Air conditioner control method, device, equipment, medium, product and air conditioner

Abstract

The invention discloses an air conditioner control method, device and equipment, a medium, a product and an air conditioner. The method comprises the steps that environment parameter information of the environment where the air conditioner is located is obtained; inputting the environment parameter information into a language model to obtain air conditioner control information output by the language model; according to the environment parameter information and the air conditioner control information, it is determined that the air conditioner control information meets a preset condition through the language model, and the preset condition represents that after the air conditioner is controlled to operate through the air conditioner control information, the temperature of the environment where the air conditioner is located meets the target temperature; and under the condition that the preset condition is met, the air conditioner is controlled to operate according to the air conditioner control information. The problem that time difference exists between the system control logic and the real comfort requirement of the human body in the related technology can be avoided, and the user experience is improved.

Description

Technical Field

[0001] The present invention belongs to the field of smart home technology, and in particular relates to an air conditioning control method, device, equipment, medium, product and air conditioner. Background Art

[0002] In today's society, air conditioners are an indispensable appliance in modern households. Air conditioners can be used to set the air outlet temperature, providing users with a stable temperature perception. However, in actual air conditioning use, humans experience a lag in their perceived comfort level. Consequently, as the air conditioner continues to deliver air, there's a time lag between the system's control logic and the body's actual comfort needs, resulting in a poor user experience. Summary of the Invention

[0003] The embodiments of the present invention provide an air conditioning control method, device, equipment, medium, product and air conditioning, which can avoid the problem of time difference between system control logic and the actual comfort needs of the human body in related technologies, and improve the user experience.

[0004] In order to solve the above problems, in a first aspect, an embodiment of the present invention discloses an air conditioning control method, the method comprising:

[0005] Obtain environmental parameter information of the environment in which the air conditioner is located;

[0006] Inputting the environmental parameter information into a language model to obtain air conditioning control information output by the language model;

[0007] determining, based on the environmental parameter information and the air conditioning control information, through the language model, that the air conditioning control information satisfies a preset condition, wherein the preset condition indicates that after the air conditioning is controlled by the air conditioning control information, the temperature of the environment in which the air conditioning is located satisfies a target temperature;

[0008] When the preset condition is met, the air conditioner is controlled to operate according to the air conditioner control information.

[0009] Optionally, determining, based on the environmental parameter information and the air-conditioning control information, by using the language model, whether the air-conditioning control information satisfies a preset condition includes:

[0010] According to the environmental parameter information and the air conditioning control information, a feature fusion network is used to obtain fusion features output by the feature fusion network;

[0011] According to the fusion feature, it is determined through the language model whether the air-conditioning control information meets a preset condition.

[0012] Optionally, obtaining the fusion features output by the feature fusion network through a feature fusion network according to the environmental parameter information and the air conditioning control information includes:

[0013] Encoding the environmental parameter information to obtain state features, and encoding the air conditioning control information to obtain control text features;

[0014] The state feature and the control text feature are input into the feature fusion network to obtain the fusion feature output by the feature fusion network.

[0015] Optionally, determining, based on the fusion feature, by using the language model whether the air conditioning control information meets a preset condition includes:

[0016] Inputting the decoded fusion feature into the language model to obtain control confirmation information output by the language model, wherein the control confirmation information is used to indicate whether the air conditioning control information meets the preset condition;

[0017] In a case where the control confirmation information indicates that the air-conditioning control information satisfies the preset condition, it is determined that the air-conditioning control information satisfies the preset condition.

[0018] Optionally, the method further includes:

[0019] If the control confirmation information indicates that the air-conditioning control information does not meet the preset condition, updating the model parameters of the language model according to the target temperature, the environmental parameter information, and the air-conditioning control information output by the language model to obtain updated first target model parameters;

[0020] Obtaining the language model after parameter update according to the first target model parameters;

[0021] Inputting the environmental parameter information into the language model after parameter update to obtain first air-conditioning control information output by the language model, until the first air-conditioning control information meets the preset condition;

[0022] The operation of the air conditioner is controlled according to the first air conditioner control information.

