Air conditioner control method, air conditioner control system and air conditioner

Abstract

The invention relates to the technical field of air conditioner control, in particular to an air conditioner control method, an air conditioner control system and an air conditioner. The method comprises the following steps: acquiring environment monitoring parameters of a plurality of regions, performing temperature and humidity change analysis and multi-region temperature and humidity spatial distribution calculation, and constructing a multi-region temperature and humidity fluctuation distribution diagram; acquiring a monitoring image of each region based on the high-definition camera, and performing cross-region personnel flow behavior prediction to generate cross-region personnel behavior prediction data; local heat flow distribution mining is carried out on the multi-region temperature and humidity fluctuation distribution map according to cross-region personnel behavior prediction data, heat flow distribution speculation is carried out, and a heat flow distribution prediction map is constructed; and an air conditioner historical monitoring log is obtained, multi-period personalized use habit evolution analysis is carried out, and therefore air conditioner use habits of different areas are obtained. Based on personalized user temperature demand analysis, more efficient and accurate intelligent air conditioner control is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of air conditioner control, and particularly to an air conditioner control method, an air conditioner control system, and an air conditioner. Background Art

[0002] In modern buildings, air conditioning systems often need to meet the temperature and humidity requirements of multiple areas simultaneously, and the usage conditions, personnel distribution, and external environmental conditions in each area are different. With the continuous influence of climate change and human activities on the indoor environment, traditional air conditioner control methods can no longer achieve real-time and precise control of each area. Especially during peak hours, some areas may require higher temperature control accuracy, while other areas can reduce energy consumption. At this time, if the air conditioning system cannot adjust and coordinate the working states of each area in real time, it will lead to energy waste and cannot provide ideal user comfort.

[0003] To meet the higher requirements of modern buildings for air conditioning systems, air conditioning systems based on intelligent control have gradually emerged, aiming to achieve optimal collaborative control of air conditioning equipment between multiple areas through advanced algorithms and technical means. Compared with traditional independent control methods, intelligent air conditioner control systems can achieve fine-tuning between areas by real-time monitoring information such as the environmental state, personnel distribution, and equipment operation status of each area, and perform dynamic optimization control according to the usage requirements of different areas. This method can not only improve the energy use efficiency of the air conditioning system but also enhance the temperature control accuracy of each area and ensure the comfort in the building.

[0004] However, despite the certain progress made by intelligent air conditioner control systems, they still face many challenges. First, the temperature and humidity fluctuations between different areas are large, and how to accurately predict and adjust these fluctuations to avoid over-regulation or inappropriate temperature control remains an urgent problem to be solved. Second, factors such as personnel flow, equipment usage, and external climate will all affect the air conditioning system. How to accurately collect, analyze this information, and perform real-time feedback and optimization in the system depends on powerful data processing capabilities and intelligent algorithms. Finally, existing air conditioner control systems often lack global coordination, and the collaborative effect between areas is insufficient, easily resulting in situations where some areas consume excessive energy or the temperature control effect is not ideal. Therefore, a more intelligent air conditioner control method is needed. Summary of the Invention

[0005] In order to solve the above technical problems, the present invention proposes an air conditioner control method, an air conditioner control system, and an air conditioner to solve at least one of the above technical problems.

[0006] To achieve the above object, the present invention provides an air conditioner control method. A high-definition camera is installed on the air conditioner, and the method includes the following steps:

[0007] Step S1: Obtain the environmental monitoring parameters of multiple regions, conduct temperature and humidity change analysis and calculate the temperature and humidity spatial distribution in multiple regions, and construct a temperature and humidity fluctuation distribution map for multiple regions;

[0008] Step S2: Based on the high-definition camera, collect the monitoring images of each region and conduct cross-regional personnel flow behavior prediction to generate cross-regional personnel behavior prediction data;

[0009] Step S3: According to the cross-regional personnel behavior prediction data, mine the local heat flow distribution of the temperature and humidity fluctuation distribution map for multiple regions, and conduct heat flow distribution speculation to construct a heat flow distribution prediction map;

[0010] Step S4: Obtain the air conditioner historical monitoring logs, conduct multi-period personalized usage habit evolution analysis, and thus obtain the air conditioner usage habits in different regions;

[0011] Step S5: Conduct in-depth analysis of the regional comprehensive temperature requirements according to the air conditioner usage habits in different regions, and generate the user personalized temperature requirement characteristics for each region;

[0012] Step S6: According to the heat flow distribution prediction map and the user personalized temperature requirement characteristics of each region, conduct adaptive local air conditioner parameter adjustment and global coordination iterative control optimization, and construct an adaptive intelligent air conditioner control model.

[0013] In the present invention, there is also provided an air conditioner control system for executing the above-mentioned air conditioner control method, including:

[0014] A temperature and humidity distribution module, configured to obtain the environmental monitoring parameters of multiple regions, conduct temperature and humidity change analysis and calculate the temperature and humidity spatial distribution in multiple regions, and construct a temperature and humidity fluctuation distribution map for multiple regions;

[0015] A behavior prediction module, configured to collect the monitoring images of each region based on the high-definition camera and conduct cross-regional personnel flow behavior prediction to generate cross-regional personnel behavior prediction data;

[0016] A flow distribution speculation module, configured to mine the local heat flow distribution of the temperature and humidity fluctuation distribution map for multiple regions according to the cross-regional personnel behavior prediction data, and conduct heat flow distribution speculation to construct a heat flow distribution prediction map;

[0017] A usage habit module, configured to obtain the air conditioner historical monitoring logs, conduct multi-period personalized usage habit evolution analysis, and thus obtain the air conditioner usage habits in different regions;

[0018] A temperature requirement module, configured to conduct in-depth analysis of the regional comprehensive temperature requirements according to the air conditioner usage habits in different regions, and generate the user personalized temperature requirement characteristics for each region;

[0019] A self-using adjustment module is used to perform adaptive local air-conditioning parameter adjustment and global coordinated iterative control optimization according to the heat flow distribution prediction diagram and the user's personalized temperature demand characteristics in each region, and construct an adaptive intelligent air-conditioning control model.

[0020] The present invention also provides an air conditioner, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the steps of the air-conditioning control method described in any one of the above are implemented.

[0021] The beneficial effects of the present invention are specifically as follows: By collecting temperature and humidity data in multiple areas, the environmental conditions of each area can be monitored in real time to ensure the accuracy of the data. This provides reliable basic data for the subsequent adjustment of the air-conditioning system. The analysis of temperature and humidity changes and the construction of spatial distribution maps make the temperature and humidity trends in each area visible, enabling managers to quickly identify which areas have large temperature and humidity fluctuations, thereby optimizing the adjustment strategy of the air-conditioning system. The temperature and humidity fluctuation distribution map helps to discover the differences in temperature and humidity fluctuations between areas, clarify which areas have more drastic temperature and humidity changes and which areas are relatively stable, thus providing data support for the dynamic adjustment of the air-conditioning system. Through image collection by high-definition cameras and prediction of personnel flow behavior, the distribution and movement trends of personnel in different areas can be understood in real time. This is crucial for air-conditioning adjustment because the distribution of personnel directly affects the heat load of the area. Cross-regional personnel behavior prediction data can help the air-conditioning system predict changes in the degree of personnel density in the area and adjust the air-conditioning temperature in advance to avoid temperature and humidity imbalance caused by personnel gathering. By combining personnel behavior with temperature and humidity data, the heat flow pattern can be further inferred. Personnel activities in different areas will cause changes in heat, and understanding these changes can help the air-conditioning system respond in real time. Based on the heat flow distribution prediction map, the air-conditioning system can adjust the temperature and humidity more precisely, avoiding energy waste caused by uneven heat flow. This step can help the air-conditioning system consider the heat distribution across regions, making the temperature and humidity changes between different regions more coordinated, thereby reducing unnecessary conflicts during air-conditioning operation. Analyzing the air-conditioning historical logs and usage habits can deeply understand the personalized needs of each area. Some areas may be more inclined to use the cooling or heating function during specific time periods. By analyzing the evolution of air-conditioning usage habits in different areas, the air-conditioning system can make predictions based on historical data to meet user needs in advance, thereby improving comfort and energy utilization efficiency. According to the air-conditioning usage habits in different areas, the temperature demand characteristics of each area can be analyzed more carefully, and the demand changes of users in each area at different time periods can be understood. This helps the air-conditioning system to make precise adjustments according to the needs of each area, avoiding discomfort and waste caused by one-size-fits-all settings. Precise matching of the user's temperature demand can improve the user's comfort with the environment and enhance the experience. Based on heat flow prediction and personalized temperature demand, the adaptive adjustment of the air-conditioning system can be realized, making the operation of the air-conditioning more flexible and intelligent. Through global coordination iterative control optimization, not only can the temperature demands of each area be met, but also the optimal allocation of energy consumption can be achieved, avoiding local over-regulation or energy waste. The adaptive model can dynamically adjust the working state of the air-conditioning according to real-time data and environmental changes, achieving more efficient and precise temperature control management. Through intelligent adjustment and optimization, the system can minimize energy consumption while ensuring a comfortable environmental quality and reducing the operating cost of the air-conditioning. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic diagram of the step flow of an air conditioner control method of the present invention;

[0023] Figure 2 It is a schematic diagram of the detailed implementation steps of step S1;

[0024] Figure 3 It is a schematic diagram of the detailed implementation steps of step S2;

[0025] Figure 4 It is a schematic diagram of the detailed implementation steps of step S3. Specific implementation manners

[0026] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0027] The embodiments of the present application provide an air conditioner control method, an air conditioner control system and an air conditioner. The execution subjects of the air conditioner control method, the air conditioner control system and the air conditioner include but are not limited to: mechanical equipment, data processing platforms, cloud server nodes, network uploading devices, etc. that can be regarded as general computing nodes of the present application. The data processing platform includes but is not limited to: at least one of an audio and image management system, an information management system, and a cloud data management system.

