Air conditioner door and window joint control method and system

Abstract

The invention relates to the technical field of ventilation regulation and control, in particular to an air conditioner door and window joint control method and system, and the method comprises the steps: information collection: constructing a multi-source data collection layer, and obtaining the following data in real time: meteorological early warning data, user operation logs, a regional climate feature library, and a user historical use region; behavior feature extraction: processing user operation logs and thermodynamic diagram data by adopting a space-time convolutional neural network to generate a three-dimensional demand map; dynamic regulation and control calculation: extracting an external window SHGC value and a wall heat conductivity coefficient based on a building BIM model, and calculating an optimal opening threshold value of a door and a window in combination with real-time meteorological data; performing cooperative control: adopting a multi-target particle swarm optimization algorithm, and generating a control instruction set under the targets of energy consumption and comfort; and an instruction is issued to the air conditioner and the electric door and window actuator. The system comprises an information acquisition module, a behavior characteristic acquisition module, a dynamic regulation and control calculation module and a cooperative control execution module. The method has the effect of reducing energy waste.

Description

Technical Field

[0001] The present application relates to the technical field of ventilation control, and in particular to a method and system for controlling air-conditioning doors and windows. Background Art

[0002] Currently, traditional air conditioning and door and window control systems typically operate independently, lacking a coordinated mechanism, resulting in inefficient energy utilization. Users must separately operate the air conditioner and open and close doors and windows. This prevents rapid response, especially during sudden environmental changes (such as rainfall, strong winds, or temperature fluctuations), which can easily lead to temperature control failure or equipment overload. Existing technologies often leave doors and windows open while air conditioning is running, causing airflow to leak out, significantly increasing building energy consumption and making it difficult to meet the demands of intelligent scenarios.

[0003] In the prior art, the air conditioning temperature is usually adjusted based on the user's temperature habits. However, the temperature adjustment for different rooms also depends on the habitual temperature, resulting in maintaining normal temperature adjustment in unoccupied rooms, causing energy waste. Summary of the Invention

[0004] In order to reduce energy waste, the present application provides an air-conditioning door and window joint control method and system.

[0005] In the first aspect, the present application provides an air-conditioning door and window joint control method and system, which adopts the following technical solutions: A method for controlling doors and windows of an air conditioner comprises the following steps: Information collection: Build a multi-source data collection layer to obtain the following data in real time: Meteorological warning data: obtain the temperature change rate, UV index and rainfall intensity in the next 3 hours; User operation log: records the air conditioner set temperature adjustment records and the door and window opening and closing time series; Regional climate characteristics database: including Köppen climate classification codes and local air conditioning energy-saving management regulations; User historical usage area: collect historical distribution map of users; Behavioral feature extraction: Spatiotemporal convolutional neural networks are used to process user operation logs and heat map data to generate a three-dimensional demand map that includes time preferences, spatial concentration, and temperature control sensitivity. Dynamic control calculation: Based on the building BIM model, the SHGC value of external windows and the thermal conductivity of the wall are extracted, and the optimal opening threshold of doors and windows is calculated in combination with real-time meteorological data. When the main living area is detected, the opening of this area is calculated first; Collaborative control execution: Using a multi-objective particle swarm optimization algorithm, a control instruction set is generated under the three objectives of energy consumption, comfort, and policy compliance; Send commands to the air conditioner and electric door and window actuators.

[0006] By adopting the above technical solutions, a dynamic control model is constructed by integrating multi-dimensional data such as meteorological warnings, user behavior, and building thermal characteristics, and through multi-dimensional big data acquisition; a spatiotemporal convolutional neural network analyzes the spatiotemporal patterns of user activities, and combines building BIM parameters (such as the SHGC value of external windows) and real-time weather to dynamically calculate the optimal door and window openings in different areas; a multi-objective optimization algorithm balances energy consumption, comfort, and policy compliance to generate globally optimal instructions.

[0007] Traditional buildings rely on whole-house ventilation and air conditioning for temperature control, resulting in energy waste. Traditional methods also rely on single-sensor data (such as temperature thresholds) and lack comprehensive analysis of user behavior patterns and building thermal characteristics, leading to rigid strategies. This solution significantly enhances environmental perception through multi-source data fusion and deep learning, making control strategies more tailored to real-world scenarios. It optimizes opening thresholds based on building thermal parameters (such as wall thermal conductivity) to reduce ineffective heat exchange. Experimental data shows improved air conditioning energy savings in summer. A multi-objective optimization algorithm responds to meteorological changes (such as sudden changes in rainfall intensity) in real time, shortening delays in strategy adjustments. Furthermore, through optimized combined control of air conditioning, doors, and windows, it reduces internal building energy consumption, improves energy utilization, and achieves energy-saving effects.

[0008] Optionally, the dynamic control calculation step also includes dividing the main living area into a first main living area and a second main living area based on historical area usage; and controlling the temperature of the second main living area so that the temperature difference between the second main living area and the historical habitual temperature is a before people are active for a long time.

[0009] By adopting the above technical solution, the system divides the area into primary and secondary areas according to the historical frequency of use. The second area maintains the difference A from the historical habitual temperature during the inactive period. The existing technology usually adopts a fixed temperature offset or unified control of the whole house, and cannot perform differentiated pre-adjustment for the secondary area. The secondary area only maintains the basic temperature difference, which reduces the invalid energy supply time compared with the constant temperature mode of the whole house.

[0010] Optionally, in the dynamic control calculation step, the temperature control of the second main living area further includes adjusting the second temperature of the second main living area when the first main living area is used according to the usage time ratio of the second main living area to the usage time of the first main living area, so that the difference between the second temperature and the historical habitual temperature is .

[0011] By adopting the above technical solution, the dynamic temperature difference is adaptively adjusted according to the usage ratio. The time it takes for the temperature to reach the standard when the user enters the second area is shortened, and it can be dynamically maintained according to actual usage conditions, further reducing energy waste and improving applicability.

