 minutes. The decision variables can include the supply air temperature T of each sub-region within the future control time domain Hc. supply (i), air supply volume G(i), fresh air ratio r(i), and system-level chilled water supply temperature T chws The objective function of the MPC model can be found in formula (5).

[0097]

[0098] in, Let H be the total cost function, and min is the function that minimizes the total cost; p The prediction time domain length is given by N; N is the number of sub-regions. Let be the actual temperature of the i-th region at time k; Set the temperature for the i-th region at time k; For the i-th region at time k, the actual concentration; Setting the i-th region at time k concentration; The total energy consumption of the system at time k can be calculated by the energy consumption models of equipment such as chillers, water pumps, and fans. This parameter can be achieved through existing technologies, and will not be described in detail in the embodiments disclosed herein. To control the input at time The variables that change (such as valve opening, fan speed, etc.) are used to solve this optimization problem, which aims to minimize the total cost. Minimize, and obtain the optimal setpoint sequence u(k) for each device within the next 15 minutes; 1. 2. 3. 4 is a weighting coefficient used to balance the importance of different cost items. This weighting coefficient can also be selected according to actual needs, which will not be elaborated here in the embodiments disclosed.

[0099] The constraints can be expressed by formulas (6) to (9):

[0100] (6)

[0101] (7)

[0102] (8)

[0103] (9)

[0104] in, Let k be the air supply volume for sub-region i at time k; This is the minimum air supply volume limit; This is the maximum air supply volume limit; This represents the fresh air ratio (the proportion of fresh air volume to total supply air volume) for sub-region i at time k. For the i-th region at time k, the actual concentration, Setting the i-th region at time k The maximum concentration; other parameters can be found above.

[0105] For example, Figure 3 A schematic diagram of a model predictive control rolling optimization according to an embodiment of the present disclosure is shown. Figure 3 The entire MPC rolling optimization process can be seen from this.

[0106] In this embodiment of the disclosure, by employing the model predictive control (MPC) method for rolling optimization, precise control of the operation of the HVAC system can be achieved, ensuring that the system remains in its optimal operating state throughout a future second time period.

[0107] S16. Based on the target setpoint sequence, regulate the HVAC system equipment.

[0108] After obtaining the target setpoint sequence, namely the supply air temperature setpoint sequence, supply air volume setpoint sequence, fresh air ratio setpoint sequence, and system-level chilled water supply temperature setpoint sequence, the HVAC system equipment can be adjusted according to the target setpoint sequence. Specifically, the system-level chilled water supply temperature setpoint sequence can be sent to the chiller and water pump to globally adjust the cooling source efficiency; the fresh air ratio setpoint sequence can be used to adjust the fresh air system of the HVAC system equipment; and the supply air temperature and supply air volume are used to control the air valves, water valves, and fan speeds in real time.

[0109] In one possible implementation, regulating the HVAC system equipment based on the target setpoint sequence includes: adjusting the fresh air system of the HVAC system equipment according to the fresh air ratio setpoint sequence; adjusting the water system of the HVAC system equipment according to the system-level chilled water supply temperature setpoint sequence when the carbon dioxide concentration in the sub-region at the target time does not meet the first preset condition; and adjusting the air system of the HVAC system equipment according to the supply air temperature setpoint sequence and the supply air volume setpoint sequence.

[0110] Specifically, the first preset condition may refer to the real-time carbon dioxide concentration in the sub-region being lower than a preset safety or comfort threshold (e.g., carbon dioxide concentration ≤ 1000 ppm). If the concentration exceeds this threshold, it is determined that the condition is not met.

[0111] In this embodiment, the opening degrees of the fresh air valve and the return air valve can be adjusted first according to the obtained fresh air ratio setpoint sequence r(i) to control the proportion of fresh air introduced. During this process, if the outdoor air enthalpy h is monitored... out Below the return air enthalpy h return Then, the preset upper limit of r(i) can be automatically adopted (i.e., the fresh air valve can be opened as wide as possible) to maximize the use of outdoor low-temperature, low-enthalpy air for free cooling, where the outdoor air enthalpy value h out and return air enthalpy h returnThis can be obtained through existing technology, and the embodiments disclosed herein will not be described in detail.