[0023] Optionally, the method further includes:

[0024] Obtain historical target control information;

[0025] updating the model parameters of the language model according to the historical target control information, the environmental parameter information, and the air-conditioning control information to obtain updated second target model parameters;

[0026] Obtaining the language model after parameter update according to the second target model parameters;

[0027] Inputting the environmental parameter information into the language model after parameter update to obtain second air-conditioning control information output by the language model, until the second air-conditioning control information meets the preset condition;

[0028] The operation of the air conditioner is controlled according to the second air conditioner control information.

[0029] Optionally, the method further includes:

[0030] Get the image of the preset area;

[0031] Performing face recognition on the image to obtain a recognition result according to a target object recognition model;

[0032] In a case where the recognition result indicates that the preset area includes the target user, the control information is used as user control information corresponding to the target user.

[0033] Optionally, the method further includes:

[0034] Obtaining historical user control information corresponding to the target user;

[0035] updating the model parameters of the language model according to the historical user control information, the environmental parameter information, and the air-conditioning control information to obtain updated third target model parameters;

[0036] Obtaining the language model after parameter update according to the third target model parameters;

[0037] Inputting the environmental parameter information into the language model after parameter update to obtain third air-conditioning control information output by the language model until the third air-conditioning control information satisfies the preset condition, the third air-conditioning control information being control information for the air-conditioning to operate in the preset area;

[0038] The operation of the air conditioner is controlled according to the third air conditioner control information.

[0039] Optionally, the preset area includes multiple sub-areas, and the method further includes:

[0040] For the air conditioning control information corresponding to each sub-area, the air conditioner is controlled to operate in the sub-area using the corresponding air conditioning control information.

[0041] In a second aspect, an embodiment of the present invention provides an air conditioning control device, the device comprising:

[0042] An acquisition module is used to obtain environmental parameter information of the environment in which the air conditioner is located;

[0043] an output module, configured to input the environmental parameter information into a language model and obtain air-conditioning control information output by the language model;

[0044] a judgment module, configured to determine, based on the environmental parameter information and the air conditioning control information, by using the language model, whether the air conditioning control information satisfies a preset condition, wherein the preset condition indicates that after the air conditioning is controlled by the air conditioning control information, the temperature of the environment in which the air conditioning is located satisfies a target temperature;

[0045] A control module is used to control the operation of the air conditioner according to the air conditioner control information when the preset conditions are met.

[0046] An embodiment of the present invention further discloses an electronic device, which includes a processor, a memory, and a computer program stored in the memory and executable on the processor; the processor implements the aforementioned method when executing the program.

[0047] An embodiment of the present invention further discloses a readable storage medium. When instructions in the readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the aforementioned method.

[0048] An embodiment of the present invention further discloses a computer program product, including instructions or transactions. When the instructions or transactions are executed by a processor in an electronic device, the electronic device executes the aforementioned method.

[0049] An embodiment of the present invention further discloses an air conditioner, which includes the above-mentioned electronic device.

[0050] The embodiments of the present invention include the following advantages:

[0051] The air-conditioning control method provided by the embodiment of the present invention obtains environmental parameter information of the environment in which the air-conditioning is located, and inputs the environmental parameter information into a language model for predicting air-conditioning control information. The language model is used to analyze the environmental parameter information and output the air-conditioning control information. At the same time, the environmental parameter information and the air-conditioning control information are combined to verify whether the air-conditioning control information meets a preset condition through the language model. The preset condition indicates that the ambient temperature meets the target temperature after the air-conditioning is controlled by the air-conditioning control information. When the preset condition is met, the air-conditioning control information is executed, thereby predicting the trend of human body temperature changes in advance through the prediction ability of the language model and generating a control indicator. The system combines the preset condition verification mechanism to compensate for the lag and shorten the control delay; and by inputting the environmental parameter information and the air-conditioning control information into the language model, the preset conditions are used to verify whether the ambient temperature meets the expectations after the control information is executed, forming a "prediction-verification-correction" closed-loop control mechanism, which can effectively deal with uncertainties such as sudden environmental changes, reduce the risk of misadjustment, and improve the robustness of the system; upgrade the air-conditioning control from passive adjustment based on current errors to active prediction based on future needs, accurately bridge the time difference between the body sensory lag and system control, and realize personalized and dynamic adaptation through data-driven, fundamentally improving the user's temperature comfort experience and system energy efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 This is a flow chart of an air conditioning control method provided by an embodiment of the present application;

