An air conditioning adaptive comfort regulation system and method thereof

Abstract

The application provides a kind of air conditioner adaptive comfort adjustment method, and is applied to air conditioning system with fresh air regulation function, and the air conditioner adaptive comfort adjustment method includes: obtaining the operating parameters of the shunt control component;Obtain the environmental parameters of the working area of the terminal air outlet device;Obtain the working parameters of the working area of the terminal air outlet device, including normal function data, current number of users and user characteristics;Obtain the meteorological parameters of the working area of the terminal air outlet device;Meteorological parameters and working parameters are used as input parameters of the Bayesian comfort model fitted by deep learning algorithm, and the corresponding comfort parameters are calculated;The difference between the environmental parameters and the comfort parameters is calculated, and the operating state of the shunt control component is controlled based on the difference.The adjustment method obtains the comfort parameters based on the constructed Bayesian comfort model, realizes the autonomous adjustment of the air conditioner comfort operating state, improves the use comfort, and optimizes the use of energy.

Description

Technical Field

[0001] This invention relates to the field of air conditioning technology, and in particular to an adaptive comfort control system for air conditioning. Background Technology

[0002] Compared to split-type air conditioners, central air conditioning systems have a higher energy efficiency ratio, and their application scenarios have expanded from their initial use in large buildings to residential homes. Residential homes, compared to large buildings such as shopping malls, hotels, and commercial office buildings, have relatively smaller application areas, fewer and more fixed users, and, as the primary living area for family members, are one of the most direct reflections of quality of life. Traditional residential central air conditioning systems rely heavily on static settings and cannot respond in real time to changes in the external environment or the activities and characteristics of users, resulting in both resource waste and a negative impact on the user experience. Summary of the Invention

[0003] In view of the shortcomings of the prior art described above, the purpose of this invention is to provide an air conditioning adaptive comfort adjustment system, which can autonomously optimize and adjust the air conditioning operating parameters to achieve personalized comfort adjustment and energy consumption optimization.

[0004] To achieve the above and other related objectives, the present invention provides the following technical solution:

[0005] An adaptive comfort adjustment method for air conditioning, applied to an air conditioning system with fresh air conditioning function, includes: a terminal air outlet device for discharging air to the working area, a fan for drawing in and pressurizing outdoor air, a heat exchange component for heat exchange of the fresh air drawn in by the fan to regulate the temperature and humidity of the fresh air, and an air-cooled heat pump for supplying a cold or heat source medium to the heat exchange component; characterized in that it further includes a flow control component connected to each of the terminal air outlet devices via pipes, the adaptive comfort adjustment method for air conditioning including:

[0006] Obtain the operating parameters of the air outlet side of the diversion control component, including temperature, humidity, and wind speed;

[0007] Obtain environmental parameters of the working area on the air outlet side of the terminal air outlet device, including temperature, humidity, and wind speed;

[0008] Obtain the operating parameters of the terminal air outlet device's working area, including normal function data of the working area, the current number of users, and personnel characteristics;

[0009] Meteorological parameters of the working area of ​​the terminal air outlet device are obtained. The meteorological parameters are accessed and read from the meteorological database with the geographical location of the working area as the keyword. The meteorological parameters include temperature, humidity and wind speed.

[0010] Using the meteorological parameters and the working parameters as input parameters for the Bayesian comfort model fitted by the deep learning algorithm, the corresponding comfort parameters are calculated.

[0011] The difference between the environmental parameter and the comfort parameter is calculated, and the operating state of the diversion control component is controlled based on the difference. If the difference is within a preset threshold range, the diversion control component is kept in operation. If the difference is on either side of the preset threshold range, the operating state of the diversion control component is adjusted based on a correction value. The correction value is equal to the correction coefficient multiplied by the difference, and the correction coefficient is estimated by a random forest algorithm.

[0012] To achieve the above technical solution, the comfort parameters corresponding to the current operating parameters and meteorological parameters are calculated based on the Bayesian comfort model fitted by a deep learning algorithm. The difference between the environmental parameters and the comfort parameters is then calculated using these comfort parameters as a baseline. This difference is used as an input parameter to control the operation of the diversion control component, making the environmental parameters closer to the comfort parameters, thereby improving the user's comfort experience. When the difference between the calculated environmental parameters and the comfort parameters is within a preset threshold range, there is no difference in human comfort experience, and the operation of the diversion control component is maintained. When the calculated difference is on either side of the preset threshold range, i.e., significantly lower or significantly higher than the comfort range, the adjusted operating parameters of the diversion control component are obtained based on a correction value, so that the environmental parameters after operation are close to the comfort parameters. The correction value is obtained by multiplying the difference by a correction coefficient, where the correction coefficient is obtained from sample experimental data using a linear interpolation algorithm.