[0028] Please refer to Figures 1 to 4 , the present invention provides an air conditioner control method, and the air conditioner control method includes the following steps:

[0029] Step S1: Obtain the environmental monitoring parameters of multiple regions, perform temperature and humidity change analysis and multi-region temperature and humidity spatial distribution calculation, and construct a multi-region temperature and humidity fluctuation distribution map;

[0030] Step S2: Collect the monitoring images of each region based on the high-definition camera, and perform cross-region personnel flow behavior prediction to generate cross-region personnel behavior prediction data;

[0031] Step S3: Mine the local heat flow distribution of the multi-region temperature and humidity fluctuation distribution map according to the cross-region personnel behavior prediction data, and perform heat flow distribution speculation to construct a heat flow distribution prediction map;

[0032] Step S4: Obtain the air conditioner historical monitoring log, perform multi-period personalized usage habit evolution analysis, and thus obtain the air conditioner usage habits of different regions;

[0033] Step S5: Perform in-depth analysis of the regional comprehensive temperature demand according to the air conditioner usage habits of different regions to generate the user personalized temperature demand characteristics of each region;

[0034] Step S6: Based on the heat flow distribution prediction map and the user's personalized temperature demand characteristics in each region, perform adaptive local air-conditioning parameter adjustment and global coordination iterative control optimization to construct an adaptive intelligent air-conditioning control model.

[0035] In the embodiment of the present invention, refer to Figure 1 , which is a schematic diagram of the step flow of an air-conditioning control method of the present invention. In this example, the steps of the method include:

[0036] Step S1: Obtain the environmental monitoring parameters of multiple regions, perform temperature and humidity change analysis and multi-region temperature and humidity spatial distribution calculation, and construct a multi-region temperature and humidity fluctuation distribution map;

[0037] In this embodiment, environmental monitoring devices are installed in each key area (such as meeting rooms, offices, corridors, etc.). These devices should be able to measure temperature and humidity in real time and have the ability to transmit data. Ensure that the selected devices have high precision and stability to improve the reliability of the data. Install temperature and humidity sensors in meeting rooms and offices respectively, and record the device model and installation location, such as "Meeting room sensor model: DHT22, installation location: east wall". Configure the monitoring devices to collect environmental parameters at regular intervals, usually recording temperature and humidity every 5 minutes. The data should be transmitted to the central database via wireless or wired means to ensure the timeliness and integrity of the data. The recording format is "timestamp, area, temperature (°C), humidity (%)", such as "2023-04-01 10:00:00, meeting room, 22.5°C, 45%". Conduct a preliminary verification of the collected data to check for outliers (such as temperature exceeding the normal range or humidity being negative). Clean the invalid data to ensure the accuracy of subsequent analysis. If the temperature recorded at a certain time is -10°C, mark it as an outlier and delete it, and record "Data cleaning completed, outlier deleted". Organize the collected environmental monitoring data by time and area, and calculate the average, maximum, and minimum values of temperature and humidity for each area. These statistical data will provide a basis for subsequent analysis of temperature and humidity changes. Calculate that the average temperature in the meeting room in the past 24 hours is 22.0°C, the maximum temperature is 24.5°C, and the minimum temperature is 20.5°C, and record it as "Meeting room temperature statistics: average 22.0°C, maximum 24.5°C, minimum 20.5°C". By comparing the temperature and humidity data at different time periods (such as morning, afternoon, and evening), analyze the temperature and humidity change trends in each area. Use time series analysis methods to identify the change patterns of temperature and humidity. If it is found that the temperature in the meeting room rises significantly in the afternoon, it can be recorded as "The temperature in the meeting room shows an obvious upward trend in the afternoon, with an average increase of 2°C". Record the results of the temperature and humidity change analysis in the database and generate a report summarizing the temperature and humidity changes in each area for subsequent decision-making. The recording format is "area, time period, temperature change, humidity change", such as "meeting room, afternoon, temperature rises 2°C, humidity drops 5%". Based on the temperature and humidity data of each area, construct a spatial distribution model. Interpolation methods (such as Kriging interpolation or inverse distance weighted interpolation) can be used to estimate the temperature and humidity changes between areas. Based on the data of known environmental monitoring points, estimate the temperature and humidity in the corridor and office, and record it as "Corridor temperature estimation: 23.0°C, humidity: 50%". Apply the interpolation algorithm in the spatial distribution model to calculate the temperature and humidity distribution in the entire monitoring area. Generate a temperature and humidity distribution table for each area and ensure the continuity and rationality of the data. If the calculated temperature and humidity in the office are 22.5°C and 48%, record it as "Office temperature and humidity distribution calculation completed: temperature 22.5°C, humidity 48%".Visualize the calculated multi - area temperature and humidity distribution data in the form of a chart to generate a temperature and humidity fluctuation distribution map. This chart should be able to visually display the temperature and humidity differences between regions for easy analysis and decision - making. Use a heat map to show the temperature and humidity distribution of each region, mark the high and low temperature and humidity areas, and record it as "The generation of the multi - area temperature and humidity fluctuation distribution map is completed."

[0038] Step S2: Based on the monitoring images collected by the high - definition camera, predict the cross - region personnel flow behavior for each region to generate cross - region personnel behavior prediction data;

[0039] In this embodiment, a suitable high-definition camera is selected to ensure that it has sufficient resolution and field of view to cover key locations in each area (such as entrances, exits, and main activity areas). The camera should have night vision capabilities to ensure effective monitoring in low-light environments. Install HD1080P cameras in the meeting room, office, and corridor respectively, record the installation location and field of view, and ensure that the cameras can clearly capture the activities of people in the area. Configure the cameras for real-time monitoring and collect monitoring images at regular intervals (such as every 5 seconds). All images should be transmitted to the central storage system via the network to ensure the real-time and integrity of the data. The recording format is "timestamp, area, image file name", such as "2023-04-01 10:00:00, meeting room, image_001.jpg". Conduct a quality assessment of the collected monitoring images to ensure that the images are clear and free of problems such as blurring or occlusion. Mark or delete images that do not meet the quality standards to ensure the accuracy of the data for subsequent analysis. If it is found that an image is blurred due to light problems, record it as "Image quality verification completed, blurred image deleted". Preprocess the collected high-definition monitoring images, including image scaling, denoising, and contrast enhancement, etc., to improve the accuracy of subsequent person detection. Uniformly adjust the resolution of each monitoring image to 640x480 and perform denoising processing, and record it as "Image preprocessing completed, image number: image_001". Apply computer vision algorithms (such as YOLO or Faster R-CNN) to detect people in the preprocessed images, identify each individual in the image and track their movement trajectories. Ensure that the detection algorithm is trained to effectively identify people under different lighting conditions. If 3 people are detected in the meeting room image, record it as "Meeting room image, number of people detected: 3". Analyze the cross-regional personnel flow behavior based on the detected people and their movement trajectories. This analysis can be combined with time series data to observe the flow frequency and direction of people between different regions. If it is detected that a person flows from the meeting room to the corridor, record it as "Person 1, flows from the meeting room to the corridor, time: 10:05". Based on historical personnel flow data, construct a cross-regional personnel behavior prediction model. Machine learning methods (such as long short-term memory network LSTM) can be used to predict future personnel flow trends. Train the model using the personnel flow data of the past week to ensure that the model can effectively capture the flow patterns and trends. Apply the constructed behavior prediction model to perform real-time prediction on the currently collected personnel flow data. Output the personnel flow situation in each area for a future period of time, including the estimated number of people flowing in and out. If the model predicts that there will be 5 people flowing into and 3 people flowing out of the meeting room in the next hour, record it as "Meeting room predicted flow: 5 people flowing in, 3 people flowing out". Record the generated cross-regional personnel behavior prediction data in the database and generate visualization charts for subsequent analysis and management. Ensure that the prediction results can be intuitively displayed for easy decision-making.Generate a predicted flow trend chart to show the personnel flow in each region, and record it as "The generation of the cross-regional personnel behavior prediction chart is completed."