[0112] Meanwhile, if the carbon dioxide concentration in the sub-region at the target time does not meet the first preset condition (it may be that the concentration exceeds the standard), the water system of the HVAC system equipment can be adjusted according to the system-level chilled water supply temperature setpoint sequence; and the air system of the HVAC system equipment can be adjusted according to the supply air temperature setpoint sequence and the supply air volume setpoint sequence.

[0113] When adjusting the water system, the optimized system-level chilled water temperature T can be used as a reference. chws Adjust the chiller unit and variable frequency water pump to follow the principle of increasing T chws Based on the principle of improving host efficiency. Specifically, T can be... chws The setpoint is sent to the chiller unit controller, and the unit responds to this higher temperature water supply demand by increasing the evaporation temperature (or condenser level, etc.), thereby significantly reducing the compressor pressure ratio and power consumption, and improving efficiency; at the same time, the variable frequency chilled water pump can adjust according to the new T... chws The system load dynamically adjusts the rotation speed, and under the premise of ensuring sufficient flow and heat exchange temperature difference, it works together to complete the water temperature regulation.

[0114] After adjusting the water system, the air system can be adjusted. Specifically, first, adjust the zone-level supply air temperature T. supply (i) The air volume G(i) setpoint is sent to the corresponding air handling unit (AHU) and variable air volume terminal (VAV). The bottom layer adopts proportional-integral (PI) control to form a fast closed loop, which adjusts the opening of the chilled and hot water valves of the air handling unit in real time to track the air supply temperature setpoint, and adjusts the opening of the terminal air valve or the fan speed to control the air supply volume, so as to achieve accurate and fast tracking of the setpoint.

[0115] To achieve global optimization and coordination, a trim & respond strategy is introduced at the upper level to slowly and dynamically optimize the system-level setpoint. By fine-tuning global parameters such as supply air static pressure or chilled water temperature and observing terminal responses, the system can be guided towards optimal energy efficiency while meeting demand. Simultaneously, a demand-controlled ventilation (DCV-PID) strategy dynamically adjusts the fresh air ratio setpoint based on real-time monitored carbon dioxide concentration using a dedicated PID controller, achieving on-demand ventilation. All control commands are processed before output, including amplitude and ramp rate limitations, to ensure smooth operation of the executing equipment and stable and reliable system response.

[0116] For example, Figure 4 A schematic diagram of a process for controlling a heating, ventilation, and air conditioning (HVAC) system device according to an embodiment of the present disclosure is shown. See also... Figure 4Based on the MPC optimization design point, optimization instructions can be conveyed. Then, the upper layer introduces the Trim & Respond strategy to perform slow dynamic optimization of the system-level setpoint. Temperature setpoints can be used for PI control of temperature, air volume, etc., and carbon dioxide setpoints can be used for DCV-PID fresh air control. The control outputs of air valves, water valves, and fresh air are all limited by amplitude and ramp rate. Safe instruction execution equipment is used.

[0117] In this embodiment of the disclosure, by prioritizing and independently adjusting the fresh air system, the water system is adjusted only when the carbon dioxide concentration in the sub-area at the target time does not meet the first preset condition. This avoids unnecessary over-operation of the water system and optimizes the overall energy efficiency of the system while ensuring health and comfort.

[0118] In one possible implementation, the method further includes: monitoring the environmental state vector and the total energy consumption of the HVAC system for each sub-region; if the environmental state vector and the total energy consumption of the HVAC system do not meet a second preset condition, constructing a parameter updater using a machine learning algorithm; and optimizing the target dynamic comfort temperature range based on the parameter updater.