[0012] Optionally, the dynamic control calculation step further includes a pre-adjustment strategy based on a user behavior prediction model, specifically: Analyze the historical usage time series of the second main living area through the LSTM network and predict the probability of users entering this area in the next 2 hours ; when When the temperature is pre-adjusted: Calculate the pre-adjustment time advance ,in is the regional thermal inertia time constant; Adjust the temperature of the second zone to the historical customary temperature at a rate of 0.4°C / min; Close doors and windows in adjacent inactive areas in a coordinated manner to reduce heat exchange.

[0013] By adopting this technical solution, traditional pre-conditioning relies on a fixed schedule and cannot adapt to changes in thermal inertia (such as fluctuations in heat loss caused by frequent opening and closing of doors and windows). LSTM prediction combines thermal inertia time parameters to implement pre-conditioning. Setting a probability threshold of 0.7 achieves a prediction accuracy of 89% (verified by the MIT dataset), and a regulation rate of 0.4°C / min balances comfort and equipment load. The coordinated closing of doors and windows in inactive areas reduces heat exchange losses, improves energy utilization, and reduces energy waste.

[0014] Optionally, in the dynamic control calculation step, a regional thermal inertia learning strategy is also included: the thermal time constant of each region is determined by a step response test : The air conditioner runs at full capacity until the temperature drops by 2°C; Record the time required for the temperature to return to the first set value; Dynamically adjust pre-adjustment time , The calculation model is as follows: ; When the frequency of opening and closing doors and windows is detected to be greater than 5 times per hour, the system will temporarily increase the Valuation 20%.

[0015] By adopting the above technical solution, the building thermal characteristic parameters are learned through step response experiments, Δt is dynamically adjusted to reduce the pre-adjustment time error, and the door and window opening and closing frequency compensation mechanism improves the model robustness. Compared with the traditional fixed-time model, energy consumption is reduced.

[0016] Optionally, the dynamic control calculation step further includes a temporary region priority competition mechanism: When N temporary active areas are detected: Calculate the heat load urgency of each area: ;in, is the current temperature difference, The most recent usage duration; The Hungarian algorithm is used to allocate air conditioning cooling capacity, giving priority to area; Maintain basic air supply volume for other areas and improve the sealing level of doors and windows.

[0017] By adopting the above technical solutions, the Hungarian algorithm is used to allocate resources when multiple regions compete, and the urgency index Quantifying the conflict level improves decision-making efficiency, and the basic air supply + sealing upgrade strategy reduces marginal energy consumption; through multi-level detection in temporary areas, it is convenient to reasonably allocate energy, so as to reduce energy loss while maintaining comfort.

[0018] Optionally, the dynamic control calculation step also includes cross-time strategy migration: Analyze historical data to generate typical scenario templates: Weekday home mode: , ; Holiday party mode: , ; Detection of a sudden change in user behavior pattern: matching historical templates with a similarity > 85%; Dynamically modify the current control strategy according to template parameters; If there is no matching template within 24 hours, strategy generation based on reinforcement learning is triggered.

[0019] By adopting the above technical solutions, the existing multi-zone control mostly adopts polling or fixed priority, which easily leads to temperature control lag in the core area; in the multi-temporary zone scenario, the urgency of each zone is calculated. (Temperature difference and the weighted value of the most recent usage duration), giving priority to >1.5. Cross-period policy migration quickly adapts to new scenarios by matching historical templates (such as holiday gathering patterns). If no matching template exists, reinforcement learning is initiated to generate a new policy. Resource allocation efficiency: The Hungarian algorithm improves cooling capacity allocation efficiency and reduces temperature control delays in core areas. The policy migration module shortens the adaptation time to new scenarios, improving system responsiveness and reducing energy waste.

[0020] Optionally, the dynamic control calculation step further includes dynamic energy efficiency threshold adjustment: Dynamically correct target temperature difference based on real-time meteorological data and regional activity intensity Allowable range: When the outdoor temperature is greater than 30℃, set The upper limit is 1.2℃; when the temporary area is detected to be used more than 3 times / hour, the The lower limit is 0.3℃; Combined with the air conditioner operating efficiency curve, optimize the compressor frequency distribution.

[0021] By adopting the above technical solution, dynamic energy efficiency threshold adjustment ( The compressor frequency is allocated according to the weight. When the outdoor temperature is greater than 30℃, the frequency is relaxed. The upper limit is 1.2℃, allowing a larger temperature difference to reduce energy consumption; when the temporary area is frequently used (>3 times / hour), tighten The lower limit is 0.3℃ to avoid energy consumption peaks caused by frequent adjustments. The compressor power is based on the regional cooling demand weight ( × area); traditional energy efficiency thresholds are fixed, which can easily lead to compressor overload or temperature control failure in extreme weather; this solution reduces energy efficiency fluctuations in hot weather and extends compressor life; compressor power is allocated on demand, reducing total energy consumption; and combined with external weather temperatures, it can maintain a suitable indoor temperature and can quickly cool down when people enter, improving living comfort.

[0022] Optionally, the dynamic control calculation step also includes a personnel distribution adaptive temperature control step: Construct a sub-meter personnel distribution matrix: Divide the area into 0.5m×0.5m grids and mark the coordinates and postures of personnel; Calculate thermal comfort impact factors: , where d is the distance between the person and the air outlet, and h is the person's sitting height; Dynamically adjust air supply parameters: determine whether the thermal comfort impact factor is greater than a first threshold, and if so, execute a first intervention step; determine whether the thermal comfort impact factor is less than a second threshold, and if so, execute a second intervention step; First intervention: Increase local air supply and lower the temperature setting; use smart curtains to adjust shading direction to reduce solar radiation interference; Second intervention: reduce the fan speed to and increase the set temperature; close the corresponding door and window air outlets to reduce heat exchange; Timing control: Update personnel distribution data every 10 minutes and optimize air supply parameter combination through PID control algorithm.

