Building fresh air linkage control system based on Internet of Things

Abstract

The invention discloses a building fresh air linkage control system based on the Internet of Things, and relates to the technical field of intelligent control linkage, a sensing acquisition module is used for acquiring building state data, analyzing the building state data according to a priority strategy, analyzing building spaces in different regions, generating region information priorities, and sending the region information priorities to a central processing unit; focusing the key change area according to the area information priority to obtain a building environment potential field graph; the cooperative control module is used for performing multi-region cooperative decision on the building environment potential field graph to obtain an expected environment parameter sequence, and adjusting the expected environment parameter sequence through an optimization algorithm to obtain a fresh air coordination instruction set; and the execution decision module is used for performing multi-objective optimization on the fresh air coordination instruction set, generating an equipment-level control instruction, issuing the control instruction to a building equipment executor and completing accurate control of the fresh air system. According to the system, an optimized and convergent fresh air closed-loop control mechanism is established, operation of cold and heat source equipment is effectively coordinated, the energy-saving effect is achieved, and energy waste is avoided.

Description

Technical Field

[0001] This invention relates to the field of technology, and in particular to a building fresh air linkage control system based on the Internet of Things. Background Technology

[0002] With the acceleration of urbanization and the increasing airtightness of modern buildings, the air quality problem inside buildings has become increasingly prominent. In order to ensure the health, comfort and work efficiency of indoor personnel, fresh air systems have become standard configurations in high-end office buildings, hotels, hospitals and residences. However, traditional fresh air control systems have obvious shortcomings in terms of intelligence, energy efficiency and management. Therefore, there is an urgent need for a new control system that links the Internet of Things with building fresh air systems.

[0003] Traditional fresh air systems are typically designed and operated as independent subsystems, lacking effective data interaction and collaborative control with other critical building systems such as HVAC systems, indoor environmental monitoring systems, security access control systems, and building automation systems. These systems cannot detect the surge in carbon dioxide concentrations in meeting rooms due to impromptu meetings, nor can they dynamically adjust the fresh air volume based on the air conditioning's cooling / heating load. This often results in energy waste by simultaneously opening windows for ventilation and applying maximum cooling. Furthermore, they cannot integrate with access control systems to reduce or shut down fresh air during non-office hours or in unoccupied areas, leading to energy waste. Most existing systems employ simple on / off or tiered control based on fixed schedules or single parameters. In addition, limited sensors cannot accurately reflect the entire building space, especially the micro-environment of functionally complex areas. The system's operational status and fault alarms still rely on manual inspections and experience-based judgment. Summary of the Invention

[0004] The technical problem solved by this invention is that traditional building fresh air systems have the problem of response lag, uneven dynamic distribution control in various areas, and are prone to defects that are incompatible with energy consumption and comfort.

[0005] To address the aforementioned technical problems, this invention provides the following technical solution: a building fresh air linkage control system based on the Internet of Things, comprising a sensing and acquisition module, a collaborative learning module, and an optimization execution module: The sensing and acquisition module is used to collect building status data, analyze the building status data according to a priority strategy, analyze the temporal change amplitude of building space in different areas, generate regional information priority, focus on key change areas according to regional information priority, and obtain a building environment potential field map. The collaborative control module is used to make multi-region collaborative decisions on the building environment potential field map, obtain the desired environmental parameter sequence, and adjust the high-order smoothing of the Bézier curve through an optimization algorithm to obtain the fresh air coordination instruction set. The execution decision module is used to perform multi-objective optimization of the fresh air coordination instruction set, generate device-level control instructions, and send the control instructions to the building equipment actuators to complete the precise control of the fresh air system.

[0006] Preferably, the sensing and acquisition module includes a data acquisition unit, a priority evaluation unit, and an environmental potential field diagram unit; The data acquisition unit includes: IoT sensors are deployed in various areas of the building to collect temperature, humidity, carbon dioxide concentration and PM2.5 concentration in each area in real time as building environment data. Infrared sensors and binocular vision people counting cameras are used to collect the number of people, their distribution and movement trajectory as people movement data. By using sensors in each area as network nodes, the collected building environment data and personnel movement data are classified and defined according to the time of collection to generate building status vectors. The IoT sensors include a high-precision carbon dioxide sensor, a temperature, humidity, and pressure combined sensor, and a laser PM2.5 sensor; The building state vector includes building environment data, personnel movement data, and time context; The time context includes time period labels and date types; The time period labels are morning, lunch break, and after get off work. The date types include weekdays and rest days.

[0007] Preferably, the priority evaluation unit is used to calculate the information priority score of each area based on the generated building status vector through temporal change amplitude analysis and dynamic weighting strategy; The time-series variation amplitude analysis includes calculating the rate of change, acceleration of change, and cumulative deviation of building environment data and personnel movement data; The rate of change is the difference between the building state vector of the current sampling period and the previous sampling period, used to reflect the instantaneous rate of change of the parameters. The change in acceleration is calculated using second-order difference to determine the trend of the rate of change relative to time, which is used to predict the risk of building deterioration. The cumulative deviation is the persistent deviation between the current building state vector and the preset comfort vector, used to describe the long-term discomfort level of the area; The dynamic weighting strategy includes: The initial time point is set at 6:00 AM, which represents the moment when the first weight is 0 and the second weight is 100%. After 1 hour, the first weight increases by 1 / 12 and the second weight decreases by 1 / 12. After 12 hours, the first weight decreases by 1 / 12 and the second weight increases by 1 / 12. The first and second weights change dynamically in a 24-hour cycle. The first weight is assigned a cumulative deviation, and the second weight is divided equally and assigned a rate of change and an acceleration of change, respectively. By normalizing and weighting the rate of change, acceleration of change, and cumulative deviation using a dynamic weighting strategy, an information priority score is obtained.

