Air supply control method and device of air conditioner and computer readable storage medium

Abstract

The invention provides an air supply control method and device of an air conditioner and a computer readable storage medium. According to the scheme, personnel position information, action state information and temperature and humidity distribution information in a current environment are acquired; performing feature fusion on the personnel position information, the action state information and the temperature and humidity distribution information to obtain a fusion feature, and determining a credibility value of the fusion feature; if the credibility value is larger than or equal to a preset threshold value, an air supply instruction is generated according to the fusion features; and adjusting air supply parameters of the air conditioner according to the air supply instruction, wherein the air supply parameters comprise air speed, air supply angle and temperature. According to the scheme, the problem that in the prior art, the accuracy of air supply positioning of an intelligent air conditioner in a complex environment is low is solved.

Description

Technical Field

[0001] This application relates to intelligent air conditioning air supply control technology, specifically, to an air conditioning air supply control method, an air conditioning air supply control device, and a computer-readable storage medium. Background Technology

[0002] Modern intelligent air conditioning systems have begun to adopt airflow control technologies based on human detection, such as using infrared sensors to locate human positions or using temperature and humidity sensors to provide zoned airflow within a room. However, as users' demands for comfort and intelligence increase, the limitations of existing intelligent air conditioning systems are gradually becoming apparent.

[0003] In existing technologies, the air supply control methods of smart air conditioners in complex environments have limitations, especially inaccurate air supply positioning. Summary of the Invention

[0004] The main objective of this application is to provide an air supply control method, an air supply control device, and a computer-readable storage medium for an air conditioner, so as to at least solve the problem of low accuracy of air supply positioning in complex environments of existing smart air conditioners.

[0005] To achieve the above objectives, according to one aspect of this application, an air supply control method for an air conditioner is provided, comprising: acquiring personnel location information, action status information, and temperature and humidity distribution information in the current environment; performing feature fusion on the personnel location information, the action status information, and the temperature and humidity distribution information to obtain fused features, and determining a confidence value of the fused features; if the confidence value is greater than or equal to a preset threshold, generating an air supply command based on the fused features; and adjusting the air supply parameters of the air conditioner according to the air supply command, wherein the air supply parameters include wind speed, air supply angle, and temperature.

[0006] Optionally, if the confidence value is greater than or equal to a preset threshold, then generating an air supply instruction based on the fusion features includes: if the confidence value is greater than or equal to the preset threshold, inputting the fusion features into a human comfort model to calculate the comfort index of the current environment; determining an air supply strategy based on the comfort index of the current environment, and generating the air supply instruction based on the air supply strategy, wherein the air supply strategy includes strategies for adjusting the wind speed, the air supply angle, and the temperature.

[0007] Optionally, the method further includes: acquiring photovoltaic system power generation data and inputting the photovoltaic system power generation data into a photovoltaic adequacy model to obtain a photovoltaic adequacy index, wherein the photovoltaic adequacy index characterizes the degree of matching between the power generation capacity of the photovoltaic system and the load demand of the air conditioner; adjusting the operating parameters of the air conditioner according to the photovoltaic adequacy index and a preset load scheduling algorithm, wherein the operating parameters include at least the compressor speed.

[0008] Optionally, the operating parameters of the air conditioner are adjusted according to the photovoltaic abundance index and a preset load scheduling algorithm. The operating parameters include at least the compressor speed, including: when the photovoltaic abundance index is greater than a first preset threshold, adjusting the compressor speed to a first speed, which is set based on the compressor's maximum speed; when the photovoltaic abundance index is less than or equal to the first preset threshold but greater than a second preset threshold, adjusting the compressor speed to the rated speed; when the photovoltaic abundance index is less than or equal to the second preset threshold, starting grid auxiliary power supply and adjusting the compressor speed and air supply intensity to a preset standard, which is determined based on the air conditioner's operating cost and comfort requirements.

[0009] Optionally, the method further includes: using millimeter-wave radar to detect the micro-motion characteristics of the human body in the current environment based on the Doppler effect, obtaining a Doppler frequency shift value, and comparing the Doppler frequency shift value with a preset frequency change threshold; using an infrared array to detect whether the body temperature of the person in the current environment is abnormal; and when the Doppler frequency shift value is greater than the preset frequency change threshold, or the body temperature of the person in the current environment is abnormal, adjusting the wind speed, the air delivery angle, and the temperature to preset wind speed range, preset air delivery angle range, and preset temperature range, respectively.

[0010] Optionally, before adjusting the wind speed, the air supply angle, and the temperature to preset wind speed range, preset air supply angle range, and preset temperature range respectively, when the Doppler frequency shift value is greater than the preset frequency change threshold, or when the body temperature of the person in the current environment is abnormal, the method further includes: matching the human micro-motion characteristics with dangerous actions in a preset typical dangerous action feature library; if the human micro-motion characteristics successfully match a first dangerous action in the preset typical dangerous action feature library, then obtaining the preset wind speed range, preset air supply angle range, and preset temperature range corresponding to the first dangerous action, wherein the preset typical dangerous action feature library includes multiple dangerous actions, and the preset typical dangerous action feature library is constructed at least based on historical user behavior data.

[0011] Optionally, the method further includes: if the confidence value is less than the preset threshold, then performing data re-verification, wherein the data re-verification includes activating a backup sensor, historical data backtracking, and time alignment compensation.

[0012] Optionally, before matching the human micro-motion characteristics with dangerous actions in a preset typical dangerous action feature library, the method further includes: acquiring the historical user behavior data and using a clustering analysis algorithm to identify dangerous behavior patterns from the historical user behavior data, wherein the historical user behavior data includes operation-related behavior data and location-related behavior data; using Failure Mode and Effects Analysis to assess the risk level of the historical user behavior data, and determining the user behavior corresponding to the risk level greater than a preset risk level as the dangerous action; and combining multiple dangerous actions into the preset typical dangerous action feature library.

[0013] According to another aspect of this application, an air supply control device for an air conditioner is provided, comprising: an acquisition unit for acquiring personnel location information, action status information, and temperature and humidity distribution information in the current environment; a fusion unit for fusing the personnel location information, action status information, and temperature and humidity distribution information to obtain fused features, and determining a confidence value of the fused features; a generation unit for generating an air supply command based on the fused features if the confidence value is greater than or equal to a preset threshold; and a first adjustment unit for adjusting the air supply parameters of the air conditioner according to the air supply command, wherein the air supply parameters include wind speed, air supply angle, and temperature.

[0014] According to another aspect of this application, a computer-readable storage medium is provided, the computer-readable storage medium including a stored program, wherein, when the program is executed, it controls the device where the computer-readable storage medium is located to perform any of the air supply control methods of the air conditioner described above.

[0015] This application's technical solution acquires personnel location information, action status information, and temperature and humidity distribution information within the current environment. It then fuses these information to obtain fused features and determines the reliability value of these fused features. If the reliability value is greater than or equal to a preset threshold, an air supply command is generated based on the fused features. The air supply parameters of the air conditioner, including wind speed, air supply angle, and temperature, are adjusted according to the air supply command. This solution integrates the personnel location, action status, and temperature and humidity distribution within the current environment. The generated feature fusion vector, combined with a dynamic weighting mechanism, ensures the accuracy and reliability of the fused information. By comparing it with a preset threshold, it intelligently determines the generation of an air supply command, thereby adjusting the air supply parameters of the air conditioner. This improves the air supply positioning accuracy of intelligent air conditioners in complex environments, thus solving the problem of low accuracy in air supply positioning of intelligent air conditioners in complex environments. Attached Figure Description

[0016] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0017] Figure 1 A hardware structure block diagram of a mobile terminal for performing an air supply control method for an air conditioner, according to an embodiment of this application, is shown.

[0018] Figure 2 A schematic flowchart of an air supply control method for an air conditioner according to an embodiment of this application is shown.

[0019] Figure 3 A flowchart illustrating the compensation mechanism of an air supply control method for an air conditioner according to an embodiment of this application is shown.

[0020] Figure 4 A system hardware architecture diagram of an air supply control system for an air conditioner provided according to an embodiment of this application is shown;

[0021] Figure 5 A flowchart illustrating a dual-buffering mechanism implementation scheme for an air supply control system for an air conditioner according to an embodiment of this application is shown.

[0022] Figure 6 A structural block diagram of an air supply control device for an air conditioner provided according to an embodiment of this application is shown.