[0013] Another aspect of the present invention provides an air conditioning adaptive comfort adjustment system, comprising: a terminal air outlet device for discharging air to a work area; a fan for drawing in and pressurizing outdoor air; a heat exchange component for heat exchange of the fresh air drawn in by the fan to regulate the temperature and humidity of the fresh air; and an air-cooled heat pump for supplying a cold or heat source medium to the heat exchange component; characterized in that it further comprises a diversion control component connected to each of the terminal air outlet devices via pipes; an internal sensing module installed on the air outlet side of the diversion control component to collect the operating parameters of the diversion control component; an external sensing module installed on the air outlet side of the terminal air outlet device to collect environmental parameters of the work area; and a human body detection module for collecting working parameters of the work area; and a computing control center communicating and interacting with the internal sensing module, the external sensing module, and the human body detection module, wherein the operating parameters include: temperature, humidity, and wind speed; the environmental parameters include: temperature, humidity, and wind speed; and the working parameters include the normal working conditions of the work area. The system includes functional data and the number of current users and their characteristics collected by the human body detection module. The computing control center stores a Bayesian comfort model fitted using a deep learning algorithm. The computing control center communicates with a meteorological database and obtains meteorological parameters for the corresponding work area based on its geographical location. These meteorological parameters include temperature, humidity, and wind speed. The received meteorological parameters and work parameters are used as input parameters to the Bayesian comfort model to calculate comfort parameters. The difference between the comfort parameters and the environmental parameters is calculated using the comfort parameters as a baseline value, and this difference is input to the diversion control component to control its operating state. If the difference is within a preset threshold range, the diversion control component maintains its operating state. If the difference is on either side of the preset threshold range, the operating state of the diversion control component is adjusted based on a correction value, where the correction value is equal to a correction coefficient multiplied by the difference. The correction coefficient is estimated using a random forest algorithm.

[0014] To achieve the above technical solution, the internal sensing module collects parameters such as temperature, humidity, and wind speed from the diversion control component; the external sensing module collects parameters such as temperature, humidity, and wind speed from the work area; and the human detection module collects the number and characteristics of personnel in the work area. The internal sensing module, external sensing module, and human detection module respectively input the collected operating parameters, environmental parameters, and working parameters to the computing control center. The computing control center accesses meteorological data using the geographical location of the work area as a keyword and obtains the meteorological parameters for the corresponding area. The computing control center calculates the comfort parameters using the meteorological parameters and working parameters as input parameters for the Bayesian comfort model, and uses the comfort parameters as a benchmark value to calculate the difference between the environmental parameters and the comfort parameters. The difference is then input to the diversion control component, and this value is used as the control parameter for the operating state.

[0015] In one embodiment of the present invention, the diversion control component includes a plurality of air outlet pipes connected to the air outlet of the fan, and the plurality of air outlet pipes are connected to a plurality of terminal air outlet devices through pipes. Each air outlet pipe is provided with an electrically controlled air volume control valve that communicates and interacts with the computing control center.

[0016] In one embodiment of the present invention, the terminal air outlet device includes a cold beam box, a condenser coil for heat exchange of airflow is provided inside the cold beam box, and a cold beam air outlet is provided on the cold beam box.

[0017] To achieve the above technical solution, the refrigerated beam box is connected to the air outlet pipe through a pipeline, thereby completing the air distribution effect. The entire refrigerated beam box has a simple structure and will not generate noise from the operation of equipment such as fans, greatly reducing wind noise. At the same time, the condenser coil is set up to perform heat exchange on the outlet air, thereby ensuring the temperature and humidity of the outlet air. Since the fresh air entering the condenser box has already been treated by the heat exchange components, the power of the condenser coil does not need to be too large, and its noise will not be too high, thus ensuring the comfort of personnel.

[0018] In one embodiment of the present invention, a temperature and humidity auxiliary adjustment component is provided on the air outlet side of the heat exchange component. The temperature and humidity auxiliary adjustment component includes a humidifier provided on the air outlet side of the heat exchange component, and an electric heater is provided at one end of the air outlet side of the humidifier.

[0019] To achieve the above technical solution, after air undergoes heat exchange through the heat exchange components, its temperature and relative humidity change. For example, in spring and autumn when dehumidification is needed, the heat exchange components cool the air, lowering its temperature and causing water vapor in the air to condense, thus reducing the amount of water vapor in the air. However, this temperature drop also leads to excessively low airflow temperature from the terminal air outlet, affecting user comfort. Therefore, the temperature of the fresh air needs to be adjusted. In summer and winter, when cooling or heating is needed, the absolute or relative humidity of the air decreases. Therefore, the humidity of the air after heat exchange needs to be adjusted, i.e., by using a humidifier. By adding water vapor to the air, the humidity of the air can be adjusted, thereby improving the comfort of personnel. The fresh air can be heated by an electric heater, thereby adjusting the temperature of the fresh air. That is, with the combined action of the humidifier and the electric heater, dehumidification can be completed while maintaining a suitable temperature in spring and summer. In autumn and winter, heating can be completed while maintaining air humidity, thereby improving the comfort of personnel. At the same time, the electric heater is located at the rear of the humidifier so that the relative humidity of the air decreases when it is heated after humidification. This prevents the temperature from dropping after long-distance transportation in the pipeline, which could lead to condensation in the pipeline and thus ensure the comfort of the users.

[0020] In one embodiment of the present invention, a housing with a main air inlet and a main air outlet is further included, wherein the heat exchange component, the temperature and humidity auxiliary adjustment component, the fan and the flow control component are arranged sequentially along the airflow direction from the air inlet to the air outlet.

[0021] To achieve the above technical solution, the enclosure design integrates the heat exchange components, temperature and humidity auxiliary regulation components, fan, and flow control components into the enclosure, facilitating subsequent installation and maintenance.

