Intelligent home environment temperature monitoring and self-adaptive regulation and control method

Abstract

The invention relates to a smart home environment temperature monitoring and self-adaptive regulation and control method, and belongs to the technical field of computers and temperature sensors. According to the method, physiological data such as the heart rate and the skin temperature of a user are collected through wearable equipment, and temperature and humidity, wind speed and radiation temperature obtained by distributed environment sensors are combined; the convolutional neural network is used to identify the activity state of the user; and a real-time thermal comfort index is calculated based on the multi-source data fusion model, and when the real-time thermal comfort index exceeds a preset comfort interval, a regulation and control instruction is dynamically generated to adjust the air supply temperature, the air speed and the direction. The system also supports a personalized preference learning and multi-user conflict coordination mechanism. According to the invention, the thermal comfort experience is improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of computer and temperature sensor, and particularly relates to an intelligent home environment temperature monitoring and self-adaptive control method. BACKGROUND

[0002] With the popularization of intelligent home systems, indoor environment temperature control technology has gradually evolved from traditional timing or temperature mode to perception and intelligence. Current mainstream adaptive temperature control schemes mostly rely on temperature sensors, infrared human body detection or simple activity recognition, and the core logic is feedback regulation based on environmental temperature and preset comfort interval.

[0003] However, human thermal comfort is not only affected by environmental parameters, but also closely related to individual metabolic heat production level. In actual living scenarios, users may be in different activity states such as sitting, walking, cooking or exercising, and their metabolic equivalent (MET) differs significantly, resulting in completely different subjective thermal perception at the same environmental temperature. Existing temperature control systems lack accurate perception of users' real-time metabolic state, making it difficult to achieve personalized thermal environment regulation in a true sense.

[0004] Dynamic temperature control based on metabolic equivalent has become a key research direction to improve thermal comfort. Metabolic equivalent, as a standardized indicator to measure human energy consumption, can effectively reflect the user's current physiological heat load. Ideally, a temperature control system should be able to adjust the target temperature setting in real time according to the MET value to match the user's heat balance demand. However, traditional MET acquisition methods rely on wearable devices or questionnaires, the former is highly invasive and has low user compliance, and the latter is time-consuming and cannot support continuous regulation, both of which are difficult to apply to intelligent home scenarios that require non-invasive and all-weather operation.

[0005] Existing technologies still have obvious deficiencies in non-contact human activity perception and metabolic state inference. Although millimeter wave radar has penetration and high-precision micro-motion detection capabilities, it has not been effectively integrated with biomechanical models to realize quantitative estimation of MET; and fuzzy control strategies are mostly used to handle uncertainties in temperature control execution, but they do not form a closed-loop linkage with dynamic metabolic input. Therefore, in complex home environments, how to accurately identify user activity intensity through non-invasive sensing means and build a metabolic-driven adaptive temperature control mechanism accordingly has become a technical problem to be solved. SUMMARY

[0006] The application provides a smart home environment temperature monitoring and adaptive control method, aiming to solve the technical problem of inconsistent user comfort in different activity states under the same environment temperature caused by the traditional temperature control system ignoring the difference in human metabolic heat production. Existing temperature control systems generally rely on fixed set temperature or control based on environmental parameters such as room temperature, humidity, time, etc., without considering the actual thermal perception needs of individual users in different activity states, resulting in energy waste and user experience decline.

[0007] The application provides a smart home environment temperature monitoring and adaptive control method, comprising: Real-time acquisition of user physiological state data through wearable physiological sensing devices, including heart rate, skin temperature, body surface humidity, and respiratory rate; Real-time acquisition of environmental state data through indoor distributed environmental sensor arrays, including air temperature, relative humidity, air flow rate, and radiation temperature; Classification and identification of the current activity state of the user through a user behavior recognition module, including sitting, standing, walking, sleeping, and high-intensity exercise; Based on the physiological state data, environmental state data, and activity state, a multi-source heterogeneous data fusion model is constructed to calculate the real-time thermal comfort index of the user; According to the deviation of the real-time thermal comfort index and the preset comfort interval threshold, an environment temperature control instruction is generated and sent to the central air conditioning system or local temperature control execution unit to dynamically adjust the supply air temperature, air speed, and air supply direction of the target area.

