Air conditioner control system and method based on smart home

Abstract

The invention relates to the technical field of intelligent control, in particular to an air conditioner control system and method based on a smart home. The air conditioning module is used for refrigerating or heating air; the image acquisition module is used for acquiring an image of a user smart home environment and defining the image as a first image; the analysis module is used for judging whether a user exists in the first image or not, identifying the identity, position and behavior of the user, retrieving a personal data packet of the user based on an analysis result, and outputting a control signal; the temperature adjusting module is used for adjusting the air supply temperature and speed of each air supply outlet; the wind direction adjusting module is used for adjusting wind directions of the air supply outlets. The operation parameters of each air supply outlet of the air conditioner are automatically adjusted based on the corresponding personal information database by recognizing the feature information such as the identity, the position and the behavior of the user, so that corresponding operation control is performed when different users are in different areas and different states, and the user experience is improved. And the air supply parameters of the air conditioner blown to the user are more suitable for the personalized requirements of the user in the current state.

Description

Technical Field

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

[0002] A smart home environment, also known as an intelligent residential environment, is a living environment integrating a variety of modern scientific and technological means. It takes a residence as a platform and uses technologies such as integrated wiring technology, network communication technology, security prevention technology, automatic control technology, and audio and video technology to integrate various facilities related to home life, thereby constructing an efficient, safe, convenient, comfortable, and environmentally friendly residential facility and family daily affairs management system.

[0003] Generally, there are multiple users of different age groups in a smart home environment. Their demands for thermal comfort are different, and the required environmental temperatures are also different. At the same time, the thermal environment required by each person in different exercise states is also different. In the existing smart home air conditioner control technology, the wind direction control mode of the air conditioner air outlet is generally a mode of the blade sweeping up and down or left and right, and the blade may also be fixed; currently, the air conditioner usually performs feedback and adjustment control according to a preset temperature. When the temperature sensor of the air conditioner detects that the return air temperature reaches the preset temperature threshold range, it enters the standby or low-power mode. However, there are differences in the thermal sensations of different people in different environments. This method is difficult to perform corresponding temperature adjustments, as well as wind direction and wind speed adjustments, according to the thermal comfort demands of different people in different behavioral states, and it is difficult to meet personalized needs.

[0004] In summary, the present invention proposes an air conditioner control system and method based on a smart home, which can perform personalized settings based on the different behavioral states of different users, automatically adjust the air supply temperature, and at the same time adjust the air supply speed and angle based on the position of the user, so as to achieve precise control of the thermal environment under different working conditions and meet the needs of human thermal comfort and energy conservation and emission reduction. Summary of the Invention

[0005] To solve the above problems, the present invention provides an air conditioner control system and method based on a smart home. By identifying feature information such as user identity, location, and behavior, and based on a corresponding personal information database, the operating parameters of each air outlet of the air conditioner are automatically adjusted, so as to perform corresponding operating control when different users are in different regions and different states, and make the air supply parameters of the air conditioner blowing towards the user more suitable for the personalized needs of the user in the current state.

[0006] To achieve the above object, the technical solution of the present invention is as follows: An air conditioner control system based on a smart home, comprising:

[0007] A database module, used to establish a user personal data packet, including at least the user's age, gender, clothing, preferences, body temperature, heart rate, and activity pattern;

[0008] An air conditioning module, which is used to cool air through a refrigeration cycle or heat air through a heating cycle. The air after heat and moisture treatment in the indoor unit is defined as the first gas;

[0009] An image acquisition module, which is used to obtain an image of the user's smart home environment and is defined as the first image;

[0010] An analysis module, which is used to receive the first image and, based on image processing technology, determine whether there is a user in the first image; when there is a user in the first image, based on face recognition technology, obtain the user identity information, which is defined as the first retrieval condition; it is also used to obtain the position information of the user in the smart home environment based on computer vision technology, which is defined as the second retrieval condition; at the same time, identify and analyze the user's behavior and clothing in the first image to obtain the third retrieval condition; then, based on the first retrieval condition, the second retrieval condition, and the third retrieval condition, perform corresponding retrieval on the user's personal data packet and output a total control signal. The total control signal includes a first control signal sent to the air conditioning module and a second control signal sent to the temperature adjustment module and the air direction adjustment module; the air conditioning module is used to operate according to the first control signal to determine the set target temperature in the room;

[0011] A temperature adjustment module, which is used to mix the air in the room with the first gas in proportion based on the second control signal and adjust the air supply temperature of the air supply outlet to reach the required temperature of the air supply outlet;

[0012] An air direction adjustment module, which is used to adjust the position and pose of the air direction guide plate in the air supply outlet based on the second control signal.

[0013] Furthermore, it further includes a visualization module for establishing a three-dimensional map of the user's smart home environment and displaying the parameters of each air supply outlet therein. The air supply outlet parameters at least include position, wind speed, temperature, and relative humidity; the analysis module is also used to obtain the expression information, skin temperature, and clothing feature information of the user in the smart home environment based on the first image, and then determine whether the air state parameters blowing towards the user meet the current user's needs; when the air state parameters do not meet the current user's needs, a third control signal is generated as feedback to control the air direction adjustment module to adjust the air direction, and at the same time, control the temperature adjustment module to adjust the air temperature and speed of the corresponding air supply outlet.

[0014] Furthermore, the temperature adjustment module includes:

[0015] An air return opening, which is arranged on one side of the air mixing chamber;

[0016] An air mixing chamber is arranged at the front end of the air outlet of the air conditioner indoor unit. The air outlet of the air conditioner indoor unit is communicated with a number of independent air mixing chambers. One side is open to inhale the first gas that is heat and moisture treated and conveyed by the air conditioning module, and a flow disturbance structure is arranged inside the air mixing chamber to promote the full and uniform mixing of the two parts of air; a certain number of air supply outlets are arranged on the other side.