[0053] Figure 2 This is a logic flow chart of an air conditioning control method provided by an embodiment of the present invention;

[0054] Figure 3 is a flow chart of another air conditioning control method provided by an embodiment of the present application;

[0055] Figure 4 is a flow chart of another air conditioning control method provided by an embodiment of the present application;

[0056] Figure 5 is a block diagram of an air conditioning control device provided in an embodiment of the present application;

[0057] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application;

[0058] Figure 7 This is a structural diagram of an air conditioner provided in an embodiment of the present application. DETAILED DESCRIPTION

[0059] The following will be combined with the accompanying drawings in the embodiments of the present application to clearly describe the technical solutions in the embodiments of the present application. Obviously, the embodiments described are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field fall within the scope of protection of this application.

[0060] The terms "first," "second," and the like in the specification and claims of this application are used to distinguish similar objects, and are not used to describe a specific order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments of this application can be implemented in an order other than that illustrated or described herein, and that the objects distinguished by "first," "second," and the like are generally of the same type, and do not limit the number of objects; for example, the first object can be one or more. In addition, the term "and / or" in the specification and claims refers to at least one of the connected objects, and the character " / " generally indicates that the objects connected are in an "or" relationship.

[0061] Before introducing the air conditioning control method, apparatus, device, medium, product, and air conditioner provided by the present disclosure, we first introduce the application scenarios involved in each embodiment of the present disclosure. The present disclosure can be applied to air conditioning control scenarios, and the air conditioning control method provided by the embodiments of the present disclosure can be applied to air conditioners or to air conditioner control terminals such as mobile terminals.

[0062] With socioeconomic development, rising incomes and improved quality of life, the purchase and use of appliances like air conditioners has become more feasible. Air conditioners, which can regulate indoor temperature and meet people's needs for a comfortable indoor temperature in different seasons, have become widely used in ordinary households.

[0063] Currently, air conditioners actively regulate indoor temperature through heat exchange systems. Their core control logic is based on user-set air outlet temperature thresholds. Cooling is activated when the ambient temperature rises above the set value, while heating is activated when it falls below, thereby maintaining a relatively stable temperature field. The theoretical basis of this control model assumes a linear positive correlation between ambient temperature and human comfort. However, in practice, this assumption differs significantly from the physiological characteristics of human thermal perception.

[0064] Since the formation of human body temperature is a complex physiological-physical coupling process, it not only depends on the ambient dry-bulb temperature, but is also affected by multiple factors such as air humidity, airflow velocity, average radiation temperature (such as direct sunlight, wall heat conduction), human metabolic rate (activity intensity), clothing thermal resistance, etc. For example, in a 24°C environment, when the humidity rises from 30% to 70%, the actual temperature perceived by the human body may be equivalent to above 26°C; and the change in the air speed at the air conditioner outlet (such as from 0.2m / s to 0.5m / s) will reduce the perceived temperature by 1-2°C by enhancing convective heat dissipation. Traditional air conditioners only use a single temperature sensor as a feedback signal, which cannot capture the comprehensive effect of the above complex environmental parameters on human thermal comfort.

[0065] More importantly, the human body's thermal perception system has a significant time lag. It takes about 10-30 seconds of nerve conduction time for the skin temperature receptors to transmit signals to the hypothalamic temperature regulation center, and the central nervous system requires additional delays to integrate multi-source sensory signals (such as skin temperature, core body temperature, sweat rate) and make regulatory decisions (such as vasoconstriction and sweat gland secretion). When the air conditioner is running continuously, this physiological lag will result in a time difference of "the ambient temperature has deviated from the comfort range, but the human body has not yet produced a subjective perception."