[0008] Preferably, the real-time acquisition of user physiological state data through wearable physiological sensing devices includes: The wearable physiological sensing devices are integrated into smart bracelets, smart clothes, or smart patches, with high-precision thermistors, capacitive humidity sensors, photoelectric plethysmography sensors, and piezoelectric respiration sensors inside; The physiological state data is acquired and transmitted to the home central processing unit through the Bluetooth Low Energy protocol.

[0009] Preferably, the real-time acquisition of environmental state data through indoor distributed environmental sensor arrays includes: At least 3 environmental perception nodes are evenly distributed at different heights and orientations in the target space, each node containing a digital temperature and humidity sensor, an ultrasonic wind speed sensor, and an infrared radiation temperature sensor; The environmental state data is acquired at a sampling period of 1 time per 2 seconds and transmitted to the home central processing unit through power line carrier communication or ZigBee protocol.

[0010] Preferably, the current activity state of the user is classified and recognized by a user behavior recognition module, including: receiving three-axis acceleration and angular velocity time series data in a continuous time as an input tensor; inputting the input tensor into a pre-trained convolutional neural network model, outputting a probability distribution of five activity states after two convolutional layers, one pooling layer and one fully connected layer; taking the maximum probability as the current activity state determination result, and the five activity states are sitting, standing, walking, sleeping and high-intensity exercise.

[0011] Preferably, based on the physiological state data, environmental state data and activity state, a multi-source heterogeneous data fusion model is constructed to calculate the real-time thermal comfort index of the user, including: calculating the basic thermal load according to the PMV-PPD thermal comfort model defined in ASHRAE Standard 55, and the metabolic rate is obtained from the ISO8996 standard according to the recognized activity state; introducing the difference between skin temperature and air temperature as a local thermal discomfort compensation term, and normalizing it through a nonlinear Sigmoid function; weighting and fusing the normalized compensation term and the PMV value to obtain a dimensionless real-time thermal comfort index.

[0012] Preferably, the difference between skin temperature and air temperature is introduced as a local thermal discomfort compensation term, and normalized through a nonlinear Sigmoid function, including: local thermal discomfort compensation term defined as skin temperature minus air temperature; substitute the formula ; normalize, the steepness coefficient, the neutral offset; the normalized result is used to correct the predicted average vote value , to obtain the real-time thermal comfort index , the local compensation weight coefficient.

[0013] Preferably, according to the deviation of the real-time thermal comfort index and the preset comfort interval threshold, an environmental temperature control instruction is generated, including: set the preset comfort interval threshold to the real-time thermal comfort index; when the real-time thermal comfort index is greater than +0.5, generate a cooling instruction; when the real-time thermal comfort index is less than -0.5, generate a heating instruction; The target air supply temperature adjustment range is determined according to a linear proportional relationship based on the real-time thermal comfort index deviation degree.

[0014] Preferably, the generated environment temperature regulation instruction further comprises: The air supply wind speed is divided into three grades of low, medium and high. Based on the three-dimensional geometric relationship between the current position coordinates of the user and the air outlet of the air conditioner, the deflection angle of the air deflector is calculated by vector projection method to determine the optimal air supply direction.

[0015] Preferably, after the central air conditioning system or the local temperature control execution unit receives the regulation instruction, the refrigerant flow is adjusted by the variable frequency compressor to change the air supply temperature, the rotating speed is adjusted by the direct current brushless fan to change the air speed, and the air deflector mechanism is driven by the stepping motor to change the air supply direction.

[0016] Preferably, the method further comprises recording the operation of the user to actively modify the set temperature within 5 minutes after the regulation, and marking the regulation event as an unsatisfied sample. The personalized metabolic rate correction coefficient of the user is updated by using an online gradient descent algorithm.

[0017] Compared with the prior art, the present application has the following advantages: 1. By fusing wearable physiological sensing data, distributed environment sensing data and high-precision user activity state recognition results, a dynamic temperature control model is constructed with individual thermal sensation as the core, breaking through the limitation of traditional temperature control systems relying only on environmental parameters. 2. The physical correction based on the standard metabolic rate and the individualized preference learning mechanism are introduced, and the difference in human body heat production under different activity states is accurately compensated. 3. A multi-user conflict coordination strategy is adopted to effectively solve the contradiction of individualized temperature control demand in a shared space. 4. The overall scheme significantly improves the thermal comfort experience of the user, and at the same time, by avoiding invalid overcooling or overheating operation, the energy consumption of the HVAC system is reduced. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 The present application is an intelligent home environment temperature monitoring and self-adaptive regulation method. Figure 2 The present application is a core principle framework diagram of a multi-source heterogeneous data fusion model. Figure 3 The present application is a logic flow framework diagram of a user behavior recognition module. Figure 4 The present application is a logic flow framework diagram of real-time thermal comfort index calculation and regulation instruction generation. Figure 5This is a schematic diagram of the multi-level interaction and data flow between the home central processing unit, wearable devices, environmental sensor array, and temperature control execution unit in this invention; Figure 6 This is a logical flowchart of the thermal comfort demand conflict coordination mechanism in multi-user scenarios in this invention. Detailed Implementation