[0017] A blower is installed in the air mixing chamber and has the function of adjustable rotation speed and air volume. It is used to extract the mixed air in the mixing chamber, provide power for air flow, ensure that the air flows to the air supply outlet and the air speed can be adjusted.

[0018] A temperature sensor and a humidity sensor are respectively arranged at the air return opening of the air mixing chamber, inside the air mixing chamber and near the air supply outlet, and are used to monitor the air temperature and humidity at each key node in real time.

[0019] A controller receives the second control signal from the analysis module and the data fed back by each temperature sensor. According to the temperature and humidity adjustment requirements included in the second control signal, combined with the real-time monitored temperature and humidity information, it accurately calculates the optimal mixing ratio of the indoor air and the first gas to reach the optimal air supply state point, and through regulating the rotation speed of the blower, it finely manages the mixing process and the air supply process, so that the air temperature, humidity and air speed sent out by each air supply outlet accurately match the required parameters.

[0020] Furthermore, it also includes a network module for connecting the intelligent home air conditioner control system with an external network, allowing users to remotely access and control the air conditioner system through a mobile device or a computer; it is also used to regularly obtain the latest weather forecast and the outdoor environment data of the user's location from the Internet, so that the system can more intelligently preset and adjust the air conditioner operation parameters.

[0021] Furthermore, the analysis module is also used to continuously analyze the long-term usage habits and preferences of users based on deep learning, and continuously optimize the control logic; it is also used to identify and learn the special needs of users under different times, seasons and weather conditions, automatically adjust the control strategy, and add the learning results to the database module.

[0022] Further, it also includes a health monitoring module. The health monitoring module is a wearable device used to monitor the user's health data, at least including heart rate and body temperature. It calculates the user's thermal comfort according to the dynamic thermal comfort model, combines the facial expression information obtained by the analysis module, and then based on the thermal comfort expectation in the user's personal data packet, performs personalized air conditioner parameter settings and sends the air conditioner parameter settings to the air conditioner module, the temperature adjustment module, and the air direction adjustment module. The temperature adjustment module is also used to monitor the air temperature and humidity in the rooms where each air outlet is located. The air direction adjustment module is also used to monitor the pose information of the air direction deflectors of each air outlet based on the pan-tilt technology. The air direction adjustment module based on the pan-tilt technology includes a gyroscope and an acceleration sensor. The gyroscope is used to real-time sense the change in the rotation angle of the air direction deflector in three-dimensional space. The acceleration sensor is used to capture the acceleration information during the movement of the deflector. The visualization module is also used to display the temperature and relative humidity of different areas in each room on a three-dimensional map.

[0023] Further, when multiple users are detected in the same public area, the analysis module is also used to: First, for each user, respectively based on their location information, accurately locate the relative position relationship between each user and the air outlet. Then, deeply analyze the behaviors, clothing, preferences of each user, as well as the body temperature and heart rate information obtained from the health monitoring module, retrieve the personal data packets of all the people in the room accordingly, comprehensively determine the target temperature and wind speed for the operation of the air conditioner module, and send a first control signal to the air conditioner module. Then, according to the preset deep reinforcement learning model, predict the personalized needs of each user for temperature and wind speed. If a user is wearing thin clothes and reading, they may prefer a relatively mild temperature and a lower wind speed. If a user has just finished exercising and is sweating profusely, they need a lower temperature and a higher wind speed to cool down quickly. If a user is wrapped in a blanket watching TV, they may require a relatively higher temperature and a gentle wind environment. Finally, according to the demand differences of different users, generate refined second control signals for the air outlets associated with each user. This signal includes the temperature value, wind speed value, and air direction adjustment instruction that the corresponding air outlet needs to adjust.

[0024] Further, the deep reinforcement learning model adopts an algorithm based on policy gradient, and the specific formula is as follows:

[0025]

[0026] In the formula, J(θ) is the objective function of the policy network, used to measure the quality of the policy π θ Here, the policy π θ is the policy for the analysis module to generate air conditioner control instructions based on user information and environmental information. θ is the parameter of the policy network, and by optimizing θ, the performance of the policy can be improved. represents the expectation calculated according to the policy π θ r t$R_t$ is the reward obtained at time $t$, which is given based on the user's satisfaction with the current air conditioner air supply. It is comprehensively evaluated by analyzing physiological and behavioral feedback such as the user's facial expression changes, skin temperature changes, and heart rate fluctuations. If the user shows a relaxed state, the skin temperature approaches the comfortable range, and the heart rate is stable, a higher reward value is given; otherwise, the reward value is lower. $\gamma$ is the discount factor, with a value between 0 and 1, used to weigh the relative importance of the current reward and future rewards. When it is close to 1, it indicates that the influence of future rewards is greater, meaning the model pays more attention to optimizing the long-term user experience, and a value of 0.95 can be taken. $T$ is the total duration of an episode, which can be understood here as the time period from when the user enters the room and turns on the air conditioner control to when the user leaves the room or no longer triggers the air conditioner adjustment requirement for a long time.

[0027] Further, to optimize the parameters $\theta$ of the policy network, the gradient ascent method is used, and its update formula is:

[0028]

[0029] Among them, $\alpha$ is the learning rate, with a value between 0 and 1, controlling the step size of each parameter update, and a value of 0.01 can be taken.