[0040] Step S3: Based on the cross-regional personnel behavior prediction data, conduct local heat flow distribution mining on the multi-regional temperature and humidity fluctuation distribution chart, and conduct speculation on the heat flow distribution to construct a heat flow distribution prediction chart;

[0041] In this embodiment, cross-regional personnel behavior prediction data is collected, including the personnel flow in each region, the predicted number of people flowing in and out. Ensure that the data is complete and without duplicates, facilitating subsequent analysis. The recording format is "region, predicted number of people flowing in, predicted number of people flowing out", such as "meeting room, 5 people flowing in, 3 people flowing out". Clean the data to remove invalid or outlier values. Analyze the personnel flow data to identify the flow patterns between different regions. Through clustering analysis or association rule mining, find the key features and patterns of personnel flow. If it is found that the flow frequency between the meeting room and the corridor is high, record it as "the personnel flow frequency between the meeting room and the corridor is high, mainly concentrated in the meeting time period". Record the analysis results in the database to ensure that the personnel flow patterns of each region are traceable. Generate a report summarizing the main characteristics of cross-regional personnel flow. The recording format is "region A, region B, flow frequency", such as "meeting room, corridor, high flow frequency". Extract data from the multi-regional temperature and humidity fluctuation distribution map obtained in the previous steps. Ensure the timeliness and accuracy of the data for heat flow analysis. Extract the temperature and humidity data of the meeting room, office, and corridor, and record it as "region, temperature, humidity", such as "meeting room, 22°C, 45%". According to the temperature and humidity fluctuation data and personnel flow data, construct a local heat flow model. The heat conduction equation and convection model can be used, considering the influence of temperature differences and personnel flow on heat flow. Set the model parameters, including the heat conduction coefficient, convection coefficient, etc., and record it as "the heat flow model parameter setting is completed: heat conduction coefficient = 0.5 W / m·K". Apply the constructed heat flow model to calculate the heat flow distribution in each region. Ensure that the calculation results can reflect the heat transfer between regions. If the calculated heat flow into the meeting room is 200 W and the outflow is 150 W, record it as "local heat flow in the meeting room: 200 W flowing in, 150 W flowing out". Select a suitable prediction method to predict the future heat flow distribution. Time series analysis or machine learning models (such as regression analysis) can be used for modeling. Use historical heat flow data to predict the heat flow trend in the next 24 hours. Based on the existing heat flow data and personnel behavior prediction data, infer the heat flow distribution. Combine the current state with the prediction model to output the future heat flow distribution. If it is predicted that 5 people will flow into the meeting room in the next 1 hour and the heat flow in will increase by 50 W, then infer that the heat flow into the meeting room will reach 250 W. Record the heat flow distribution inference results in the database and generate a heat flow distribution prediction map to facilitate the intuitive display of the heat flow in each region. Generate a heat flow prediction map showing the heat distribution in each region, including the heat flowing in and out, and record it as "the heat flow distribution prediction map generation is completed".

[0042] Step S4: Obtain the historical monitoring logs of the air conditioner, conduct an analysis of the evolution of personalized usage habits in multiple time periods, and thus obtain the air conditioner usage habits in different regions;

[0043] In this embodiment, determine the log storage location of the air conditioner monitoring system to ensure access to the required historical monitoring data. This data should include information such as the operating status, set temperature, actual temperature, usage time, etc. of the air conditioners in each area. The format of the logged data is "timestamp, area, operating status, set temperature, actual temperature", such as "2023-04-01 10:00:00, meeting room, on, 22°C, 21°C". Extract the historical monitoring logs of the air conditioners for a past period (such as the past month) from the monitoring system. Store the extracted data in a database for subsequent analysis and query. The number of extracted records is 500, and the record format is "area, monitoring log file name, number of records", such as "meeting room, ac_log_2023.csv, number of records 500". Clean the data of the extracted historical monitoring logs to remove duplicate, invalid, or abnormal records. Ensure the integrity and accuracy of the data to prevent interference with subsequent analysis. If a record with a temperature of -10°C is found, mark it as abnormal and delete it, and record "data cleaning completed, outlier deleted". Divide the historical monitoring logs into multiple time periods (such as morning, afternoon, evening, etc.) to analyze the air conditioner usage habits in different time periods. Ensure that the definition of each time period is reasonable for comparison. Define the time periods as "morning (6:00 - 12:00), afternoon (12:00 - 18:00), evening (18:00 - 24:00)". Statistically analyze the air conditioner usage in each area during different time periods, including usage frequency, average set temperature, and operating time, etc. Statistical analysis methods (such as counting method and mean calculation) can be used to obtain these data. The usage frequency counted during the morning time period in the meeting room is 10 times, and the average set temperature is 22°C, which is recorded as "meeting room morning usage frequency: 10 times, average set temperature: 22°C". According to the statistical results, analyze the air conditioner usage habits in different areas to find personalized characteristics. Some areas may be more inclined to use the air conditioner during specific time periods or have a preference for specific set temperatures. If it is found that the usage frequency of the air conditioner in the meeting room in the afternoon is significantly higher than other time periods, record it as "the meeting room has a high air conditioner usage frequency in the afternoon, and the preferred set temperature is 22°C".

[0044] Step S5: Deeply analyze the comprehensive temperature requirements of each area based on the air conditioner usage habits in different areas, and generate the user's personalized temperature requirement characteristics for each area;

[0045] In this embodiment, the air-conditioning usage habit data of each area is collected, which includes information such as the usage frequency, average set temperature, and actual operating status of each area at different time periods. These data are integrated together for comprehensive analysis. The record format is "area, time period, usage frequency, average set temperature, actual temperature", such as "meeting room, morning, usage frequency 10 times, average set temperature 22°C, actual temperature 21°C". The collected data is cleaned to remove duplicate and invalid records to ensure the integrity and accuracy of the data. Special attention is paid to the rationality of the time period division to ensure the effectiveness of subsequent analysis. If it is found that the usage frequency within a certain time period is 0 times, this record is considered for deletion to prevent interference with the subsequent analysis results. The cleaned data is stored in the database and prepared for analysis. Ensure that the usage habit data of each area can accurately reflect its air-conditioning usage situation. The number of sorted data records is 500, and the record format is "area, air-conditioning usage habit data preparation completed". According to the air-conditioning usage habit data of different areas, a comprehensive temperature demand analysis model is constructed. A multiple linear regression model can be used, considering multiple factors affecting temperature demand, such as personnel flow, usage frequency, and time period. The model parameters are set as "set temperature = α * usage frequency + β * number of people + γ * time period influence", where α, β, and γ are coefficients to be determined. The comprehensive temperature demand analysis model is trained and verified using historical data. Ensure that the model can accurately predict the temperature demand of each area, and the verification method can use cross-validation or K-fold validation. After training the model with historical data, the prediction accuracy is verified in the test set, and the record is "model training completed, verification accuracy rate is 85%". Using the trained model, the temperature demand of each area is calculated, and the personalized temperature demand characteristics of each area at different time periods are output. If the model output result of the meeting room is "set temperature 23°C", it is recorded as "meeting room personalized temperature demand characteristics: 23°C".

[0046] Step S6: According to the heat flow distribution prediction map and the user personalized temperature demand characteristics of each area, perform adaptive local air-conditioning parameter adjustment and global coordination iterative control optimization to construct an adaptive intelligent air-conditioning control model.

[0047] In this embodiment, the heat inflow and outflow data of each region are extracted from the previously constructed heat flow distribution prediction map. These data will be used to evaluate the heat state and temperature requirements within the region. The recording format is "region, predicted inflow heat (W), predicted outflow heat (W)", such as "meeting room, inflow 200W, outflow 150W". The user personalized temperature requirement characteristic data of each region are collected. These characteristics are based on previous analysis and reflect the user's expectations for air conditioner settings. The recording format is "region, personalized temperature requirement (°C)", such as "meeting room, personalized temperature requirement 23°C". The heat flow data and personalized temperature requirement characteristics are integrated and stored in the database to ensure the integrity and traceability of the data for subsequent adjustment and optimization. The integrated data is recorded as "region, inflow heat, outflow heat, personalized requirement", such as "meeting room, 200W, 150W, 23°C". According to the extracted heat flow data and the user's personalized temperature requirement characteristics, the local air conditioner parameter adjustment targets are set. These targets should be adjusted around achieving comfort and energy-saving effects. The target is set as "adjust the actual temperature of the meeting room to 23°C and keep the inflow heat greater than the outflow heat". An adaptive control strategy is adopted to dynamically adjust the air conditioner settings of each region according to the current environmental conditions and the predicted heat flow situation. This can be achieved through a PID controller or a fuzzy control algorithm. If the current temperature of the meeting room is 24°C and the personalized requirement is 23°C, the set temperature of the air conditioner can be lowered to 22°C to meet the requirement. After the local adjustment is implemented, the temperature changes and heat flow states of each region are monitored in real time. According to the feedback data, the adjustment effect is evaluated and necessary adjustments are made. If the temperature of the meeting room drops to 23°C after adjustment, it is recorded as "the adjustment effect of the meeting room is good, and the temperature reaches the setting". On the basis of the local adjustment, the global optimization target is set to ensure the coordination between different regions of the entire system. The global target should include minimizing energy consumption and maximizing comfort. The global target is set as "while maintaining the temperature of all regions within the personalized requirement range, minimize the energy consumption as much as possible". An iterative optimization algorithm (such as a genetic algorithm or a particle swarm optimization) is used to optimize the global adjustment parameters. Through multiple iterations, the best combination of air conditioner settings is found to reach the global optimal state. The air conditioner settings of the meeting room and the corridor are adjusted through the algorithm, so that the overall energy consumption is reduced by 5% while the user satisfaction is improved. During the optimization process, the results of each iteration are recorded, including the temperature, energy consumption and user feedback of each region. According to the evaluation results, the optimization strategy is further adjusted to achieve the final goal. The energy consumption reduction of the meeting room after optimization is recorded in the format of "region, energy consumption before optimization, energy consumption after optimization", such as "meeting room, energy consumption 100W, energy consumption after optimization 80W".