[0019] refer to Figures 1 to 6 This invention provides a method for monitoring and adaptively controlling the temperature of a smart home environment. Its core lies in constructing a dynamic temperature control model based on individual thermal sensation by integrating wearable physiological sensor data, distributed environmental perception data, and high-precision user activity state recognition results. This solves the technical problem of inconsistent user comfort under the same ambient temperature due to traditional temperature control systems neglecting differences in human metabolic heat production. The following will elaborate on the specific implementation of this method.

[0020] The method includes: Step S1: Collect the user's physiological state data in real time through wearable physiological sensing devices; Step S2: Acquire environmental status data in real time through an indoor distributed environmental sensor array; Step S3: Classify and identify the user's current activity status through the user behavior recognition module; Step S4: Based on the physiological state data, environmental state data, and activity state, construct a multi-source heterogeneous data fusion model to calculate the user's real-time thermal comfort index; Step S5: Based on the deviation between the real-time thermal comfort index and the preset comfort range threshold, generate an ambient temperature control command and send the control command to the central air conditioning system or local temperature control execution unit to dynamically adjust the air supply temperature, wind speed and air supply direction of the target area.

[0021] In step S1, the wearable physiological sensing device is a miniature sensing unit integrated into a smart bracelet, smart clothing, or smart patch. This miniature sensing unit contains a high-precision thermistor for measuring skin temperature, a capacitive humidity sensor for measuring surface humidity, a photoplethysmography (PPG) sensor for measuring heart rate, and a piezoelectric respiration sensor for measuring respiratory rate. All physiological signals are sampled at a frequency greater than 10 times per second and transmitted to a home central processing unit via Bluetooth Low Energy protocol.

[0022] The home central processing unit filters, denoises, and performs baseline correction on the received raw physiological signals to ensure signal quality meets the requirements of subsequent modeling. Skin temperature is measured from 32°C to 42°C with an accuracy of ±0.1°C; body surface humidity, expressed as a percentage of relative humidity, is measured from 10% to 90% with an accuracy of ±2%; heart rate is measured from 40 to 200 beats per minute with an accuracy of ±1 beat; and respiratory rate is measured from 6 to 40 beats per minute with an accuracy of ±0.5 beats per minute.

[0023] All physiological parameters are stored in the local cache with timestamp synchronization and uploaded to the main memory of the home central processing unit at fixed intervals.

[0024] In step S2, the indoor distributed environmental sensor array consists of no fewer than three environmental sensing nodes, each evenly distributed at different heights and orientations within the target space to cover the three-dimensional space of the user's usual residence area. Each node includes a digital temperature and humidity sensor, an ultrasonic anemometer, and an infrared radiation temperature sensor.

[0025] Digital temperature and humidity sensors are used to measure air temperature and relative humidity. The air temperature measurement range is 0 to 50 degrees Celsius with an accuracy of ±0.2 degrees Celsius, and the relative humidity measurement range is 5% to 95% with an accuracy of ±1.5%. Ultrasonic anemometers are used to measure air velocity, with a measurement range of 0 to 5 meters per second and an accuracy of ±0.1 meters per second. Infrared radiation temperature sensors are used to measure the average radiation temperature of surrounding object surfaces, with a measurement range of 0 to 60 degrees Celsius and an accuracy of ±0.5 degrees Celsius.

[0026] Each node establishes a bidirectional data link with the home central processing unit via wired power line carrier communication or wireless ZigBee protocol, with environmental data sampling every 2 seconds. The home central processing unit performs spatial weighted averaging on the data from each node, with the weighting coefficients determined based on the reciprocal of the Euclidean distance between the node and the user's current location to highlight the impact of near-field environmental parameters on the user's thermal perception.