[0030] Further, the air conditioner control method based on smart home includes the following steps: Step 1, information acquisition and analysis. The image of the smart home environment is acquired through the image acquisition module, and then the analysis module identifies the image information, determines whether there is a user, and acquires the user identity information, location information, and behavior information. At the same time, the analysis module and the health monitoring module acquire the user's facial expression, clothing characteristics, and health data, and calculate the dynamic thermal comfort according to the dynamic thermal comfort model to obtain the corresponding comprehensive information;

[0031] Step 2, demand matching. Based on the user's personal data packet in the database module, combined with the comprehensive information in Step 1, and jointly with the latest weather forecast and outdoor environment data obtained by the network module, the operation parameters of the temperature adjustment module, air direction adjustment module, and air conditioner module are automatically adjusted in real time; and continuously collect the special needs of the user under different conditions to optimize the control logic based on deep learning, and add the learning results to the user's personal data packet;

[0032] Step 3, visualization and remote control. Use the visualization module to establish a three-dimensional map of the user's smart home environment, and display the temperature and relative humidity of each room, the parameters of each air supply outlet, and the operation parameters of the air conditioner in this three-dimensional map. At the same time, the user can remotely access and control the entire air conditioner system through the network module via a remote terminal.

[0033] The above scheme has the following beneficial effects:

[0034] 1. Compared with the prior art, this solution improves the structure of the household air conditioner, sets up an air mixing chamber outside the air outlet of the indoor unit of the air conditioner, and comprehensively applies advanced technologies such as image acquisition, face recognition, computer vision, and deep learning. It can achieve accurate identification of user identities, real-time tracking of positions and behaviors, and in-depth analysis of user preferences and habits, enabling the air conditioning system to perform intelligent adjustment according to the personalized needs of users and provide more comfortable and considerate services.

[0035] 2. This solution not only considers the information in the user's personal data packet but also obtains the latest weather forecast, outdoor environmental parameters, and indoor environmental parameters in real time through the network module. This enables the air conditioning system to automatically adjust the operating parameters according to environmental factors such as different seasons, weather conditions, and indoor-outdoor temperature differences, ensuring that the indoor environment always remains in the most suitable state for the user.

[0036] 3. This solution, by introducing a health monitoring module, can monitor the user's health data (such as heart rate, body temperature, etc.) in real time, calculate the dynamic thermal comfort based on the dynamic thermal comfort model (DTS), obtain the expression information through the analysis module, and then comprehensively analyze the user's thermal comfort according to the dynamic thermal comfort and expression information. According to the thermal comfort expectations in the pre-stored user personal data packet, it makes an intelligent decision on the best operating strategy to meet the user's personalized needs under the current working conditions, and accurately adjusts the air conditioning operating parameters based on these data to ensure the user's comfort and health. This not only improves the user's personalized experience but also reflects the deep concern for the user's health. Based on the presence of people in the room, the initial set temperature of the air conditioning module is set according to the status of individuals or multiple people, and then the air supply temperature and speed of each air outlet are set according to the individual's status and preferences, reflecting the personalized differences of users and improving the comfort of different individuals in the multi-user shared area, solving the problem that the existing air conditioning control cannot effectively balance the thermal comfort of multiple users.

[0037] 4. In this solution, the visualization module enables users to intuitively view the 3D map of the smart home environment, as well as information such as the temperature and relative humidity of each room, air supply outlet parameters, and air conditioning operating parameters. At the same time, the addition of the network module allows users to access and control the entire air conditioning system at any time through a remote terminal, not only improving the user's usage efficiency but also bringing a more flexible, convenient, and comfortable air conditioning usage experience to the user.

[0038] 5. This solution can achieve precise cooling and heating while ensuring the comfort of each user by precisely adjusting the air conditioning parameters and real-time monitoring of the indoor temperature and relative humidity, effectively reducing energy consumption. This not only helps to reduce the user's electricity bill but also conforms to the current social trend of green environmental protection, energy conservation, and carbon reduction.

[0039] Additional aspects and advantages of the present invention will be given in part in the following description, become apparent in part from the following description, or be learned through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 Schematic diagram of the control system for an air conditioner based on a smart home according to an embodiment of the control system and method of the present invention;

[0041] Figure 2 Schematic diagram of the method steps for an air conditioner control system and method based on a smart home according to an embodiment of the present invention;

[0042] Figure 3 Schematic diagram of air treatment in an air conditioner control system and method based on a smart home according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0043] The technical solutions of the present invention will be described clearly and completely below with reference to the accompanying drawings. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0044] In the description of the present invention, it should be noted that the orientation or positional relationship indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention. In addition, the terms "first", "second", and "third" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0045] In the description of the present invention, it should be noted that unless otherwise clearly defined and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0046] The following will be further described in detail through specific embodiments:

[0047] Embodiment 1:

[0048] As Figure 1 、 Figure 2 、 Figure 3As shown: An air conditioner control system based on smart home, including:

[0049] A database module for establishing user personal data packets, including at least the user's age, gender, clothing, body temperature, heart rate, preferences, and activity patterns. The database module stores the personal data packets of each user, including key information such as age, gender, preferences, and activity patterns, which are obtained through user registration, historical behavior data collection, and possibly user surveys, thereby improving the accuracy and reliability of subsequent control of the operating parameters of the air conditioner. The activity pattern refers to the behavior and activity rules of the user at different time periods in a day or a week, including but not limited to: daily living time (such as wake-up time, meal time, working time, rest time, sleep time, etc.), location movement (the frequency and time of movement between different areas in the home (bedroom, living room, kitchen, bathroom, study, etc.)), activity intensity (which can be divided into static (such as reading, watching TV), mild (such as tidying up the house, taking a walk), moderate (such as yoga, jogging), and heavy (such as fitness, heavy physical labor), etc.) according to the user's physical activity), social activities (whether there are often guests visiting, and the frequency and duration of these activities), and special habits (such as regular exercise, meditation, gardening, etc. personal hobbies). Thus, according to the user's activity pattern, the operating state of the air conditioner is intelligently adjusted, such as preheating / pre-cooling the room when the user is about to wake up, reducing energy consumption when the user goes out for a short time, and automatically shutting down to reduce energy consumption when the user goes out for a long time.