[0008] Preferably, the environmental potential field map unit is used to focus on key change areas based on the priority score of regional information to obtain a building environmental potential field map; Building a building area graph structure based on spatiotemporal attention network. Where node V represents each area of ​​the building, and edge This represents the spatial adjacency relationship between different regions. Each node stores a potential field value, which is generated by superimposing the basic environmental potential field and the priority adjustment potential field. The basic environmental potential field of each region is calculated by adding the recovered potential field value and the disturbance potential field value. The expression for calculating the recovered potential field value is as follows: ; in, To recover the potential field gain coefficient, For the current area Building environment data, The building environment data in the preset comfort vector, To restore the potential field value; ; in, The value of the disturbance potential field. The gain coefficient of the interference potential field. The sources of interference include densely populated areas and equipment heat sources. For the region With interference source Differences in building environment data between them The range of influence of the interference potential field; Each basic environmental potential value is multiplied by the information priority score of the current region to obtain the adjusted potential value. Based on the interaction between adjacent regions, the potential field is propagated through the building area graph structure. A graph convolutional network is then used to aggregate the potential values ​​of adjacent regions and update the potential value of each region. The potential field value of each region is assigned to the corresponding node to generate a building environment potential field map.

[0009] Preferably, the collaborative control module includes a multi-region collaborative decision-making unit and an instruction generation unit; The multi-regional collaborative decision-making unit includes: Based on the potential field values ​​corresponding to each node in the building environment potential field diagram, a multi-regional collaborative decision-making framework is constructed, which includes a regional priority ranking and a fresh air balance negotiation algorithm. The regional priority ranking includes calculating the priority ranking value of each region based on the potential field values ​​in the environmental potential field map. The expression for calculating the priority ranking value is as follows: ; in, For the region The priority ranking value, For the region The potential field value, For the region The potential field gradient in space, and For the weighting coefficients, satisfying The weighting coefficients are obtained using the entropy weighting method, and the processing logic of the entropy weighting method is as follows: Extract the potential field values ​​and gradients of all regions, and construct... A potential field value matrix is ​​obtained, wherein the first column of the potential field value matrix is ​​the potential field value, and the second column is the potential field gradient. The potential field value matrix is ​​then standardized using z-score processing. The information entropy of the potential field value and the potential field gradient are calculated separately. Based on the information entropy, a difference coefficient is calculated, and then the difference coefficient is normalized to obtain the weighting coefficients. and .

[0010] Preferably, the processing logic of the fresh air balance negotiation algorithm is as follows: The initial fresh air demand is calculated based on the potential field value and priority ranking value of each region. The expression for calculating the initial fresh air demand is as follows: ; in, For the region The initial fresh air demand, Basic air volume, As a demand gain coefficient, the initial fresh air demand of each region is sent to the central coordinator for aggregation. The total fresh air demand is calculated. If the total fresh air demand is less than the system capacity, the initial fresh air demand is retained and the fresh air delivery command is executed. If the total fresh air demand is greater than the system capacity, the central coordinator initiates a demand adjustment mechanism. The processing logic of the demand adjustment mechanism is as follows: Set a regional adjustment coefficient, which includes a first adjustment coefficient and a second adjustment coefficient. Calculate the average priority ranking value of all regions. If the priority ranking value of the current region is higher than the average value, assign the first adjustment coefficient to the current region. Multiply the first adjustment coefficient by the initial fresh air demand to obtain the first air volume allocation value. If the priority ranking value of the current area is less than the average value, assign a second adjustment coefficient to the current area, multiply the second adjustment coefficient by the initial fresh air demand to obtain the second air volume allocation value, and save the first air volume allocation value and the second air volume allocation value as the air volume allocation value.

[0011] Preferably, the environmental parameter optimization and instruction generation unit includes: Based on the potential field value and wind volume distribution value, the expected environmental parameter vector for each region is calculated. The expression for calculating the expected environmental parameter vector is as follows: ; in, For the region At any moment The expected environmental parameter vector, To preset the comfort vector, The spatial gradient of the potential field value. The time integral of the potential field value The rate of change of the potential field value. , and For adaptive control coefficients, the desired environmental parameter vector is converted into desired environmental parameters and stored. A high-order continuous fit is performed on the desired environmental parameter vector using a fifth-order Bézier curve. The optimal control time point is calculated through an optimization algorithm, and the optimal control time point and the desired environmental parameter vector are saved in the fresh air coordination instruction set. The fresh air coordination instruction set includes area number, air volume allocation value, and desired environmental parameters.

[0012] Preferably, the execution decision module includes: The system obtains the current fresh air coordination instruction set, constructs a comprehensive objective function, dynamically adjusts the air volume allocation value of the fresh air coordination instruction set, converts the optimized air volume allocation value into device-level control instructions, and sends the control instructions to the corresponding building equipment actuators to complete the control of the fresh air system.