[0027] In step S3, the user behavior recognition module uses fused data from a triaxial accelerometer and a gyroscope as its input source. The triaxial accelerometer and gyroscope are integrated into a wearable physiological sensing device to continuously collect motion data of the user's torso or limbs.

[0028] The user behavior recognition module receives acceleration and angular velocity time-series data over 30 consecutive seconds, forming an input tensor of dimension 300 x 6 (10 frames per second, 300 frames in 30 seconds, each frame containing triaxial acceleration and triaxial angular velocity). This input tensor is fed into a pre-trained convolutional neural network model, which contains two convolutional layers, a max-pooling layer, and a fully connected layer.

[0029] The first convolutional layer uses 16 one-dimensional convolutional kernels of size 5, and the second convolutional layer uses 32 one-dimensional convolutional kernels of size 3. The activation function for both layers is the modified linear unit.

[0030] The pooling layer employs max pooling with a window size of 2. The fully connected layer outputs a dimension of 5, corresponding to 5 activity states: sitting, standing, walking, sleeping, and high-intensity exercise. The model output is a probability distribution for each activity state, and the state with the highest probability is selected as the current activity state determination. The convolutional neural network model was trained and validated on an action database containing no fewer than 10,000 subjects before deployment, achieving a classification accuracy greater than 95%.

[0031] The home central processing unit triggers a behavior recognition process every 30 seconds and stores the recognition results in a structured message format in the user state context buffer.

[0032] In step S4, the multi-source heterogeneous data fusion model is a hybrid model based on a combination of physical mechanisms and data-driven approaches. This hybrid model first calculates the basic thermal load according to the PMV-PPD thermal comfort model defined in ASHRAE Standard 55. The inputs to the PMV model include air temperature, mean radiant temperature, relative humidity, air velocity, clothing thermal resistance, and metabolic rate.

[0033] The thermal resistance of the clothing is set to a default value of 0.6clo, suitable for typical home wear. The initial metabolic rate value is obtained from the ISO8996 standard based on the activity status identified in step S3: 1.0 MET for sitting, 1.2 MET for standing, 2.5 MET for walking, 0.7 MET for sleeping, and 4.0 MET for high-intensity exercise.

[0034] Based on this, the model introduces the difference between skin temperature and ambient temperature as a compensation term for local thermal discomfort. Defined as skin temperature minus air temperature, its physical meaning lies in reflecting the driving force of heat exchange between the body surface and the environment. When When the value is too high, even if the overall PMV value is within the comfortable range, users may still experience localized overheating or overcooling. To quantify this effect, a nonlinear sigmoid function is introduced. Normalization is performed: ; This is the steepness coefficient, with a value of 0.5; This is a neutral offset, with a value of 2 degrees Celsius. The final real-time thermal comfort index. Calculated by the following formula: ; To predict the average vote count, This is the local compensation weighting coefficient, with a value of 0.8. This is a dimensionless numerical value, ranging from -2 to +2, with a value of 0 indicating a thermally neutral state. Upon receiving updated physiological, environmental, and activity status data, the home central processing unit immediately invokes this fusion model to calculate the TCI and writes the result to the thermal comfort state register.

[0035] In step S5, the preset comfort zone threshold is a real-time thermal comfort index between -0.5 and +0.5. When the calculated real-time thermal comfort index is greater than +0.5, the user is determined to be in a state of thermal discomfort, and the system generates a cooling command; when the real-time thermal comfort index is less than -0.5, the user is determined to be in a state of cold discomfort, and the system generates a heating command.

[0036] The specific parameters of the control command include the target supply air temperature adjustment range, the supply air velocity level, and the deflection angle of the air guide vanes. Temperature adjustment range Determined based on a linear proportional relationship according to the deviation of the real-time thermal comfort index: ; This is the sensitivity coefficient, with a value of 2 degrees Celsius per unit. The fan speed is divided into three levels: low, medium, and high, corresponding to fan speeds of 1000 rpm, 1500 rpm, and 2000 rpm, respectively. The deflection angle of the air guide vane is calculated in real time based on the user's spatial coordinates and the geometric relationship between the air conditioner's vents and the air outlet. The home central processing unit obtains the user's current location through the indoor positioning module, combines it with the three-dimensional coordinates of the air conditioner's vents, and uses vector projection to calculate the optimal airflow direction, ensuring that the airflow directly acts on the user's skin without causing discomfort from direct airflow.