[0119] It can continuously collect and monitor the system's environmental state vector E actual The key performance indicators (KPIs) are calculated from (i, t) and the total energy consumption of the HVAC system. The key performance indicators may include the comfort compliance rate, energy consumption per unit area, and IAQ compliance rate. The comfort compliance rate can be found in formula (10), the energy consumption per unit area can be found in formula (11), and the IAQ compliance rate can be found in formula (12).

[0120] (10)

[0121] in, The comfort level compliance rate is represented by N, where N is the number of sub-regions, T is the total number of discrete sampling times within the evaluation period, and t represents all discrete sampling times within the evaluation period. For actual temperature measurement, This represents the total number of "region-time" samples.

[0122] (11)

[0123] in, The total energy actually consumed by the system during the evaluation period. The total building area is EUI (Energy Use Intensity), which is the energy consumption per unit area.

[0124] (12)

[0125] in, This represents the actual carbon dioxide concentration. Set an upper limit for carbon dioxide concentration. This refers to the IAQ (carbon dioxide) compliance rate over time; other parameters are described above.

[0126] The second preset condition can be a key performance indicator, namely whether the comfort compliance rate, energy consumption per unit area, and IAQ compliance rate have reached preset target thresholds. The preset target thresholds can be set according to actual needs. For example, if the comfort compliance rate is lower than 95% and / or the energy consumption per unit area exceeds the preset benchmark value by 20%, and / or the IAQ compliance rate is lower than 90%, the second preset condition can be considered not met.

[0127] When the environmental state vector and the total energy consumption of the HVAC system do not meet the second preset condition, a parameter updater can be constructed using machine learning algorithms. Specifically, a parameter updater can be constructed using machine learning algorithms (such as reinforcement learning or Bayesian optimization) based on the environmental state vector and KPI evaluation results. Based on the parameter updater, the formula for optimizing the target dynamic comfort temperature range is obtained. The formula for optimizing the target dynamic comfort temperature range can be found in formula (13).

[0128] (13)

[0129] in, For parameter updater, EUI represents the energy consumption per unit area, indicating the comfort level compliance rate. The environmental state vectors collected historically and the total energy consumption of the HVAC system are used as the basis for this calculation. , For the set of parameters to be updated, It can include α in formulas (2) and (3). 1、 α 2、 α 3、 α4, It can include β in formulas (2) and (3). 1、 β 2、 β 3、 β4.

[0130] In one possible implementation, the model parameters of MPC can also be optimized based on the parameter updater, as shown in formula (14).

[0131] (14)

[0132] in, This is the updated parameter vector of the MPC model. For parameter updater, Set the value for the target dynamic comfort temperature range. This is the actual temperature; other parameters can be found in the previous text.

[0133] In this embodiment, the system strategy can be continuously self-optimized through periodic online updates, allowing the control performance to approach the global optimum over long-term operation. The evaluation cycle and update cycle can be determined according to actual needs. For example, the update cycle can be a higher frequency such as five minutes or fifteen minutes to achieve the best real-time comfort within the building, and the evaluation cycle can be consistent with the update cycle.

[0134] In this embodiment, real-time pedestrian flow data at a target time is determined by collecting image information from a network of cameras within the building. Based on the pedestrian flow data, multiple sub-regions of the building at the target time are divided. The pedestrian density trend of each sub-region during a first preset time period after the target time is predicted based on the pedestrian density over a continuous historical period prior to the target time. Based on the pedestrian density trend, a first set of influencing parameters, and the basic comfort temperature ranges of multiple functional areas within the building, a target dynamic comfort temperature range for each sub-region at the target time is determined. Based on the target dynamic comfort temperature range, a target setpoint sequence for the HVAC system equipment in each sub-region during a second time period after the target time is obtained. The HVAC system equipment is then adjusted based on the target setpoint sequence. This enables more precise and intelligent control of the HVAC system within a transportation hub building.