[0066] The PID (proportional-integral-differential) control algorithm of traditional air conditioners relies on the "current temperature error" for adjustment and is a typical hysteresis correction system. This control logic cannot predict the delayed effect of human thermal perception, which inevitably leads to a time mismatch between control actions and actual needs: when the system detects temperature deviation and initiates adjustment, the actual comfort needs of the human body may have entered the next stage, which ultimately manifests as large indoor temperature fluctuations (±2-3°C), frequent manual adjustment of set points by users (an average of 3-5 times a day), and energy waste (hysteresis adjustment causes the compressor start and stop frequency to increase by 15%-20%). In addition, different user groups (such as the elderly, children, and the infirm) have significant differences in sensitivity to temperature changes. The traditional fixed threshold control mode cannot meet personalized comfort needs, further exacerbating the experience defects caused by the hysteresis of physical perception.

[0067] In order to solve the above problems, the present disclosure provides an air-conditioning control method, device, equipment, medium, product and air-conditioning, which obtains environmental parameter information of the environment in which the air-conditioning is located, and inputs the environmental parameter information into a language model for predicting air-conditioning control information, uses the language model to analyze the environmental parameter information and outputs the air-conditioning control information, and at the same time combines the environmental parameter information with the air-conditioning control information, verifies whether the air-conditioning control information meets a preset condition through the language model, and the preset condition indicates that the ambient temperature meets the target temperature after the air-conditioning is controlled by the air-conditioning control information. When the preset condition is met, the air-conditioning control information is executed, thereby predicting the human body temperature change in advance through the prediction ability of the language model. The system can predict the temperature of the air conditioner and generate control instructions based on the trend, and compensate for the lag in combination with the preset condition verification mechanism to shorten the control delay. It can also input the environmental parameter information and air-conditioning control information into the language model, and use the preset conditions to verify whether the ambient temperature meets the expectations after the control information is executed, forming a "prediction-verification-correction" closed-loop control mechanism to effectively deal with uncertainties such as sudden environmental changes, reduce the risk of misadjustment, and improve the robustness of the system. It can upgrade the air-conditioning control from passive adjustment based on current errors to active prediction based on future needs, accurately bridging the time difference between the body sensory lag and system control, and realize personalized and dynamic adaptation through data-driven, fundamentally improving the user's temperature comfort experience and system energy efficiency.

[0068] Method Example

[0069] The air conditioning control method provided in the embodiment of the present application is described in detail below through specific embodiments and their application scenarios in conjunction with the accompanying drawings.

[0070] Figure 1 This is a flow chart of an air conditioning control method provided by an embodiment of the present application. Figure 1 As shown, the method can be applied to an air conditioner, and can also be applied to a control terminal of the air conditioner such as a mobile terminal.

[0071] The method may include the following steps.

[0072] In step 101, environmental parameter information of the environment in which the air conditioner is located is obtained.

[0073] The air conditioner may be an onboard air conditioner installed inside the vehicle. The environmental parameter information may include the temperature inside and outside the vehicle where the air conditioner is installed, the current operating status of the vehicle, and the current weather information.

[0074] In this step, a distributed temperature sensing network can be set up inside the vehicle to achieve accurate measurement of the three-dimensional temperature field inside the vehicle, including multi-point temperature values ​​of the driver's seat (such as 0.3m from the seat surface), the co-driver's seat and the rear passenger area. For example, a micro-thermocouple array with an accuracy of ±0.5°C can be used, combined with an infrared thermal imager to non-contact monitor the temperatures of key thermal interfaces such as window glass and seat surfaces, to analyze the airflow organization in the vehicle (such as the impact of the wind direction of the air-conditioning outlet on the temperature distribution) and local thermal comfort of the human body (such as the body perception deviation caused by insufficient temperature in the foot area).

[0075] In the case of obtaining the temperature outside the vehicle, a plurality of temperature sensors may be provided outside the vehicle for respectively measuring the ambient air temperature and the headwind temperature when the vehicle is traveling, thereby obtaining the temperature outside the vehicle.