[0023] By adopting the above technical solution, the personnel distribution grid is constructed in real time through the fusion positioning of millimeter wave radar and infrared thermal imaging. Value (related to the air outlet distance d and the sitting height h of the personnel), Increase the air volume (1.5m / s) and lower the temperature by 0.3℃ in areas where people gather (people gather), and use curtains to block the sun. Reduce the wind speed in the crowded area (edge zone) to 0.6m / s and close doors and windows. Existing air supply strategies mostly use uniform air supply, which leads to overheating in crowded areas and overcooling in edge zones, and the wind blows directly on people, causing discomfort. This solution reduces temperature fluctuations in crowded areas and improves the thermal comfort index. By reducing the air supply intensity in edge zones, energy savings are increased.

[0024] In a second aspect, the present application provides a system that adopts the following technical solutions: A system includes the following modules: Information collection module: used to build a multi-source data collection layer to obtain meteorological warning data, user operation logs, regional climate feature database, and user historical usage areas in real time; Behavior feature extraction module: The input end is connected to the output end of the information collection module to generate a three-dimensional demand map that includes time period preference, spatial concentration, and temperature control sensitivity; Dynamic control calculation module: The input end is connected to the output end of the behavior feature extraction module, and is used to calculate the optimal opening threshold of doors and windows based on real-time meteorological data; Collaborative control execution module: The input end is connected to the output end of the dynamic control calculation module, and is used to generate a control instruction set and send instructions to the air conditioner and electric door and window actuators.

[0025] By adopting the above technical solution, each module works together through a standardized interface: the information acquisition module inputs multi-source data in real time, the behavior feature extraction module generates a three-dimensional demand map, the dynamic control module calculates the optimal strategy, and the collaborative execution module sends instructions to the terminal device; the traditional system has a high degree of functional coupling and poor scalability, making it difficult to integrate new sensors or algorithms; it supports rapid access to new sensors (such as PM2.5 monitoring), shortening the upgrade cycle; and the modular design reduces the impact of single-point failures.

[0026] In summary, this application includes at least one of the following beneficial technical effects: 1. Through multi-source data fusion, the limitations of traditional single-threshold control are overcome, achieving a leap from "passive response" to "active prediction." Coupled analysis of building thermal parameters and user behavior data can identify hidden heat loss paths and guide users in improving building insulation. Sub-meter occupant positioning and thermal comfort factor calculation enable precise air supply, solving the energy waste problem caused by uniform air supply. Long-term recording of occupant distribution heat maps can optimize building space layout. 2. Balancing multiple objectives, including energy consumption, comfort, and policy compliance, to avoid excessively sacrificing other indicators for a single objective (e.g., significantly reducing comfort for energy savings). Combining policy compliance constraints (e.g., air conditioning ≥ 26°C in summer) with user habit learning reduces the frequency of manual user intervention. 3. Mechanisms such as temperature rise anomaly diagnosis and equipment failure downgraded operation ensure that the system maintains basic services during extreme weather or hardware failures. Energy efficiency loss data recorded during failures can be used to guide equipment maintenance priorities and improve operation and maintenance efficiency. DETAILED DESCRIPTION

[0027] The application is described in further detail below.

[0028] This embodiment discloses a method for controlling doors and windows of an air conditioner.

[0029] Example 1: An air-conditioning door and window joint control method, comprising the following steps: Information collection: Build a multi-source data collection layer to obtain the following data in real time: Meteorological warning data: obtain the temperature change rate, UV index and rainfall intensity in the next 3 hours; User operation log: records the air conditioner set temperature adjustment records and the door and window opening and closing time series; Regional climate characteristics database: including Köppen climate classification codes and local air conditioning energy-saving management regulations; User historical usage area: collect historical distribution map of users; Specifically, deploy an IoT sensor network, integrate the meteorological API interface, door and window status sensors, and air conditioning operation record modules. Specifically, obtain the temperature change rate (℃ / h), ultraviolet index (UVI), and rainfall intensity (mm / h) in real time over the next three hours through the Meteorological Bureau API. An operation recording chip is embedded in the air conditioner remote control to record temperature setting value change events by timestamp (for example, setting 26°C at 2023-08-2014:35). The building BIM system exports the solar heat gain coefficient (SHGC) value of the exterior windows and the thermal conductivity coefficient (W / m²·K) of the wall. Bluetooth beacon positioning technology is used to generate a user's historical thermal distribution map.

[0030] Behavioral feature extraction: Spatiotemporal convolutional neural networks are used to process user operation logs and heat map data to generate a three-dimensional demand map that includes time preferences, spatial concentration, and temperature control sensitivity. Specifically, a spatiotemporal convolutional neural network (ST-CNN) model is constructed: Input layer: user operation log matrix (time × temperature setting value) superimposed on heat map (spatial coordinates × residence time); Convolutional layer: uses 3×3×3 three-dimensional convolution kernel to extract spatiotemporal features; Output layer: Generates a three-dimensional demand map (time period preference coefficient, spatial aggregation index, and temperature control sensitivity level); For example, analysis found that the user concentration in the living room from 19:00 to 21:00 every Wednesday reached 0.92 (maximum value 1). During this period, the temperature control sensitivity was increased to Level 3, and the system prioritized ensuring temperature stability in this area.

[0031] The extraction of spatiotemporal features improves the accuracy of behavioral pattern recognition and the matching degree between temperature control instructions and actual user needs.

[0032] Dynamic control calculation: Based on the building BIM model, the SHGC value of external windows and the thermal conductivity of the wall are extracted, and the optimal opening threshold of doors and windows is calculated in combination with real-time meteorological data. When the main living area is detected, the opening of this area is calculated first; Specifically, a building thermodynamics model is built based on the EnergyPlus engine to solve differential equations in real time: ,in is the real-time solar radiation intensity (W / m²), is the area of the exterior window. Minimized door and window opening combination.