[0022] In one embodiment of the present invention, a mounting frame is provided in the housing, the fan is a centrifugal fan and is located above the mounting frame, an L-shaped baffle is provided on the air outlet side of the centrifugal fan, and the flow control component is located on the L-shaped baffle; a water baffle is provided on the air outlet side of the humidifier, and a water baffle is provided on the water baffle that slopes downward away from the humidifier.

[0023] To achieve the above technical solution, the L-shaped wind deflector design enables the bending design of the air duct inside the box, thereby compressing the volume of the box and allowing it to be directly installed on the ceiling of the room, thus meeting the needs of small users; the water baffle design prevents the water vapor sprayed by the humidifier from overflowing into the duct of the entire air conditioning system.

[0024] In one embodiment of the present invention, the air conditioning adaptive comfort adjustment system uses the air conditioning adaptive comfort adjustment method described above to control the operating state of the diversion control component.

[0025] As described above, the air conditioning adaptive comfort adjustment system and the air conditioning adaptive comfort adjustment method provided by the present invention have the following beneficial effects:

[0026] 1. Based on the constructed Bayesian comfort model, comfort parameters are calculated using operating parameters and meteorological parameters as input parameters. The difference between environmental parameters and comfort parameters is calculated based on the comfort parameters, and this difference is used as the input parameter to control the operating state of the diversion control component, so that the environmental parameters are close to the comfort parameters, thereby improving the user's comfort. Furthermore, the operating state of the diversion control component is adjusted in a timely manner based on changes in operating parameters and meteorological parameters, thereby optimizing energy use.

[0027] 2. The temperature and humidity of the fresh air are uniformly regulated by the heat exchange components. In conjunction with the diversion control components, the fresh air is delivered to different terminal air outlets according to the comfort parameters. On the one hand, this enables adaptive adjustment of the operation requirements of different areas based on working parameters and meteorological parameters. On the other hand, the functional design of centralized fresh air treatment and distribution can meet the needs of home use. In actual home applications, the centralized fresh air treatment section can be installed in areas with low daily activity or low noise requirements, such as kitchens, balconies, or bathrooms, while the terminal air outlets can be installed in bedrooms, living rooms, or studies to ensure low noise in the environment when people are resting, thereby improving the comfort of users. Attached Figure Description

[0028] Figure 1 This is a structural logic diagram of an embodiment of this application.

[0029] Figure 2 This is a schematic diagram of the overall structure of an embodiment of this application.

[0030] Figure 3 This is a schematic diagram of the overall structure of the centralized fresh air treatment unit in the embodiments of this application.

[0031] Figure 4 This is a schematic diagram of the internal structure of the centralized fresh air treatment unit in the embodiments of this application.

[0032] Figure 5 This is a side view of the internal structure of the centralized fresh air treatment unit in the embodiments of this application.

[0033] Figure 6 This is a schematic diagram of the structure of the terminal air outlet device in the embodiments of this application.

[0034] Explanation of the technical feature labels in the attached drawings:

[0035] 1. Centralized fresh air treatment unit; 11. Housing; 12. Air filter assembly; 13. Heat exchange assembly; 14. Temperature and humidity auxiliary regulation assembly; 141. Humidifier; 142. Water baffle; 143. Air guide plate; 144. Electric heater; 15. Centrifugal fan; 16. Diversion control assembly; 161. Air outlet duct; 162. Electrically controlled air volume control valve; 17. Mounting bracket; 18. L-shaped air baffle; 19. Condensate tray; 2. Terminal air outlet device; 21. Cooling beam box; 22. Condensate coil; 3. Air-cooled heat pump; 4. Self-circulating cooling beam device; 41. Drum-type constant air volume fan; 42. Coil heat exchanger; 5. Buried coil. Detailed Implementation

[0036] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification.

[0037] It should be understood that the structures, proportions, sizes, etc., illustrated in the accompanying drawings of this specification are merely for illustrative purposes to aid those skilled in the art and are not intended to limit the scope of the invention. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in proportions, or adjustments to size, without affecting the effectiveness and purpose of the invention, should still fall within the scope of the technical content disclosed in this invention. Furthermore, the terms such as "upper," "lower," "left," "right," "middle," and "one" used in this specification are merely for clarity and are not intended to limit the scope of the invention. Changes or adjustments to their relative relationships, without substantially altering the technical content, should also be considered within the scope of the invention.

[0038] Example 1:

[0039] Please see Figure 1-2 This invention provides an adaptive comfort control system for air conditioning, comprising: a terminal air outlet device 2 for discharging air to the work area; a fan for drawing in and pressurizing outdoor air; a heat exchange component 13 for heat exchange of the fresh air drawn in by the fan to regulate the temperature and humidity of the fresh air; and an air-cooled heat pump 3 for supplying a cold or heat source medium to the heat exchange component 13; further comprising a flow control component 16 connected to the terminal air outlet device 2 via pipes; an internal sensing module installed on the air outlet side of the flow control component 16 to collect the operating parameters of the flow control component 16; and an ambient temperature sensor module installed on the air outlet side of the terminal air outlet device 2 to collect the ambient temperature of the work area. The system includes an external sensing module for environmental parameters and a human detection module, as well as a computing control center that communicates and interacts with the internal sensing module, the external sensing module, and the human detection module. The computing control center stores a Bayesian comfort model fitted based on a deep learning algorithm. The computing control center communicates and interacts with a meteorological database and obtains meteorological parameters for the corresponding work area based on the geographical location of the work area. It uses the received meteorological parameters and work parameters as input parameters for the Bayesian comfort model to calculate comfort parameters. It uses the comfort parameters as a benchmark value to calculate the difference between the environmental parameters and the model, and inputs the difference to the diversion control component 16 to control the operating status.