[0135] Application scenario examples

[0136] Figure 5 This diagram illustrates the flow of an architecture for a zoned and time-based intelligent control method for a transportation hub building HVAC system integrating multi-source sensing, according to an embodiment of this disclosure. Figure 5 As shown, the process includes:

[0137] Step 1 (S1): Data Acquisition and Fusion

[0138] By collecting information on pedestrian flow, environment, and flights through cameras and sensor networks, and after filtering and fusion processing, the system outputs real-time pedestrian density, dwell time, and precise environmental conditions for each zone.

[0139] Specifically, through a network of cameras deployed in various areas of the terminal, indoor and outdoor temperature and humidity monitoring, A solar radiation sensor network is established and connected to the airport flight information database to collect real-time data on passenger flow, environment, and operations. The aforementioned multi-source heterogeneous data undergoes time synchronization, filtering, noise reduction, and data fusion processing to output the passenger flow density ρ for each zone at time t. occupancy (i,t), average time of stay for people t dwell (i,t) and the environmental state vector E(i,t) = [T in (i,t), RH in (i,t),Solar in (i,t), T out (t), RH out (t), Solar out (t),...].

[0140] Step 2 (S2): Pedestrian Flow Feature Extraction and Trend Prediction

[0141] Based on visual models, real-time pedestrian flow features are extracted and combined with exogenous variables such as flights and weather to predict pedestrian flow trends in the next 30-120 minutes.

[0142] Specifically, the video stream of S1 can be analyzed in real time based on computer vision models such as YOLO and DeepSORT to extract the coordinates and movement trajectories of people in each zone. This is combined with external disturbance variables X from S1, such as flight takeoff and landing information and outdoor weather data. exogenous (t), using time series models (such as LSTM, GRU) to predict the trend of crowd density in each zone within the future time Δt (30–120 minutes), see formula (1).

[0143] Step 3 (S3): Dynamic Comfort and Health Goal Generation

[0144] Based on the basic comfort range of functional areas, and combined with real-time and predicted pedestrian density, dwell time, season and solar radiation, the comfort temperature range of each zone is dynamically adjusted, and the indoor air quality (IAQ) control target is determined simultaneously.

[0145] Specifically, based on the basic comfort temperature range [T] of different functional areas i (such as check-in area, waiting area, and commercial area) of the terminal in different seasons. base min (j), T base max (j)], combining S1 and S2 to obtain real-time and predicted crowd density ρ and crowd dwell time t dwell Solar radiation intensity in Indoor wind speed v in Dynamically adjust the comfort temperature setting range for each zone [T] set min(i, t), T set max (i, t)].

[0146] The basic comfort zones for different functional areas are as follows:

[0147]

[0148] The modified model can be quantified as formula (2) and formula (3).

[0149] Simultaneously, indoor air quality (IAQ) control targets, such as carbon dioxide concentration setpoints, are determined based on predicted pedestrian traffic. For details, please refer to formula (4).

[0150] Step 4 (S4): Multi-objective setpoint optimization based on MPC (upper-level decision-making)

[0151] A rolling optimization model of Model Predictive Control (MPC) with a prediction time domain of 1 hour is established. Internally, it uses the system model to predict future loads and optimizes the supply air parameters, water temperature setpoints, etc. of each zone within the next 15 minutes with "comfort deviation + total system energy consumption" as the core objective.

[0152] Specifically, a Model Predictive Control (MPC) rolling optimization model is established with a prediction time domain (Hp) of 1 hour and a control time domain (Hc) of 15 minutes. This model uses the predicted pedestrian flow trend in S2 and the dynamic comfort and health goals generated in S3 as feedforward inputs. By solving a constrained multi-objective optimization problem, the optimal operating setpoint of the system is determined. Decision variables include the supply air temperature T of each zone within the future Hc. supply (i), air supply volume G(i), fresh air ratio r(i), and system-level chilled water supply temperature T chws For the specific formula, please refer to formula (5).

[0153] Step 5 (S5): Device collaborative execution based on energy saving priority (lower-level execution)

[0154] The optimized setpoints from the upper layer are distributed to the operational layer, enabling robust local control through PI control, Trim & Respond, and DCVPD. Each actuator adjusts air supply and water valves according to the setpoints, ensuring smooth and reliable operation through amplitude limiting and ramp limiting.