[0076] When obtaining the current operating status of the vehicle, the on-board CAN bus can be connected to obtain real-time operating status data, such as through wheel speed sensors or GPS signal solutions to evaluate the amount of fresh air penetration when the vehicle is driving, and to determine the working status of the air-conditioning compressor by the engine speed, as well as to monitor whether there is air leakage through gaps through door sensors based on the door and window switch status, thereby triggering the correction of the heat load in the vehicle.

[0077] In order to obtain the weather information at the current moment, the Internet of Vehicles module can be connected to the real-time meteorological data API to obtain the weather information of the current geographic location.

[0078] In step 102, the environmental parameter information is input into a language model to obtain air-conditioning control information output by the language model.

[0079] Among them, the language model is trained to predict the control information of the air conditioner.

[0080] In an embodiment of the present application, the language model may include a large language model, which refers to a deep learning model with a self-supervision mechanism trained using a large amount of text data. It can generate natural language text or understand the meaning of language text by calling data in a database through semantic analysis. Among them, the database can contain massive natural language data and can contain training set data of a large language model, which can be obtained from the Internet through a data acquisition program or downloaded from a data set website. The large language model can handle a variety of natural language tasks, such as text classification, question and answer, dialogue, etc. The large language model used in the embodiment of the present application can be some common large language models. For these large language models, the character string corresponding to the air conditioning control information prediction algorithm formula of the air conditioner can be input into the model, and the meaning of each symbol in the formula can also be input into the large language model, so that the air conditioning control information prediction algorithm formula of the air conditioner and the physical meaning of each symbol in the formula are stored in the database related to the large language model, so that the large language model has the ability to predict the control information of the air conditioner.

[0081] In this embodiment, after obtaining the current environmental parameter information, it can first be input into a language model. The language model then performs semantic analysis on the input environmental parameter information to understand the meaning and interrelationships of the individual parameters. For example, when inputting "indoor temperature 28°C, humidity 60%, outdoor temperature 32°C," the model can identify the physical meaning represented by these parameters. The language model then matches the input environmental parameter information with existing knowledge based on stored algorithm formulas and learned mapping relationships, and performs inference calculations. For example, based on the current temperature and humidity information and combined with a human comfort model, it calculates the appropriate air conditioner temperature setpoint and operating mode. After inference calculations, the large language model outputs air conditioner control information, such as "adjust the air conditioner setpoint to 24°C, turn on cooling mode, and adjust the fan speed to medium." This control information is derived from the analysis of environmental parameters and the model's predictions, and is designed to ensure that the air conditioner operates in a manner that meets the user's comfort needs.

[0082] In step 103, based on the environmental parameter information and the air-conditioning control information, it is determined through the language model whether the air-conditioning control information meets a preset condition.

[0083] The preset condition indicates that after the air conditioner is controlled to operate through the air conditioner control information, the temperature of the environment in which the air conditioner is located meets the target temperature.

[0084] In this step, the verification of whether the air conditioning control information meets the preset conditions can be achieved by constructing a closed-loop prediction mechanism of "environment-control-temperature response" through a language model.

[0085] Optionally, the environmental parameter information (such as current temperature, humidity, human activity, etc.) and the air-conditioning control information (such as temperature setting value, air volume, compressor frequency, etc.) can be fused and encoded first to form an input vector containing spatiotemporal features, which is then input into the physical heuristic prediction subnetwork integrated in the language model.

[0086] Then, the subnetwork embeds simplified thermodynamic equations (such as the room heat balance equation) with the deep learning architecture and combines historical environmental state data to perform time series prediction of the temperature evolution trajectory after executing the control instructions, generating a temperature prediction sequence for the next H time steps.

[0087] Among them, this precondition can be verified by double constraints:

[0088] Temperature deviation constraint, that is, the difference between the predicted temperature and the target temperature must be within the allowable range (such as ±0.5°C) to ensure control accuracy;

[0089] The temperature change rate constraint limits the temperature fluctuation range per unit time (such as -0.3℃ / min to +0.3℃ / min), which is in line with the human body's thermal adaptation law.