[0048] In this embodiment, refer to Figure 2, which is a schematic diagram of the detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include:

[0049] Performing real-time environmental perception based on multiple sensor points to obtain environmental monitoring parameters of multiple regions;

[0050] Identifying the temperature and humidity change characteristics according to the environmental monitoring parameters;

[0051] Performing sequential temperature and humidity change discrete fitting for each region on the temperature and humidity change characteristics to construct temperature and humidity change curves for different regions;

[0052] Performing multi-region temperature and humidity spatial distribution calculation on the environmental monitoring parameters to obtain the regional temperature and humidity spatial distribution characteristics;

[0053] Based on the regional temperature and humidity spatial distribution characteristics, performing dynamic mapping modeling on the temperature and humidity change curves to construct a multi-region temperature and humidity fluctuation distribution map.

[0054] In this embodiment, high-precision temperature and humidity sensors are arranged in multiple key areas (such as meeting rooms, offices, laboratories, etc.). Each sensor should have the functions of data acquisition and wireless transmission to achieve real-time monitoring and data sharing. Two sensors are selected to be arranged in each area, one close to the window to monitor external impacts, and the other close to the air conditioner outlet to monitor the internal air quality. Configure the sensors to collect temperature and humidity data once per minute and upload the data to the central control system in real time. Ensure the integrity and accuracy of the data for subsequent analysis. The format of the recorded data is "timestamp, area, temperature (°C), humidity (%)", such as "2023-04-01 10:00:00, meeting room, 22.5°C, 45%". Preprocess the collected real-time data to eliminate outliers and noise. Statistical methods can be used to identify and eliminate unreasonable extreme values to ensure the quality of the data. If the temperature in the meeting room suddenly increases to 30°C and the humidity drops to 10% during a certain data collection, it can be identified as an outlier and eliminated. By analyzing the temperature and humidity data of each area, identify the change characteristics. This can be achieved by calculating the mean, standard deviation, and change rate of the data to help identify obvious temperature and humidity change trends. If the temperature in the meeting room rises from 22°C to 24°C and the humidity drops from 45% to 40% within one hour, it is recorded as "temperature rises 2°C, humidity drops 5%". Record the identified temperature and humidity change characteristics in the database to ensure that the change characteristics of each area are traceable. This will provide basic data for subsequent fitting analysis. The recording format is "area, characteristic type, change amplitude", such as "meeting room, temperature change, +2°C; humidity change, -5%". Discretize the temperature and humidity data of each area to facilitate the construction of time series change curves. The data can be averaged by setting a time window (such as five minutes). Calculate the average of the temperature and humidity values of the meeting room every five minutes and record it as "time period, average temperature, average humidity", such as "10:00-10:05, temperature 22.5°C, humidity 45%". Use methods such as polynomial fitting or spline curves to fit the discretized temperature and humidity data to construct the temperature and humidity change curves of each area. Ensure that the fitting curve can accurately reflect the temperature and humidity change trend. If the equation obtained after the cubic polynomial fitting of the temperature and humidity data in the meeting room is T(t) = 0.1t3 - 0.5t2 + 22, it is recorded as "meeting room temperature fitting equation". Record the results of the fitting curve in the database and generate a visual chart for easy understanding and analysis. Ensure that the temperature and humidity change curves of each area are clearly visible. Generate a temperature and humidity change diagram of the meeting room, marking the fitting curve and actual data points. Integrate the temperature and humidity data of multiple areas to form a comprehensive database. This database should contain the real-time temperature and humidity information of all areas to facilitate spatial distribution calculations. Integrate the data of the meeting room, office, and laboratory to form a data set in the format of "area, time, temperature, humidity".Select a suitable spatial distribution calculation method, such as Kriging interpolation or inverse distance weighting method, to calculate the spatial distribution characteristics of temperature and humidity in different regions. These methods can estimate the temperature and humidity values in unknown regions based on known points. Use the inverse distance weighting method to calculate the weighted average temperature of adjacent regions to ensure that the temperature and humidity distribution in each region can be accurately estimated. Record the calculated spatial distribution characteristics of regional temperature and humidity in the database and generate a spatial distribution map. Ensure that the map can clearly display the temperature and humidity distribution in different regions. Generate a heat map to mark the temperature and humidity distribution in each region for easy observation and analysis. Based on the spatial distribution characteristics of regional temperature and humidity, select a suitable dynamic mapping modeling method, such as time series analysis or dynamic system modeling. This will help predict future temperature and humidity change trends. Use the time series analysis method to predict the temperature and humidity changes in each region within the next hour. Based on the collected temperature and humidity data and spatial distribution characteristics, establish a dynamic mapping model and verify it. Compare with the actual data and adjust the model parameters to improve the prediction accuracy. If the prediction model performs well in the actual temperature and humidity changes, it can be recorded as "The dynamic mapping model has high accuracy and conforms to the actual situation". Generate a fluctuation distribution map of the dynamic mapping results to show the temperature and humidity fluctuation characteristics in different regions. Ensure that the map can intuitively reflect the temperature and humidity change trends in each region. Generate a fluctuation distribution map to show the temperature and humidity fluctuations in each region and mark the important change points for easy analysis and decision-making.

[0055] In this embodiment, refer to Figure 3 , which is a schematic diagram of the detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include:

[0056] Collect monitoring images of each region based on the high-definition camera;

[0057] Perform global brightness optimization on the monitoring images to generate globally brightness-optimized monitoring images;

[0058] Perform dynamic image blur elimination on the globally brightness-optimized monitoring images to generate dynamically blur-optimized images;

[0059] Perform depth vision detection on the dynamically blur-optimized images to mark the people in different regions;

[0060] Predict the cross-regional personnel flow behavior of the people in different regions to generate cross-regional personnel behavior prediction data.

[0061] In this embodiment, high-definition cameras are arranged in each key area (such as meeting rooms, corridors, offices, etc.) to ensure full coverage of the monitoring area. Each camera should have high definition and night vision capabilities to adapt to different environmental lighting conditions. Select to install a high-definition camera with 12 million pixels in the meeting room and ensure that its viewing angle can cover the entire room, recorded as "The installation of the meeting room camera is completed, 2023-04-01". Configure the camera to collect multiple frames of images per second and transmit the image data to the central processing system in real time. Ensure the high quality and stability of the image data for subsequent processing. Set the camera to collect images at a speed of 30 frames per second, and the recording format is "timestamp, area, image file name", such as "2023-04-01 10:00:00, meeting room, image_001.jpg". Monitor the environment of each area in real time to ensure the integrity of the image data. Store the collected image data on a secure server and perform regular backups to ensure that the data is not lost. Record the image storage situation of each area to ensure that all image files have been successfully uploaded to the server. Perform global brightness analysis on each monitored image to evaluate the brightness distribution and contrast of the image. By calculating the average brightness value and standard deviation of the image, identify the images that need to be optimized. If the average brightness of a certain image is 50 (in the range of 0-255), and the standard deviation is 15, brightness adjustment may be required. Use image processing algorithms (such as histogram equalization) to optimize the global brightness of low-brightness images. This process can improve the contrast of the image, making the details in the image more clearly visible. Apply histogram equalization to process an image of the meeting room, and record it as "The average brightness of the optimized image is 100" after processing. Store the optimized image and record the key data during the processing. Generate a comparison chart to show the effects before and after optimization for subsequent analysis. The recording format is "area, brightness before processing, brightness after processing", such as "meeting room, 50 before processing, 100 after processing". Perform dynamic blur detection on the images after global brightness optimization to identify the blurred areas caused by movement. Edge detection algorithms (such as the Canny algorithm) can be used to evaluate the clarity of the image. If it is detected that the edge information in a certain image is blurred, deblurring processing is required. Use deblurring algorithms (such as Wiener filtering or blind deblurring) to process the blurred image to restore clarity. This process should ensure reducing the impact caused by motion blur. Apply Wiener filtering to process a blurred image, and record it as "The clarity of the image after dynamic blur optimization is improved" after processing. Store the image after dynamic blur optimization and record the processing results. Generate a comparison chart to show the effects before and after deblurring for subsequent analysis and verification. The recording format is "area, clarity before deblurring, clarity after deblurring", such as "meeting room, clarity before deblurring 0.3, clarity after deblurring 0.8". Select a suitable deep learning model (such as YOLO or Mask R-CNN) for person detection.The model should be trained to identify people in different areas and be able to accurately mark them in real-time images. Use the trained YOLOv5 model to ensure its high accuracy and real-time performance, and record it as "Person detection model loaded successfully". Input the image optimized for dynamic blur into the deep vision detection model for person recognition and marking. The detection results should show the number of people and their location information in each area. Three people were detected in the conference room image, and it was recorded as "Conference room, 3 people detected, location marking completed". Store the detection results and generate a visualization image to show the distribution of people in each area. Ensure that the markings are clear for subsequent analysis and decision-making. The recording format is "Area, number of detected people, marking status", such as "Conference room, 3 people, marking completed". According to the deep vision detection results, organize the personnel flow data in different areas. Record the entry and exit status and time of each person in each area for analyzing the flow behavior. If 2 people leave and 1 person enters the conference room, it should be recorded as "Conference room, timestamp, number of people leaving 2, number of people entering 1". Based on the flow data, select a suitable prediction model (such as time series analysis or Markov model) to predict the cross-regional personnel flow behavior. The model should be able to identify the trends and patterns of personnel flow. Build a time series model based on historical flow data to predict the personnel flow in the conference room and office in the next 30 minutes. Record the prediction results in the database to ensure that the personnel flow prediction for each area is traceable. Generate a report to show the cross-regional personnel behavior prediction data. The recording format is "Area, prediction time, expected number of people leaving, expected number of people entering", such as "Conference room, expected number of people leaving 2 and entering 1 after 30 minutes".