[0040] Specifically, the basic idea of ​​the Bayesian comfort model is to utilize the conditional dependencies between variables, train the network parameters based on historical data, and then perform probabilistic inference to predict future values. Its prediction principle can be summarized in the following steps:

[0041] 1) Construct the network structure. Based on the causal relationships between variables, determine the nodes and directed edges of the network to form a DAG graph.

[0042] 2) Learning parameters. Given training data D={X1,X2,........,Xn}, maximize the marginal likelihood of the data with respect to the current network structure, that is, learn the CPT conditional probability through the formula: P[Xi|pa(Xi)]=argmax_P.P(D|G,P); where G is the network structure, P is all CPT parameters, and the parameter learning method is Bayesian method.

[0043] 3) Inference and Prediction. Given the evidence e and the query variable Y, the posterior probability P(Y|e) is calculated using Bayes' rule, i.e.: P(Y|e)=α.P(e|Y)P(Y)=α.∑yP(e|y)P(y); where α is the normalization factor. Through inference and prediction, the posterior distribution or most likely value of variable Y can be predicted.

[0044] 4) Evaluate the prediction. Use metrics such as accuracy and recall to evaluate the prediction performance and optimize the model.

[0045] This application utilizes the Bayesian Network Toolbox (BNT) in Matlab software to construct a Bayesian comfort model. Meteorological parameters and operational parameters serve as variable nodes in the Bayesian network. Specifically, meteorological parameters include, but are not limited to, temperature, humidity, and wind speed; operational parameters include, but are not limited to, the number and characteristics of personnel in the work area, and routine functional data of the work area. The Bayesian Network Toolbox in Matlab is used to achieve Bayesian network structure learning, parameter learning, and network inference.

[0046] In Bayesian networks, learning involves training on a large number of samples to explore the implicit relationships within the network, i.e., solving for the conditional probabilities between node variables. The sample data used in the Bayesian learning process is called the training set data. The data used to verify the rationality and accuracy of the prediction model is called the test set data, which compares the actual values ​​of the test set data with the predicted values ​​for verification.

[0047] The training and test set data used in this embodiment are derived from user usage data over a sufficiently long period, specifically one year. The actual values ​​in the test set are user-defined parameters. The Bayesian comfort model in this embodiment is trained and fitted based on one year of user usage data. Its variables implicitly reflect the relationship between users' personalized comfort settings; therefore, this Bayesian comfort model has personalized predictive capabilities.

[0048] The internal sensing module and external sensing module in this embodiment include a temperature and humidity sensor and a wind speed sensor. The internal sensing module is used to collect the operating parameters of the diversion control component 16, including temperature, humidity and wind speed. The external sensing module is used to collect the environmental parameters of the working area, including temperature, humidity and wind speed.

[0049] The human body detection module in this embodiment uses a space occupancy sensor based on AI vision algorithms, such as the VS121 sensor. It analyzes human characteristics through AI vision algorithms to achieve people counting and area coverage.

[0050] Furthermore, in this embodiment, the computing control center is deployed on a cloud server, which can be a remote public cloud server or a self-built local private cloud server. The computing control center communicates with the meteorological database and can access and read meteorological data within the database. The specific access rules are as follows: the computing control center accesses and reads meteorological data corresponding to the geographical location of the work area in the meteorological database using the geographical location of the work area as the keyword; wherein, the geographical location of the work area can be collected by the locator built into the external sensing module, or it can be input and stored by the user through the control panel.

[0051] The shunt control component 16 operates based on the received difference. Specifically, it maintains operation when the difference is within a preset threshold range, and adjusts its operation when the difference is on either side of the preset threshold range. The preset threshold range is set according to the general human sensitivity to comfort.

[0052] If the difference falls within a preset threshold range, the operating state of the diversion control component is adjusted based on a correction value. The correction value equals the correction coefficient multiplied by the difference, and the correction coefficient is estimated using a random forest algorithm. Random forest is an ensemble learning method that improves the accuracy and stability of a model by constructing multiple decision trees and combining their predictions. In regression problems, random forest obtains the final prediction result by averaging the predictions of multiple decision trees. This embodiment uses the Matlab random forest tool to estimate the correction coefficient K, and the main steps can be summarized as follows:

[0053] 1) Data preparation. Based on the fitted Bayesian comfort model, set up a sufficient number of training and test datasets.

[0054] 2) Create a random forest model. Use the TreeBagger function in Matlab to create a random forest model, specifying the number of trees. In this example, the number of trees is set to 1000. For example:

[0055] numTrees = 1000; % Set the number of decision trees;

[0056] model = TreeBagger(numTrees, X(idxTrain,:), Y(idxTrain));

[0057] Here, X(idxTrain,:) and Y(idxTrain) represent the feature matrix and target variable of the training set, respectively.