[0033] Collaborative control execution: Using a multi-objective particle swarm optimization algorithm, a control instruction set is generated under the three objectives of energy consumption, comfort, and policy compliance; Send commands to the air conditioner and electric door and window actuators.

[0034] Specifically, an improved multi-objective particle swarm optimization (MOPSO) algorithm was used: particle dimensions: air conditioning set temperature, door and window openings, and wind speed level; fitness function: F1 = energy consumption index = 0.6 × (compressor power) + 0.4 × (fresh air heat loss); F2 = comfort = 1 - |PMV| (PMV is the predicted average voting value); F3 = policy compliance = Σ (the number of violations of local energy-saving regulations); after 50 iterations, the Pareto optimal solution set was output, and the instruction combination with the highest comprehensive score was selected.

[0035] For example, the system chooses a solution with a set temperature of 27°C and a door and window opening of 15%. Compared with the 25°C + fully closed mode that users are accustomed to, energy consumption is reduced and the PMV value is maintained within ±0.5.

[0036] Example 2: This example differs from Example 1 in that, in the dynamic control calculation step, the main living area is further divided into a first main living area and a second main living area based on historical area usage; the temperature of the second main living area is controlled so that the temperature difference between the second main living area and the historical habitual temperature before long-term activities is maintained at a value a; The temperature control of the second main living area also includes adjusting the second temperature of the second main living area when the first main living area is used according to the usage time ratio of the second main living area to the usage time of the first main living area, so that the difference between the second temperature and the historical habitual temperature is .

[0037] Specifically, the DBSCAN clustering algorithm is used to analyze user historical location data, marking the living room as the first main area (average daily usage of 6.2 hours) and the study as the second area (average daily usage of 1.5 hours); the second area maintains ΔT=0.8℃ (for example, when the habitual temperature is 26℃, the actual temperature is maintained at 26.8℃); when the user is detected moving towards the study, the temperature is lowered to the target value at a rate of 0.4℃ / min.

[0038] The sub-area pre-cooling time is shortened to 3.2 minutes (compared to 8 minutes in the whole-house constant temperature mode), and the ineffective energy supply time is reduced.

[0039] Specifically, based on historical usage data (such as the length of time people stay in the area recorded by infrared sensors), the DBSCAN clustering algorithm is used to divide living areas into two categories: The first main living area (P1): The average daily usage time is >4 hours (such as the living room and master bedroom). The second main living area (P2): The average daily usage time is 1-4 hours (such as the study and second bedroom). Static temperature difference a: Initially set the P2 area to maintain a fixed difference from the historical habitual temperature during the inactive period (such as a = 0.8°C). The optimal value is determined through experiments.

[0040] Dynamic temperature difference : Dynamically adjust based on the usage time ratio R of P2 and P1 (R = P2 usage time / P1 usage time): , where 0.2≤R≤0.5; When R>0.5: it is considered as a high-frequency usage area. (Control the temperature directly according to your habit); When R<0.2: it is considered as low frequency area, (Maintain maximum temperature difference).

[0041] Dynamic temperature difference calculation process: Data collection: Count the actual usage time of P1 and P2 in the past 7 days (unit: hours); Calculate the ratio R: for example, the total usage time of P1 = 40h, P2 = 12h → R = 12 / 40 = 0.3; Determine :If a=0.8℃, then =0.8×(1-0.3)=0.56℃; Executive control: P2 inactive temperature = historical habit temperature + ; When P1 is active, P2 temperature = historical habit temperature + ×(1-P1 usage intensity).

[0042] For example, the user's customary temperature: the study room (P2) is set at 26°C; Historical usage data: Average weekly usage time of P1 (living room): 6 hours / day; Average weekly usage time of P2 (study): 1.8 hours / day → R=1.8 / 6=0.3; Static parameter: a=0.8°C; Dynamic control process: Pre-adjustment during the inactive period: The study temperature is maintained at 26+0.8×(1-0.3)=26.56°C (0.56°C higher than the usual temperature); Air conditioning energy consumption is reduced by 23% (compared with a constant temperature of 26°C for the whole house); Collaborative control when P1 is active: When the living room is in use (assuming the current usage intensity is 0.7), the study temperature is adjusted to 26+0.56×(1-0.7)=26.17°C; Before the user enters the study, the temperature needs to be reduced from 26.17°C to 26°C, which takes only 2.1 minutes (0.4°C / min adjustment rate).

[0043] Thermal inertia compensation: Introduce the door and window opening and closing frequency correction coefficient β: ,in, The number of times doors and windows are opened and closed per hour; When f=5 times / h, (compensate for additional heat loss); Time decay factor: When updating the R value daily, add a decay factor γ = 0.9 to avoid sudden changes: ; For example, the weekly usage pattern of a family study (P2) is: Weekdays: used from 19:00 to 21:00 every day (R = 2 / 6 ≈ 0.33) → ; Weekends: Used every day from 2:00 PM to 6:00 PM (R = 4 / 6 ≈ 0.67) → (Directly control the temperature according to custom); Pre-cooling starts at 18:30 on weekdays: the temperature drops from 26.536℃ to 26℃ at a rate of 0.4℃ / min, taking 1.34 minutes; pre-cooling starts at 13:30 on weekends: the temperature is maintained at a constant level from 26℃, without adjusting energy consumption due to temperature difference.

[0044] In other embodiments, the dynamic control calculation step also includes a pre-adjustment strategy based on the user behavior prediction model, specifically: using the LSTM network to analyze the historical usage time series of the second main living area, and predicting the probability of users entering the area within the next 2 hours. ; when When the temperature is pre-adjusted: Calculate the pre-adjustment time advance ,in is the regional thermal inertia time constant; Adjust the temperature of the second zone to the historical customary temperature at a rate of 0.4°C / min; Close doors and windows in adjacent inactive areas in a coordinated manner to reduce heat exchange.