[0023] The above figures include the following reference numerals:

[0024] 102. Processor; 104. Memory; 106. Transmission device; 108. Input / output device. Detailed Implementation

[0025] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0026] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present application.

[0027] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this application described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0028] As described in the background section, existing methods for controlling the air supply of smart air conditioners in complex environments have limitations, particularly inaccurate air supply positioning. To address the issue of low accuracy in air supply positioning of smart air conditioners in complex environments, embodiments of this application provide an air supply control method, an air supply control device, and a computer-readable storage medium.

[0029] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0030] The methods and embodiments provided in this application can be executed on a mobile terminal, computer terminal, or similar computing device. Taking running on a mobile terminal as an example, Figure 1 This is a hardware structure block diagram of a mobile terminal for an air conditioning air supply control method according to an embodiment of the present invention. Figure 1 As shown, a mobile terminal may include one or more ( Figure 1 Only one is shown in the diagram. A processor 102 (which may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.) and a memory 104 for storing data are also shown. The mobile terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the mobile terminal described above. For example, the mobile terminal may also include components that are more... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0031] The memory 104 can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the air supply control method of the air conditioner in this embodiment of the invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, thereby implementing the above-described method. The memory 104 may include high-speed random access memory and non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the above-described networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof. The transmission device 106 is used to receive or send data via a network. Specific examples of the above-described networks may include wireless networks provided by the mobile terminal's communication provider. In one example, the transmission device 106 includes a network interface controller (NIC), which can be connected to other network devices via a base station to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0032] This embodiment provides an air supply control method for an air conditioner that runs on a mobile terminal, computer terminal, or similar computing device. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Also, although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than that shown here.

[0033] Figure 2 This is a schematic flowchart of an air supply control method for an air conditioner according to an embodiment of this application. Figure 2 As shown, the method includes the following steps:

[0034] Step S201: Obtain personnel location information, action status information, and temperature and humidity distribution information in the current environment;

[0035] Specifically, millimeter-wave radar is used for human body localization. Leveraging the characteristics of its high-frequency electromagnetic waves, it can accurately detect the location of people even in environments with insufficient light or numerous obstacles. Infrared arrays can provide thermal images of the human body, thus inferring the approximate location of people. Especially when people are stationary or minimally active, the heat source localization of infrared sensors is more intuitive and effective. Millimeter-wave radar not only provides location information but can also analyze the speed and direction of human movement through the Doppler effect, thereby identifying the person's activity state, such as sitting, walking, or sleeping. Infrared arrays, combined with changes in temperature distribution, can also help determine whether a person is in motion. Comprehensive analysis of this information allows the air conditioning system to understand the activity patterns of people in the current environment, thereby formulating appropriate airflow strategies.

[0036] Temperature and humidity sensor arrays are distributed across different locations within the room, enabling real-time monitoring of the room's temperature and humidity distribution and the creation of a temperature and humidity distribution map. This distributed monitoring method not only captures temperature and humidity changes caused by localized heat sources but also reflects the overall temperature and humidity environment of the room, making it crucial for zoned airflow or overall temperature control strategies.

[0037] By integrating millimeter-wave radar, infrared arrays, and temperature and humidity sensors, the system can efficiently and accurately acquire the precise location and activity status (such as sitting, walking, or sleeping) of people in the current environment, as well as the distribution of indoor temperature and humidity. This not only enhances the sensitivity to human activity and environmental changes but also provides stable and reliable fused data as a basis for decision-making. This comprehensive and refined environmental perception capability lays the foundation for the subsequent intelligent generation of air supply commands closely related to human comfort and health.

[0038] Step S202: The above-mentioned personnel location information, above-mentioned action status information and above-mentioned temperature and humidity distribution information are fused to obtain fused features, and the confidence value of the above-mentioned fused features is determined.

[0039] Specifically, personnel location information comes from millimeter-wave radar or infrared sensors, providing coordinates and position data of personnel in space. Motion status information is derived from sensor-detected data analysis, including whether the person is sitting, walking, or asleep. Temperature and humidity distribution information is real-time temperature and humidity data provided by a group of temperature and humidity sensors, reflecting the thermal comfort level of the environment. This raw information from different sensors is preprocessed and converted into a unified feature representation. Then, a Graph Convolutional Network (GCN) algorithm is used for feature-level fusion to generate a fused feature that comprehensively reflects the environmental state and personnel behavior. This fused feature is represented by a fused feature vector.

[0040] After the fused features are generated, the quality of these fused data is further evaluated. The confidence value is determined through a dynamic weighting mechanism, reflecting the consistency, stability, and accuracy of the fused features under the current environmental conditions. The confidence value is calculated based on multiple factors, such as data consistency between sensors, the rate of environmental change, and the robustness of the algorithm, ensuring that only high-quality, high-reliability fused features can be used to generate air delivery commands. For example, the confidence value of millimeter-wave radar data is calculated and denoted as the first confidence value, and the confidence value of infrared array data is calculated and denoted as the second confidence value. Weights are assigned to the first and second confidence values ​​respectively; that is, the first and second confidence values ​​are weighted and summed to obtain a comprehensive confidence value, i.e., the confidence value of the fused feature. The weights can be determined based on the sensor's accuracy, the impact of environmental factors (such as lighting conditions) on the sensor, and the real-time nature of the data. The initial weights are the factory calibration values; for example, the initial weight for the confidence value of millimeter-wave radar data is 0.6, and the initial weight for the confidence value of infrared array data is 0.4. Based on environmental step compensation, the weight of millimeter-wave radar data reliability is increased when the temperature is above 25°C. If the radar signal-to-noise ratio is less than 15dB, the weight is automatically multiplied by 0.5. A dynamic adjustment mechanism is used to change the weight to adapt to sensor performance under different environments. For example, the weight of millimeter-wave radar increases under bright light or obstructed conditions, and decreases otherwise.

[0041] By following these steps, a more comprehensive and accurate understanding of the environment can be obtained, thereby guiding the adjustment of air supply parameters. This design not only enhances the air conditioner's intelligent response capabilities but also ensures the reliability of control decisions, avoiding system misjudgments caused by abnormal data from a single sensor. This further improves the accuracy and comfort of the air conditioner's air supply control, providing users with a more refined and personalized air supply experience.

[0042] Step S203: If the confidence value is greater than or equal to the preset threshold, then generate an air supply instruction based on the fusion features.

[0043] Specifically, if the confidence value of the calculated and optimized fused features reaches or exceeds a preset threshold, the current environmental perception data is considered sufficiently reliable to guide the adjustment of air supply parameters. This means that specific air supply instructions will be generated based on the fused features (which include comprehensive information on personnel location, movement status, and temperature and humidity distribution). The preset threshold is used to distinguish between data reliability and uncertainty. When the confidence value is lower than the preset threshold, the data quality is questioned, leading to additional data verification or re-collection measures to ensure that inappropriate control decisions are not made due to erroneous or incomplete data. When the confidence value of the fused features is greater than or equal to the preset threshold, this information is considered to accurately reflect the current environmental state, including the precise location of personnel, activity patterns (such as sitting, walking, sleeping), and indoor temperature and humidity distribution. Based on the fused features, a pre-trained human comfort model is invoked to calculate and determine the most suitable air supply instructions for the current environmental conditions and personnel status.

[0044] The setting of preset thresholds is primarily based on the reliability and accuracy requirements of environmental sensing data, as well as considerations for user comfort and safety. In air conditioning air supply control methods, the determination of preset thresholds is based on key factors such as sensor data quality and the robustness of the fusion algorithm. Specifically, this includes sensor characteristics, such as measurement errors, response times, and anti-interference capabilities of millimeter-wave radar, infrared arrays, and temperature and humidity sensors. It also considers the stability of sensor performance under different environmental conditions (such as light, temperature changes, and obstruction). For example, millimeter-wave radar has better penetration and anti-obstruction capabilities than infrared sensors in complex environments, but infrared sensors may have higher accuracy in human body positioning. Finally, it considers the accuracy and consistency of the fusion algorithm (such as G-CNN networks) in processing multi-source information. Finally, it considers the test results of the fusion effect, including the stability and correlation of the fused features under various scenarios. For example, in practical applications, the preset threshold is set at a point that balances the above considerations, such as 0.8, meaning that the air supply command will only be executed when the reliability of the fused features reaches at least 80%. The selection of this preset threshold is based on performance in testing, ensuring that only highly reliable data triggers control actions, thus avoiding unnecessary airflow adjustments while ensuring a rapid and accurate response in critical moments. This mechanism enables high-efficiency and energy-saving operation while guaranteeing user comfort and safety.