[0013] Preferably, the processing logic for constructing a comprehensive objective function and dynamically adjusting the air volume allocation value of the fresh air coordination control instruction set is as follows: Based on the fan law, energy consumption costs are calculated according to the air volume allocation value, fan power and real-time electricity price. The comfort deviation is obtained by calculating the difference between the current building environment data and the expected environmental parameters. The comfort deviation includes temperature deviation, humidity deviation, carbon dioxide concentration deviation and PM2.5 deviation. The comfort deviation and energy consumption cost are used as sub-objectives of the comprehensive objective function and calculated in a weighted manner to obtain equipment-level control commands.

[0014] Preferably, the device-level control instructions include optimized airflow allocation values, control instructions, and auxiliary information values; The optimized air volume allocation includes the optimal air volume setting value for each area, the total air volume requirement, and the optimization convergence status. The control commands include fan speed setpoint, valve opening commands for each zone, and cooling / heating power requirements. The auxiliary information values ​​include the estimated energy consumption and operating cost, and the optimization solution status, which includes success, failure, and timeout.

[0015] The beneficial effects of this invention: This invention achieves significant benefits in many aspects by constructing an intelligent building fresh air linkage control system. Firstly, the system uses an IoT sensor network deployed in various zones of the building to collect real-time and accurate environmental data such as temperature, humidity, carbon dioxide concentration, and PM2.5 concentration. It then combines infrared sensors and binocular vision cameras to acquire the number, distribution, and movement trajectories of people, forming a complete building state vector. This provides a solid data foundation for system decision-making. The system employs time-series variation amplitude analysis to calculate the rate of change, acceleration of change, and cumulative deviation of environmental parameters. Furthermore, a dynamic weighting strategy adaptively adjusts the weight coefficients of each indicator according to different time periods, such as weekday commuting, thereby generating an accurate reflection of regional emergency situations. Based on the information priority scoring of degree, this method can intelligently identify sudden environmental changes and continuous comfort issues. At the collaborative control level, the system innovatively introduces the artificial potential field method into building environment modeling, constructing a building environment potential field map that includes gravitational and repulsive potential fields. The graph structure intuitively displays the environmental state of each area, and the dynamic propagation and aggregation of potential field values ​​are realized based on the spatiotemporal graph attention network. On this basis, the system generates the expected environmental parameter trajectory of each area through a multi-area collaborative decision-making mechanism, combined with priority ranking and distributed negotiation algorithms, and uses a fifth-order Bézier curve for high-order smooth fitting to ensure the continuity and stability of control commands. This effectively solves the problem of multi-area resource competition and realizes the optimized allocation of fresh air resources. Attached Figure Description

[0016] Figure 1 This is a basic flowchart of a building fresh air linkage control system based on the Internet of Things, provided as an embodiment of the present invention. Detailed Implementation

[0017] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0018] Example, refer to Figure 1 This paper presents an IoT-based building ventilation linkage control system, including a sensing and acquisition module, a collaborative learning module, and an optimization execution module. The sensing and acquisition module is used to collect building status data, analyze the building status data according to the priority strategy, and perform time-series change amplitude analysis on building space in different areas to generate regional information priority. Based on the regional information priority, the module focuses on key change areas to obtain a building environment potential field map. The collaborative control module is used to make multi-region collaborative decisions on the potential field map of the building environment, obtain the desired environmental parameter sequence, and adjust the high-order smoothing of the Bézier curve through the optimization algorithm to obtain the fresh air coordination instruction set. The execution decision module is used to perform multi-objective optimization of the fresh air coordination instruction set, generate device-level control instructions, and send the control instructions to the building equipment actuators to complete the precise control of the fresh air system.

[0019] In this embodiment, a high-precision environmental perception system is constructed using IoT sensor networks and spatiotemporal attention mechanisms. This solves the problem of one-sided and lagging environmental state perception in traditional building ventilation systems, facilitating a leap from single-point monitoring to full-area situational awareness. Through multi-agent collaborative decision-making and Bezier curve smoothing optimization technology, the system analyzes the poor coordination of multi-area ventilation distribution and frequent equipment oscillations, which helps to achieve dynamic balanced resource allocation and stable system operation. By adopting a dual-objective optimization model of power saving and comfort deviation, equipment-level control commands are generated, resolving the contradiction between energy efficiency and comfort. This helps to significantly reduce system energy consumption while ensuring indoor environmental quality. Finally, through a closed-loop intelligent control architecture of perception-decision-execution, a fundamental transformation of building ventilation systems from passive response to active optimization is achieved, providing an innovative solution for building a green, low-carbon, smart, and comfortable building environment.

[0020] The sensing and acquisition module includes a data acquisition unit, a priority evaluation unit, and an environmental potential field diagram unit; The data acquisition unit includes: IoT sensors are deployed in various areas of the building to collect temperature, humidity, carbon dioxide concentration and PM2.5 concentration in each area in real time as building environment data. Infrared sensors and binocular vision people counting cameras are used to collect the number of people, their distribution and movement trajectory as people movement data. By using sensors in each area as network nodes, the collected building environment data and personnel movement data are classified and defined according to the time of collection to generate building status vectors. IoT sensors include high-precision carbon dioxide sensors, temperature, humidity and pressure combined sensors, and laser PM2.5 sensors; Building state vectors include building environment data, personnel movement data, and time context; The time context includes time period labels and date types; The time slots are labeled as morning, lunch break, and after get off work. Date types include weekdays and rest days.