[0155] Specifically, the upper-level optimized setpoint u(k) generated by S4 is sent down to the operation layer of the building automation system. First, the fresh air system is adjusted: based on the optimized fresh air ratio r(i), the opening of the fresh air valve is adjusted. When the outdoor air enthalpy h... out Below the return air enthalpy h returnAt that time, the upper limit of r(i) is used to maximize the utilization of free cooling. Secondly, the water system is adjusted: based on the optimized system-level chilled water temperature T... chws Adjust the chiller unit and variable frequency water pump to follow the principle of increasing T chws To improve the efficiency of the main unit. Secondly, regulate the air system: adjust the zone-level setpoint T. supply (i) and G(i) are issued to each air conditioning unit and variable air volume terminal, driving actuators such as air valves and water valves through a hybrid strategy including PI control, Trim & Respond (optimal response), and demand-controlled ventilation (DCV-PID). PI control is primarily used to control the opening degree of water valves and air valves. Trim & Respond (optimal response) mainly handles the coordination problem of "contention" between multiple local controllers and insufficient global resources. Demand-controlled ventilation (DCV-PID) focuses on addressing the issue of supplying fresh air on demand. All control actions are subject to amplitude and ramp rate limits to ensure a smooth and reliable system response.

[0156] Step 6 (S6): System Feedback and Data Recording

[0157] Continuously collect actual system operation data, calculate key performance indicators (KPIs), including comfort compliance rate, energy consumption per unit area, and IAQ compliance rate, and compare and analyze them with the targets set in S3 and S4.

[0158] Specifically, continuously collect actual system operation data, including environmental data for each partition. actual (i, t) and total system energy consumption P actual (t). To calculate the key performance indicators (KPIs), please refer to formulas (10) to (12).

[0159] Step 7 (S7): Model self-learning and parameter optimization. Based on the long-term running data and KPI evaluation results of S6, the parameters of the dynamic comfort model in S3 and the parameters of the MPC model in S4 are updated online using machine learning algorithms to achieve continuous self-optimization of the system strategy.

[0160] Specifically, based on the long-term operational data and KPI evaluation results recorded in S6, a parameter updater g is constructed using machine learning algorithms (such as reinforcement learning or Bayesian optimization). For the dynamic comfort model in S3, its weight coefficients are optimized. See formula (13) for the MPC model in S4, calibrate its system model parameters. For details, please refer to formula (14).

[0161] In this embodiment of the disclosure, the system strategy is continuously self-optimized through periodic online updates, so that the control performance continuously approaches the global optimum during long-term operation.

[0162] In this embodiment, real-time video analysis accurately senses the population density and dwell time in each zone, and combined with environmental parameters such as solar radiation, dynamically generates and optimizes the temperature and fresh air setpoints for each zone. This effectively solves the "oversupply" or "undersupply" problems caused by traditional fixed setpoint control, enabling the air conditioning and ventilation loads to precisely match the actual needs of the area, ensuring a comfortable and healthy environment while avoiding energy waste at the source. Secondly, by using the predicted population flow trend as a feedforward signal to the model predictive controller, the system can adjust its operating status in advance and smooth out load fluctuations. Simultaneously, following the energy-saving priority execution strategy of "fresh air-water system-air system," it prioritizes the use of free cooling and optimizes the efficiency of the main unit, ensuring that upper-level optimization commands are executed in a manner with the highest global energy efficiency, thereby significantly reducing the overall energy consumption of the system.