[0045] In summary, the generation of an air supply command is only triggered when the confidence value of the fused features is greater than or equal to a preset threshold, thereby guiding the air conditioner to make corresponding adjustments. This mechanism ensures the accuracy of control decisions and avoids invalid or harmful air supply adjustments caused by data quality issues, making it a key element in improving user experience and achieving personalized air supply control.

[0046] Step S204: Adjust the air supply parameters of the air conditioner according to the above air supply command. The above air supply parameters include wind speed, air supply angle and temperature.

[0047] Specifically, the air supply command includes a suggested fan speed based on the current environment and user status. For example, when the user is resting, the fan speed will be reduced to avoid discomfort caused by direct strong airflow; while under conditions of movement or in a hot environment, the fan speed will be increased to enhance the cooling effect. The fan speed parameter in the command is converted into a control signal for the air conditioner's internal fan, adjusting the motor speed and thus changing the actual air volume delivered. Airflow angle adjustment ensures that the airflow is delivered in the intended direction, avoiding direct airflow onto the body or ensuring that the airflow reaches the area requiring cooling. Based on the user's position and activity level, the angle of the air conditioner's air deflector is adjusted to face or avoid the user, achieving the best airflow effect without affecting user comfort. In emergency mode, if a user is detected to have fallen, a specific air supply mode will be switched, such as a roof reflection mode, to avoid direct airflow onto the user and cause secondary injury. The air supply command also includes adjusting to a suitable air supply temperature. Temperature adjustment involves the air conditioner's cooling or heating functions, sending control signals to the compressor, evaporator, or heating element to adjust their operating status to achieve the desired air supply temperature. In some scenarios, such as an elderly person's bedroom at night or a gym, temperature regulation is crucial for maintaining a comfortable and safe environment.

[0048] In this way, key air supply parameters such as wind speed, air supply angle, and temperature can be dynamically adjusted based on received air supply commands, achieving intelligent, efficient, and user-friendly air supply control. This mechanism not only improves the convenience and comfort of air conditioning use but also effectively saves energy and reduces unnecessary energy consumption, demonstrating the dual value of smart home appliances in improving quality of life and saving energy and reducing emissions.

[0049] This embodiment integrates the location, movement status, and temperature and humidity distribution of personnel within the current environment. The resulting fusion features ensure the accuracy and reliability of the fused information. Based on the fused environmental features and their reliability, and by comparing them with preset thresholds, the system autonomously decides whether to generate an air supply command. This allows for precise adjustment of the air conditioner's fan speed, air supply angle, and temperature, ensuring users enjoy a comfortable and personalized air supply experience in various situations. This improves the air supply positioning accuracy of smart air conditioners in complex environments, thus solving the problem of low accuracy in air supply positioning under such conditions.

[0050] In the specific implementation process, if the confidence value is greater than or equal to the preset threshold, an air supply instruction is generated based on the fusion features, including: if the confidence value is greater than or equal to the preset threshold, the fusion features are input into the human comfort model to calculate the comfort index of the current environment; an air supply strategy is determined based on the comfort index of the current environment, and the air supply instruction is generated based on the air supply strategy, wherein the air supply strategy includes strategies for adjusting the wind speed, the air supply angle, and the temperature.

[0051] Specifically, when the reliability value of the environmental perception data is greater than or equal to a preset threshold, it indicates that the currently collected information on personnel location, action status, and temperature and humidity distribution is sufficiently accurate and reliable. These fused features are then input into a pre-trained human comfort model. This human comfort model is based on an improved PMV-PPD (Predicted Mean Vote-Predicted Percentage of Dissatisfied) model, comprehensively considering the influence of factors such as temperature, humidity, and wind speed on human perception to calculate the comfort index of the current environment.

[0052] Based on the calculated comfort index, it can be determined whether the current environment meets the standards for human comfort. If there is a deviation, corresponding air supply strategies need to be developed to improve comfort. Air supply strategies include adjusting wind speed, adjusting the air supply angle, or changing the air supply temperature, aiming to bring the indoor environment to or close to the ideal state of human comfort, such as reducing direct airflow, improving the comfort of the airflow, and adjusting the distribution of warm and cool air. Once the air supply strategy is determined, air supply instructions are generated based on this strategy. These instructions contain specific parameters for adjusting wind speed, air supply angle, and temperature.

[0053] In summary, when the credibility of the fused features meets the requirements, the comfort level of the current environment is calculated using a human comfort model, and an air supply strategy is formulated and executed accordingly. This enables the dynamic adjustment of air supply parameters based on the indoor environment and the state of the users, ensuring that users experience optimal comfort under various conditions. This mechanism not only enhances the intelligence of the air conditioner but also effectively improves energy efficiency, reduces unnecessary energy consumption, and provides users with a more humanized and environmentally friendly user experience.

[0054] In some embodiments of this application, the above method further includes: acquiring photovoltaic system power generation data and inputting the photovoltaic system power generation data into a photovoltaic adequacy model to obtain a photovoltaic adequacy index, wherein the photovoltaic adequacy index characterizes the degree of matching between the power generation capacity of the photovoltaic system and the load demand of the air conditioner; adjusting the operating parameters of the air conditioner according to the photovoltaic adequacy index and a preset load scheduling algorithm, wherein the operating parameters include at least the compressor speed.

[0055] Specifically, real-time monitoring of photovoltaic (PV) system power generation data involves the actual output of electricity converted from the solar energy captured by the PV panels. Acquiring PV system power generation data is crucial for assessing the availability of renewable energy, specifically whether PV power generation capacity is sufficient and how much clean energy can be provided for smart air conditioning. The acquired PV system power generation data is input into a PV sufficiency model, which calculates the Grille Power Index (GPI). This index reflects the degree of matching between the PV system's power generation capacity and the immediate energy demand of the air conditioning system. The GPI calculation considers factors such as the real-time power generation of the PV system, the system efficiency coefficient, the air conditioning load demand, and the potential grid auxiliary power supply capacity, making it a key indicator for dynamic energy management.

[0056] The operating parameters of the air conditioner, including the compressor speed, are adjusted based on the calculated photovoltaic sufficiency index and a preset load scheduling algorithm. Other operating parameters include airflow intensity, which is quantified based on several core parameters: wind speed (m / s), average wind speed at 1 meter from the air outlet, and air volume (m³ / s). 3 / h), air volume per unit time, airflow organization index, and air supply coverage uniformity (0-1.0). The core actuators for air supply intensity are the air guide plate mechanism, stepper motor control angle (0-180°), duct damper, electric adjustment opening (0-100%), and variable frequency fan PWM speed regulation.

[0057] By acquiring real-time photovoltaic (PV) system power generation data and inputting it into a PV sufficiency model, a PV sufficiency index is calculated to quantify the matching relationship between PV energy supply and the immediate energy demand of air conditioning. Based on the PV sufficiency index and a preset load scheduling algorithm, the core operating parameters of the air conditioning system, especially the compressor speed, can be dynamically adjusted to ensure that clean energy is prioritized when PV energy is abundant, and that the load is rationally planned when PV power generation is limited to avoid energy waste. This energy-coordinated control strategy based on the PV sufficiency index not only effectively improves the energy utilization rate of the air conditioning system and reduces operating costs, but also promotes the effective integration and use of clean energy. In other words, it achieves dual optimization of energy utilization efficiency and user experience in the field of air conditioning control.

[0058] Furthermore, the operating parameters of the air conditioner are adjusted according to the aforementioned photovoltaic abundance index and preset load scheduling algorithm. The operating parameters include at least the compressor speed, including: when the photovoltaic abundance index is greater than a first preset threshold, adjusting the compressor speed to a first speed, which is set based on the compressor's maximum speed; when the photovoltaic abundance index is less than or equal to the first preset threshold but greater than a second preset threshold, adjusting the compressor speed to the rated speed; when the photovoltaic abundance index is less than or equal to the second preset threshold, starting grid auxiliary power supply and adjusting the compressor speed and air supply intensity to preset standards, which are determined based on the air conditioner's operating cost and comfort requirements.