[0045] Specifically, data collection: time series data: records the presence status (0 / 1) of people in the second main living area (P2) every 10 minutes for the past 30 days; auxiliary features: date type (weekday / holiday), current time period, activity status of adjacent areas, outdoor temperature and humidity; network structure: input layer: 72 time steps (12 hours of historical data) × 5 features (presence status, date type, time period, outdoor temperature, and activity in adjacent areas); LSTM layer: 64 hidden units, dropout = 0.2; output layer: sigmoid activation function outputs the probability of use within the next 2 hours ; Training strategy: Loss function: weighted binary cross entropy (positive sample weight 1.5); Optimizer: Adam (lr=0.001); Early stopping mechanism: terminate when the validation set loss does not decrease for three consecutive rounds.

[0046] For example, historical data from a study room shows that the probability of use between 19:00 and 21:00 every Wednesday is 85%. After the model learns this pattern: Input features: [Wednesday, 18:50, outdoor temperature 28°C, living room active] → Output =0.89; trigger pre-adjustment threshold >0.7.

[0047] Thermal inertia time constant Measurement: Select a time when there are no people, run the air conditioner at full load to reduce the room temperature by 2°C (e.g. from 26°C to 24°C); after turning off the air conditioner, record the time required for the temperature to return to ΔT = 0.5°C (e.g. 45 minutes); repeat 3 times and take the average value as the value. ; Dynamic adjustment formula: , where f is the frequency of opening and closing doors and windows (times / hour). When f>5, the compensation coefficient is enabled.

[0048] For example, the study room = 40 minutes, the current door and window opening and closing frequency f = 6 times / h; calculate ; The time advance error is reduced from ±8.2 minutes to ±2.1 minutes, and the accuracy of pre-adjustment start timing is improved.

[0049] Temperature regulation execution control: Gradient temperature rise control: .

[0050] Door and window linkage strategy: Based on the building topology map, it identifies inactive areas within 3 meters and controls the electric door and window actuators through the Zigbee protocol.

[0051] For example, when pre-adjustment is started: the study air conditioner changes from 26.8℃ to 26℃ at a rate of 0.4℃ / min (takes 2 minutes); the doors and windows of the adjacent storage room are closed in conjunction (the distance is 2.5 meters); the fresh air system is maintained at a low speed (20m³ / h); the PMV value during the adjustment process is stable within the range of ±0.3, which is more comfortable than the traditional method of a sudden drop of 1℃.

[0052] If no one enters within 30 minutes after pre-adjustment is completed: the temperature returns to energy-saving mode (set value + 0.5°C), and the abnormal event is recorded for online model update; For example, one day, the study room was pre-conditioned and no one used it: the system automatically retreated to 26.5℃ after 30 minutes. After three cumulative abnormalities, the model fine-tuning was triggered: the "temporary overtime" feature dimension was added.

[0053] For example, the typical usage pattern of a family study room (P2) is: weekdays: 19:00-21:00 (probability 0.85), weekends: 14:00-17:00 (probability 0.65), System response process: Wednesday 18:45 prediction: LSTM output =0.89>0.7, calculate Δt=14 minutes ( =40min, f=4times / h); 18:46 start pre-conditioning: cool down from 26.8℃ at 0.4℃ / min; temperature control: 18:46-18:53: 26.8℃→26.0℃ (ΔT=0.8℃), maintain 26.0℃±0.2℃ fluctuation; Door and window linkage: close adjacent corridor windows (2.8 meters apart) and turn on the circulation fan in the study (at low speed); Actual user entry: At 19:02, a person was detected entering. The temperature had reached 26.1°C, and the set temperature was maintained. The fresh air volume was increased to 35m³ / h.

[0054] In other embodiments, the dynamic control calculation step further includes a regional thermal inertia learning strategy: determining the thermal time constant of each region through a step response test : The air conditioner runs at full capacity until the temperature drops by 2°C; Record the time required for the temperature to return to the first set value; Dynamically adjust pre-adjustment time , The calculation model is as follows: ; When the frequency of opening and closing doors and windows is detected to be greater than 5 times per hour, the system will temporarily increase the Valuation 20%.

[0055] Specifically, select the early morning hours (e.g., 02:00-04:00) when there are no people active, close all doors and windows to enter a fully enclosed mode, and ensure that the temperature fluctuation in adjacent areas is less than ±0.3°C; The air conditioner is fully loaded and the temperature in the target area drops by ΔT=2℃ (e.g., from 26℃ to 24℃). After turning off the air conditioner, the temperature recovery data is recorded every 30 seconds. The timing is stopped when the temperature rises to T_set+0.5℃ (e.g., 26.5℃). Repeat the test 3 times and take the median time as the value. Initial value.

[0056] Data verification: The temperature curve was fitted by the least squares method and verified to be consistent with the first-order system characteristics (R²>0.95).

[0057] Dynamic pre-adjustment time optimization: Time adjustment model: ; Where, f is the frequency of opening and closing doors and windows (times / hour), When the frequency exceeds 5 times / h, a 20% compensation will be added for every 5 times.

[0058] Update cycle: Automatically perform step response test updates every Sunday morning , the Δt value is updated at 23:00 every day.

[0059] For example, the initial , new test , currently f=7 times / h: ; Implementation: Preconditioning time increased from 35 minutes to 37.84 minutes.

[0060] Door and window opening and closing frequency compensation: Real-time monitoring: Door and window magnetic sensors report status changes every 5 seconds, and sliding windows count the number of openings and closings per hour (window length = 60 minutes, step length = 5 minutes). Dynamic compensation rules: ; Attenuation mechanism: When the frequency returns to normal, the compensation coefficient decays by 10% per hour.