[0021] A heterogeneous sensor network is deployed across various areas of the building, including workstations, meeting rooms, and corridors. Binocular vision cameras are deployed in meeting rooms to count the number of people entering and exiting, map their movement trajectories, and determine whether someone is entering or leaving. Sensors measuring carbon dioxide concentration, temperature, humidity, and PM2.5 are located on the central ceiling of the meeting room to monitor the environmental quality of the core area. Each sensor group in each area is treated as a smart node, and data is aggregated to an edge gateway via an IoT protocol. The gateway assigns a unified timestamp to all data and associates it with spatial location information. All data at a specific time t and a specific zone i are integrated into a structured building state vector: Building State Vector = {Environmental Data: [Temperature, Humidity, ...}} [PM2.5], Personnel data: [Real-time number of people, personnel distribution heat map, movement speed], Time context: [Time period label, date type]}, The accurate data from the data acquisition unit avoids malfunctions caused by sensor errors, thereby improving comfort and energy efficiency. The generated state vector serves as a unified and standardized system state description, which is beneficial for subsequent state control.

[0022] The priority evaluation unit is used to calculate the information priority score of each area based on the generated building status vector through time-series change magnitude analysis and dynamic weighting strategy; The time-series variation magnitude analysis includes calculating the rate of change, acceleration of change, and cumulative deviation of building environment data and personnel movement data; The rate of change is the difference between the building state vector of the current sampling period and the previous sampling period, used to reflect the instantaneous rate of change of the parameters; The rate of change of acceleration is calculated using second-order difference to predict the risk of building deterioration. The cumulative deviation is the persistent deviation between the current building state vector and the preset comfort vector, used to describe the long-term discomfort level of the area; The dynamic weighting strategy includes: The initial time point is set at 6:00 AM, which represents the moment when the first weight is 0 and the second weight is 100%. After 1 hour, the first weight increases by 1 / 12 and the second weight decreases by 1 / 12. After 12 hours, the first weight decreases by 1 / 12 and the second weight increases by 1 / 12. The first and second weights change dynamically in a 24-hour cycle. The first weight is assigned a cumulative deviation, and the second weight is divided equally and assigned a rate of change and an acceleration of change, respectively. By normalizing and weighting the rate of change, acceleration of change, and cumulative deviation using a dynamic weighting strategy, an information priority score is obtained.

[0023] The rate of change is obtained by calculating the Euclidean distance difference between the building state vector of the current sampling period and the previous sampling period. The rate of change is used to capture instantaneous changes in the environment, such as a sudden surge in carbon dioxide concentration caused by a sudden influx of people, to ensure timely response to emergencies. Based on the concept of differential equations, the instantaneous rate of change is approximated by the difference between different boundaries, which meets the real-time requirements of the control system. For example, when the carbon dioxide concentration in the conference room rises from 600ppm to 1200ppm in 5 minutes, the rate of change is 120ppm / min, which triggers a high priority. The cumulative deviation is obtained by calculating the persistent deviation between the current state and the comfort target. The integration time window T is set to 10 minutes. The cumulative deviation is used to identify long-term environmental problems. A weighted adaptive mechanism based on time period and date type is constructed, which can automatically adjust the working mode according to the time pattern of building use, so that the system behavior is more in line with actual needs.

[0024] The environmental potential field map unit is used to focus on key change areas based on the priority score of regional information to obtain the building environmental potential field map; Building a building area graph structure based on spatiotemporal attention network. Where node V represents each area of ​​the building, and edge This represents the spatial adjacency relationship between different regions. Each node stores a potential field value, which is generated by superimposing the basic environmental potential field and the priority adjustment potential field. The basic environmental potential field of each region is calculated by adding the recovered potential field value and the disturbance potential field value. The expression for calculating the recovered potential field value is as follows: ; in, To recover the potential field gain coefficient, For the current area Building environment data, The building environment data is used as a preset comfort vector. To restore the potential field value; ; in, The value of the disturbance potential field. The gain coefficient of the interference potential field. The sources of interference include densely populated areas and equipment heat sources. For the region With interference source Differences in building environment data between them The range of influence of the interference potential field; Each basic environmental potential value is multiplied by the information priority score of the current region to obtain the adjusted potential value. Based on the interaction between adjacent regions, the potential field is propagated through the building area graph structure. A graph convolutional network is then used to aggregate the potential values ​​of adjacent regions and update the potential value of each region. The potential field value of each region is assigned to the corresponding node to generate a building environment potential field map.

[0025] The building area graph structure is constructed, which abstracts the building space into a graph structure model. Nodes represent independent control areas and edges represent spatial adjacency relationships. The purpose is to establish a computer-processable topological framework, provide a data structure foundation for subsequent potential field calculations, realize the mapping from physical space to digital space, establish a unified expression paradigm for multi-area collaborative control, and describe complex spatial relationships through the definition of nodes and edges. The recovery potential field and the disturbance potential field are calculated separately. The recovery potential field uses a quadratic form function to represent the difference between the current state and the ideal state, while the disturbance potential field uses a distance function to describe the influence intensity of the disturbance source, quantifying the degree of comfort deviation of the environmental state and the influence of the disturbance factors. From a top-level design perspective, a mathematical model of the environmental situation is realized, transforming the abstract concept of comfort into a concrete numerical indicator. Its principle draws on the potential field theory in physics, describing the energy state of the system through mathematical functions. The potential field map is dynamically updated according to the sensor data update cycle to reflect the latest status of the building environment. The building environment potential field map constructed by the above method not only includes the environmental status information of each area, but also highlights key change areas through priority scoring, and considers the mutual influence between areas through potential field propagation. This provides an intuitive and comprehensive situational awareness foundation for the subsequent collaborative control module. In the building environment, we can use the negative gradient of the potential field as a control command, such as adjusting the air supply volume and temperature setpoint. In the collaborative control module, the potential field map will be used to generate corresponding control strategies.