[0059] Specifically, when the photovoltaic (PV) sufficiency index is greater than the first preset threshold, it indicates that the current power generation capacity of the PV system significantly exceeds the air conditioning load demand, indicating an energy surplus state. In this case, the compressor speed is reduced to a first speed, which is a lower value set based on the compressor's maximum speed, such as 90% of the maximum speed. This adjustment strategy not only makes full use of solar energy and reduces reliance on grid power, but also extends the compressor's lifespan, reduces operating noise, and provides a quieter and more comfortable operating environment. When the PV sufficiency index is less than or equal to the first preset threshold but greater than the second preset threshold, it means that PV energy can basically cover the air conditioning load demand at the current moment, but the energy ratio is close to the critical point. At this time, the compressor speed is adjusted to the rated speed to ensure the normal cooling or heating effect of the air conditioner, while utilizing the remaining capacity of PV energy to achieve efficient energy utilization. If the PV sufficiency index is less than or equal to the second preset threshold, it indicates that PV energy is insufficient to meet the air conditioning load demand, and the system enters grid-assisted power supply mode. While activating the grid auxiliary power supply, the compressor speed and air supply intensity are adjusted to the preset standard. This standard balances the air conditioning operating cost with the user's comfort needs, ensuring that basic cooling or heating services can still be provided when energy is scarce, while avoiding unnecessary energy consumption and reducing overall operating costs.

[0060] In summary, a hierarchical energy management strategy was adopted to achieve precise control of air conditioner compressor speed and intelligent startup of grid-assisted power supply under different photovoltaic (PV) sufficiency conditions. This significantly improved energy efficiency and system operating economy while meeting user comfort requirements. By combining the PV sufficiency index with load scheduling algorithms, intelligent adjustment of air conditioner operating parameters was achieved, effectively improving energy utilization efficiency and user satisfaction. This dynamic load scheduling based on the PV sufficiency index not only optimizes the energy utilization mode of air conditioners but also provides technical support for building a green and intelligent energy ecosystem.

[0061] In other embodiments of this application, the method further includes: using millimeter-wave radar to detect the micro-motion characteristics of the human body in the current environment based on the Doppler effect, obtaining the Doppler frequency shift value, and comparing the Doppler frequency shift value with a preset frequency change threshold; using an infrared array to detect whether the body temperature of the person in the current environment is abnormal; and adjusting the wind speed, the air delivery angle, and the temperature to preset wind speed range, preset air delivery angle range, and preset temperature range, respectively, when the Doppler frequency shift value is greater than the preset frequency change threshold or the body temperature of the person in the current environment is abnormal.

[0062] Specifically, millimeter-wave radar is used to detect subtle movements of the human body. (f d >50Hz is considered a fall. By analyzing Doppler frequency shift values, subtle human movements, such as falls and fainting, are identified. d v is the Doppler frequency shift value. r Here, f0 represents the radial velocity, f0 = 60 GHz is the radar center frequency, and c is the speed of light. The obtained Doppler frequency shift value is compared with a preset frequency change threshold. If the Doppler frequency shift value exceeds the frequency change threshold, a potential emergency is considered to have occurred, and corresponding air supply parameter adjustments will be taken immediately. Simultaneously, the body temperature of people in the current environment is detected using an infrared array to determine if there are any abnormally high or low temperatures, which may indicate poor health or other unexpected conditions. For example, if a body temperature change ΔT > 2°C for 3 seconds, an abnormal body temperature is determined.

[0063] When either of the above two scenarios occurs—that is, the Doppler frequency shift value exceeds the preset frequency change threshold or the person's body temperature is abnormal—the air supply parameters, including wind speed, air supply angle, and temperature, will be adjusted. Adjusting the wind speed to the preset range typically involves reducing the wind speed to avoid secondary injury from strong winds. Switching the air supply angle to the preset range usually means adjusting the air supply direction so that it does not blow directly onto the person. For example, in roof reflection mode, the air guide plate angle is adjusted to approximately 60° to avoid direct airflow. Depending on the specific situation, the temperature will be adjusted to the preset temperature range. For instance, if the ambient temperature is low, the indoor temperature will be moderately increased; conversely, it will be moderately decreased or kept unchanged to achieve a comfortable and safe effect. See Table 1 for the dynamic adjustment strategy of air supply parameters.

[0064] Table 1

[0065]

[0066] By integrating millimeter-wave radar and infrared array technology, the system can monitor the micro-motion characteristics and body temperature of the human body in the current environment in real time, and then make emergency responses. Specifically, when the millimeter-wave radar detects that the Doppler frequency shift value of a human body exceeds a preset frequency change threshold, or when the infrared array detects an abnormal body temperature, it immediately adjusts the airflow speed to a preset comfortable range, changes the airflow angle to avoid direct airflow onto the human body, and adjusts the temperature to a suitable level, ensuring that the user will not be harmed by strong airflow or inappropriate temperature. This mechanism greatly improves the response speed and processing efficiency of smart air conditioners in emergency situations. It can not only identify potential health risks in a timely manner, but also quickly make appropriate adjustments to the airflow parameters, effectively maintaining the user's comfort and safety, and providing comprehensive technical support for smart air conditioners in high-precision airflow control and user health monitoring.

[0067] Furthermore, before adjusting the wind speed, air supply angle, and temperature to preset wind speed range, preset air supply angle range, and preset temperature range respectively, when the Doppler frequency shift value is greater than the preset frequency change threshold, or when the body temperature of the person in the current environment is abnormal, the method further includes: matching the human micro-motion characteristics with dangerous actions in a preset typical dangerous action feature library; if the human micro-motion characteristics successfully match a first dangerous action in the preset typical dangerous action feature library, then obtaining the preset wind speed range, preset air supply angle range, and preset temperature range corresponding to the first dangerous action, wherein the preset typical dangerous action feature library includes multiple dangerous actions, and the preset typical dangerous action feature library is constructed at least based on historical user behavior data.

[0068] Specifically, when the Doppler frequency shift detected by the millimeter-wave radar exceeds a preset frequency change threshold, or when the infrared array detects abnormal body temperature, the air supply parameters are not immediately adjusted. Instead, the detected micro-movement characteristics of the human body are first compared with a preset database of typical dangerous actions. This database contains a variety of known dangerous action patterns and is constructed based on historical user behavior data, health monitoring data, and environmental change data. It covers typical behavioral characteristics in emergency situations such as falls and fainting.

[0069] If a person's micro-movement characteristics match a pre-defined typical dangerous action in a database of primary dangerous actions, the system will identify this action type and retrieve the corresponding preset wind speed range, preset airflow angle range, and preset temperature range from the database. The air conditioner will then adjust the wind speed, airflow angle, and temperature based on these parameters to adapt to the current emergency situation, reduce potential harm, and maintain indoor comfort.

[0070] By introducing a pre-set database of typical dangerous actions, the ability to identify and respond to emergencies is significantly enhanced. When abnormal human micro-movements or body temperature are detected, instead of directly adjusting the airflow parameters, these characteristics are first intelligently matched with various dangerous actions in the database. Once a specific primary dangerous action is identified, the corresponding pre-set fan speed range, pre-set airflow angle range, and pre-set temperature range are invoked based on the action type to respond to the emergency in the most suitable way. This feature database, built based on historical user behavior data and combined with artificial intelligence pattern recognition technology, enables smart air conditioners to make more accurate and personalized decisions when facing complex environmental changes and user behaviors. This not only improves the response speed and efficiency to sudden health events but also significantly enhances users' sense of security and comfort.

[0071] Furthermore, before matching the aforementioned human micro-motion characteristics with dangerous actions in the preset typical dangerous action feature library, the method further includes: acquiring the aforementioned historical user behavior data, and using a clustering analysis algorithm to identify dangerous behavior patterns from the aforementioned historical user behavior data, wherein the aforementioned historical user behavior data includes operational behavior data and location-based behavior data; using Failure Mode and Effects Analysis to assess the risk level of the aforementioned historical user behavior data, and identifying user behaviors corresponding to the aforementioned risk levels that are greater than the preset risk level as the aforementioned dangerous actions; and combining multiple of the aforementioned dangerous actions into the preset typical dangerous action feature library.