[0050] An air conditioner module for cooling air through a refrigeration cycle or heating air through a heating cycle. The air after heat and moisture treatment in the indoor unit is defined as the first gas.

[0051] An image acquisition module for acquiring images of the user's smart home environment, defined as the first image. The image acquisition method is preferably a high-resolution camera, which is installed at key positions, such as the living room, bedroom, etc., to capture the user and their activities. The image acquisition module is responsible for receiving the video stream from the camera and converting it into image data for analysis.

[0052] An analysis module, configured to receive a first image and, based on image processing technology, determine whether there is a user in the first image; when there is a user in the first image, obtain user identity information based on face recognition technology, defined as a first retrieval condition; is also used to obtain the position information of the user in the smart home environment based on computer vision technology, defined as a second retrieval condition; at the same time, identify and analyze the user's behavior and clothing in the first image to obtain a third retrieval condition; then, based on the first retrieval condition, the second retrieval condition, and the third retrieval condition, perform corresponding retrieval on the user's personal data packet and output a total control signal, where the total control signal includes a first control signal sent to the air conditioner module and a second control signal sent to the temperature adjustment module and the wind direction adjustment module; the air conditioner module is configured to operate at a target temperature set in the room during operation according to the first control signal. In addition, the first control signal may also include the wind speed when the air conditioner module blows air. For example, if the user is identified as liking a lower temperature and being about 28 years old, and is currently lying on the sofa in the living room watching TV, then the system may output a control signal to lower the temperature setting value in the living room. This calculation process will comprehensively consider factors such as the structure of the room, the position of the air supply outlet (the position of the air supply outlet can be extended to a suitable position by installing an air duct), the position of the user, and the user's preferences.

[0053] The temperature regulation module is used to mix the air in the room with the first gas in proportion based on the second control signal (according to the current thermal environment and the user's personalized thermal comfort requirements), and adjust the supply air temperature of each air supply outlet to the required temperature of that air supply outlet; the temperature regulation module is also used to monitor the air temperature in the room where each air supply outlet is located. For example, when there are multiple users in the living room, first, the personal data packets of all the people in the room are retrieved accordingly, and the set temperature and wind speed for the operation of the air conditioning module are comprehensively determined, that is, the first control signal is sent to the air conditioning module; combined with the indoor environment parameters after the air conditioner runs, then the personalized needs of the users are collected. For example, user A needs a slightly warmer environment to relax and rest. The temperature regulation module calculates the required supply air temperature and speed for user A according to this requirement, and at the same time calculates based on the initial supply air temperature and the air temperature in the living room to obtain the mixing ratio of the two, and correspondingly adjusts the corresponding air valve (a type of electric control valve) in the air mixing chamber. At the same time, this ratio will also change in real time with the air temperature in the living room, so that the temperature near user A in the living room quickly reaches the preset value. User B may be exercising and has a lower temperature requirement. Then, through the cooperation of the subsequent air direction regulation module and the temperature regulation module, the supply of lower temperature air can be completed. Moreover, for the collection of the temperature and humidity information of each room, it is detected by a temperature and humidity sensor. For the selection of its model, the MIUI intelligent temperature and humidity meter can be preferably selected, which not only facilitates the user to directly view the temperature and humidity of the room where they are located, but also facilitates the transmission of the collected temperature and humidity data to the analysis module for subsequent analysis and control.

[0054] The air direction regulation module is used to adjust the position and pose of the air direction guide plate in the air supply outlet based on the second control signal. The second control signal includes the position of the air supply outlet where the user is located, as well as the desired air supply direction and wind speed of the user. Based on the parsed information, the air direction regulation module will use the built-in algorithm to calculate the optimal position and pose that the air direction guide plate should be adjusted to. After calculating the optimal position and pose, the air direction regulation module will adjust the angle and position of the air direction guide plate through a driving mechanism (such as a motor or a stepper motor). This process is usually fast and accurate to ensure that the air supply direction and speed can quickly meet the user's needs. During the adjustment process, the air direction regulation module will also monitor the air supply effect in real time and make fine adjustments according to the actual situation. At the same time, it will also feedback the adjusted position and pose information to the analysis module so that the system can continuously optimize and adjust the air supply strategy to blow the air at the required temperature towards the user at an appropriate angle.

[0055] It also includes a visualization module for establishing a three-dimensional map of the user's smart home environment based on three-dimensional modeling technology and displaying various air supply outlet parameters therein. The air supply outlet parameters at least include position, wind speed, temperature, and humidity. Furthermore, the user can intuitively view information such as room temperature and humidity, air supply outlet parameters, and air conditioner operation parameters through this three-dimensional map. The visualization module is also used to display the temperature and humidity of each room area on the three-dimensional map and display the air conditioner operation parameters. The air conditioner operation parameters at least include the target temperature, supply air temperature, humidity, and speed set in the room.

[0056] The analysis module is also used to obtain the user's facial expression information, skin temperature, and clothing feature information in the smart home environment based on the first image, calculate the dynamic thermal comfort based on physiological parameters such as heart rate and body temperature, and then comprehensively judge whether the air state parameters blowing towards the user meet the current user's needs; when the air state parameters do not meet the current user's needs, a third control signal is generated as feedback to control the wind direction adjustment module to adjust the wind direction, and at the same time, control the temperature adjustment module to adjust the supply air temperature and speed of the corresponding air supply outlet. For example, if the user looks very hot (such as blushing and sweating), the system may automatically adjust the air conditioner parameters to lower the indoor temperature. Similarly, if the user is wearing thick clothes, the system may also adjust the temperature accordingly.