[0026] The collaborative control module includes a multi-region collaborative decision-making unit and an instruction generation unit; Multi-regional collaborative decision-making units include: Based on the potential field values ​​corresponding to each node in the building environment potential field diagram, a multi-region collaborative decision-making framework is constructed, which includes a regional priority ranking and fresh air balance negotiation algorithm. Regional priority ranking involves calculating the priority value for each region based on the potential field values ​​in the environmental potential field diagram. The expression for calculating the priority value is as follows: ; in, For the region The priority ranking value, For the region The potential field value, For the region The potential field gradient in space, and For the weighting coefficients, satisfying The weighting coefficients are obtained using the entropy weighting method, and the processing logic of the entropy weighting method is as follows: Extract the potential field values ​​and gradients of all regions, and construct... The potential field value matrix is ​​obtained, with the first column representing the potential field values ​​and the second column representing the potential field gradient. The matrix is ​​then standardized using z-scores. The information entropy of the potential field values ​​and gradient is calculated separately. Based on the information entropy, the difference coefficient is calculated and then normalized to obtain the weighting coefficients. and .

[0027] The regional priority ranking method is based on node data in the environmental potential field graph. Priority values ​​are obtained through weighted calculations. The purpose of regional priority ranking is to integrate the potential field value and its spatial gradient—two key indicators—into a single evaluation value. The potential field value reflects the urgency of the current environmental state, while the spatial gradient reflects the diffusion trend of environmental problems. Combining these two factors allows for a more comprehensive assessment of regional importance. The entropy weight method aims to objectively determine the relative importance of each feature, avoiding biases caused by subjective assignment. The fresh air balance negotiation algorithm, based on the priority ranking value, uses a distributed optimization method to calculate the air volume allocation value for each region. This process is achieved through iterative optimization. The objective function comprehensively considers the total demand deviation and inter-regional differences. Constraints include the system's total air volume limit and regional air volume range. First, each region calculates its initial fresh air demand based on its local potential field value and priority ranking value. The central coordinator aggregates the demand and initiates an adjustment mechanism when limits are exceeded, achieving resource tilting by introducing a regional adjustment coefficient. Finally, adjacent regions perform local coordination to ensure smooth air volume allocation. This method achieves a complete closed loop from environmental perception to resource allocation, ensuring rapid response to urgent needs and improving the overall efficiency of system operation.

[0028] The processing logic of the fresh air balance negotiation algorithm is as follows: The initial fresh air demand is calculated based on the potential field value and priority ranking value of each region. The expression for calculating the initial fresh air demand is as follows: ; in, For the region The initial fresh air demand, Basic air volume, As the demand gain coefficient, the initial fresh air demand of each region is sent to the central coordinator for aggregation. The total fresh air demand is calculated. If the total fresh air demand is less than the system capacity, the initial fresh air demand is retained and the fresh air delivery command is executed. If the total fresh air demand is greater than the system capacity, the central coordinator activates the demand adjustment mechanism. The processing logic of the demand adjustment mechanism is as follows: Set the area adjustment coefficient, which includes a first adjustment coefficient and a second adjustment coefficient, to determine the area priority. Calculate the average priority ranking value of all areas. If the priority ranking value of the current area is higher than the average, assign the first adjustment coefficient to the current area. Multiply the first adjustment coefficient by the initial fresh air demand to obtain the first air volume allocation value. If the priority ranking value of the current area is less than the average value, assign a second adjustment coefficient to the current area, multiply the second adjustment coefficient by the initial fresh air demand to obtain the second air volume allocation value, and save the first air volume allocation value and the second air volume allocation value as the air volume allocation value.

[0029] The first adjustment factor is set to be greater than 1, and the second adjustment factor is set to be less than 1. Assume an office building has four areas: A, B, C, and D. The system parameters are set as follows: base airflow of 100 m³ / h, demand gain factor of 0.5, first adjustment factor of 1.2, second adjustment factor of 0.8, and total system capacity limited to 400 m³ / h. The initial demand table for building potential field values ​​is as follows: ; The initial demand is calculated to be 1235 m³ / h. The average value of the priority ranking is 4.175. Therefore, the potential field value of region A should be adjusted to 7.2 > 4.175, and the air volume allocation value should be 460 × 1.2 = 552. Region B: Potential field value = 5.1 > 4.175, air volume distribution value = 355 × 1.2 = 426; Region C: Potential field value = 2.9 < 4.175, air volume distribution value = 245 × 0.8 = 196; Region D: Potential field value = 1.5 < 4.175, air volume distribution value = 175 × 0.8 = 140; After the adjustment, high-priority areas A and B will receive more air volume resources, while the air volume requirements of low-priority areas C and D will be appropriately limited. By calculating the potential field value and priority ranking value in real time, the system can dynamically identify the areas with the most urgent environmental needs, ensuring that limited resources are prioritized for key areas. Secondly, by adopting a hierarchical adjustment mechanism, the system can both guarantee the basic needs of high-priority areas and reasonably limit the resource occupation of low-priority areas, thereby achieving optimal resource allocation. In addition, through the total amount judgment and automatic adjustment mechanism of the central coordinator, the system can flexibly adjust the allocation strategy according to the real-time capacity status to ensure the stable operation of the system.