[0062] In this embodiment, the specific steps for predicting the cross-regional personnel flow behavior of people in different areas to generate cross-regional personnel behavior prediction data are as follows:

[0063] Perform image frame cutting on the people in different areas to extract multiple person image frames;

[0064] Perform dynamic optical flow movement tracking on the person image frames to generate the temporal movement trajectories of each person;

[0065] Calculate the movement frequency, direction, and residence time of the people in the temporal movement trajectories;

[0066] Perform cross-regional trajectory recognition on the temporal movement trajectories to calculate the cross-regional trajectory frequency;

[0067] Perform regional flow topology structure analysis based on the cross-regional trajectory frequency, person movement frequency, direction, and residence time to generate a regional flow topology network;

[0068] Calculate the probability of personnel cross - regional transfer for the regional flow topology network to obtain the personnel transfer probability between regions;

[0069] Conduct in - depth analysis of cross - regional flow based on the personnel transfer probability between regions, and construct the cross - regional personnel flow pattern;

[0070] Predict the personnel flow behavior based on the cross - regional personnel flow pattern to generate cross - regional personnel behavior prediction data.

[0071] In this embodiment, the marked monitoring images are obtained from the depth vision detection system, and these images already contain the position information of the personnel within the area. Ensure that the images are clear and the personnel markings are accurate for cutting processing. Record a monitoring image of a meeting room, mark the positions of 3 personnel, and record it as "meeting room, image file name, number of personnel 3". Use computer vision technology (such as OpenCV) to cut the image frames of each marked personnel. According to the bounding box coordinates output by the detection model, extract each individual personnel image. If the detected bounding box coordinates of a personnel are (x1, y1, x2, y2), then the corresponding personnel image is generated through the cutting operation and recorded as "cut successfully, personnel image number 1". Store the cut personnel images in the database and record the relevant information for subsequent processing. Ensure that the storage path and number of each personnel image are clearly visible. The recording format is "personnel number, cut image file name", such as "personnel 1, person_image_1.jpg". Select a suitable optical flow algorithm (such as the Lucas-Kanade method) for dynamic tracking of the personnel images. Ensure that the algorithm can effectively process the motion information of the images and extract the movement trajectories of each personnel. Set parameters to adapt to the movement speed of the personnel in the meeting room to ensure the tracking accuracy. Apply the dynamic optical flow algorithm to each cut personnel image, calculate the position of each personnel at each moment in real time, and generate its temporal movement trajectory. Record the position information at each time point. If the movement trajectory points of a personnel within the monitoring time are [(x1, y1), (x2, y2),...], then the generated trajectory is recorded as "personnel 1, number of movement trajectory points 5". Store the generated temporal movement trajectory in the database and generate a visualization chart to intuitively display the movement trajectory of each personnel. The recording format is "personnel number, trajectory points", such as "personnel 1, trajectory points [(x1, y1), (x2, y2)]". Select a suitable calculation method to evaluate the movement frequency, direction, and stay time of each employee during the monitoring period. This can be achieved by analyzing the time interval and position change of the trajectory points. Set a time window (such as every minute) to count the movement of the personnel. According to the temporal movement trajectory, calculate the movement frequency (such as the number of movements per minute), movement direction (calculated based on the trajectory change), and stay time (the time at the same position) of each personnel within the specified time. If a personnel stays in the meeting room for 5 minutes, the movement frequency is 2 times, and the direction is southeast, then record it as "personnel 1, stay time 5 minutes, movement frequency 2 times, direction southeast". Record the calculation results in the database for subsequent analysis. Ensure that the movement behavior data of each employee is traceable. The recording format is "personnel number, movement frequency, direction, stay time", such as "personnel 1, frequency 2, direction southeast, stay time 5 minutes". Define the boundaries of each monitoring area to facilitate identifying whether personnel move across regions. The boundary lines of each region can be marked using a coordinate system.Record the boundary coordinates of the meeting room and the corridor for subsequent analysis of the movement trajectories. Analyze the movement trajectories of each person to identify cross-regional movement behaviors. If the trajectory points change from the coordinates of one region to another, it is determined as cross-regional movement. If a person moves from the meeting room to the corridor, record it as "Person 1, cross-regional movement, starting point meeting room, ending point corridor". Statistically analyze the cross-regional trajectory frequencies between each region and record the number of cross-regional movements of people within each region. If Person 1 makes 3 cross-regional movements during the monitoring period, record it as "Person 1, cross-regional movement frequency 3 times". Based on the cross-regional movement data, construct the topological structure of regional flows. Each region is regarded as a node, and cross-regional movement is regarded as an edge connecting these nodes. The flow relationship between the meeting room and the corridor can be defined as an edge and recorded as "meeting room - corridor". Use network analysis tools (such as Gephi or NetworkX) to construct the regional flow topological network to ensure that each node and edge can accurately reflect the flow relationship between regions. The constructed topological network includes the meeting room, office, and corridor, and record it as "Topological network construction completed". Visualize the constructed regional flow topological network to ensure that the flow relationship is clearly visible for subsequent analysis and decision-making. Generate a regional flow topological map to indicate the flow relationship and frequency between each region. Select a suitable probability calculation method to evaluate the transfer probability of people between different regions. A Markov chain model can be used for modeling. Set the state transition matrix and record the transfer frequencies between each region. Calculate the transfer probability between each region based on the cross-regional movement frequency data to ensure that the calculation results can reflect the real situation of people's transfer behaviors. If the transfer frequency from the meeting room to the corridor is 3 times and the total number of transfers is 10 times, the transfer probability is 0.3. Record the calculated transfer probability in the database and generate a summary report for subsequent analysis. The record format is "Region A to Region B, transfer probability", such as "Meeting room to corridor, transfer probability 0.3". Based on the transfer probability data, conduct in-depth analysis of the cross-regional personnel flow patterns. Identify the key patterns and trends of personnel flow. If it is found that people are more inclined to move from the meeting room to the corridor, record it as "Personnel flow pattern: high flow frequency from meeting room to corridor". Record the analysis results in the database for subsequent applications. Ensure that the analysis results of each flow pattern are traceable. The record format is "Flow pattern, description", such as "Flow pattern, high flow frequency from meeting room to corridor". Based on the identified flow patterns, construct a prediction model for personnel movement behaviors. Time series prediction or machine learning methods can be used for modeling. Use historical flow data to construct an ARIMA model and record it as "Personnel flow prediction model construction completed". Apply the constructed prediction model to predict the personnel flow situation in the future for a certain period. Ensure that the prediction results can reflect the real dynamics of personnel flow. Record the predicted personnel flow situation in the meeting room and the corridor within the next 30 minutes and record it as "Prediction result: It is predicted that 5 people will flow into the meeting room and 3 people will flow out".Record the prediction results in the database and generate a summary report for subsequent decision-making and optimization. Ensure the accuracy and traceability of the prediction data. The recording format is "area, expected number of inflows, expected number of outflows", such as "meeting room, expected 5 inflows, 3 outflows".

[0072] In this embodiment, refer to Figure 4 , which is a schematic diagram of the detailed implementation steps of step S3. In this embodiment, the detailed implementation steps of step S3 include:

[0073] Conduct cross-regional air heat circulation analysis based on cross-regional personnel behavior prediction data to generate a cross-regional air heat circulation effect;

[0074] Perform heat circulation dynamics evolution on the cross-regional air heat circulation effect to generate regional heat circulation dynamics logic;

[0075] Conduct real-time heat flow trend analysis on the multi-regional temperature and humidity fluctuation distribution map to generate a real-time heat flow trend;

[0076] Mine the local heat flow distribution of different regions based on the regional heat circulation dynamics logic for the real-time heat flow trend to obtain the local heat flow distribution characteristics of different regions;

[0077] Speculate on the heat flow distribution at multiple time points for the local heat flow distribution characteristics to construct a heat flow distribution prediction map.