[0058] 3) Make predictions. Use the test set data to make predictions by calling the model's predict method to predict the target variable, for example: yPred = predict(model, X(idxTest,:)); where X(idxTest,:) represents the feature matrix of the test set.

[0059] 4) Evaluate the model. Evaluate the model's performance using the mean squared error, for example:

[0060] mse = mean((Y(idxTest) - yPred).^2);

[0061] Here, Y(idxTest) represents the true target variable of the test set.

[0062] 5) Adjust parameters. Adjust the parameters of the random forest based on the model's performance to optimize the model's performance.

[0063] 6) Estimate the value of the correction coefficient K. Estimate the value of K through multiple training and prediction iterations: , where ki is the coefficient of each decision tree.

[0064] Please see Figure 2-6 The embodiments of the present invention also include a housing 11 with a main air inlet and a main air outlet, a fan, a heat exchange component 13, a temperature and humidity auxiliary adjustment component 14, and a flow diversion control component 16. The heat exchange component 13, the temperature and humidity auxiliary adjustment component 14, the fan, and the flow diversion control component 16 are arranged sequentially along the airflow direction from the air inlet to the air outlet. The housing 11 and the structural components within the housing 11 constitute the centralized fresh air treatment unit 1. The entire centralized fresh air treatment unit 1 is integrated into a single housing 11, facilitating installation.

[0065] Specifically, the housing 11 is equipped with a mounting bracket 17, and the fan is a centrifugal fan 15, which is positioned above the mounting bracket 17. An L-shaped baffle 18 is provided on the air outlet side of the centrifugal fan 15, and the flow control component 16 is mounted on the L-shaped baffle 18. The entire fresh air centralized treatment unit 1 adopts a bending design for its ductwork, compressing the overall volume of the fresh air centralized treatment unit 1 so that it can be directly installed in the room ceiling, thereby meeting the needs of small users.

[0066] Specifically, the temperature and humidity auxiliary adjustment component 14 is located on the air outlet side of the heat exchange component 13, including a humidifier 141 located on the air outlet side of the heat exchange component 13. An electric heater 144 is provided at one end of the air outlet side of the humidifier 141, and a water baffle 142 is provided on the air outlet side of the humidifier 141. A water baffle plate is provided on the water baffle 142 that slopes downward away from the humidifier 141. The water baffle plate is provided to prevent the water vapor sprayed by the humidifier 141 from overflowing into the duct of the entire air conditioning system.

[0067] Specifically, the diversion control component 16 includes several air outlet pipes 161 connected to the air outlet of the fan. Several air outlet pipes 161 are connected to several terminal air outlet devices 2 through pipes. Each air outlet pipe 161 is equipped with an electrically controlled air volume control valve 162 that communicates and interacts with the computing control center.

[0068] The terminal air outlet device 2 includes a cold beam box 21 with a cold beam air outlet. The cold beam box 21 contains a condenser coil 22 for heat exchange of the airflow. The cold beam box 21 is connected to the air outlet duct 161 in the centralized fresh air treatment unit 1 via a pipe, thus achieving air distribution. The entire cold beam box 21 has a simple structure and does not generate noise from operating equipment such as fans, significantly reducing wind noise. Simultaneously, the condenser coil 22 performs further heat exchange on the outlet air, ensuring the outlet air temperature and humidity. Because the fresh air entering the condenser box has already been treated by the centralized fresh air treatment unit 1, the power of the condenser coil 22 does not need to be excessive, and its noise level will not be too high, thus ensuring the comfort of users.

[0069] The air-cooled heat pump 3 is connected to a self-circulating cooling beam device 4, which includes a drum-type constant air volume fan 41 and a coil heat exchanger 42 connected to the air-cooled heat pump 3. While ensuring the outlet air pressure through the drum-type constant air volume fan 41, the coil heat exchanger 42 regulates the temperature and humidity of the airflow. This is suitable for rooms such as mahjong rooms or e-sports rooms that require rapid temperature control but have low noise requirements, thereby improving their adaptability.

[0070] The air-cooled heat pump 3 is also connected to a buried coil 5 via a pipe, which can be buried in the floor to form underfloor heating or underfloor cooling, thereby improving its adaptability. Since the buried coil 5 is connected to the inlet and outlet pipes of the air-cooled heat pump 3, the high-temperature cooling medium after the cooling medium in the fresh air centralized treatment unit 1 cools the cold air in spring and autumn can flow into the buried coil 5 to reheat and dry the room, while lowering the temperature of the cooling medium. This allows the air-cooled heat pump 3 to reach the preset temperature of the cooling medium with lower energy consumption, thereby achieving energy saving.

[0071] The internal sensing module is installed on the air outlet side of the diversion control component, while the external sensing module and the human body detection module are installed on the air outlet side of the terminal air outlet device.