[0030] The environmental parameter optimization and instruction generation unit includes: Based on the potential field value and wind volume distribution value, the expected environmental parameter vector for each region is calculated. The expression for calculating the expected environmental parameter vector is as follows: ; in, For the region At any moment The expected environmental parameter vector, To preset the comfort vector, The spatial gradient of the potential field value. The time integral of the potential field value The rate of change of the potential field value. , and For adaptive control coefficients, the desired environmental parameter vector is converted into desired environmental parameters and stored. A high-order continuous fit is performed on the desired environmental parameter vector using a fifth-order Bézier curve. The optimal control time point is calculated through an optimization algorithm, and the optimal control time point and the desired environmental parameter vector are saved in the fresh air coordination instruction set. The fresh air coordination instruction set includes area number, air volume allocation value, and desired environmental parameters.

[0031] First, the desired environmental parameter sequence is refined. Based on the dynamic characteristics of the potential field, a multi-dimensional parameter sequence containing the current state, change trend and cumulative effect is constructed. By introducing the spatial influence of the potential field gradient, the historical cumulative effect of the potential field integral and the dynamic prediction of the potential field change rate, a parameter trajectory with time dimension is generated. This process adopts an adaptive control mechanism to automatically adjust the control parameters according to the severity of environmental changes, so as to ensure a balance between rapid response and stable operation. Next, the adjustment curve is smoothed and optimized. The system uses a fifth-order Bézier curve to perform a high-order continuous fit on the generated parameter sequence. The optimal control point is calculated through an optimization algorithm, ensuring that the fitted curve closely follows the original sequence while maintaining the continuity of the first derivative (rate of change) and the second derivative (acceleration). This process effectively eliminates parameter jumps, ensures a smooth transition of environmental parameters such as temperature and humidity, and strictly limits the maximum rate of change within a comfortable range, avoiding frequent equipment adjustments and personnel discomfort caused by sudden parameter changes. Finally, feasibility verification and adjustments were performed under multiple constraints. The system comprehensively addressed three types of key constraints: equipment capacity constraints ensured that the airflow allocated to each area was within the physical limits of the fans and valves; physical space constraints verified pressure balance based on the duct network model to prevent local overpressure or underpressure; and time response constraints considered system thermal inertia and airflow propagation delay, setting reasonable time buffers for parameter changes. Through an iterative correction process, the system ensured that the generated instructions met both environmental regulation requirements and the actual operating limitations of all physical equipment and systems.

[0032] The system ultimately outputs a fresh air coordination instruction set, which includes precise airflow setpoints, target temperature and humidity parameters and their smooth change trajectories for each area, as well as corresponding equipment control sequences, providing a complete solution for precise control at the execution layer.

[0033] The decision-making execution module includes: The system obtains the current fresh air coordination instruction set, constructs a comprehensive objective function, dynamically adjusts the air volume allocation value of the fresh air coordination instruction set, converts the optimized air volume allocation value into device-level control instructions, and sends the control instructions to the corresponding building equipment actuators to complete the control of the fresh air system.

[0034] By constructing a comprehensive objective function and performing dynamic optimization, this module achieves refined energy consumption management. The system can automatically seek the best balance between energy consumption and comfort based on real-time environmental needs and external conditions (such as electricity price signals). For example, during peak electricity price periods, the system will adjust the air volume distribution appropriately to prioritize reducing energy consumption while ensuring basic comfort, thereby effectively saving operating costs. By further optimizing the fresh air coordination command set, accurate tracking of environmental parameters in each area is ensured. It can dynamically respond to disturbances such as changes in personnel and equipment start-up / shutdown, quickly adjusting temperature, humidity, and other parameters. Key parameters such as concentration are maintained within a comfortable range, providing users with a stable, healthy, and comfortable indoor environment, avoiding common problems such as excessive cold, excessive heat, or stuffy air that are common under traditional control methods; Introducing a comprehensive objective function for multi-objective optimization transforms system decision-making from relying on empirical rules to model-based scientific computation. This gives the system greater robustness and adaptability, enabling it to cope with complex and dynamically changing building environments and automatically generate the current optimal control strategy under different operating conditions and constraints.

[0035] The processing logic for dynamically adjusting the airflow allocation values ​​of the fresh air coordination control command set, based on the constructed comprehensive objective function, is as follows: Based on the fan law, energy consumption costs are calculated according to the air volume allocation value, fan power and real-time electricity price. The comfort deviation is obtained by calculating the difference between the current building environment data and the expected environmental parameters. The comfort deviation includes temperature deviation, humidity deviation, carbon dioxide concentration deviation and PM2.5 deviation. The comfort deviation and energy consumption cost are used as sub-objectives of the comprehensive objective function and calculated in a weighted manner to obtain equipment-level control commands.