[0072] Specifically, the process begins by acquiring a large amount of historical user behavior data. This data includes operational behaviors (such as accidental operation of the air conditioner, repeated button presses, and long presses) and location-related behaviors (such as moving close to the air vent, putting hands into the fan area, and children climbing). Clustering analysis algorithms are then used to process this historical user behavior data to identify behavioral patterns with similar characteristics and potential risks. These patterns indicate safety hazards or health risks faced by users under specific conditions. After identifying potential dangerous behavioral patterns, Failure Mode and Effects Analysis (FMEA) is used to assess the risk level of each behavior. FMEA is a tool that systematically identifies potential failure modes in the product design phase or process and assesses their potential impact. The severity, probability of occurrence, and detection difficulty of each dangerous behavior are assessed, and a risk level is calculated based on this assessment. User behaviors with risk levels higher than a preset risk level are considered dangerous actions that pose a substantial threat to user safety and are selected to build a dangerous action feature library, i.e., a preset typical dangerous action feature library.

[0073] The establishment of a pre-defined database of typical dangerous actions is based at least on historical user behavior data, but it is not limited to this. It will continue to absorb new data and user feedback, and continuously improve and update the list of dangerous actions through iterative learning and manual review to adapt to the ever-changing usage environment and user needs.

[0074] By integrating in-depth analysis and risk assessment mechanisms of historical user behavior data, the intelligent air conditioner's ability to identify and respond to potentially dangerous actions has been significantly enhanced. First, a large amount of historical user behavior data, categorized by operation and location, was collected and analyzed. Clustering analysis algorithms were used to identify a series of behavioral patterns that may pose safety risks. Then, using Failure Mode and Effects Analysis (FMEA), these behavioral patterns underwent detailed risk level assessments, filtering out user behaviors with risk levels exceeding preset risk levels and defining them as dangerous actions requiring emergency response. Finally, all identified dangerous actions were integrated to construct a preset typical dangerous action feature library, serving as the core for subsequent dynamic monitoring and decision support. This process effectively improves the intelligent air conditioner's self-adaptability and safety performance in complex environments, ensuring that in emergencies such as falls or fainting, the air supply parameters can be adjusted quickly and accurately to avoid or mitigate potential harm, while maintaining indoor comfort and safety. This feature library construction strategy based on big data analysis and risk assessment provides solid technical support for intelligent air conditioners.

[0075] In some embodiments of this application, the method further includes: if the confidence value is less than the preset threshold, then data re-verification is performed, wherein the data re-verification includes activating a backup sensor, historical data backtracking, and time alignment compensation.

[0076] Specifically, see the flowchart for the compensation mechanism. Figure 3If the confidence value is less than the preset threshold (0.8), it means that the main sensor data is interfered with or of poor quality. The data re-verification process will be automatically initiated to ensure the accuracy and safety of the air supply decision. Data re-verification includes key steps such as activating backup sensors, historical data backtracking, and time alignment compensation. Regarding activating backup sensors, backup environmental sensing sensors, such as additional millimeter-wave radar, infrared sensors, or temperature and humidity sensors, are immediately activated to obtain more environmental data, increasing the diversity and reliability of data sources. Regarding historical data backtracking, historical environmental sensing data from the most recent period will be backtracked. The continuity and stability of this data will be used to assist in the current environmental assessment to overcome the impact of transient data interference. Regarding time alignment compensation, to ensure accurate synchronization of data between different sensors, time alignment compensation will be performed. By compensating for possible time differences between sensors, the synchronization and consistency of multi-source data are guaranteed, thereby improving the accuracy of data fusion. Typical scenarios include millimeter-wave radar false alarms (such as pets running), infrared sensor obstruction, and interference from sudden changes in environmental temperature and humidity.

[0077] like Figure 3 As shown, if the confidence value is less than a preset threshold, the strategy will be downgraded, switching to a conservative air supply mode and disabling dynamic tracking to avoid making overly aggressive air supply adjustments based on potentially inaccurate data, ensuring that the user's basic comfort is not affected. If the confidence value is less than the preset threshold, manual confirmation is required by pushing an alarm message to the user, requesting the user to manually confirm the current environmental status, or the user can manually adjust the air supply parameters, such as wind speed, temperature, and air supply angle, based on their own experience. This step incorporates user feedback, enhancing the accuracy of the system's decision-making.

[0078] If, after activating the above compensation mechanism, the reliability of the data is improved through manual verification or data re-verification, the environmental conditions will be reassessed and the air supply parameters adjusted. If sufficient reliability is still not achieved after data re-verification, the conservative air supply mode will be maintained until new data or user confirmation can provide more reliable environmental perception information.

[0079] The compensation mechanism is a strategy employed by intelligent air conditioning systems to address the degradation of environmental perception data quality. Through three levels—strategy degradation, manual confirmation, and data re-verification—it ensures that air supply control maintains high accuracy and security even in complex or disruptive environments, reflecting the human-centered design of smart home technology in safeguarding user experience.

[0080] In some embodiments of this application, personalized airflow preference adjustments are achieved through user behavior learning, providing a customized comfort experience. User behavior tracking and analysis: Deep learning technology is introduced to record and analyze users' airflow preferences over a long period, such as personal preferences for wind speed, temperature, and humidity, as well as changes in these preferences across different time periods and activity levels. Preference prediction model: Based on historical user data, a preference prediction model is built to anticipate potential user airflow needs, enabling proactive service. Users do not need to make frequent manual adjustments; airflow parameters are automatically adjusted according to personal preferences, providing a more considerate and comfortable experience. By learning user behavior, the air conditioning system becomes more intelligent, capable of predicting and meeting users' personalized needs, thus improving user satisfaction.

[0081] In other embodiments of this application, the dynamic allocation strategy of photovoltaic energy and grid power is optimized, and energy storage technology is combined to maximize the efficiency of clean energy utilization. Based on the photovoltaic air conditioning system, an energy storage optimization algorithm is developed to dynamically adjust the charging and discharging strategy of the energy storage device according to the real-time fluctuations of photovoltaic power generation and user electricity demand. When photovoltaic power generation is insufficient to meet the air conditioning load, the device intelligently participates in grid interaction, rationally dispatches energy storage to release power, and reduces dependence on traditional electricity. Through energy storage and dynamic dispatch, the efficiency of photovoltaic energy utilization is improved, achieving greener and lower-carbon energy management. This reduces unnecessary purchases of grid power, lowers user electricity costs, alleviates grid pressure during peak periods, and promotes the overall balance and stability of the energy system.

[0082] This application also provides an air supply control system for an air conditioner; see the system hardware architecture diagram below. Figure 4 The system consists of three core modules: an environmental perception module, an intelligent decision-making module, and an execution control module. The system adopts a distributed architecture design, using the G-Link communication protocol to achieve data interaction between modules, with a communication rate of 100Mbps and latency controlled within 5ms.

[0083] The environmental perception module includes a millimeter-wave radar array, an infrared sensor group, and a high-precision temperature and humidity matrix. The data from each sensor is preprocessed by an STM32H743 microcontroller, with sampling frequencies of 60Hz for radar, 30Hz for infrared, and 10Hz for temperature and humidity.

[0084] The data fusion algorithm is implemented on the environment-aware controller (Xilinx Zynq UltraScale+ MPSoC), and the specific process includes:

[0085] (1) Spatiotemporal alignment:

[0086] A double-buffering mechanism is employed to address the time difference issue in multi-sensor sampling. A unified coordinate system is established. Where v_r is the radar velocity, ΔT is the infrared temperature difference, and k is the Gree calibration coefficient (0.35±0.02).

[0087] (2) Feature-level fusion:

[0088] Developing a fusion network based on G-CNN:

[0089]

[0090] This code implements feature-level fusion of millimeter-wave radar and infrared data. It extracts the spatiotemporal features of radar signals using a 1D convolutional layer (GL_Conv1d), while simultaneously processing infrared thermal imaging data using a fully connected layer. Finally, the two types of features are concatenated along the channel dimension to form a 128+128=256-dimensional fused feature vector.

[0091] The dual-buffering architecture aims to address the sampling time difference issue among millimeter-wave radar (60Hz), infrared array (30Hz), and temperature and humidity matrix (10Hz). A flowchart of the implementation scheme can be found here. Figure 5 The key implementation steps are as follows:

[0092] (1) Hardware-level time synchronization:

[0093] Global clock source: A clock chip is used to synchronize the clocks of all sensors via the PTP protocol, with an error of <1μs.

[0094] Trigger signal: The radar sends a synchronization pulse every 16.67ms (60Hz), and the infrared and temperature and humidity sensors sample on the falling edge of the pulse.

[0095] (2) Double buffer management strategy, as shown in Table 2.

[0096] Table 2

[0097]

[0098]

[0099] G-CNN employs a multi-branch convolutional structure, with each branch processing data from different sensors or modalities (such as images, temperature, vibration, etc.), and performing feature extraction and fusion through a shared weight mechanism. Finally, the fused feature vector is output through a fully connected layer. See Table 3 for specific network configurations.