[0072] A fan draws in outdoor air, pressurizes it, and sends it to the flow control component 16. A heat exchange component 13 exchanges heat with the fresh air drawn in by the fan to regulate its temperature and humidity. The flow control component 16 is connected to several terminal air outlets 2 via pipes. The flow control component 16 operates based on adjusted operating parameters calculated using difference and correction values ​​to send the regulated fresh air to different terminal air outlets 2. When the air passes through the heat exchange component 13 for heat exchange, its temperature and relative humidity change. For example, in spring and autumn when dehumidification is needed, the heat exchange component 13 cools the air, lowering its temperature and causing water vapor in the air to condense, thus reducing the amount of water vapor. However, this also lowers the temperature, resulting in excessively low airflow temperature from the terminal air outlets 2, affecting user comfort. Therefore, the temperature of the fresh air needs to be adjusted. In summer and winter... When cooling or heating is required, the absolute or relative humidity of the air will decrease. Therefore, it is necessary to adjust the humidity of the air after heat exchange. This is achieved by adding water vapor to the air through a humidifier 141, thereby assisting in adjusting the humidity of the gas and improving the comfort of the user. The fresh air can be heated by an electric heater 144, thereby adjusting the temperature of the fresh air. In other words, with the combined action of the humidifier 141 and the electric heater 144, dehumidification can be completed while maintaining a suitable temperature in spring and summer. In autumn and winter, heating can be completed while maintaining air humidity, thereby improving the comfort of the user. At the same time, the electric heater 144 is located at the rear end of the humidifier 141 so that after the air is humidified, its relative humidity decreases when the temperature is raised. This prevents the temperature from dropping after long-distance transportation in the pipeline, which could lead to condensation in the pipeline and thus ensure the comfort of the user.

[0073] The specific adjustment process of this application is as follows: The computing control center uses the received working parameters and meteorological parameters as input parameters, calculates the comfort parameters based on the Bayesian comfort model, and calculates the difference between the environmental parameters and the comfort parameters as the benchmark. The computing control center inputs the difference to the diversion control component. The diversion control component uses the difference as the input parameter, obtains the correction value and the corrected operating parameters according to the correction formula, and the electrically controlled air volume control valve sends air to the terminal air outlet device using the corrected operating parameters as the working parameters.

[0074] The fresh air treatment process is as follows: after the air-cooled heat pump 3 supplies a cold or heat source to the fresh air centralized treatment unit 1, the fresh air is heated or cooled to adjust its temperature and humidity. Then, the treated fresh air is delivered to different terminal air outlet devices 2 through the diversion control component to meet the air demand of the terminal space. At this time, it is not necessary to set up an independent fresh air treatment device in each terminal air outlet device 2, nor is it necessary to lay out a corresponding cooling duct system connected to the air-cooled heat pump 3, thereby reducing the complexity of the entire air conditioning structure. At the same time, the terminal air outlet devices 2 do not need to perform fresh air treatment, thus eliminating the need to install treatment equipment for adjusting the outlet air temperature, humidity, wind speed, and wind pressure. Therefore, the entire terminal air outlet device 2 has a simple structure and low space occupancy. The small size facilitates installation and eliminates the need for on-site treatment of fresh air in the terminal air outlets 2, thereby reducing noise and ensuring user comfort. During operation, the fresh air from the air conditioning unit is processed in the centralized fresh air treatment unit 1 and then distributed to different terminal air outlets 2 by the flow control component 16 using optimized operating parameters calculated according to the Bayesian comfort model. In practical home applications, the centralized fresh air treatment unit 1 can be installed in areas with low daily activity levels or low noise requirements, such as kitchens, balconies, or bathrooms, while the terminal air outlets 2 can be installed in bedrooms, living rooms, or studies to ensure low noise levels during rest and improve user comfort.

[0075] Example 2:

[0076] The present invention also provides an air conditioning adaptive comfort adjustment method, applied to the air conditioning adaptive comfort adjustment system of Embodiment 1, the air conditioning adaptive comfort adjustment method comprising:

[0077] S100. Obtain the operating parameters of the diversion control component; these parameters are acquired through an internal sensing module installed on the air outlet side of the diversion control component, including but not limited to temperature, humidity, and wind speed.

[0078] S200. Obtain environmental parameters of the working area of ​​the terminal air outlet device; these parameters are collected by an external sensor module installed on the air outlet side of the terminal air outlet device, including temperature, humidity, and wind speed, but not limited to these.

[0079] S300: Obtain the working parameters of the working area of ​​the terminal air outlet device, including the normal function data of the working area, the current number of users and their characteristics; this is collected by the human body detection module installed on the air outlet side of the terminal air outlet device. The normal function data of the working area can be customized by the user, such as the elderly's room, children's room, study room, etc.

[0080] S400: Obtain meteorological parameters of the working area of ​​the terminal air outlet device; the meteorological parameters are accessed and read by the computing control center using the geographical location of the working area as the keyword to read the meteorological data of the corresponding geographical location in the meteorological database; the geographical location of the working area can be collected by the locator built into the external sensing module, or it can be entered and stored by the user through the control panel; the meteorological parameters include temperature, humidity, and wind speed, but are not limited to these.

[0081] S500. Using the meteorological parameters and the working parameters as input parameters for the Bayesian comfort model fitted by the deep learning algorithm, the corresponding comfort parameters are calculated.

[0082] Specifically, the basic idea of ​​the Bayesian comfort model is to utilize the conditional dependencies between variables, train the network parameters based on historical data, and then perform probabilistic inference to predict future values. Its prediction principle can be summarized in the following steps:

[0083] 1) Construct the network structure. Based on the causal relationships between variables, determine the nodes and directed edges of the network to form a DAG graph.