[0036] By incorporating energy costs and comfort deviations into the objective function and performing weighted optimization, the system can achieve refined energy management while ensuring environmental quality. During peak electricity consumption periods, the system can automatically and appropriately adjust operating parameters to significantly reduce energy consumption while maintaining basic comfort, achieving the optimal balance between energy efficiency and comfort; comprehensively considering temperature, humidity, and other factors... Environmental parameters such as concentration and PM2.5 are managed through a unified comfort deviation calculation framework to ensure that all environmental dimensions can be effectively controlled.

[0037] Equipment-level control commands include optimized airflow allocation values, control commands, and auxiliary information values; Optimized air volume allocation includes the optimal air volume setpoint for each area, total air volume requirement, and optimization convergence status. Control commands include fan speed setpoint, valve opening commands for each zone, and cooling / heating power requirements; The auxiliary information values ​​include the estimated energy consumption and operating costs, as well as the optimization solution status, which includes success, failure, and timeout.

[0038] By outputting the optimal air volume setpoint and total air volume demand for each area, the system achieves a complete closed loop from macro-level total control to micro-level regional allocation, ensuring that air volume resources can be accurately allocated to the areas that need them most. At the same time, the coordinated output of the fan speed setpoint and the opening command of the air valve in each area avoids the equipment action conflicts and resource consumption common in traditional systems, significantly improving the overall efficiency of the system.

[0039] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0040] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A building fresh air linkage control system based on the Internet of Things, characterized in that, It includes a sensing and acquisition module, a collaborative control module, and an execution and decision-making module: The sensing and acquisition module is used to collect building status data, analyze the building status data according to a priority strategy, analyze the temporal change amplitude of building space in different areas, generate regional information priority scores, and mark key change areas according to the regional information priority scores to obtain a building environment potential field map. The collaborative control module is used to make multi-region collaborative decisions on the building environment potential field map, obtain the desired environmental parameter sequence, and adjust the high-order smoothing of the Bézier curve through an optimization algorithm to obtain the fresh air coordination instruction set. The execution decision module is used to perform multi-objective optimization of the fresh air coordination instruction set to obtain device-level control instructions, and then send the control instructions to the building equipment actuators to complete the precise control of the fresh air system.

2. The building fresh air linkage control system based on the Internet of Things as described in claim 1, characterized in that, The sensing and acquisition module includes a data acquisition unit, a priority evaluation unit, and an environmental potential field diagram unit. The data acquisition unit includes: IoT sensors are deployed in various areas of the building to collect temperature, humidity, carbon dioxide concentration and PM2.5 concentration in each area in real time as building environment data. Infrared sensors and binocular vision people counting cameras are used to collect the number of people, their distribution and movement trajectory as people movement data. By using sensors in each area as network nodes, the collected building environment data and personnel movement data are classified and defined according to the time of collection to generate building status vectors. The IoT sensors include a high-precision carbon dioxide sensor, a temperature, humidity, and pressure combined sensor, and a laser PM2.5 sensor; The building state vector includes building environment data, personnel movement data, and time context; The time context includes time period labels and date types; The time period labels are morning, lunch break, and after get off work. The date types include weekdays and rest days.

3. The building fresh air linkage control system based on the Internet of Things as described in claim 2, characterized in that, The priority evaluation unit is used to calculate the information priority score of each area based on the generated building status vector through time-series change amplitude analysis and dynamic weighting strategy. The time-series variation amplitude analysis includes calculating the rate of change, acceleration of change, and cumulative deviation of building environment data and personnel movement data; The rate of change is the difference between the building state vector of the current sampling period and the previous sampling period, used to reflect the instantaneous rate of change of the parameters. The change in acceleration is calculated using second-order difference to determine the trend of the rate of change relative to time, which is used to predict the risk of building deterioration. The cumulative deviation is the persistent deviation between the current building state vector and the preset comfort vector, used to describe the long-term discomfort level of the area; The dynamic weighting strategy includes: The initial time point is set at 6:00 AM, which represents the moment when the first weight is 0 and the second weight is 100%. After 1 hour, the first weight increases by 1 / 12 and the second weight decreases by 1 / 12. After 12 hours, the first weight decreases by 1 / 12 and the second weight increases by 1 / 12. The first and second weights change dynamically in a 24-hour cycle. The first weight is assigned a cumulative deviation, and the second weight is divided equally and assigned a rate of change and an acceleration of change, respectively. By normalizing and weighting the rate of change, acceleration of change, and cumulative deviation using a dynamic weighting strategy, an information priority score is obtained.

4. The building fresh air linkage control system based on the Internet of Things as described in claim 3, characterized in that, The environmental potential field map unit is used to mark key change areas according to the priority score of regional information to obtain the building environmental potential field map; Building a building area graph structure based on spatiotemporal attention network. Where node V represents the various areas of the building, and edge This represents the spatial adjacency relationship between different regions. Each node stores a potential field value, which is generated by superimposing the basic environmental potential field and the priority adjustment potential field. The basic environmental potential field of each region is calculated by adding the recovered potential field value and the disturbance potential field value. The expression for calculating the recovered potential field value is as follows: ； in, To recover the potential field gain coefficient, For the current area Building environment data, The building environment data is used as a preset comfort vector. To restore the potential field value; ； in, The value of the disturbance potential field. The gain coefficient of the interference potential field. The sources of interference include densely populated areas and equipment heat sources. For the region With interference source Differences in building environment data between them The range of influence of the interference potential field; Each basic environmental potential value is multiplied by the information priority score of the current region to obtain the adjusted potential value. Based on the interaction between adjacent regions, the potential field is propagated through the building area graph structure. A graph convolutional network is then used to aggregate the potential values ​​of adjacent regions and update the potential value of each region. The potential field value of each region is assigned to the corresponding node to generate a building environment potential field map.