[0100] Table 3

[0101]

[0102] Key parameter settings during training:

[0103] 1. Dataset:

[0104] Source: Multimodal sensor data (e.g., infrared images, temperature, vibration signals, sound signals, etc.). Preprocessing: Image data: Normalized to the [0,1] range, randomly cropped and horizontally flipped for enhancement. Time series data: Standardized and segmented using a sliding window.

[0105] The split ratio is: Training set:Validation set:Test set = 70%:15%:15%.

[0106] 2. Loss function:

[0107] Classification task: Cross Entropy Loss. Regression task: Mean Squared Error (MSE) or Smooth L1 Loss.

[0108] 3. Optimizer:

[0109] Optimization algorithm: AdamW. Initial learning rate: 1e-4, using cosine annealing scheduling. Weight decay: 1e-5 (to prevent overfitting).

[0110] 4. Regularization strategies:

[0111] Dropout: Add Dropout to the fully connected layer (probability set to 0.5). Batch Normalization: Used to accelerate training and improve the model's generalization ability.

[0112] 5. Training details:

[0113] Batch Size: 64 (adjustable based on GPU memory). Number of Epochs: 100–150 (terminated based on early stopping strategy on the validation set). Evaluation Metrics: Classification task: Accuracy, F1-score. Regression task: MAE, RMSE.

[0114] G-CNN employs a multi-branch shared weights + attention mechanism in feature-level fusion, as detailed below:

[0115] 1. Multi-branch structure: Each branch corresponds to a modality (e.g., image, temperature, vibration), and its features are extracted separately. 2. Feature alignment: The feature maps output by each branch are spatially aligned (e.g., through interpolation or downsampling). 3. Attention fusion: Channel attention mechanisms (e.g., SE Block) or spatial attention mechanisms (e.g., CBAM) are introduced to dynamically weight and fuse the features of each modality. 4. Fuded features: The weighted features are input into the fully connected layer for final decision-making.

[0116] The generation of air supply strategies involves the following key technologies:

[0117] (1) Human comfort model:

[0118] Establish an improved PMV-PPD model Where M is the human metabolic rate, ΔT eff This is the output of the effective temperature difference algorithm.

[0119] (2) Actuator control:

[0120] A three-loop control strategy is adopted:

[0121] Outer loop: Comfort deviation control (PID parameters: Kp = 0.8, Ki = 0.05, Kd = 0.12). Middle loop: Wind speed fuzzy control. Inner loop: Motor vector control (FOC algorithm).

[0122] Basis for determining basic PID parameters:

[0123] Initial parameters (Kp = 0.8, Ki = 0.05, Kd = 0.12). Experimental platform: "Human-Environment" simulation chamber.

[0124] Setting method: Among them, T u = 12.5s (system oscillation period), K sys =0.4 (air conditioning system gain), T d = 3.2s (equivalent delay time).

[0125] Dynamic adjustment strategy:

[0126] Environmental conditions are adaptive, see Table 4.

[0127] Table 4

[0128] Environmental factors Parameter adjustment rules Technical Implementation Room temperature change rate >1℃ / min Kp×1.2, Kd×0.8 Call the algorithm to predict temperature trends Humidity >70% RH Ki×1.5 Activate humidity compensation module Photovoltaic power supply Kp×0.7 (inhibits frequent regulation) Adjusted dynamically according to the GPI index

[0129] Technical verification for adjustment:

[0130] For a comparison of the experimental data, please refer to Table 5.

[0131] Table 5

[0132] Scene Fixed PID comfort compliance rate Adaptive PID comfort compliance rate Elderly bedroom at night 82％ 97％ Gym 76％ 93％ Meeting room 88％ 95％

[0133] Dynamic adjustment effect:

[0134] Overshoot suppression: When the temperature drops sharply, the adaptive PID reduces the overshoot from 8.3% to 2.1%.

[0135] Energy optimization: 19% energy savings in conference room scenarios (by reducing unnecessary compressor start-ups and shutdowns).

[0136] Photovoltaic synergistic optimization:

[0137] Achieve dynamic allocation of photovoltaic power:

[0138] (1) Establish the photovoltaic adequacy model: GPI=(P_PV(t)) / (P_comp+P_fan)×η_GL, where η_GL is the system efficiency coefficient (measured 0.92).

[0139] (2) Load scheduling algorithm (preset load scheduling algorithm):

[0140]

[0141] The algorithm dynamically adjusts the compressor speed based on the Photovoltaic Adequacy Index (GPI). When photovoltaic power is abundant (GPI>1.2), the compressor operates at 90% of its maximum speed; when power is moderate, it maintains the rated speed; and when power is insufficient, it activates grid auxiliary power supply. Energy optimization is achieved through tiered control.

[0142] The logic for threshold division is shown in Table 6.

[0143] Table 6

[0144]

[0145] The following is a detailed explanation of the criteria, research basis, and methods for ensuring the coverage and accuracy of the pre-defined database of typical dangerous actions:

[0146] 1. User behavior data collection: Recording user behavior in multiple scenarios (including but not limited to: gestures, body posture, proximity distance, trigger frequency, and usage habits). Utilizing sensors such as cameras, millimeter-wave radar, and infrared arrays to collect user behavior data and form a raw dataset.

[0147] 2. Compliant with ergonomics and safety standards: Referencing industry safety standards; considering potential risks during the use of air conditioning products (such as high-temperature air outlets, foreign object inhalation into the fan, accidental activation of the emergency stop button, etc.); defining the boundaries and classifications of "dangerous actions" from the perspective of user safety.

[0148] 3. Classification System Construction: Various dangerous actions are divided into several categories, such as: Operational: accidental touch, repeated button press, long press, accidental start, etc. Location: body near the air outlet, hand into the fan area, children climbing, etc. Environmental: prolonged operation in high temperature environment, abnormal air circulation due to poor ventilation, etc. Equipment: abnormal vibration, excessive noise, abnormal indicator lights, etc.

[0149] User behavior data analysis: Based on data analysis of a large number of real user scenarios, high-frequency and high-risk operation patterns are extracted; typical dangerous behavior patterns are identified through machine learning methods such as cluster analysis and decision trees.

[0150] Human-Computer Interaction Research: Introducing Human-Computer Interaction (HCI) theory to analyze users' operational intentions and behavioral trajectories regarding air conditioning products in different contexts; combining cognitive psychology models to predict potential user errors or inappropriate behaviors.

[0151] Safety Engineering and Risk Assessment: The FMEA (Failure Mode and Effects Analysis) method is used to assess the risk level of various operations; a risk score is given for each dangerous action, and those actions are prioritized for inclusion in the feature library as high-risk actions.

[0152] Multimodal data fusion: Integrates information from multiple sources such as video, voice, infrared, millimeter-wave radar, temperature and humidity to build a more comprehensive user behavior model; improves the ability to identify dangerous actions in complex scenarios.

[0153] Methods to ensure coverage and identification accuracy:

[0154] 1. Feature engineering optimization:

[0155] 1) Extract features for each dangerous action, including: spatial location (coordinates, distance), time series (action duration, frequency), action type (gesture, posture, behavior pattern), and environmental parameters (temperature, humidity, light).

[0156] 2) Construct multidimensional feature vectors for training the recognition model.

[0157] 2. Machine learning model training: Deep learning models (such as LSTM, CNN, Transformer) are used for behavior classification; transfer learning is used to transfer existing behavior recognition models to air conditioning scenarios; an incremental learning mechanism is introduced to support the continuous expansion and updating of dangerous action features.

[0158] 3. Verification and Testing Mechanisms: Simulation Testing: Simulate various dangerous action scenarios to verify the model's recognition capabilities; Real-world Verification: Deploy the system in a real environment, collect feedback, and optimize the model; A / B Testing: Compare the performance of different models on the same dataset and select the optimal solution.

[0159] 4. Dynamic update mechanism: Establish a feature database update mechanism to promptly supplement or adjust features based on newly emerging user behaviors or security risks; support remote OTA upgrades to ensure continuous system optimization.

[0160] Examples of practical applications (some dangerous actions) can be found in Table 7.

[0161] Table 7

[0162]

[0163]

[0164] This application also provides an air supply control device for an air conditioner. It should be noted that the air supply control device for an air conditioner in this application can be used to execute the air supply control method for an air conditioner provided in this application. This device is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0165] The following describes the air supply control device for an air conditioner provided in the embodiments of this application.