[0057] The analysis module is also used to continuously analyze the user's long-term usage habits and preferences based on deep learning and continuously optimize the control logic. For example, with the support of deep learning technology, the analysis module can analyze the user's usage habits in the past few months or even years and find that the user always sets the air conditioner temperature to 24°C around 10 pm and raises the temperature to 26°C before 6 am to get ready to wake up. Based on such habits, the analysis module will automatically adjust the control logic to ensure that the air conditioner temperature is lowered to the user's preferred 24°C at 10 pm every night and slowly raised to 26°C before 6 am to provide the most comfortable sleeping environment.

[0058] In addition, the analysis module can also identify and learn the user's special needs under different times, seasons, and weather conditions based on deep learning and automatically adjust the control strategy. For example, in the hot summer, when the user frequently adjusts the air conditioner wind speed and temperature, the analysis module will realize that the user may be experiencing discomfort caused by high temperature and automatically adjust the control strategy to provide stronger cooling effect. At the same time, the analysis module can also adjust the control strategy in advance according to the weather forecast information. For example, when the forecast shows that there will be heavy rain, strong wind and other weather or seasonal changes (such as the beginning of winter), the analysis module will predict that the indoor temperature may drop and start the air conditioner in advance for preheating to ensure that the user can enjoy a comfortable environment immediately after returning home.

[0059] In specific use, the analysis module is also used to, when multiple users are detected in the same common area, first calculate the indoor temperature that satisfies most users according to the thermal comfort requirements of the multiple users, and output a temperature set value to the air conditioning module. Next, for each user, based on their location information, accurately locate the relative position relationship between each user and the air supply outlet. Then, deeply analyze the behavior, clothing, preferences of each user, as well as the body temperature and heart rate information obtained from the health monitoring module, and combine with the indoor environmental parameters to predict the personalized needs of each user for temperature and wind speed according to the preset deep reinforcement learning model. If a user is wearing thin clothes and reading, they may prefer a relatively mild temperature and a lower wind speed. If a user has just finished exercising and is sweating profusely, they need a lower temperature and a larger wind speed to cool down quickly. If a user is watching TV wrapped in a blanket, they may require a relatively high temperature and a gentle wind environment. Finally, according to the demand differences of different users, refined second control signals are generated for the air supply outlets associated with each user, and the signal includes the temperature value, wind speed value, and wind direction adjustment instruction that the corresponding air supply outlet needs to adjust.

[0060] The deep reinforcement learning model adopts an algorithm based on policy gradient, and the specific formula is as follows:

[0061]

[0062] In the formula, J(θ) is the objective function of the policy network, which is used to measure the quality of the policy π θ Here, the policy π θ is the policy for the analysis module to generate air conditioning control instructions based on user information and environmental information. θ is the parameter of the policy network, and the performance of the policy is improved by optimizing θ. denotes taking the expectation according to the policy π θ ; r t is the reward obtained at time t. The reward is given according to the user's satisfaction with the current air conditioning supply, and is comprehensively evaluated by analyzing the physiological and behavioral feedback such as the user's facial expression changes, skin temperature changes, and heart rate fluctuations. If the user shows a relaxed state, the skin temperature approaches the comfortable range, and the heart rate is stable, a higher reward value is given, otherwise the reward value is lower; γ is the discount factor, with a value between 0 and 1, which is used to weigh the relative importance of the current reward and future rewards. When it is close to 1, it indicates that the influence of future rewards is greater, meaning that the model pays more attention to optimizing the long-term user experience, and can take the value of 0.95; T is the total duration of an episode, which can be understood here as the time period from when the user enters the room and turns on the air conditioning control to when the user leaves the room or no longer triggers the air conditioning adjustment demand for a long time.

[0063] To optimize the parameter θ of the policy network, the gradient ascent method is adopted, and its update formula is:

[0064]

[0065] Among them, α is the learning rate, with a value between 0 and 1, which controls the step size of each parameter update and can take the value of 0.01.

[0066] In the actual operation scenario of the smart home air conditioning system, when there are multiple users (such as User A, User B, and User C) in the same public area (such as the living room), the analysis module begins to play a key role.

[0067] First of all, the image acquisition module uses high-resolution cameras distributed at key positions in the living room to continuously collect environmental images and transmit them to the analysis module. Based on mature image processing technology, the analysis module quickly analyzes the images and accurately identifies the position information of each user in the living room. For example, by comparing the human contours in the image with the spatial coordinates of the living room, it is determined that User A is in the sofa area, and the nearest air outlet to him is the No. 3 air outlet on the east wall; User B is in the reading corner near the window, and the corresponding nearest air outlet is the No. 1 air outlet on the south wall; User C is playing on the carpet in the center of the living room, and the associated air outlet is the No. 5 air outlet in the center of the ceiling, thus completing the accurate positioning of the relative position relationship between the user and the air outlet.

[0068] Next, the analysis module deeply mines multi-dimensional information to predict user needs. On the one hand, it carefully analyzes the images transmitted by the image acquisition module again, and uses advanced image recognition algorithms to identify the user's behaviors (such as reading, exercising, watching TV, etc.) and clothing (thin clothes, thick coats, sports equipment, etc.) characteristics; on the other hand, it obtains the user's body temperature and heart rate data in real time from the health monitoring module, which updates the data to the analysis module every few seconds through wearable devices such as smart bracelets or smart patches worn by the user. At the same time, the indoor environmental parameters (including the current room temperature, humidity, light intensity, etc.) are also collected in real time, which are provided by temperature and humidity sensors, light sensors, etc. installed in the corners of the living room and are all gathered to the analysis module.