[0078] In this embodiment, cross-regional personnel behavior prediction data is collected, including the personnel flow, temperature and humidity changes, and heat distribution in each region. These data should cover different time periods and working conditions for comprehensive analysis. Record the personnel flow data in meeting rooms, offices, and corridors in the format of "region, number of people flowing, time period", such as "meeting room, 5 people flowing in, 3 people flowing out". Calculate the cross-regional air heat circulation effect based on the personnel flow data and the temperature and humidity distribution within the region. The heat circulation equation can be used, considering the influencing factors of heat transfer, such as temperature difference, flow rate, etc. If the temperature in the meeting room is 24°C and the temperature in the corridor is 22°C, the heat circulation effect caused by personnel flow can be calculated as Q = k * A * (T1 - T2), where Q is the heat flow, k is the heat conduction coefficient, and A is the contact area. Record the calculated cross-regional air heat circulation effect in the database and generate a visualization chart to intuitively display the heat circulation situation between different regions. Generate a heat circulation effect diagram to mark the heat transfer situation between regions. Based on the cross-regional air heat circulation effect, construct a heat circulation dynamics model. This model should be able to describe the time evolution and spatial distribution of heat circulation, considering the storage and dissipation of heat. Use the heat transfer equation and the principle of energy conservation to construct a time-space heat circulation dynamics model. Use numerical simulation methods (such as the finite element method or the finite difference method) to solve the heat circulation dynamics model and simulate the heat circulation evolution process under different conditions. Set the initial conditions as the temperature distribution in each region in the simulation and record the evolution of heat circulation at different time points. Record the simulation results in the database and generate a heat circulation dynamics logic diagram for more in-depth analysis and understanding. Generate a heat flow evolution diagram to show the heat circulation state and trend at different time periods. Obtain the temperature and humidity fluctuation distribution maps of multiple regions from the real-time monitoring system. These maps should be able to reflect the heat distribution within each region for trend analysis. Record the real-time temperature and humidity distribution in meeting rooms, offices, and corridors in the format of "region, temperature, humidity". Analyze the real-time heat flow trend in each region based on the obtained temperature and humidity fluctuation distribution maps. The heat flow trend can be evaluated by calculating the temperature change rate and the heat change amount. If the temperature in the meeting room rises by 2°C within 1 hour while the temperature in the corridor remains unchanged, the heat flow trend towards the corridor can be inferred. Record the analysis results in the database and generate a report to show the real-time heat flow trend. Ensure that the heat flow situation in each region is clearly visible. The recording format is "region, heat flow trend", such as "meeting room, heat flowing towards the corridor". Define the characteristics of local heat flow according to the heat circulation dynamics logic. Consider the influence of local temperature and humidity changes and personnel flow on heat distribution. Define the local flow characteristics as "temperature change amplitude, humidity change amplitude, and personnel flow frequency". Mine the local heat flow distribution for each region and analyze the heat flow characteristics within different regions. In-depth analysis can be carried out through heat distribution maps and flow trajectories.Analyze the heat distribution in the meeting room and find that the temperature in some areas rises due to the gathering of people. Record the characteristics of local heat flow in the database and generate a visualization graph to show the heat flow distribution in different areas. Generate a local heat distribution map to show the areas with temperature rise and the corresponding personnel distribution in the meeting room. Collect heat flow data at different time points for multi-time point distribution speculation. Ensure the integrity of the time series of the data for analysis. Record the temperature and humidity data of each area at different time points in the format of "time, area, temperature, humidity". Based on the multi-time point data, construct a heat flow distribution prediction model. Time series analysis methods can be used for modeling to predict the future heat distribution. Use the ARIMA model to fit the historical heat flow data and predict the future heat flow trend. Record the prediction results in the database and generate a heat flow distribution prediction graph for subsequent decision support and optimization. Generate a heat flow distribution prediction graph for the next 3 hours to show the expected temperature changes and heat flow in each area.

[0079] In this embodiment, step S4 includes the following steps:

[0080] Obtain the historical monitoring log of the air conditioner; calculate the air conditioner usage frequency for the historical monitoring log of the air conditioner and extract the air conditioner usage frequencies of multiple areas;

[0081] Calculate the activity frequencies, time periods, and number of people in different stages of the historical monitoring log of the air conditioner;

[0082] Based on the air conditioner usage frequencies, activity frequencies, time periods, and number of people in different stages, conduct user behavior pattern analysis to obtain the user behavior patterns in different areas;

[0083] Extract the time-series air conditioner operation parameters according to the historical monitoring log of the air conditioner;

[0084] Conduct multi-period personalized usage habit evolution analysis on the user behavior patterns in different areas according to the time-series air conditioner operation parameters, so as to obtain the air conditioner usage habits in different areas.

[0085] In this embodiment, historical monitoring logs are obtained from the air-conditioning control system or monitoring devices. These logs should include information such as air-conditioning usage records, temperature settings, operating hours, etc. for each area. Ensure the integrity and accuracy of the log data for subsequent analysis. The record format is "timestamp, area, operating status, set temperature", e.g., "2023-04-01 10:00:00, meeting room, on, 22°C". Sort out the obtained monitoring logs, arrange them in chronological order, and remove duplicate or invalid records. Ensure the clarity and consistency of the data before analysis. Remove duplicate air-conditioning on records within the same time period to ensure that only one valid record is retained for each time point. Store the sorted historical monitoring logs in a database for subsequent querying and analysis. Ensure that each record has a clear identifier for subsequent use. The record format is "area, monitoring log file name, number of records", e.g., "meeting room, ac_log_2023.csv, number of records 500". Determine the calculation method for the air-conditioning usage frequency. It can be defined as the number of times the air-conditioning is turned on within a specified time period, or the number of people using the air-conditioning in a specific area. Set the usage frequency as the number of times the air-conditioning is turned on per hour. Calculate the air-conditioning usage frequency for each area by analyzing the historical monitoring logs. Statistical analysis methods (such as the counting method) can be used to count the usage situations in different time periods for each area. If the meeting room is turned on 3 times within one hour, record it as "meeting room, usage frequency 3 times / hour". Record the calculated usage frequency in the database and generate a visualization chart to intuitively display the air-conditioning usage situations in different areas. Generate a bar chart to show the air-conditioning usage frequencies in each area for easy comparison and analysis. Define different activity phases according to the air-conditioning usage situation (such as working hours, rest hours, meeting hours, etc.). Ensure that the definition of each phase is reasonable for frequency and number statistics. Define working hours as 9:00 - 18:00 from Monday to Friday every week. Statistically analyze the air-conditioning usage frequency, time period, and number of users within each activity phase. Analyze the historical monitoring logs and record the air-conditioning usage situations within each phase. If the number of people using the air-conditioning in the meeting room during the meeting time period (e.g., 10:00 - 11:00) is 10, record it as "meeting room, number of users during meeting time period 10". Record the statistical results in the database to ensure that the activity frequency, time period, and number of people in each phase are traceable. Generate a summary report for subsequent analysis. The record format is "area, activity phase, usage frequency, time period, number of people", e.g., "meeting room, meeting time period, usage frequency 5, number of people 10". Define the key features of user behavior patterns, such as usage frequency, activity phase, and number of people. Select a suitable analysis method (such as clustering analysis or association rule mining) to identify user behavior patterns in different areas. Select the K-means clustering algorithm to perform clustering analysis on user behavior. Use historical data to analyze user behavior in different areas. Identify different user groups and their air-conditioning usage habits through clustering analysis.If the analysis result shows that the user behavior pattern in the meeting room is "high-frequency use, concentrated in the meeting time period", it is recorded as "Meeting room user behavior pattern: high-frequency, meeting concentrated". Record the behavior pattern analysis results in the database to ensure that the user behavior patterns in each area are traceable. Generate a report to show the user behavior patterns in different areas. Define the time-series air-conditioning operation parameters according to the historical monitoring logs, including operation time, set temperature, actual temperature, operation status, etc. Extract these parameters for subsequent analysis. Extract the set temperature and actual temperature records of the meeting room within a certain time period. Organize the extracted time-series operation parameters into a table format to ensure the completeness of the records at each time point. Store them in the database for subsequent analysis. The record format is "timestamp, area, set temperature, actual temperature", such as "2023-04-01 10:00:00, meeting room, 22°C, 23°C". Record the extracted time-series air-conditioning operation parameters in the database to ensure the integrity of the data. Generate a visualization chart to show the air-conditioning operation conditions in different time periods. Generate a line chart to show the changes in the set temperature and actual temperature of the meeting room. Define the characteristics of personalized usage habits based on the extracted time-series operation parameters, such as set temperature preference, operation time preference, etc. These characteristics will be used to analyze the evolution of users' usage habits. Define the set temperature preference as the average set temperature of the user in different time periods. Use the time-series data to analyze the evolution of users' usage habits in different time periods. Time series analysis methods can be used to observe the changing trends of the set temperature and actual temperature. If it is found that the average set temperature of the user gradually decreases in summer, it is recorded as "The set temperature preference gradually decreases in summer". Record the results of the habit evolution analysis in the database to ensure that the evolution of usage habits in each area is traceable. Generate a report to show the air-conditioning usage habits in different areas. The record format is "area, usage habit description", such as "meeting room, the set temperature preference gradually decreases".

[0086] In this embodiment, step S5 includes the following steps:

[0087] Perform detail enlargement processing on multiple personnel image frames and optimize the resolution to obtain resolution-optimized image frames;

[0088] Perform in-depth semantic recognition of the personnel's clothing situation on the resolution-optimized image frames to generate the real-time clothing situation of each person;

[0089] Perform personalized somatosensory temperature sensitivity calculation on the real-time clothing situation to generate the somatosensory temperature sensitivity of each person;

[0090] Perform in-depth analysis of the regional comprehensive temperature requirements on the somatosensory temperature sensitivity according to the air-conditioning usage habits in different regions to generate the personalized temperature requirement characteristics of users in each region.