[0166] Figure 6 This is a schematic diagram of an air supply control device for an air conditioner according to an embodiment of this application. Figure 6 As shown, the device includes an acquisition unit 10, a fusion unit 20, a generation unit 30, and a first adjustment unit 40. The acquisition unit is used to acquire personnel location information, action status information, and temperature and humidity distribution information in the current environment; the fusion unit is used to fuse the aforementioned personnel location information, action status information, and temperature and humidity distribution information to obtain fused features, and determine the confidence value of the fused features; the generation unit is used to generate an air supply command based on the fused features if the confidence value is greater than or equal to a preset threshold; the first adjustment unit is used to adjust the air supply parameters of the air conditioner according to the air supply command, the air supply parameters including wind speed, air supply angle, and temperature.

[0167] This embodiment integrates the location, movement status, and temperature and humidity distribution of personnel within the current environment. The resulting fusion features ensure the accuracy and reliability of the fused information. Based on the fused environmental features and their reliability, and by comparing them with preset thresholds, the system autonomously decides whether to generate an air supply command. This allows for precise adjustment of the air conditioner's fan speed, air supply angle, and temperature, ensuring users enjoy a comfortable and personalized air supply experience in various situations. This improves the air supply positioning accuracy of smart air conditioners in complex environments, thus solving the problem of low accuracy in air supply positioning under such conditions.

[0168] In the specific implementation process, the above-mentioned generation unit includes an input module and a generation module. The input module is used to input the above-mentioned fused features into the human comfort model if the above-mentioned confidence value is greater than or equal to the above-mentioned preset threshold, and calculate the comfort index of the current environment; the generation module is used to determine the air supply strategy according to the above-mentioned comfort index of the current environment, and generate the above-mentioned air supply command based on the above-mentioned air supply strategy, the above-mentioned air supply strategy includes the strategy of adjusting the above-mentioned wind speed, the above-mentioned air supply angle and the above-mentioned temperature.

[0169] In summary, when the credibility of the fused features meets the requirements, the comfort level of the current environment is calculated using a human comfort model, and an air supply strategy is formulated and executed accordingly. This enables the dynamic adjustment of air supply parameters based on the indoor environment and the state of the users, ensuring that users experience optimal comfort under various conditions. This mechanism not only enhances the intelligence of the air conditioner but also effectively improves energy efficiency, reduces unnecessary energy consumption, and provides users with a more humanized and environmentally friendly user experience.

[0170] In some embodiments of this application, the above-mentioned device further includes an acquisition unit and a second adjustment unit. The acquisition unit is used to acquire photovoltaic system power generation data and input the photovoltaic system power generation data into a photovoltaic adequacy model to obtain a photovoltaic adequacy index, wherein the photovoltaic adequacy index characterizes the degree of matching between the power generation capacity of the photovoltaic system and the load demand of the air conditioner; the second adjustment unit is used to adjust the operating parameters of the air conditioner according to the photovoltaic adequacy index and a preset load scheduling algorithm, wherein the operating parameters include at least the compressor speed.

[0171] By acquiring real-time photovoltaic (PV) system power generation data and inputting it into a PV sufficiency model, a PV sufficiency index is calculated to quantify the matching relationship between PV energy supply and the immediate energy demand of air conditioning. Based on the PV sufficiency index and a preset load scheduling algorithm, the core operating parameters of the air conditioning system, especially the compressor speed, can be dynamically adjusted to ensure that clean energy is prioritized when PV energy is abundant, and that the load is rationally planned when PV power generation is limited to avoid energy waste. This energy-coordinated control strategy based on the PV sufficiency index not only effectively improves the energy utilization rate of the air conditioning system and reduces operating costs, but also promotes the effective integration and use of clean energy. In other words, it achieves dual optimization of energy utilization efficiency and user experience in the field of air conditioning control.

[0172] Furthermore, the aforementioned second adjustment unit includes a first adjustment module, a second adjustment module, and a third adjustment module. The first adjustment module is used to adjust the compressor speed to a first speed, which is set based on the compressor's maximum speed, when the photovoltaic abundance index is greater than a first preset threshold. The second adjustment module is used to adjust the compressor speed to a rated speed when the photovoltaic abundance index is less than or equal to the first preset threshold but greater than a second preset threshold. The third adjustment module is used to activate grid auxiliary power supply and adjust the compressor speed and airflow intensity to preset standards when the photovoltaic abundance index is less than or equal to the second preset threshold. These preset standards are determined based on the air conditioner's operating costs and comfort requirements.

[0173] Through a tiered energy management strategy, precise control of air conditioner compressor speed and intelligent startup of grid-assisted power supply are achieved under varying photovoltaic (PV) sufficiency conditions. This significantly improves energy efficiency and system operating economy while meeting user comfort needs. By combining the PV sufficiency index with load scheduling algorithms, intelligent adjustment of air conditioner operating parameters is realized, effectively enhancing energy utilization efficiency and user satisfaction. This dynamic load scheduling based on the PV sufficiency index not only optimizes the energy utilization mode of air conditioners but also provides technical support for building a green and intelligent energy ecosystem.

[0174] In other embodiments of this application, the above-mentioned device further includes a first detection unit, a second detection unit, and a third adjustment unit. The first detection unit is used to detect the micro-motion characteristics of the human body in the current environment using millimeter-wave radar based on the Doppler effect, obtain the Doppler frequency shift value, and compare the Doppler frequency shift value with a preset frequency change threshold; the second detection unit is used to detect whether the body temperature of the person in the current environment is abnormal using an infrared array; the third adjustment unit is used to adjust the wind speed, the air delivery angle, and the temperature to preset wind speed range, preset air delivery angle range, and preset temperature range, respectively, when the Doppler frequency shift value is greater than the preset frequency change threshold, or when the body temperature of the person in the current environment is abnormal.

[0175] By integrating millimeter-wave radar and infrared array technology, the system can monitor the micro-motion characteristics and body temperature of the human body in the current environment in real time, and then make emergency responses. Specifically, when the millimeter-wave radar detects that the Doppler frequency shift value of a human body exceeds a preset frequency change threshold, or when the infrared array detects an abnormal body temperature, it immediately adjusts the airflow speed to a preset comfortable range, changes the airflow angle to avoid direct airflow onto the human body, and adjusts the temperature to a suitable level, ensuring that the user will not be harmed by strong airflow or inappropriate temperature. This mechanism greatly improves the response speed and processing efficiency of smart air conditioners in emergency situations. It can not only identify potential health risks in a timely manner, but also quickly make appropriate adjustments to the airflow parameters, effectively maintaining the user's comfort and safety, and providing comprehensive technical support for smart air conditioners in high-precision airflow control and user health monitoring.

[0176] Furthermore, the aforementioned device also includes a matching unit, used to match the human micro-motion characteristics with dangerous actions in a preset typical dangerous action feature library before adjusting the wind speed, the air supply angle, and the temperature to preset wind speed range, preset air supply angle range, and preset temperature range, respectively, when the Doppler frequency shift value is greater than the preset frequency change threshold, or when the body temperature of the person in the current environment is abnormal. If the human micro-motion characteristics are successfully matched with the first dangerous action in the preset typical dangerous action feature library, the preset wind speed range, the preset air supply angle range, and the preset temperature range corresponding to the first dangerous action are obtained. The preset typical dangerous action feature library includes a variety of dangerous actions and is constructed based at least on historical user behavior data.

[0177] By introducing a pre-set database of typical dangerous actions, the ability to identify and respond to emergencies is significantly enhanced. When abnormal human micro-movements or body temperature are detected, instead of directly adjusting the airflow parameters, these characteristics are first intelligently matched with various dangerous actions in the database. Once a specific primary dangerous action is identified, the corresponding pre-set fan speed range, pre-set airflow angle range, and pre-set temperature range are invoked based on the action type to respond to the emergency in the most suitable way. This feature database, built based on historical user behavior data and combined with artificial intelligence pattern recognition technology, enables smart air conditioners to make more accurate and personalized decisions when facing complex environmental changes and user behaviors. This not only improves the response speed and efficiency to sudden health events but also significantly enhances users' sense of security and comfort.