[0069] At this time, the algorithm based on policy gradient of the preset deep reinforcement learning model starts to operate. The J(θ) function in the model serves as the core evaluation criterion of the policy network, constantly measuring the quality of the air conditioning control instruction policy π θ generated by the current analysis module. For example, in the initial stage, the analysis module randomly generates a set of air conditioning control instructions for each user's air outlet, including the preliminary setting values of temperature, wind speed, and wind direction, which is an initial form of the policy π θ and θ, as the parameters of the policy network, affects the generation logic of the instructions.

[0070] At each time step t, the system will conduct multi-faceted physiological and behavioral monitoring of the user according to the current air conditioning operation state and the actual environment where the user is located to evaluate the reward r tFor example, for user A, if they are reading while wearing thin clothes, with a calm facial expression, skin temperature stable in a relatively comfortable range, and heart rate also remaining steady, according to the pre-set comprehensive evaluation rules, a relatively high reward value is given, such as r t = 0.8; while for user B who has just finished exercising and is sweating profusely, if the air temperature blown out by the current air outlet is high and the wind speed is low, resulting in their body temperature not being effectively reduced and heart rate remaining high, and the expression is slightly anxious, then a lower reward value is given, such as r t = 0.2. These real-time reward values are fed back into the deep reinforcement learning model to participate in subsequent policy optimization.

[0071] The discount factor γ = 0.95 plays a role in balancing current and future rewards, meaning that when the model optimizes the policy, it not only focuses on the immediate feedback of the current user but also considers the coherence and optimization potential of the user experience in a subsequent period. For example, when the analysis module adjusts the wind speed of a certain air outlet, although the current user may not immediately show a significant improvement in satisfaction, in the long run, this adjustment helps to maintain the stability of the overall indoor thermal comfort environment, and the model will reasonably weigh the potential benefits brought by this adjustment according to γ.

[0072] The entire regulation process starts from when the user enters the room and turns on the air conditioner until the user leaves the room or no longer triggers the air conditioner adjustment requirement for a long time. This is a complete episode, and the duration is recorded as T. During this process, in order to continuously optimize the parameters θ of the policy network, the gradient ascent method is used. The learning rate α = 0.01, which determines the step size of each parameter update. For example, when the model receives a series of reward signals generated by user feedback, according to the formula of the gradient ascent method slowly but steadily adjusts θ. If after a certain adjustment, it is found that the subsequent generated air conditioner control instructions can keep more users in a comfortable state for a longer time and obtain a higher total reward, it means that the parameter adjustment direction is correct, and the model continues to optimize in this direction; otherwise, the optimization direction is finely tuned, and through continuous iteration, the control instructions generated by the analysis module become more and more accurate.

[0073] Finally, through repeated training and optimization of the deep reinforcement learning model, the analysis module can generate refined second control signals for the air supply vents associated with each user according to the personalized needs differences of different users (the target temperature initially set indoors for the first control signal is 26°C). For example, for the No. 3 air supply vent near User A, a second control signal is generated that includes adjusting the temperature to 22°C, the wind speed to 0.8 m / s, and the wind direction to blow slightly downward; for the No. 1 air supply vent corresponding to User B, an instruction is issued to set the temperature to 20°C, increase the wind speed to 2 m / s, and blow the wind horizontally straight; the No. 5 air supply vent where User C is located receives a control signal with the temperature set to 21°C, the wind speed of 0.5 m / s, and the wind direction spreading in a circular pattern, so as to ensure that different users can enjoy an air-conditioning environment that meets their own needs in different areas of the living room.

[0074] It also includes a network module for connecting the smart home air-conditioning control system to the external network, allowing users to remotely access and control the air-conditioning system through mobile devices or computers. For example, users can remotely access and control the air-conditioning system through mobile devices such as smartphones, tablets, or laptops. At the same time, users can also adjust the air-conditioning operation parameters at any time when they are out, ensuring that they can enjoy a comfortable environment when they get home; it is also used to regularly obtain the latest weather forecast and outdoor environment data of the user's location from the Internet, such as temperature, humidity, air quality index, etc., so that the system can more intelligently preset and adjust the air-conditioning operation parameters.

[0075] The analysis module is also used to continuously analyze the long-term usage habits and preferences of users based on deep learning, and continuously optimize the control logic; it is also used to identify and learn the special needs of users under different times, seasons, and weather conditions, and automatically adjust the control strategy.

[0076] It also includes a health monitoring module for monitoring user health data, at least including heart rate and body temperature, and then based on the information in the user's personal data packet, performing personalized air-conditioning parameter settings. For the monitoring of heart rate and body temperature, users can wear smart health monitoring devices for real-time monitoring, such as devices like REDMI Watch 5.

[0077] The wind direction adjustment module also monitors the pose information of the wind direction deflectors at each air outlet based on pan-tilt technology. As an important technology in the field of intelligent monitoring and automatic control, pan-tilt technology endows the wind direction adjustment module with higher flexibility and accuracy. Through the built-in gyroscope, acceleration sensor, and high-precision motor drive system, the pan-tilt can stably track and record every subtle movement of the wind direction deflector, ensuring not only the stability and accuracy of the wind direction deflector during adjustment but also enabling the system to obtain and feedback the current pose information of the wind direction deflector in real time. For example, when the analysis module issues an adjustment instruction according to the user's preferences and current environmental conditions, the wind direction adjustment module will immediately activate the pan-tilt technology to accurately position the wind direction deflector of the target air outlet. Subsequently, through a calculation and calibration process, the pan-tilt will drive the wind direction deflector to be adjusted according to the preset angle and position, and during the entire adjustment process, the pan-tilt will continuously monitor and record the pose information of the wind direction deflector to ensure that the adjustment result meets the user's expectations.