[0084] 2) Learning parameters. Given training data D={X1,X2,........,Xn}, maximize the marginal likelihood of the data with respect to the current network structure, that is, learn the CPT conditional probability through the formula: P[Xi|pa(Xi)]=argmax_P.P(D|G,P); where G is the network structure, P is all CPT parameters, and the parameter learning method is Bayesian method.

[0085] 3) Inference and Prediction. Given the evidence e and the query variable Y, the posterior probability P(Y|e) is calculated using Bayes' rule, i.e.: P(Y|e)=α.P(e|Y)P(Y)=α.∑yP(e|y)P(y); where α is the normalization factor. Through inference and prediction, the posterior distribution or most likely value of variable Y can be predicted.

[0086] 4) Evaluate the prediction. Use metrics such as accuracy and recall to evaluate the prediction performance and optimize the model.

[0087] This application utilizes the Bayesian Network Toolbox (BNT) in Matlab software to construct a Bayesian comfort model. Meteorological parameters and operational parameters serve as variable nodes in the Bayesian network. Specifically, meteorological parameters include, but are not limited to, temperature, humidity, and wind speed; operational parameters include, but are not limited to, the number and characteristics of personnel in the work area, and routine functional data of the work area. The Bayesian Network Toolbox in Matlab is used to achieve Bayesian network structure learning, parameter learning, and network inference.

[0088] In Bayesian networks, learning involves training on a large number of samples to explore the implicit relationships within the network, i.e., solving for the conditional probabilities between node variables. The sample data used in the Bayesian learning process is called the training set data. The data used to verify the rationality and accuracy of the prediction model is called the test set data, which compares the actual values ​​of the test set data with the predicted values ​​for verification.

[0089] The training and test set data used in this embodiment are derived from user usage data over a sufficiently long period, specifically one year. The actual values ​​in the test set are user-defined parameters. The Bayesian comfort model in this embodiment is trained and fitted based on one year of user usage data. Its variables implicitly reflect the relationship between users' personalized comfort settings; therefore, this Bayesian comfort model has personalized predictive capabilities.

[0090] S600. Calculate the difference between the environmental parameters and the comfort parameters, and control the operating state of the diversion control component based on the difference. Specifically, if the difference is within a preset threshold range, the diversion control component remains in operation; if the difference is on either side of the preset threshold range, the operating state of the diversion control component is adjusted. The adjustment method is based on a correction value, where the correction value equals a correction coefficient multiplied by the difference, and the correction coefficient is estimated using a random forest algorithm.

[0091] Specifically, random forest is an ensemble learning method that improves the accuracy and stability of a model by constructing multiple decision trees and combining their predictions. In regression problems, random forest obtains the final prediction by averaging the predictions of multiple decision trees. This embodiment uses the Matlab random forest tool to estimate the correction coefficient K, and the main steps can be summarized as follows:

[0092] 1) Data preparation. Based on the fitted Bayesian comfort model, set up a sufficient number of training and test datasets.

[0093] 2) Create a random forest model. Use the TreeBagger function in Matlab to create a random forest model, specifying the number of trees. In this example, the number of trees is set to 1000. For example:

[0094] numTrees = 1000; % Set the number of decision trees;

[0095] model = TreeBagger(numTrees, X(idxTrain,:), Y(idxTrain));

[0096] Here, X(idxTrain,:) and Y(idxTrain) represent the feature matrix and target variable of the training set, respectively.

[0097] 3) Make predictions. Use the test set data to make predictions by calling the model's predict method to predict the target variable, for example: yPred = predict(model, X(idxTest,:)); where X(idxTest,:) represents the feature matrix of the test set.

[0098] 4) Evaluate the model. Evaluate the model's performance using the mean squared error, for example:

[0099] mse = mean((Y(idxTest) - yPred).^2);

[0100] Here, Y(idxTest) represents the true target variable of the test set.

[0101] 5) Adjust parameters. Adjust the parameters of the random forest based on the model's performance to optimize the model's performance.

[0102] 6) Estimate the value of the correction coefficient K. Estimate the value of K through multiple training and prediction iterations: , where ki is the coefficient of each decision tree.

[0103] Based on the correction coefficient K obtained by random forest estimation, the correction value of the operating parameters of the diversion control component 16 to achieve the comfort parameters is obtained, and the operating parameters of the diversion control component 16 are adjusted according to the correction value so that the difference between the environmental parameters and the comfort parameters is within the preset threshold range.

[0104] Based on the above S100-S600, the air conditioner completes adaptive comfort adjustment, so that when the meteorological parameters and operating parameters change significantly, the air conditioner obtains appropriate comfort parameters based on the Bayesian comfort model, and obtains adjustment correction values ​​based on the comfort parameters and the current corresponding environmental parameters, so as to adjust and correct the operating parameters of the diversion control component 16, thereby making the environmental parameters of the working area tend to the comfort parameters, and providing users with a good comfort experience.