[0178] Furthermore, the aforementioned device also includes an identification unit, an evaluation unit, and a combination unit. The identification unit acquires the aforementioned historical user behavior data before matching the aforementioned human micro-motion characteristics with dangerous actions in a preset typical dangerous action feature library, and uses a clustering analysis algorithm to identify dangerous behavior patterns from the aforementioned historical user behavior data. The aforementioned historical user behavior data includes operational behavior data and location-based behavior data. The evaluation unit uses Failure Mode and Effects Analysis (FMEA) to evaluate the risk level of the aforementioned historical user behavior data and identifies user behaviors corresponding to risk levels greater than a preset risk level as the aforementioned dangerous actions. The combination unit combines multiple of the aforementioned dangerous actions into the preset typical dangerous action feature library.

[0179] By integrating in-depth analysis and risk assessment mechanisms of historical user behavior data, the intelligent air conditioner's ability to identify and respond to potentially dangerous actions has been significantly enhanced. First, a large amount of historical user behavior data, categorized by operation and location, was collected and analyzed. Clustering analysis algorithms were used to identify a series of behavioral patterns that may pose safety risks. Then, using Failure Mode and Effects Analysis (FMEA), these behavioral patterns underwent detailed risk level assessments, filtering out user behaviors with risk levels exceeding preset risk levels and defining them as dangerous actions requiring emergency response. Finally, all identified dangerous actions were integrated to construct a preset typical dangerous action feature library, serving as the core for subsequent dynamic monitoring and decision support. This process effectively improves the intelligent air conditioner's self-adaptability and safety performance in complex environments, ensuring that in emergencies such as falls or fainting, the air supply parameters can be adjusted quickly and accurately to avoid or mitigate potential harm, while maintaining indoor comfort and safety. This feature library construction strategy based on big data analysis and risk assessment provides solid technical support for intelligent air conditioners.

[0180] In some embodiments of this application, the above-mentioned device further includes a re-verification unit, which is used to perform data re-verification if the confidence value is less than the preset threshold. The data re-verification includes activating a backup sensor, historical data backtracking, and time alignment compensation.

[0181] The compensation mechanism is a strategy employed by intelligent air conditioning systems to address the degradation of environmental perception data quality. Through three levels—strategy degradation, manual confirmation, and data re-verification—it ensures that air supply control maintains high accuracy and security even in complex or disruptive environments, reflecting the human-centered design of smart home technology in safeguarding user experience.

[0182] The aforementioned air conditioning air supply control device includes a processor and a memory. The aforementioned acquisition unit, fusion unit, generation unit, first adjustment unit, etc., are all stored as program units in the memory, and the processor executes the aforementioned program units stored in the memory to achieve the corresponding functions. All of the above modules are located in the same processor; or, the above modules are located in different processors in any combination.

[0183] The memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0184] This invention provides a computer-readable storage medium including a stored program, wherein, when the program is executed, it controls the device containing the computer-readable storage medium to perform the air supply control method of the air conditioner.

[0185] This invention provides a processor for running a program, wherein the program executes the air supply control method of the air conditioner.

[0186] This invention provides an electronic device, including a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps of the air conditioning air supply control method described above. The device described herein can be a server, PC, PAD, mobile phone, etc.

[0187] This application also provides a computer program product that, when executed on a data processing device, is adapted to perform the steps of initializing the air supply control method of the air conditioner described above.

[0188] It is obvious to those skilled in the art that the modules or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. They can be implemented using computer-executable program code, and thus can be stored in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those described herein, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0189] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0190] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0191] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0192] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0193] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0194] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0195] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0196] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0197] It should also be noted that 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. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0198] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for controlling the air supply of an air conditioner, characterized in that, include: Acquire information on the location, movement status, and temperature and humidity distribution of personnel in the current environment; The personnel location information, the action status information, and the temperature and humidity distribution information are fused to obtain fused features, and the confidence value of the fused features is determined. If the confidence value is greater than or equal to a preset threshold, then an air supply instruction is generated based on the fusion features; The air supply parameters of the air conditioner are adjusted according to the air supply command, and the air supply parameters include wind speed, air supply angle and temperature.

2. The method according to claim 1, characterized in that, If the confidence value is greater than or equal to a preset threshold, then an air supply instruction is generated based on the fusion features, including: If the confidence value is greater than or equal to the preset threshold, the fusion feature is input into the human comfort model to calculate the comfort index of the current environment. An air supply strategy is determined based on the comfort index of the current environment, and an air supply command is generated based on the air supply strategy. The air supply strategy includes strategies for adjusting the wind speed, the air supply angle, and the temperature.

3. The method according to claim 1, characterized in that, The method further includes: Acquire photovoltaic system power generation data and input the photovoltaic system power generation data into the photovoltaic adequacy model to obtain the photovoltaic adequacy index. The photovoltaic adequacy index characterizes the degree of matching between the power generation capacity of the photovoltaic system and the load demand of the air conditioner. The operating parameters of the air conditioner are adjusted according to the photovoltaic abundance index and the preset load scheduling algorithm, and the operating parameters include at least the compressor speed.

4. The method according to claim 3, characterized in that, The operating parameters of the air conditioner are adjusted according to the photovoltaic abundance index and the preset load scheduling algorithm. These operating parameters include at least the compressor speed, and include: If the photovoltaic abundance index is greater than a first preset threshold, the compressor speed is adjusted to a first speed, which is set according to the compressor's maximum speed. If the photovoltaic abundance index is less than or equal to the first preset threshold and greater than the second preset threshold, the compressor speed is adjusted to the rated speed. If the photovoltaic abundance index is less than or equal to the second preset threshold, the grid auxiliary power supply is activated, and the compressor speed and air supply intensity are adjusted to a preset standard, which is determined based on the operating cost and comfort requirements of the air conditioner.

5. The method according to claim 1, characterized in that, The method further includes: The micro-motion characteristics of the human body in the current environment are detected by millimeter-wave radar based on the Doppler effect, and the Doppler frequency shift value is obtained. The Doppler frequency shift value is then compared with a preset frequency change threshold. An infrared array is used to detect whether the body temperature of people in the current environment is abnormal. If the Doppler frequency shift value is greater than the preset frequency change threshold, or if the body temperature of the person in the current environment is abnormal, the wind speed, the air delivery angle, and the temperature will be adjusted to preset wind speed range, preset air delivery angle range, and preset temperature range, respectively.

6. The method according to claim 5, characterized in that, Before adjusting the wind speed, the air delivery angle, and the temperature to preset wind speed range, preset air delivery angle range, and preset temperature range respectively, when the Doppler frequency shift value is greater than the preset frequency change threshold, or when the body temperature of the personnel in the current environment is abnormal, the method further includes: The human body micro-motion characteristics are matched with dangerous actions in a preset typical dangerous action feature library. If the human body micro-motion characteristics successfully match the first dangerous action in the preset typical dangerous action feature library, then the preset wind speed range, the preset air supply angle range, and the preset temperature range corresponding to the first dangerous action are obtained. The preset typical dangerous action feature library includes a variety of dangerous actions, and the preset typical dangerous action feature library is constructed based on at least historical user behavior data.

7. The method according to claim 1, characterized in that, The method further includes: If the confidence value is less than the preset threshold, data re-verification is performed. The data re-verification includes activating backup sensors, historical data backtracking, and time alignment compensation.

8. The method according to claim 6, characterized in that, Before matching the human micro-motion features with dangerous actions in a preset database of typical dangerous actions, the method further includes: The historical user behavior data is acquired, and a clustering analysis algorithm is used to identify dangerous behavior patterns from the historical user behavior data, wherein the historical user behavior data includes operation-related behavior data and location-related behavior data; Failure Mode and Effects Analysis is used to assess the risk level of the historical user behavior data, and user behaviors corresponding to risk levels greater than a preset risk level are identified as dangerous actions. The various dangerous actions are combined into the preset typical dangerous action feature library.

9. An air supply control device for an air conditioner, characterized in that, include: The acquisition unit is used to acquire personnel location information, action status information, and temperature and humidity distribution information in the current environment. The fusion unit is used to fuse the personnel location information, the action status information, and the temperature and humidity distribution information to obtain fused features, and to determine the confidence value of the fused features; The generation unit is used to generate an air supply instruction based on the fusion features if the confidence value is greater than or equal to a preset threshold. The first adjustment unit is used to adjust the air supply parameters of the air conditioner according to the air supply command. The air supply parameters include wind speed, air supply angle and temperature.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device containing the computer-readable storage medium to perform the air supply control method of any one of claims 1 to 8.

Images