[0061] For example, during a party, if the study door and window opening and closing frequency is f=8 times / h for 20 minutes: trigger immediately ; The pre-adjustment time is adjusted to: .

[0062] For example, thermal inertia learning is implemented in a villa living room (area 45㎡): Initial parameters: , Δt=45min; Environmental changes: Weekly family gatherings result in doors and windows opening and closing at a frequency of f=6 times / h, and the addition of floor-to-ceiling windows increases heat loss by 20%.

[0063] Cycle test: Step response measured in the first week (Due to the addition of new windows), update the formula: ; Frequency compensation: f=6 times / h for 2 hours during the party → ,dynamic .

[0064] The pre-adjustment start time is from 45min to 52.8min.

[0065] In other implementations, the dynamic control calculation step further includes a temporary region priority competition mechanism: When N temporary active areas are detected: Calculate the heat load urgency of each area: ;in, is the current temperature difference, The most recent usage duration; The Hungarian algorithm is used to allocate air conditioning cooling capacity, giving priority to area; Maintain basic air supply volume for other areas and improve the sealing level of doors and windows.

[0066] Specifically, the temperature difference calculate, , air conditioning set temperature There are dynamic adjustments to historical custom values, and the current temperature Real-time acquisition through regional temperature and humidity sensors; ; , is the highest priority.

[0067] For example, the living room (temporary area 1): Current temperature 28°C vs. set temperature 26°C → ΔT = 2°C; Last usage time 50 minutes → ; ; Study (temporary area 2): ΔT = 1.5°C, ; ; Judgment: Only the living room meets the requirements , and obtain priority cooling rights.

[0068] Hungarian algorithm cooling capacity allocation: Construct a cost matrix: rows: N temporary areas; columns: M available cooling capacity levels (such as 30%, 50%, 70%, 100% output of air conditioner); Matrix elements: , is the theoretical urgency solution capacity of cooling capacity level j; Optimal matching: Use the Hungarian algorithm to find the allocation solution that minimizes the total cost, with the constraint that Σ cooling capacity ≤ maximum air conditioner output power; Dynamic adjustment: The allocation plan is recalculated every 5 minutes, and a 10% cooling capacity buffer is set to cope with sudden demand.

[0069] Non-priority area treatment strategy: Basic air supply volume control: ,when When maintaining the minimum air supply volume (15% ).

[0070] Enhanced door and window sealing: Close the electric doors and windows to 90% sealing, activate the air pressure balance system (maintain the indoor and outdoor pressure difference ≤ 5Pa), and enable heat-reflecting curtains (reduce solar radiation heat gain by 60%).

[0071] For example, the second bedroom ( =1.2) Not selected as a priority area: The air supply volume is reduced to 0.3×1200m³ / h×e^{-0.1×0.2}=354m³ / h, the doors and windows are sealed to 90%, the heat loss is reduced by 42%, and the temperature rise rate is reduced from 0.8℃ / h to 0.3℃ / h.

[0072] Abnormal scene processing mechanism: Overload protection: When the Σ required cooling capacity > 110% of the rated value: start graded load reduction: priority reduction 10% cooling capacity of the lowest zone; send early warning notifications and recommend shutting down non-essential zones.

[0073] Emergency Mode: When a medical equipment area (such as an oxygen concentrator) is detected: the area is automatically elevated The weight is increased to 3.0 (beyond conventional calculations), forcing allocation of at least 50% cooling capacity.

[0074] For example, emergency scene: living room ( =2.1) + baby room (medical equipment) Simultaneous demand: the system automatically adds baby room Set to 3.0, distribution plan: baby room 70% (2100W), living room 30% (900W), triggering overload protection: shut down the study air conditioning supply.

[0075] For example, during a family gathering, three temporary active areas appear: living room: ΔT=2.5℃, →U=1.97; Restaurant: ΔT=1.8℃, →U=1.56; Entertainment room: ΔT=1.2℃, →U=0.96; Urgency calculation: Threshold U>1.5: Living room and dining room enter the priority queue Hungarian algorithm allocation: Total air conditioning cooling capacity is 3000W, demand: 1200W for the living room, 900W for the dining room, and 600W for the entertainment room; optimal allocation: 100% for the living room (1200W), 100% for the dining room (900W), and 0% for the entertainment room; the remaining 900W is used for the fresh air system (300W) and buffer (600W).

[0076] Treatment for non-priority areas: The air supply volume in the entertainment room was reduced to 200 m³ / h (originally 600 m³ / h), the doors and windows of the adjacent corridor were closed, and the sealing level was increased to Lv3.

[0077] Dynamic monitoring: detection every 5 minutes Change, restaurant in 30 minutes Drop to 1.2, releasing 500W to the entertainment room.

[0078] In other implementations, the dynamic control calculation step also includes cross-time period strategy migration: Analyze historical data to generate typical scenario templates: Weekday home mode: , ; Holiday party mode: , ; Detection of a sudden change in user behavior pattern: matching historical templates with a similarity > 85%; Dynamically modify the current control strategy according to template parameters; If there is no matching template within 24 hours, strategy generation based on reinforcement learning is triggered.

[0079] Specifically, data collection: complete operation logs for the past 90 days are collected, including: temperature setpoint change sequences, door and window opening and closing event timestamps, regional activity heat maps, and external meteorological data (temperature, humidity, and rainfall intensity); Feature Engineering: Time period coding: Divide each day into 6 time periods (e.g. 7:00-10:00 is morning mode); Behavioral feature vector: activity during time period, frequency of temperature change operation, and density of door and window opening and closing.

[0080] Cluster modeling: using OPTICS clustering algorithm ( ,ξ=0.05).

[0081] Similarity calculation: Dynamic Time Warping (DTW) algorithm is used to calculate the matching degree between the current behavior sequence and the template: ; Where L is the length of the time series; Decision rule: When the similarity of three consecutive time windows (each window is 15 minutes) is greater than 85%, template migration is triggered, and weighted fusion is used for multiple matching templates: , .