[0163] The embodiments disclosed herein are not limited to improvements on a single algorithm or device, but rather represent a complete solution. The architecture begins with the synchronous fusion processing of heterogeneous data from multiple sources, including airport cameras, environmental sensors, and flight information, and uses this data to drive predictions of the spatiotemporal distribution of passenger flow and environmental trends. The prediction results are then used as feedforward signals to dynamically generate zoned and time-specific comfort and indoor air quality targets. Subsequently, model predictive control (MPC) is used for system-level rolling optimization calculations to derive the optimal setpoint sequence, which is ultimately distributed to various levels of execution equipment through a collaborative strategy based on energy-saving priorities. The performance of the entire system is continuously monitored and fed back to the model parameters for self-learning optimization, forming a continuously self-improving intelligent closed loop.

[0164] This disclosure breaks through the traditional fixed setpoint control mode and creatively establishes a quantitative relationship model between human flow, environment, and comfort. This model acknowledges that densely populated areas have higher metabolic heat, areas with longer dwell times are more sensitive to temperature fluctuations, and areas with strong solar radiation have greater sensible heat loads. Therefore, the modified model proposed in this disclosure, as shown in formulas (2) and (3), allows the system to intelligently "predict" changes in load and comfort requirements in each area and adjust control targets in advance. This not only improves the control accuracy of comfort but also avoids energy waste caused by unreasonable setpoints, making it a key algorithm for achieving "on-demand control."

[0165] This disclosure embodiment incorporates a highly efficient energy-saving decision-making mechanism at the execution layer. The strategy stipulates: First, based on the optimized fresh air ratio and comparing the outdoor and return air enthalpy values, it determines whether to maximize the use of free cooling, prioritizing energy saving by adjusting the fresh air valve. Second, based on the system-level chilled water temperature setpoint, it adjusts water-side equipment such as chiller units and variable frequency pumps, adhering to the efficient principle of "maximizing chilled water supply temperature while meeting demand." Finally, the zone-level supply air temperature and air volume settings are distributed to each air conditioning unit and variable air volume terminal. During this process, "Trim & Respond" is used to resolve resource contention among multiple terminals, combined with "Demand Controlled Ventilation (DCV)" to ensure precise, on-demand fresh air supply. This collaborative strategy ensures the entire system always operates in the most energy-efficient mode, translating optimization results into actual energy savings.

[0166] Figure 6 This diagram illustrates a block diagram of a zone-based, time-based intelligent control device for a HVAC system in a transportation hub building, based on an embodiment of this disclosure, incorporating multi-source sensing. Figure 6 As shown, the device 20 includes:

[0167] The pedestrian flow data acquisition module 21 is used to determine the real-time pedestrian flow data inside the building at a target time by collecting image information through the camera network inside the building.

[0168] The sub-region division module 22 is used to divide the building into multiple sub-regions at the target time based on the real-time pedestrian flow data.

[0169] The crowd density trend acquisition module 23 is used to predict the crowd density trend of each sub-region in the first preset time period after the target time based on the crowd density in the continuous historical time period before the target time of each sub-region.

[0170] The target dynamic comfort temperature range acquisition module 24 is used to determine the target dynamic comfort temperature range of each sub-region at the target time based on the population density trend, the first set of influencing parameters and the basic comfort temperature range of multiple functional areas in the building.

[0171] The target setpoint sequence acquisition module 25 is used to obtain the target setpoint sequence of the HVAC system equipment in the second time period after the target time for each sub-region, based on the target dynamic comfort temperature range.

[0172] The control module 26 is used to control the HVAC system equipment based on the target setpoint sequence.

[0173] In some embodiments, the functions or modules of the apparatus provided in this disclosure can be used to perform the methods described in the above method embodiments. The specific implementation can be referred to the description of the above method embodiments, and for the sake of brevity, it will not be repeated here.

[0174] This disclosure also provides a non-volatile computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the above-described method.

[0175] This disclosure also provides a computer program product, including a computer program or a non-volatile computer-readable storage medium carrying the computer program, wherein the computer program, when executed by a processor, implements the steps of the above method.