[0091] In this embodiment, the labeled person image frames are obtained from the depth vision detection system. Each image frame should contain the complete image information of the person, ensuring that the image is clearly visible. Record the person image frames in the meeting room, labeled as "Meeting room, Person Image Frame No. 1". Use image processing algorithms (such as bilinear interpolation or nearest neighbor interpolation) to perform detail expansion processing on each image frame. This processing can enhance the details in the image, making the clothing situation of the person more obvious. Expand the size of each image frame to 1.5 times the original, and record it as "Image Frame No. 1, Detail Expansion Processing Completed". Optimize the resolution of the expanded image frame to improve the image quality. Super-resolution reconstruction techniques (such as SRCNN or ESPCN) can be used to enhance the clarity of the image. The resolution of the original image is 640x480, and it is increased to 1280x960 after super-resolution processing, and recorded as "Image Frame No. 1, Resolution Optimization Completed". Select a suitable deep learning model (such as Mask R-CNN or YOLO) for deep semantic recognition of the clothing situation of the person. Ensure that the model is trained to accurately identify different types of clothing and accessories. Load the trained YOLOv5 model to ensure its high accuracy and real-time performance, and record it as "Clothing Situation Recognition Model Loaded Completed". Input the image frame with optimized resolution into the deep semantic recognition model for automatic recognition of the clothing situation of the person. The recognition results should include information such as the clothing type and color of each person. If it is recognized that a person is wearing a blue shirt and black pants, then record it as "Person Image Frame No. 1, Clothing Situation: Blue Shirt, Black Pants". Store the recognition results in the database and generate a visual display. Ensure that the output information is clear for subsequent analysis and use. The record format is "Person Number, Clothing Situation Description", such as "Person 1, Clothing Situation: Blue Shirt, Black Pants". Define the calculation method of the body temperature sensitivity according to the clothing situation of the person. The body temperature can be affected by various factors, including clothing type, environmental temperature, and humidity, etc. A body temperature model (such as a thermal comfort model) can be used to calculate the body temperature of each person. Calculate the personalized body temperature sensitivity according to parameters such as the clothing situation, environmental temperature, and humidity of the person. Use a formula (such as the PMV model) to evaluate the body temperature. If the clothing type of the person is "thin shirt", the external environmental temperature is 25°C, and the humidity is 60%, then calculate its body temperature sensitivity as "moderately sensitive". Record the calculated body temperature sensitivity in the database to ensure that the sensitivity data of each person is traceable. The record format is "Person Number, Body Temperature Sensitivity", such as "Person 1, Moderately Sensitive". Define the analysis method of the comprehensive temperature requirement for the area, considering factors such as the body temperature sensitivity of the person, the air conditioner usage frequency, and the historical temperature data in the area, etc. Set the comprehensive temperature requirement as "the weighted average of the person sensitivity and the area average temperature".Analyze the user's personalized temperature demand characteristics of each area according to the recorded body temperature sensitivity and the air conditioning usage habits of the area. Statistical analysis methods (such as the weighted average method) can be used for calculation. If there are 10 users in the meeting room and the sensitivities are "moderately sensitive" and "low sensitive" respectively, the comprehensive temperature demand of the meeting room is calculated as "24°C" according to the weights. Record the analysis results of the comprehensive temperature demand in the database and generate a visualization chart for easy understanding and analysis. Generate a pie chart to show the comprehensive temperature demand characteristics of each area for subsequent optimization of the air conditioning settings.

[0092] In this embodiment, step S6 includes the following steps:

[0093] Calculate the real-time air conditioning parameters of multiple current areas;

[0094] Calculate the optimal air conditioning parameters for each area one by one according to the heat flow distribution prediction map and the user's personalized temperature demand characteristics of each area, and generate the optimal air conditioning parameters for each area;

[0095] Based on the real-time air conditioning parameters of the multiple areas, calculate the parameter deviation corresponding to the area for the optimal air conditioning parameter area, and generate the dynamic air conditioning parameter deviation for each area;

[0096] Based on the dynamic air conditioning parameter deviation, perform adaptive local air conditioning parameter adjustment and global coordinated iterative control optimization to construct an adaptive intelligent air conditioning control model.

[0097] In this embodiment, the operating parameters of the air conditioners in each area are obtained in real time from the air conditioner monitoring system. These parameters include the current temperature, humidity, set temperature, wind speed, and operating status, etc., to ensure a comprehensive understanding of the status of the air conditioning equipment. The real-time parameters of the meeting room are recorded as "Area: Meeting Room, Current Temperature: 22°C, Humidity: 45%, Set Temperature: 20°C, Wind Speed: High". The collected real-time air conditioner parameters are sorted out and stored in the database for subsequent calculation and analysis. The data should be classified according to the area to ensure accurate information and easy query. The recording format is "Area, Current Temperature, Humidity, Set Temperature, Wind Speed", such as "Meeting Room, 22°C, 45%, 20°C, High". The real-time air conditioner parameters of multiple current areas are recorded in the database to ensure that the status of each area is traceable. A real-time report is generated to display the operating status of the air conditioners in each area. The generated report shows that "the current temperature in the meeting room is 22°C, the set temperature is 20°C, and adjustment is required". Based on the heat flow distribution prediction map, the heat flow conditions in each area are analyzed. Combining with the user's personalized temperature demand characteristics, adaptive air conditioner parameters are formulated. If the heat flow prediction map shows that the temperature in the meeting room is relatively high and the user's demand is 22°C, then the optimal set temperature may be 21°C. The optimal air conditioner parameters for each area are calculated, including the set temperature, operating mode (such as cooling or heating), wind speed, etc. An optimization algorithm (such as genetic algorithm or particle swarm optimization) is used to ensure that the parameters of each area achieve the best effect. The goal is set as "adjust the temperature in the meeting room to 21°C and maintain the humidity at 45%", and the required wind speed and operating time are calculated. The calculated optimal air conditioner parameters are recorded in the database, and a visualization chart is generated for subsequent operation and management. A chart is generated to show the optimal parameter settings for each area, including the set temperature, wind speed, and operating mode. A suitable calculation method is selected to evaluate the deviation between the current real-time air conditioner parameters and the optimal air conditioner parameters. A simple difference calculation method can be used. The deviation is set as "Current Parameter - Optimal Parameter". The deviation between the real-time air conditioner parameters and the optimal parameters of each area is calculated to generate dynamic air conditioner parameter deviation data. This data will be used for subsequent adjustment and optimization. If the current temperature in the meeting room is 22°C and the optimal setting is 21°C, the deviation is "+1°C", and it is recorded as "Dynamic Air Conditioner Parameter Deviation in Meeting Room: +1°C". The dynamic air conditioner parameter deviation is recorded in the database to ensure that the deviation data of each area is traceable. A deviation analysis report is generated for subsequent adjustment. The recording format is "Area, Dynamic Deviation", such as "Meeting Room, +1°C". According to the dynamic air conditioner parameter deviation, an adaptive adjustment strategy is defined. Considering the magnitude and direction of the deviation, corresponding adjustment measures (such as increasing or decreasing the set temperature) are formulated. If the deviation is +1°C, the set temperature can be considered to be adjusted to 20°C. Local parameter adjustment is performed on the air conditioners in each area. The above adjustments are achieved through the control system to ensure that the air conditioners can quickly respond to real-time demands. The set temperature of the meeting room is adjusted from 22°C to 20°C, and the adjustment process is recorded.After local adjustment, based on the feedback information of each area, global coordination control is carried out. The air-conditioning parameters of multiple areas are optimized as a whole through an iterative optimization algorithm (such as model predictive control). If the temperature in the meeting room still does not reach the target value after adjustment, continue iterative optimization and adjust the air-conditioning parameters of other areas. An adaptive intelligent air-conditioning control model is constructed, integrating real-time data, optimal parameters, dynamic deviation and adjustment strategies. Ensure that the model can respond to environmental changes and user needs in real time. The model should include input (real-time parameters, user needs), processing (optimization algorithm, adjustment strategy) and output (adjusted air-conditioning parameters). The control model is trained with historical data to ensure its effectiveness and accuracy in different scenarios. The cross-validation method is used to evaluate the model performance. The air-conditioning usage data of the past year is used for training to verify the adaptability of the model in different seasons. The constructed adaptive intelligent air-conditioning control model is recorded in the database, and a model performance report is generated for subsequent optimization and management. The recording format is "model name, performance index", such as "adaptive control model, accuracy 95%".