[0037] After receiving the control command, the central air conditioning system or local temperature control execution unit adjusts the refrigerant flow rate through the variable frequency compressor to change the air supply temperature, adjusts the speed of the DC brushless fan to change the wind speed, and drives the air guide mechanism through the stepper motor to change the air supply direction.

[0038] The response delay for all actions is less than 3 seconds. During continuous control, the minimum time interval between two adjacent temperature adjustment commands is no less than 30 seconds to avoid system oscillation. This time interval is enforced by the command scheduler. Any newly generated temperature adjustment command that is less than 30 seconds old is temporarily stored in the command queue and issued only after the time window is met.

[0039] Furthermore, the home central processing unit incorporates a user-individualized thermal comfort preference learning module. This module records the user's manual feedback after each adjustment. If the user actively modifies the set temperature within 5 minutes of adjustment, this adjustment event is marked as an unsatisfactory sample. The system extracts the corresponding data from the unsatisfactory samples. The user's personalized metabolic rate correction coefficient is updated using an online gradient descent algorithm based on the user's metabolic rate value, activity status, environmental parameters, and original control commands.

[0040] The initial value of the personalized metabolic rate correction coefficient adopts the ISO8996 standard value, and is subsequently dynamically adjusted based on historical feedback data, with an adjustment step of 5% each time and an adjustment upper limit of ±30% of the standard value. For example, if a user frequently lowers the temperature while sitting, the system infers that their actual metabolic rate is greater than 1.0 MET, and gradually increases the correction coefficient to 1.15 MET.

[0041] In multi-user scenarios, the method also includes a conflict coordination mechanism. When multiple users exist in the same space and their real-time thermal comfort indices are outside the comfort range and their adjustment directions are opposite, the system calculates the degree of thermal discomfort for each user. absolute value .like If the difference is less than 0.2, priority will be given to meeting the needs of those with the greater absolute value of thermal discomfort; if the difference is less than 0.2, a weighted average strategy will be used to generate a compromise control instruction. Weighting coefficients Related to user age and health status: The weighting coefficient for elderly users and those with chronic diseases is set at 1.5, for ordinary adult users at 1, and for children at 1.2. This is a compromise target. Calculated by the following formula: ; The system is based on The control commands are recalculated to achieve Pareto optimization of the thermal comfort needs of multiple users.

[0042] The implementation of the above method relies on a complete system architecture. This system includes wearable physiological sensing devices, an indoor distributed environmental sensor array, a user behavior recognition module, a multi-source heterogeneous data fusion model, a real-time thermal comfort index calculation unit, a control command generator, a central air conditioning system or local temperature control execution unit, and a home central processing unit. The home central processing unit, as the core control node, runs a real-time operating system and has multi-threaded task scheduling capabilities to ensure the time synchronization of each data acquisition, processing, and control stage.

[0043] All communication links employ encrypted transmission protocols to ensure user privacy and security. Upon system startup, a self-test program is executed to verify the connectivity and functional integrity of each sensor and actuator. During normal operation, the system generates an operation log every minute, recording key parameters and events for fault diagnosis and performance optimization.

[0044] In summary, this embodiment achieves accurate perception and dynamic response to individual user thermal sensations by strictly following the sequence of steps S1 to S5. Each step in the method includes detailed sub-steps, parameter definitions, algorithm formulas, and exception handling mechanisms to ensure the feasibility and robustness of the technical solution. Although the system part is not excessively elaborated, the necessary components and their interaction logic have been disclosed, meeting the requirements for full disclosure in patents.

[0045] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0046] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for monitoring and adaptively controlling ambient temperature in smart homes, characterized in that, include: Wearable physiological sensing devices collect users' physiological status data in real time, including heart rate, skin temperature, body surface humidity, and respiratory rate. The environmental status data is acquired in real time through an indoor distributed environmental sensor array, including air temperature, relative humidity, air velocity, and radiation temperature. The user behavior recognition module classifies and identifies the user's current activity state, which includes sitting, standing, walking, sleeping, and high-intensity exercise. Based on the physiological state data, environmental state data, and activity status, a multi-source heterogeneous data fusion model is constructed to calculate the user's real-time thermal comfort index. Based on the deviation between the real-time thermal comfort index and the preset comfort range threshold, an ambient temperature control command is generated and sent to the central air conditioning system or local temperature control execution unit to dynamically adjust the supply air temperature, wind speed and air supply direction of the target area.