[0176] Figure 7 A block diagram of an electronic device according to an embodiment of the present disclosure is shown. For example, device 1900 may be provided as a server or terminal device. (Refer to...) Figure 7 The apparatus 1900 includes a processing component 1922, which further includes one or more processors, and memory resources represented by memory 1932 for storing instructions, such as application programs, that can be executed by the processing component 1922. The application programs stored in memory 1932 may include one or more modules, each corresponding to a set of instructions. Furthermore, the processing component 1922 is configured to execute instructions to perform the methods described above.

[0177] Device 1900 may also include a power supply component 1926 configured to perform power management of device 1900, a wired or wireless network interface 1950 configured to connect device 1900 to a network, and an input / output interface 1958 (I / O interface). Device 1900 can operate on an operating system, such as Windows Server, stored in memory 1932. TM macOS X TM Unix TM Linux TM FreeBSD TM Or similar.

[0178] In an exemplary embodiment, a non-volatile computer-readable storage medium is also provided, such as a memory 1932 including computer program instructions that can be executed by a processing component 1922 of the device 1900 to perform the above-described method.

[0179] Computer-readable storage media can be tangible devices capable of holding and storing programs / instructions used by instruction execution devices. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0180] The computer program (or computer-readable program instructions) described herein can be downloaded from a computer-readable storage medium to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage medium in the respective computing / processing device.

[0181] The computer program (or computer program instructions) used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing state information from the computer-readable program instructions to implement various aspects of this disclosure.

[0182] Various aspects of this disclosure are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions.

[0183] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that, when executed by the processor of the computer or other programmable data processing apparatus, they create means for implementing the functions / actions specified in one or more blocks of the flowchart and / or block diagram. These computer-readable program instructions can also be stored in a computer-readable storage medium that causes a computer, programmable data processing apparatus, and / or other device to operate in a particular manner; thus, the computer-readable medium storing the instructions comprises an article of manufacture that includes instructions for implementing aspects of the functions / actions specified in one or more blocks of the flowchart and / or block diagram.

[0184] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to perform the functions / actions specified in one or more boxes of a flowchart and / or block diagram.

[0185] The above description is merely an exemplary embodiment of the present invention and is not intended to limit the scope of protection of the present invention, which is determined by the appended claims.

[0186] The term “exemplary” as used herein means “serving as an example, embodiment, or illustration.” Any embodiment illustrated herein as “exemplary” is not necessarily to be construed as superior to or better than other embodiments.

[0187] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0188] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of an instruction containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than those shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0189] The various embodiments of this disclosure have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or technical improvements to the embodiments in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A method for zoned and time-based intelligent control of HVAC systems in transportation hub buildings, integrating multi-source sensing, characterized in that, include: By collecting image information from the network of cameras inside the building, real-time pedestrian flow data inside the building at a target time can be determined; Based on the real-time pedestrian flow data, multiple sub-areas of the building at the target time are divided; Based on the pedestrian density in each sub-region during consecutive historical time periods before the target time, predict the pedestrian density trend of each sub-region in the first preset time period after the target time. Based on the aforementioned pedestrian density trend, the first set of influencing parameters, and the basic comfort temperature ranges of multiple functional areas within the building, the target dynamic comfort temperature range for each sub-area at the target time is determined. Based on the target dynamic comfort temperature range, the target setpoint sequence of the HVAC system equipment in each sub-region during the second time period after the target time is obtained. Based on the target setpoint sequence, the HVAC system equipment is controlled.

2. The method according to claim 1, characterized in that, The step of predicting the population density trend of each sub-region in the first preset time period after the target time, based on the population density in the continuous historical time period before the target time of each sub-region, includes: Obtain the first external disturbance variable at the target time, wherein the external disturbance variable includes weather data and flight time series data; The pedestrian density of each sub-region during a continuous historical time period before the target time and the first external disturbance variable are input into the prediction model to obtain the pedestrian density trend of each sub-region during a first preset time period after the target time.