[0078] Embodiment 2:

[0079] The difference from Embodiment 1 lies in that the air conditioner control method based on smart home is as follows:

[0080] Step 1, information acquisition and analysis: The image acquisition module acquires the image of the smart home environment, and then the analysis module identifies the image information, determines whether there is a user, and obtains the user identity information, location information, and behavior information. At the same time, the health monitoring module acquires the user's expression, clothing characteristics, and health data, and calculates the dynamic thermal comfort. These information are weighted and combined to obtain the corresponding comprehensive information;

[0081] Step 2, demand matching: Based on the user personal data packet in the database module, combined with the comprehensive information in Step 1, and jointly with the latest weather forecast, outdoor environmental data, and indoor environmental parameters obtained by the network module, the operation parameters of the temperature adjustment module, the operation parameters of the wind direction adjustment module, and the operation parameters of the air conditioner module are automatically adjusted in real time; and continuously collect the special needs of users under different conditions to optimize the control logic based on deep learning, and add the learning results to the user personal data packet;

[0082] Step 3, visualization and remote control: Use the visualization module to establish a three-dimensional map of the user's smart home environment, and display the temperature, humidity, parameters of each air outlet, and air conditioner operation parameters of each room in the three-dimensional map. At the same time, the user can use the network module to remotely access and control the entire air conditioner system through a remote terminal.

[0083] Obviously, the above embodiments are merely examples given for clear illustration and not limitations on the implementation manners. For those of ordinary skill in the art, other different forms of changes or alterations can be made based on the above description. It is not necessary and impossible to enumerate all implementation manners here. And the obvious changes or alterations derived therefrom still fall within the protection scope of this invention.

Claims

1. The air conditioning control system based on smart home is characterized by: include: A database module, used to establish a user personal data package, including at least the user's age, gender, clothing, preferences, body temperature, heart rate and activity pattern; The air conditioning module is used to cool the air through a refrigeration cycle or heat the air through a heating cycle, and the air after the heat and humidity treatment in the indoor unit is defined as the first gas; An image acquisition module, used to acquire an image of a user's smart home environment, defined as a first image; An analysis module, configured to receive the first image and determine whether a user exists in the first image based on image processing technology; When there is a user in the first image, the user identity information is obtained based on face recognition technology, which is defined as the first search condition; It is also used to obtain the location information of the user in the smart home environment based on computer vision technology, which is defined as the second search condition; at the same time, the user behavior and clothing are identified and analyzed on the first image to obtain the third search condition; based on the first search condition, the second search condition and the third search condition, the user's personal data packet is searched accordingly, and a total control signal is output, wherein the total control signal includes a first control signal sent to the air conditioning module and a second control signal sent to the temperature adjustment module and the wind direction adjustment module; the air conditioning module is used to determine the target temperature set in the room according to the first control signal; A temperature adjustment module, for mixing the air in the room with the first gas in proportion based on the second control signal, and adjusting the air supply temperature of each air supply outlet to reach the required temperature of the air supply outlet; The wind direction adjustment module is used to adjust the position of the wind guide plate in the air supply port based on the second control signal.

2. The air conditioning control system based on smart home according to claim 1, characterized in that: It also includes a visualization module for establishing a three-dimensional map of the user's smart home environment, and displaying the parameters of each air outlet therein, wherein the air outlet parameters include at least position, wind speed, temperature and relative humidity; the analysis module is also used to obtain the facial expression information, skin temperature and clothing feature information of the user in the smart home environment based on the first image, and then determine whether the state parameters of the air being blown to the user meet the needs of the current user; when the air state parameters do not meet the needs of the current user, a third control signal is generated as feedback to control the wind direction adjustment module to adjust the wind direction, and at the same time, the temperature adjustment module is controlled to adjust the air temperature and speed of the corresponding air outlet.

3. The air conditioning control system based on smart home according to claim 2, characterized in that: The temperature regulating module comprises: A return air inlet, the return air inlet is arranged on one side of the air mixing chamber; The air mixing chamber is arranged at the front end of the air outlet of the air conditioner indoor unit. The air outlet of the air conditioner indoor unit is connected to a plurality of independent air mixing chambers. One side of the opening inhales the first gas which is heat-humidified and transported by the air conditioning module, and a turbulent flow structure is arranged inside the air mixing chamber to promote the full and uniform mixing of the two parts of air. A certain number of air supply ports are arranged on the other side. The fan is installed in the air mixing chamber and has the function of adjusting the speed and air volume. It is used to extract the mixed air in the mixing chamber, provide power for the air flow, ensure that the air flows to the air outlet and can adjust the wind speed; Temperature sensors and humidity sensors are respectively arranged at the return air outlet of the air mixing chamber, in the air mixing chamber and near the air supply outlet to monitor the air temperature and humidity of each key node in real time; The controller receives the second control signal from the analysis module and the data fed back by each temperature sensor, and accurately calculates the optimal mixing ratio of indoor air and the first gas to achieve the optimal air supply state point based on the temperature and humidity adjustment requirements contained in the second control signal and the temperature and humidity information monitored in real time. The controller also finely controls the mixing process and the air supply process by adjusting the fan speed, so that the air temperature, humidity and wind speed delivered from each air outlet accurately match the required parameters.