2. The method for monitoring and adaptively controlling the temperature of a smart home environment according to claim 1, characterized in that, Real-time collection of users' physiological state data through wearable physiological sensing devices, including: The wearable physiological sensing device is integrated into a smart bracelet, smart clothing or smart patch, and is equipped with a high-precision thermistor, a capacitive humidity sensor, a photoplethysmography pulse wave sensor and a piezoelectric respiration sensor. The physiological state data is acquired and transmitted to the home central processing unit via Bluetooth Low Energy protocol.

3. The method for monitoring and adaptively controlling the temperature of a smart home environment according to claim 2, characterized in that, Real-time environmental status data is acquired through an indoor distributed environmental sensor array, including: No fewer than three environmental sensing nodes are evenly distributed at different heights and orientations in the target space. Each node includes a digital temperature and humidity sensor, an ultrasonic wind speed sensor, and an infrared radiation temperature sensor. The environmental status data is acquired at a sampling period of once every 2 seconds and transmitted to the home central processing unit via power line carrier communication or ZigBee protocol.

4. The method for monitoring and adaptively controlling the temperature of a smart home environment according to claim 3, characterized in that, The user behavior recognition module categorizes and identifies the user's current activity status, including: It receives triaxial acceleration and angular velocity time-series data over a continuous time period as input tensors; The input tensor is input into a pre-trained convolutional neural network model, which outputs a probability distribution of five activity states after passing through two convolutional layers, one pooling layer, and one fully connected layer. The activity state with the highest probability is taken as the current activity state determination result. The five activity states are sitting, standing, walking, sleeping, and high-intensity exercise.

5. The method for monitoring and adaptively controlling the temperature of a smart home environment according to claim 4, characterized in that, Based on the aforementioned physiological state data, environmental state data, and activity status, a multi-source heterogeneous data fusion model is constructed to calculate the user's real-time thermal comfort index, including: The baseline heat load was calculated based on the PMV-PPD thermal comfort model defined by ASHRAE Standard 55, and the metabolic rate was obtained from the ISO 8996 standard based on the identified activity status. The difference between skin temperature and air temperature is introduced as a local thermal discomfort compensation term and normalized using a nonlinear Sigmoid function. The normalized compensation term is weighted and fused with the PMV value to obtain a dimensionless real-time thermal comfort index.

6. The method for monitoring and adaptively controlling the temperature of a smart home environment according to claim 5, characterized in that, The difference between skin temperature and air temperature is introduced as a compensation term for local thermal discomfort, and normalized using a nonlinear Sigmoid function, including: Local thermal discomfort compensation item Defined as skin temperature minus air temperature; Substitute into the formula ; Normalize, This is the steepness coefficient. This is the neutral offset; normalization results Used to correct the predicted average vote value The value is used to obtain the real-time thermal comfort index. , This represents the local compensation weighting coefficient.

7. The method for monitoring and adaptively controlling the temperature of a smart home environment according to claim 6, characterized in that, Based on the deviation between the real-time thermal comfort index and the preset comfort range threshold, an ambient temperature control command is generated, including: Set the preset comfort zone threshold as the real-time thermal comfort index; When the real-time thermal comfort index is greater than +0.5, a cooling command is generated; when the real-time thermal comfort index is less than -0.5, a heating command is generated. The target supply air temperature adjustment range is determined according to the deviation of the real-time thermal comfort index in a linear proportional relationship.

8. The method for monitoring and adaptively controlling the temperature of a smart home environment according to claim 7, characterized in that, The generation of ambient temperature control instructions also includes: The air supply speed is divided into three levels: low, medium, and high. Based on the user's current location coordinates and the three-dimensional geometric relationship between the air conditioner's air outlet, the deflection angle of the air guide plate is calculated using the vector projection method to determine the optimal air delivery direction.

9. The method for monitoring and adaptively controlling the temperature of a smart home environment according to claim 8, characterized in that, After receiving the control command, the central air conditioning system or local temperature control execution unit adjusts the refrigerant flow rate through the variable frequency compressor to change the air supply temperature, adjusts the speed of the DC brushless fan to change the wind speed, and drives the air guide mechanism through the stepper motor to change the air supply direction.

10. The method for monitoring and adaptively controlling the temperature of a smart home environment according to claim 9, characterized in that, The method also includes recording user actions to actively modify the set temperature within 5 minutes after adjustment, and marking such adjustment events as unsatisfactory samples; The user's personalized metabolic rate correction coefficient is updated using an online gradient descent algorithm.

Images