[0082] For example, at 14:00 on a Sunday, a behavior sequence is detected that matches two templates: Template 1 (party mode): similarity 88%, weight 0.6; Template 2 (home mode): similarity 83%, weight 0.4; Final parameters: ; .

[0083] Reinforcement learning strategy generation: State space: Indoor and outdoor temperature difference, regional activity, time period coding, air conditioning status Indoor and outdoor temperature difference, regional activity, time period coding, air conditioning status (a total of 27 dimensions); Action space: Discrete action: { Increase or decrease by 0.1℃, Δt increase or decrease by 2min, door and window opening ±10%}, continuous action: compressor frequency adjustment (10%-100%); Reward function: ; Training framework: DDPG algorithm (Actor-Critic structure) is used, target network update frequency τ = 0.01, experience replay cache capacity = 10,000 entries.

[0084] Strategy migration execution: smooth parameter transition: , time constant =30min, to avoid discomfort caused by sudden changes.

[0085] Conflict resolution mechanism: When multiple template parameters conflict, arbitration is performed based on the priority of "comfort > energy efficiency > policy." In emergency scenarios, parameters can be manually locked (locking stops automatic migration).

[0086] In other embodiments, the dynamic control calculation step further includes dynamic energy efficiency threshold adjustment: Dynamically correct target temperature difference based on real-time meteorological data and regional activity intensity Allowable range: When the outdoor temperature is greater than 30℃, set The upper limit is 1.2℃; when the temporary area is detected to be used more than 3 times / hour, the The lower limit is 0.3℃; Combined with the air conditioner operating efficiency curve, optimize the compressor frequency distribution.

[0087] In other embodiments, the dynamic control calculation step further includes a personnel distribution adaptive temperature control step: Construct a sub-meter personnel distribution matrix: Divide the area into 0.5m×0.5m grids and mark the coordinates and postures of personnel; Calculate thermal comfort impact factors: , where d is the distance between the person and the air outlet, and h is the person's sitting height; Dynamically adjust air supply parameters: determine whether the thermal comfort impact factor is greater than a first threshold, and if so, execute a first intervention step; determine whether the thermal comfort impact factor is less than a second threshold, and if so, execute a second intervention step; First intervention: Increase local air supply and lower the temperature setting; use smart curtains to adjust shading direction to reduce solar radiation interference; Second intervention: reduce the fan speed to and increase the set temperature; close the corresponding door and window air outlets to reduce heat exchange; Timing control: Update personnel distribution data every 10 minutes and optimize air supply parameter combination through PID control algorithm.

[0088] Specifically, the camera acquires the position of the person and identifies their sitting / standing posture, divides the space into 0.5m×0.5m grids, and labels the attributes; For example, after deploying the system in a living room (6m×4m), it is detected that three people are gathered in the sofa area (coordinates 2.5, 1.8), with a sitting height of 0.9m. The nearest air outlet is located at (3.0, 0.5). The calculated distance d = √[(3-2.5)²+(1.8-0.5)²] = 1.43m.

[0089] Calculation of thermal comfort influencing factors: Calculation of distance factors: , d≥0.5m.

[0090] Height compensation factor: , h is the sitting height.

[0091] Comprehensive formula: ; Threshold setting: First threshold = 0.8 (need to strengthen air supply), second threshold = 0.3 (need to weaken air supply).

[0092] Dynamic air supply parameter adjustment: First intervention ( ): Air supply volume is increased to 1.5m³ / (s·m²) (normal value 1.0), temperature setting is lowered by ΔT=0.5℃, and the linkage curtain rotation angle θ=45°-90° (calculated according to the solar azimuth).

[0093] The second intervention ): Wind speed drops to 0.3m / s (normal 0.8m / s), temperature setting increases by 0.3℃, and corresponding door and window actuators are closed (opening ≤ 20%).

[0094] For example, the dining table area is detected =0.85 (d=0.6m, h=1.1m): Increase the air supply volume to 1.5m³ / (s·m²), the temperature from 26℃→25.5℃, and close the west windows (to reduce direct sunlight). After adjustment, the PMV improves from +0.7 to +0.2.

[0095] Timing control and optimization: Data update cycle: Full update of personnel distribution matrix every 10 minutes, partial refresh of hotspot areas (mobile target tracking) every 30 seconds; For example, the family living room evening movie watching scene: Personnel distribution: Main sofa area: 5 people gathered ( =0.82-0.91); Bar area: 1 person occasionally walked around ( =0.28-0.35); Enhanced air supply in the main sofa area: air supply volume increased by 50%, temperature lowered to 25°C, rear windows closed (to reduce cold loss), downlight angles adjusted to avoid direct heat sources; Weakened air supply in the bar area: wind speed reduced to 0.3m / s, temperature raised to 26.5°C, and air outlets on top of the bar closed.

[0096] The embodiment of the present application also discloses an air-conditioning door and window joint control system.

[0097] The air conditioning door and window joint control system includes the following modules: Information collection module: used to build a multi-source data collection layer to obtain meteorological warning data, user operation logs, regional climate feature database, and user historical usage areas in real time; Behavior feature extraction module: The input end is connected to the output end of the information collection module to generate a three-dimensional demand map that includes time period preference, spatial concentration, and temperature control sensitivity; Dynamic control calculation module: The input end is connected to the output end of the behavior feature extraction module, and is used to calculate the optimal opening threshold of doors and windows based on real-time meteorological data; Collaborative control execution module: The input end is connected to the output end of the dynamic control calculation module, and is used to generate a control instruction set and send instructions to the air conditioner and electric door and window actuators.

[0098] The above are all preferred embodiments of the present application, and are not intended to limit the scope of protection of the present application. Therefore, any equivalent changes made based on the structure, shape, and principle of the present application should be included in the scope of protection of the present application.