4. The air conditioning control system based on smart home according to claim 3, characterized in that: It also includes a network module for connecting the smart home air-conditioning control system to the external network, allowing users to remotely access and control the air-conditioning system through mobile devices or computers; it is also used to regularly obtain the latest weather forecast and outdoor environment data in the user's area from the Internet so that the system can more intelligently preset and adjust the air-conditioning operating parameters.

5. The air conditioning control system based on smart home according to claim 4, characterized in that: The analysis module is also used to continuously analyze users' long-term usage habits and preferences based on deep learning, and continuously optimize the control logic; it is also used to identify and learn users' special needs under different times, seasons, and weather conditions, automatically adjust control strategies, and add learning results to the database module.

6. The air conditioning control system based on smart home according to claim 5, characterized in that: It also includes a health monitoring module, which is a wearable device for monitoring user health data, including at least heart rate and body temperature, calculating the user's thermal comfort according to a dynamic thermal comfort model, and combining the facial expression information obtained by the analysis module, and then based on the thermal comfort expectations in the user's personal data package, performing personalized air-conditioning parameter settings, and sending the air-conditioning parameter settings to the air-conditioning module, the temperature adjustment module and the wind direction adjustment module; the temperature adjustment module is also used to monitor the air temperature and humidity in the room where each air outlet is located; the wind direction adjustment module is also used to monitor the position information of the wind guide plates of each air outlet based on the pan-tilt technology; the wind direction adjustment module based on the pan-tilt technology includes a gyroscope and an accelerometer, the gyroscope is used to sense the change in the rotation angle of the wind guide plate in three-dimensional space in real time; the accelerometer is used to capture the acceleration information during the movement of the guide plate; the visualization module is also used to display the temperature and relative humidity of different areas of each room on a three-dimensional graph.

7. The air conditioning control system based on smart home according to claim 6, characterized in that: The analysis module is also used for, when multiple users are detected in the same public area, first, for each user, based on their location information, accurately locating the relative position relationship between each user and the air outlet; then, deeply analyzing each user's behavior, clothing, preferences and body temperature and heart rate information obtained from the health monitoring module, correspondingly searching the personal data packets of all people in the room, comprehensively determining the target temperature and wind speed of the air-conditioning module, and sending a first control signal to the air-conditioning module; then predicting each user's personalized needs for temperature and wind speed according to the preset deep reinforcement learning model; if the user is reading in thin clothes, he may prefer a milder temperature and a lower wind speed; if the user has just finished exercising and is sweating profusely, he needs a lower temperature and a higher wind speed to cool down quickly; if the user is wrapped in a blanket watching TV, he may need a relatively high temperature and a breeze environment; finally, according to the differences in needs of different users, a refined second control signal is generated for the air outlet associated with each user, and the signal includes the temperature value, wind speed value and wind direction adjustment instruction that need to be adjusted for the corresponding air outlet.

8. The air conditioning control system based on smart home according to claim 7, characterized in that: The deep reinforcement learning model adopts an algorithm based on policy gradient. The specific formula is as follows: Where J(θ) is the objective function of the policy network, which is used to measure the policy π θ The strategy here is π θ That is, the analysis module generates the strategy of air conditioning control instructions based on user information and environmental information. θ is the parameter of the strategy network. The performance of the strategy is improved by optimizing θ. According to the strategy π θ Seek expectation; r t is the reward obtained at time t. The reward is given based on the user's satisfaction with the current air supply of the air conditioner. It is comprehensively evaluated by analyzing the user's expression changes, skin temperature changes, heart rate fluctuations and other physiological and behavioral feedback. If the user is relaxed, the skin temperature is close to the comfortable range, and the heart rate is stable, a higher reward value will be given, otherwise a lower reward value will be given; γ is a discount factor, which ranges from 0 to 1. It is used to weigh the relative importance of current rewards and future rewards. When it is close to 1, it indicates that future rewards have a greater influence, which means that the model pays more attention to long-term user experience optimization. It can be taken as 0.

95. T is the total duration of an episode, which can be understood as the time period from when the user enters the room and turns on the air conditioning to when the user leaves the room or the air conditioning adjustment demand is not triggered for a long time.

9. The air conditioning control system based on smart home according to claim 8, characterized in that: In order to optimize the parameters θ of the policy network, the gradient ascent method is used, and its update formula is: Among them, α is the learning rate, which ranges from 0 to 1 and controls the step size of each parameter update. The value can be 0.

01.

10. The air conditioning control method based on smart home is characterized in that: Using the system as described in any one of claims 1 to 9, Including the following steps: Step 1: Information acquisition and analysis: The image of the smart home environment is acquired through the image acquisition module, and then the analysis module identifies the image information, determines whether there is a user, and obtains the user's identity information, location information and behavior information. At the same time, the analysis module and the health monitoring module obtain the user's expression, clothing characteristics and health data, and calculate the dynamic thermal comfort according to the dynamic thermal comfort model to obtain the corresponding comprehensive information; Step 2: demand matching: based on the user's personal data package in the database module, combined with the comprehensive information in step 1, and the latest weather forecast and outdoor environment data obtained by the network module, the operating parameters of the temperature control module, wind direction control module and air conditioning module are automatically adjusted in real time; and the special needs of users under different conditions are continuously collected to optimize the control logic based on deep learning, and the learning results are added to the user's personal data package; Step three, visualization and remote control, use the visualization module to create a three-dimensional map of the user's smart home environment, and display the temperature and relative humidity of each room, the parameters of each air outlet and the air conditioning operation parameters in the three-dimensional map. At the same time, the user can use the network module to remotely access and control the entire air conditioning system through a remote terminal.

Images