5. The building fresh air linkage control system based on the Internet of Things as described in claim 4, characterized in that, The collaborative control module includes a multi-regional collaborative decision-making unit and an instruction generation unit; The multi-regional collaborative decision-making unit includes: Based on the potential field values ​​corresponding to each node in the building environment potential field diagram, a multi-regional collaborative decision-making framework is constructed, which includes a regional priority ranking and a fresh air balance negotiation algorithm. The regional priority ranking includes calculating the priority ranking value of each region based on the potential field values ​​in the environmental potential field map. The expression for calculating the priority ranking value is as follows: ； in, For the region The priority ranking value, For the region The potential field value, For the region The potential field gradient in space, and For the weighting coefficients, satisfying The weighting coefficients are obtained using the entropy weighting method, and the processing logic of the entropy weighting method is as follows: Extract the potential field values ​​and gradients of all regions, and construct... A potential field value matrix is ​​obtained, wherein the first column of the potential field value matrix is ​​the potential field value, and the second column is the potential field gradient. The potential field value matrix is ​​then standardized using z-score processing. The information entropy of the potential field value and the potential field gradient are calculated separately. Based on the information entropy, a difference coefficient is calculated, and then the difference coefficient is normalized to obtain the weighting coefficients. and .

6. The building fresh air linkage control system based on the Internet of Things as described in claim 5, characterized in that, The processing logic of the fresh air balance negotiation algorithm is as follows: The initial fresh air demand is calculated based on the potential field value and priority ranking value of each region. The expression for calculating the initial fresh air demand is as follows: ； in, For the region The initial fresh air demand, Basic air volume, As a demand gain coefficient, the initial fresh air demand of each region is sent to the central coordinator for aggregation. The total fresh air demand is calculated. If the total fresh air demand is less than the system capacity, the initial fresh air demand is retained and the fresh air delivery command is executed. If the total fresh air demand is greater than the system capacity, the central coordinator initiates a demand adjustment mechanism. The processing logic of the demand adjustment mechanism is as follows: Set a regional adjustment coefficient, which includes a first adjustment coefficient and a second adjustment coefficient. Calculate the average priority ranking value of all regions. If the priority ranking value of the current region is higher than the average value, assign the first adjustment coefficient to the current region. Multiply the first adjustment coefficient by the initial fresh air demand to obtain the first air volume allocation value. If the priority ranking value of the current area is less than the average value, assign a second adjustment coefficient to the current area, multiply the second adjustment coefficient by the initial fresh air demand to obtain the second air volume allocation value, and save the first air volume allocation value and the second air volume allocation value as the air volume allocation value.

7. A building fresh air linkage control system based on the Internet of Things as described in claim 6, characterized in that, The environmental parameter optimization and instruction generation unit includes: Based on the potential field value and wind volume distribution value, the expected environmental parameter vector for each region is calculated. The expression for calculating the expected environmental parameter vector is as follows: ； in, For the region At any moment The expected environmental parameter vector, To preset the comfort vector, The spatial gradient of the potential field value. The time integral of the potential field value The rate of change of the potential field value. , and For adaptive control coefficients, the desired environmental parameter vector is converted into desired environmental parameters and stored. A high-order continuous fit is performed on the desired environmental parameter vector using a fifth-order Bézier curve. The optimal control time point is calculated through an optimization algorithm, and the optimal control time point and the desired environmental parameter vector are saved in the fresh air coordination instruction set. The fresh air coordination instruction set includes area number, air volume allocation value, and desired environmental parameters.

8. A building fresh air linkage control system based on the Internet of Things as described in claim 7, characterized in that, The execution decision module includes: The system obtains the current fresh air coordination instruction set, constructs a comprehensive objective function, dynamically adjusts the air volume allocation value of the fresh air coordination instruction set, converts the optimized air volume allocation value into device-level control instructions, and sends the control instructions to the corresponding building equipment actuators to complete the control of the fresh air system.

9. A building fresh air linkage control system based on the Internet of Things as described in claim 8, characterized in that, The processing logic for dynamically adjusting the airflow allocation values ​​of the fresh air coordination control command set, based on the constructed comprehensive objective function, is as follows: Based on the fan law, energy consumption costs are calculated according to the air volume allocation value, fan power and real-time electricity price. The comfort deviation is obtained by calculating the difference between the current building environment data and the expected environmental parameters. The comfort deviation includes temperature deviation, humidity deviation, carbon dioxide concentration deviation and PM2.5 deviation. The comfort deviation and energy consumption cost are used as sub-objectives of the comprehensive objective function and calculated in a weighted manner to obtain equipment-level control commands.

10. A building fresh air linkage control system based on the Internet of Things as described in claim 9, characterized in that, The device-level control commands include optimized airflow distribution, control commands, and auxiliary information values; The optimized air volume allocation includes the optimal air volume setting value for each area, the total air volume requirement, and the optimization convergence status. The control commands include fan speed setpoint, valve opening commands for each zone, and cooling / heating power requirements. The auxiliary information values ​​include the estimated energy consumption and operating cost, and the optimization solution status, which includes success, failure, and timeout.

Images