[0098] In the present invention, there is also provided an air-conditioning control system for performing the above-mentioned air-conditioning control method, including:

[0099] A temperature and humidity distribution module, configured to obtain environmental monitoring parameters of multiple areas, perform temperature and humidity change analysis and multi-area temperature and humidity spatial distribution calculation, and construct a multi-area temperature and humidity fluctuation distribution map;

[0100] A behavior prediction module, configured to collect monitoring images of each area based on the high-definition camera, and perform cross-area personnel flow behavior prediction to generate cross-area personnel behavior prediction data;

[0101] A flow distribution speculation module, configured to perform local heat flow distribution mining on the multi-area temperature and humidity fluctuation distribution map according to the cross-area personnel behavior prediction data, and perform heat flow distribution speculation to construct a heat flow distribution prediction map;

[0102] A usage habit module, configured to obtain the air-conditioning historical monitoring log, perform multi-period personalized usage habit evolution analysis, so as to obtain the air-conditioning usage habits of different areas;

[0103] A temperature demand module, configured to perform in-depth analysis of the regional comprehensive temperature demand according to the air-conditioning usage habits of different areas, and generate the user personalized temperature demand characteristics of each area;

[0104] A self-usage adjustment module, configured to perform adaptive local air-conditioning parameter adjustment and global coordination iterative control optimization according to the heat flow distribution prediction map and the user personalized temperature demand characteristics of each area, and construct an adaptive intelligent air-conditioning control model.

[0105] The present invention also provides an air conditioner, which includes a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps of the air conditioner control method described in any one of the above are implemented.

[0106] Therefore, from any perspective, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the application document are intended to be encompassed within the present invention.

[0107] Those skilled in the art will clearly understand that for the convenience and brevity of description, the specific working processes of the systems, systems, and units described above refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0108] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it is stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application essentially, or the part that contributes to the prior art, or all or part of the technical solution, is embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The foregoing storage medium includes: various media such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disc that store program codes.

[0109] As described above, these are only specific implementation manners of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but will conform to the broadest scope consistent with the principles and novel features invented herein.

Claims

1. An air conditioning control method, characterized in that: The air conditioner is provided with a high-definition camera, and the steps include: Step S1: Acquire environmental monitoring parameters of multiple regions, perform temperature and humidity change analysis and multi-region temperature and humidity spatial distribution calculation, and construct a multi-region temperature and humidity fluctuation distribution map; Step S2: Based on the high-definition camera, monitoring images of each area are collected, and cross-region personnel flow behavior prediction is performed to generate cross-region personnel behavior prediction data; Step S3: mining the local heat flow distribution of the multi-region temperature and humidity fluctuation distribution map based on the cross-region personnel behavior prediction data, and inferring the heat flow distribution to construct a heat flow distribution prediction map; Step S4: Obtain the historical monitoring log of the air conditioner, and perform a multi-period personalized usage habit evolution analysis to obtain the air conditioner usage habits in different regions; Step S5: Perform in-depth analysis of regional comprehensive temperature requirements based on air conditioning usage habits in different regions, and generate user personalized temperature requirement characteristics for each region; Step S6: Adaptive local air conditioning parameter adjustment and global coordinated iterative control optimization are performed according to the heat flow distribution prediction map and the personalized temperature demand characteristics of users in each area to build an adaptive intelligent air conditioning control model.

2. The air conditioning control method according to claim 1, characterized in that: The specific steps of step S1 are: Real-time environmental perception based on multiple sensor points to obtain environmental monitoring parameters of multiple areas; Identify temperature and humidity change characteristics according to the environmental monitoring parameters; Perform discrete fitting of the temperature and humidity change characteristics in each region to construct temperature and humidity change curves for different regions; Performing multi-region temperature and humidity spatial distribution calculation on the environmental monitoring parameters to obtain regional temperature and humidity spatial distribution characteristics; Based on the regional temperature and humidity spatial distribution characteristics, the temperature and humidity change curve is dynamically mapped and modeled to construct a multi-region temperature and humidity fluctuation distribution map.

3. The air conditioning control method according to claim 1, characterized in that: The specific steps of step S2 are: Collect monitoring images of each area based on the high-definition camera; Performing global brightness optimization on the monitoring image to generate a global brightness optimized monitoring image; Performing dynamic image blur elimination on the global brightness optimized monitoring image to generate a dynamic blur optimized image; Perform deep vision detection on motion blur optimized images to mark people in different areas; Predict the cross-regional personnel flow behavior of people in different regions to generate cross-regional personnel behavior prediction data.

4. The air conditioning control method according to claim 3, characterized in that: The specific steps of predicting the cross-regional personnel flow behavior of personnel in different regions to generate cross-regional personnel behavior prediction data are as follows: Performing image frame cutting processing on the persons in the different areas to extract multiple person image frames; Performing dynamic optical flow motion tracking on the person image frame to generate a temporal motion trajectory of each person; Calculate the movement frequency, direction and dwell time of personnel in the time series movement trajectory; Identify the cross-regional trajectories of people based on the time-series movement trajectory and calculate the cross-regional trajectory frequency; Performing a regional flow topology structure analysis based on the cross-region trajectory frequency, personnel movement frequency, direction, and residence time to generate a regional flow topology network; Calculate the probability of personnel transfer across regions on the regional flow topology network to obtain the probability of personnel transfer between regions; Based on the inter-regional personnel transfer probability, an in-depth analysis of inter-regional flow is conducted to construct the inter-regional personnel flow law; Based on the cross-regional personnel flow patterns, personnel flow behavior is predicted to generate cross-regional personnel behavior prediction data.

5. The air conditioning control method according to claim 1, characterized in that: The specific steps of step S3 are: Conduct cross-regional air heat circulation analysis based on cross-regional personnel behavior prediction data to generate cross-regional air heat circulation effects; Conduct thermal flow dynamics evolution on cross-regional air thermal flow effects and generate regional thermal flow dynamics logic; Conduct real-time heat flow trend analysis on multi-region temperature and humidity fluctuation distribution maps to generate real-time heat flow trends; According to the regional heat flow dynamics logic, the local heat flow distribution is mined based on the real-time heat flow trend to obtain the local heat flow distribution characteristics in different regions; The local heat flow distribution characteristics are used to infer the heat flow distribution at multiple time points, and a heat flow distribution prediction map is constructed.

6. The air conditioning control method according to claim 1, characterized in that: The specific steps of step S4 are: Obtaining a historical monitoring log of an air conditioner; calculating the air conditioner usage frequency of the historical monitoring log of the air conditioner, and extracting the air conditioner usage frequency of multiple areas; Calculate the activity frequency, time period and number of people at different stages of the air conditioner historical monitoring log; Performing user behavior pattern analysis based on the air conditioner usage frequency, activity frequency at different stages, time periods, and number of people to obtain user behavior patterns in different areas; Extracting time-series air conditioning operation parameters based on air conditioning historical monitoring logs; According to the time-series air-conditioning operation parameters, the user behavior patterns in different regions are analyzed for the evolution of personalized usage habits over multiple periods, thus obtaining the air-conditioning usage habits in different regions.

7. The air conditioning control method according to claim 1, characterized in that: The specific steps of step S5 are: Performing detail enlargement processing on multiple personnel image frames and optimizing resolution to obtain resolution-optimized image frames; Perform deep semantic recognition of people's clothing conditions on the resolution-optimized image frame to generate real-time clothing conditions for each person; Performing personalized body temperature sensitivity calculation on the real-time wearing situation, thereby generating body temperature sensitivity of each person; According to the air conditioning usage habits in different regions, the perceived temperature sensitivity is deeply analyzed for regional comprehensive temperature demand, and the user personalized temperature demand characteristics of each region are generated.

8. The air conditioning control method according to claim 1, characterized in that: The specific steps of step S6 are: Calculate the current real-time air conditioning parameters of multiple areas; Calculate the optimal air conditioning parameters for each area based on the heat flow distribution prediction map and the personalized temperature demand characteristics of users in each area, and generate the optimal air conditioning parameters for each area; Calculating the regional corresponding parameter deviation of the optimal air-conditioning parameter region based on the real-time air-conditioning parameters of the multiple regions, and generating a dynamic air-conditioning parameter deviation for each region; Based on the dynamic air-conditioning parameter deviation, adaptive local air-conditioning parameter adjustment is performed, and global coordinated iterative control optimization is performed to build an adaptive intelligent air-conditioning control model.

9. An air conditioning control system, characterized in that: Used to execute the air conditioning control method according to claim 1, comprising: The temperature and humidity distribution module is used to obtain environmental monitoring parameters of multiple regions, analyze temperature and humidity changes, calculate the spatial distribution of temperature and humidity in multiple regions, and construct a temperature and humidity fluctuation distribution map in multiple regions; A behavior prediction module is used to collect monitoring images of each area based on the high-definition camera and predict the cross-regional personnel flow behavior to generate cross-regional personnel behavior prediction data; The flow distribution inference module is used to mine the local heat flow distribution of the multi-region temperature and humidity fluctuation distribution map based on the cross-region personnel behavior prediction data, and to infer the heat flow distribution and construct a heat flow distribution prediction map; The usage habit module is used to obtain the historical monitoring logs of air conditioners and conduct multi-period personalized usage habit evolution analysis to obtain the air conditioner usage habits in different regions; The temperature demand module is used to conduct in-depth analysis of regional comprehensive temperature demand based on the air conditioning usage habits of different regions, and generate personalized temperature demand characteristics for users in each region; The self-adjustment module is used to perform adaptive local air-conditioning parameter adjustment and global coordinated iterative control optimization according to the heat flow distribution prediction map and the personalized temperature demand characteristics of users in each area, and to build an adaptive intelligent air-conditioning control model.

10. An air conditioner, comprising a memory and a processor, wherein a computer program is stored in the memory, characterized in that: When the processor executes the computer program, the steps of the air conditioning control method according to any one of claims 1 to 8 are implemented.

Images