3. The method according to claim 1, characterized in that, The first set of influencing parameters includes the average dwell time of people at the target time and the environmental state vector within the sub-region. The environmental state vector includes indoor temperature, indoor humidity, indoor solar radiation intensity, indoor wind speed, indoor humidity, outdoor temperature, outdoor humidity, outdoor solar radiation intensity, and outdoor wind speed. Based on the pedestrian density trend, the first set of influencing parameters, and the basic comfort temperature ranges of multiple functional areas within the building, the target dynamic comfort temperature range for each sub-area at the target time is obtained, including: The population density trend and the parameters in the first set of influencing parameters are respectively input into their respective correction functions to obtain multiple correction values ​​for the basic comfort temperature range. Assign corresponding weight coefficients to the multiple correction values ​​and perform a weighted summation to obtain the weighted correction value; The weighted correction amount is added to the maximum and minimum values ​​of the basic comfort temperature range to obtain the target dynamic comfort temperature range for each sub-region at the target time.

4. The method according to claim 1, characterized in that, The method further includes: The outdoor carbon dioxide concentration and the predicted number of people in each sub-area at the target time are obtained. The predicted number of people is obtained by performing people identification and shift time series data prediction on the image information. The target carbon dioxide concentration setting is obtained based on the outdoor carbon dioxide concentration and the predicted number of people.

5. The method according to claim 4, characterized in that, The step of obtaining the target setpoint sequence of HVAC system equipment for each sub-region in the second time period after the target time, based on the target dynamic comfort temperature range, includes: With the objective of satisfying constraints, rolling optimization is performed using model predictive control (MPC) to obtain the sequence of target setpoints. The constraints include: The actual temperature of each sub-region is kept within its corresponding target dynamic comfort temperature range. The actual carbon dioxide concentration in each sub-region shall not exceed its corresponding target carbon dioxide concentration setting value; To minimize the total energy consumption of the HVAC system; The target setpoint sequence includes a supply air temperature setpoint sequence, a supply air volume setpoint sequence, a fresh air ratio setpoint sequence, and a system-level chilled water supply temperature setpoint sequence.

6. The method according to claim 5, characterized in that, The step of regulating the HVAC system equipment based on the target setpoint sequence includes: Adjust the fresh air system of the HVAC system equipment according to the fresh air ratio setpoint sequence; When the carbon dioxide concentration in the sub-region at the target time does not meet the first preset condition. Adjust the water system of the HVAC system equipment according to the system-level chilled water supply temperature setpoint sequence; The air system of the HVAC system equipment is adjusted according to the supply air temperature setpoint sequence and the supply air volume setpoint sequence.

7. The method according to any one of claims 1 to 6, characterized in that, The method further includes: Monitor the environmental state vector and total energy consumption of the HVAC system in each of the sub-regions; If the environmental state vector and the total energy consumption of the HVAC system do not meet the second preset condition, a parameter updater is constructed using a machine learning algorithm. Based on the parameter updater, the target dynamic comfort temperature range is optimized.

8. A zoned and time-based intelligent control device for the HVAC system of a transportation hub building, integrating multi-source sensing, characterized in that, The device includes: The pedestrian flow data acquisition module is used to determine the real-time pedestrian flow data inside the building at a target time by collecting image information through the camera network inside the building; The sub-region division module is used to divide the building into multiple sub-regions at the target time based on the real-time pedestrian flow data. The crowd density trend acquisition module is used to predict the crowd density trend of each sub-region in the first preset time period after the target time based on the crowd density of each sub-region in the continuous historical time period before the target time. The target dynamic comfort temperature range acquisition module is used to determine the target dynamic comfort temperature range of each sub-region at the target time based on the population density trend, the first set of influencing parameters and the basic comfort temperature range of multiple functional areas in the building. The target setpoint sequence acquisition module is used to obtain the target setpoint sequence of HVAC system equipment in the second time period after the target time for each sub-region, based on the target dynamic comfort temperature range. The control module is used to control the HVAC system equipment based on the target setpoint sequence.

9. A non-volatile computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer program product comprising a computer program, or a non-volatile computer-readable storage medium carrying a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images