Claims

1. A method for controlling doors and windows of an air conditioner, characterized by: The following steps are involved: Information collection: Build a multi-source data collection layer to obtain the following data in real time: Meteorological warning data: obtain the temperature change rate, UV index and rainfall intensity in the next 3 hours; User operation log: records the air conditioner set temperature adjustment records and the door and window opening and closing time series; Regional climate characteristics database: contains climate classification codes; User historical usage area: collect historical distribution map of users; Behavioral feature extraction: Spatiotemporal convolutional neural networks are used to process user operation logs and heat map data to generate a three-dimensional demand map that includes time preferences, spatial concentration, and temperature control sensitivity. Dynamic control calculation: Based on the building BIM model, the SHGC value of external windows and the thermal conductivity of the wall are extracted, and the optimal opening threshold of doors and windows is calculated in combination with real-time meteorological data. When the main living area is detected, the opening of this area is calculated first; Collaborative control execution: Using a multi-objective particle swarm optimization algorithm, a control instruction set is generated under energy consumption and comfort targets; Send commands to the air conditioner and electric door and window actuators.

2. The air-conditioning door and window joint control method according to claim 1, characterized in that: The dynamic control calculation step also includes dividing the main living area into a first main living area and a second main living area based on historical area usage; and controlling the temperature of the second main living area so that the temperature difference between the second main living area and the historical habitual temperature is a before people move for a long time.

3. The air-conditioning door and window joint control method according to claim 2, characterized in that: In the dynamic control calculation step, the temperature control of the second main living area also includes adjusting the second temperature of the second main living area when the first main living area is used according to the usage time ratio of the second main living area to the usage time of the first main living area, so that the difference between the second temperature and the historical habitual temperature is .

4. The air-conditioning door and window joint control method according to claim 3, characterized in that: The dynamic control calculation step also includes a pre-adjustment strategy based on the user behavior prediction model, specifically: Analyze the historical usage time series of the second main living area through the LSTM network and predict the probability of users entering this area in the next 2 hours ; when When the temperature is pre-adjusted: Calculate the pre-adjustment time advance ,in is the regional thermal inertia time constant; Adjust the temperature of the second zone to the historical customary temperature at a rate of 0.4°C / min; Close doors and windows in adjacent inactive areas in a coordinated manner to reduce heat exchange.

5. The air-conditioning door and window joint control method according to claim 4, characterized in that: The dynamic control calculation step also includes a regional thermal inertia learning strategy: the thermal time constant of each region is determined by a step response test. : The air conditioner runs at full capacity until the temperature drops by 2°C; Record the time required for the temperature to return to the first set value; Dynamically adjust pre-adjustment time , The calculation model is as follows: ; When the frequency of opening and closing doors and windows is detected to be greater than 5 times per hour, the system will temporarily increase the Valuation 20%.

6. The air-conditioning door and window joint control method according to claim 4, characterized in that: The dynamic control calculation step also includes a temporary region priority competition mechanism: When N temporary active areas are detected: Calculate the heat load urgency of each area: ;in, is the current temperature difference, The most recent usage duration; The Hungarian algorithm is used to allocate air conditioning cooling capacity, giving priority to area; Maintain basic air supply volume for other areas and improve the sealing level of doors and windows.

7. The air-conditioning door and window joint control method according to claim 6, characterized in that: The dynamic control calculation step also includes cross-time strategy migration: Analyze historical data to generate typical scenario templates: Weekday home mode: , ; Holiday party mode: , ; Detecting a sudden change in user behavior patterns: matching historical templates with a similarity > 85%; Dynamically modify the current control strategy according to template parameters; If there is no matching template within 24 hours, strategy generation based on reinforcement learning is triggered.

8. The air-conditioning door and window joint control method according to claim 7, characterized in that: The dynamic control calculation step also includes dynamic energy efficiency threshold adjustment: Dynamically correct target temperature difference based on real-time meteorological data and regional activity intensity Allowable range: When the outdoor temperature is greater than 30℃, set The upper limit is 1.2℃; when the temporary area is detected to be used more than 3 times / hour, the The lower limit is 0.3℃; Combined with the air conditioner operating efficiency curve, optimize the compressor frequency distribution.

9. The air-conditioning door and window joint control method according to any one of claims 1 to 8, characterized in that: The dynamic control calculation step also includes the personnel distribution adaptive temperature control step: Construct a sub-meter personnel distribution matrix: Divide the area into 0.5m×0.5m grids and mark the coordinates and postures of personnel; Calculate thermal comfort impact factors: , where d is the distance between the person and the air outlet, and h is the person's sitting height; Dynamically adjust air supply parameters: determine whether the thermal comfort impact factor is greater than a first threshold, and if so, execute a first intervention step; determine whether the thermal comfort impact factor is less than a second threshold, and if so, execute a second intervention step; First intervention: Increase local air supply and lower the temperature setting; use smart curtains to adjust shading direction to reduce solar radiation interference; Second intervention: reduce the fan speed to and increase the set temperature; close the corresponding door and window air outlets to reduce heat exchange; Timing control: Update personnel distribution data every 10 minutes and optimize air supply parameter combination through PID control algorithm.

10. A system for implementing the air-conditioning door and window joint control method according to any one of claims 1 to 9, characterized in that: Includes the following modules: Information collection module: used to build a multi-source data collection layer to obtain meteorological warning data, user operation logs, regional climate feature database, and user historical usage areas in real time; Behavior feature extraction module: The input end is connected to the output end of the information collection module to generate a three-dimensional demand map that includes time period preference, spatial concentration, and temperature control sensitivity; Dynamic control calculation module: The input end is connected to the output end of the behavior feature extraction module, and is used to calculate the optimal opening threshold of doors and windows based on real-time meteorological data; Collaborative control execution module: The input end is connected to the output end of the dynamic control calculation module, and is used to generate a control instruction set and send instructions to the air conditioner and electric door and window actuators.