[0105] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. An adaptive comfort adjustment method for air conditioning, applied to an air conditioning system with fresh air conditioning function, comprising: The system comprises: a terminal air outlet device for discharging air into a work area; a fan for drawing in and pressurizing outdoor air; a heat exchange component for heat exchange of the fresh air drawn in by the fan to regulate the temperature and humidity of the fresh air; and an air-cooled heat pump for supplying a cold or heat source medium to the heat exchange component; characterized in that it further includes a flow control component connected to each of the terminal air outlet devices via pipes; the air conditioning adaptive comfort adjustment method includes: Obtain the operating parameters of the air outlet side of the diversion control component, including temperature, humidity, and wind speed; Obtain environmental parameters of the working area on the air outlet side of the terminal air outlet device, including temperature, humidity, and wind speed; Obtain the operating parameters of the terminal air outlet device's working area, including normal function data of the working area, the current number of users, and personnel characteristics; Meteorological parameters of the working area of ​​the terminal air outlet device are obtained. The meteorological parameters are accessed and read from the meteorological database with the geographical location of the working area as the keyword. The meteorological parameters include temperature, humidity and wind speed. Using the meteorological parameters and the working parameters as input parameters for the Bayesian comfort model fitted by the deep learning algorithm, the corresponding comfort parameters are calculated. The difference between the environmental parameter and the comfort parameter is calculated, and the operating state of the diversion control component is controlled based on the difference. If the difference is within a preset threshold range, the diversion control component is kept in operation. If the difference is on either side of the preset threshold range, the operating state of the diversion control component is adjusted based on a correction value. The correction value is equal to the correction coefficient multiplied by the difference, and the correction coefficient is estimated by a random forest algorithm.

2. An adaptive comfort control system for air conditioning, characterized in that, include: A terminal air outlet device for discharging air into a work area includes a fan for drawing in and pressurizing outdoor air, a heat exchange component for heat exchange of the fresh air drawn in by the fan to regulate the temperature and humidity of the fresh air, and an air-cooled heat pump for supplying a cold or heat source medium to the heat exchange component. The device is characterized by further including a flow control component connected to each terminal air outlet device via pipes, an internal sensing module installed on the outlet side of the flow control component to collect its operating parameters, an external sensing module installed on the outlet side of the terminal air outlet device to collect environmental parameters of the work area, and a human body detection module to collect working parameters of the work area, as well as a computing control center that communicates and interacts with the internal sensing module, the external sensing module, and the human body detection module. The operating parameters include temperature, humidity, and wind speed; the environmental parameters include temperature, humidity, and wind speed; and the working parameters include normal functional data of the work area and data obtained through the human body detection module. The module collects the current number of users and their characteristics; the computing control center stores a Bayesian comfort model fitted based on a deep learning algorithm; the computing control center communicates and interacts with a meteorological database and obtains meteorological parameters corresponding to the work area based on the geographical location of the work area, the meteorological parameters including temperature, humidity, and wind speed; the received meteorological parameters and work parameters are used as input parameters for the Bayesian comfort model to calculate comfort parameters; the difference between the environmental parameters and the comfort parameters is calculated based on the comfort parameters, and the difference is input to the diversion control component to control its operating state; if the difference is within a preset threshold range, the diversion control component is kept in operating state; if the difference is on either side of the preset threshold range, the operating state of the diversion control component is adjusted based on a correction value, the correction value being equal to a correction coefficient multiplied by the difference, the correction coefficient being estimated using a random forest algorithm.

3. The air conditioning adaptive comfort adjustment system according to claim 2, characterized in that, The diversion control component includes several air outlet pipes connected to the air outlet of the fan. The several air outlet pipes are connected to several terminal air outlet devices through pipes. Each air outlet pipe is equipped with an electrically controlled air volume control valve that communicates and interacts with the computing control center.

4. The air conditioning adaptive comfort adjustment system according to claim 3, characterized in that, The terminal air outlet device includes a cold beam box, which is equipped with a condenser coil for heat exchange of the airflow, and a cold beam air outlet is provided on the cold beam box.

5. An air conditioning adaptive comfort adjustment system according to claim 4, characterized in that, A temperature and humidity auxiliary adjustment component is provided on the air outlet side of the heat exchange component. The temperature and humidity auxiliary adjustment component includes a humidifier provided on the air outlet side of the heat exchange component, and an electric heater is provided at one end of the air outlet side of the humidifier.

6. An air conditioning adaptive comfort adjustment system according to claim 5, characterized in that, It also includes a housing with a main air inlet and a main air outlet, wherein the heat exchange component, temperature and humidity auxiliary adjustment component, fan and flow control component are arranged sequentially along the airflow direction from the air inlet to the air outlet.

7. An air conditioning adaptive comfort adjustment system according to claim 6, characterized in that, The housing is equipped with a mounting frame, and the fan is a centrifugal fan. The centrifugal fan is positioned above the mounting frame, and an L-shaped baffle is provided on the air outlet side of the centrifugal fan. The flow control component is positioned on the L-shaped baffle. A water baffle is provided on the air outlet side of the humidifier, and a water baffle is provided on the water baffle that slopes downward away from the humidifier.

8. An air conditioning adaptive comfort adjustment system according to claim 7, characterized in that, The air conditioning adaptive comfort adjustment system uses the air conditioning adaptive comfort adjustment method described in claim 1 to control the operating state of the diversion control component.

Images