[0090] When the prediction sequence satisfies the above dual constraints, the air-conditioning control information can be determined to be valid; when the prediction sequence does not satisfy the above dual constraints, the language model can adjust parameters, generate a new control scheme through gradient optimization or beam search, and re-perform prediction verification until the preset conditions are met.

[0091] In this way, when the language model includes a large language model, its cross-domain knowledge reasoning capabilities can extract implicit control rules (such as the tendency for condensation to form due to large temperature differences in winter) and optimize prediction boundary conditions. Simultaneously, by parsing target temperature instructions expressed in natural language (such as "comfort zone"), the semantics are converted into quantitative verification thresholds, enhancing the system's adaptability to fuzzy requirements. The entire verification process combines data-driven learning capabilities with prior knowledge of physical laws to form an intelligent "prediction-verification-correction" decision-making closed loop. This ensures that the output control information not only meets the physical capabilities of the device but also dynamically adapts to environmental changes and human comfort needs, providing reliability assurance for subsequent control execution.

[0092] In step 104, when the preset condition is met, the air conditioner is controlled to operate according to the air conditioner control information.

[0093] The above technical solution is adopted to obtain the environmental parameter information of the environment in which the air conditioner is located, and input the environmental parameter information into the language model used to predict the air conditioner control information. The language model is used to analyze the environmental parameter information and output the air conditioner control information. At the same time, the environmental parameter information and the air conditioner control information are combined to verify whether the air conditioner control information meets the preset condition through the language model. The preset condition indicates that the ambient temperature meets the target temperature after the air conditioner is controlled by the air conditioner control information. When the preset condition is met, the air conditioner control information is executed, thereby predicting the trend of human body temperature change in advance and generating control instructions through the prediction ability of the language model. A conditional verification mechanism is set up to compensate for lag and shorten control delay; and by inputting both environmental parameter information and air-conditioning control information into the language model, preset conditions are used to verify whether the ambient temperature meets expectations after the control information is executed, forming a "prediction-verification-correction" closed-loop control mechanism to effectively respond to uncertain factors such as sudden environmental changes, reduce the risk of misadjustment, and improve system robustness; upgrade air-conditioning control from passive adjustment based on current errors to active prediction based on future needs, accurately bridging the time difference between physical sensation lag and system control, and at the same time achieve personalized and dynamic adaptation through data-driven, fundamentally improving user temperature comfort experience and system energy efficiency.

[0094] In some embodiments, as Figure 2 As shown, the above step 103 can be implemented in the following manner.

[0095] First, according to the environmental parameter information and the air-conditioning control information, a feature fusion network is used to obtain fusion features output by the feature fusion network.

[0096] In this step, the feature fusion network receives the environmental parameter vector X (such as temperature, humidity, number of people, etc.) and the air conditioning control vector U (such as set temperature, air volume, operating mode, etc.) as input.

[0097] For example, we can first perform differentiated feature extraction on two types of data: for the temporal dynamic characteristics of environmental parameters, a one-dimensional convolutional neural network (1D-CNN) or a Transformer encoder is used to extract spatiotemporal features (such as temperature change trends, and the correlation pattern between humidity and personnel density); for the discrete logical characteristics of control parameters, the embedding layer is used to map categorical variables (such as the operating mode "cooling / heating") into low-dimensional dense vectors, and continuous variables (such as air volume level) are normalized.

[0098] Then, the correlation between the two types of features can be established through the cross-attention mechanism. For example, the attention weight matrix of environmental features and control features can be calculated to identify key influencing factors (such as the dominant role of wind volume parameters on the temperature drop rate in a high temperature environment), and preliminary fusion features can be generated through feature concatenation or gated fusion.

[0099] In order to further capture high-order nonlinear relationships, the network can also introduce a multi-layer perceptron (MLP) module to transform the fused features through nonlinear activation functions (such as ReLU) to generate high-level features containing "environment-control" coupling semantics, such as composite features that represent the "synergistic effect of low temperature setting and dehumidification mode in high humidity scenarios."

[0100] Optionally, during training, the feature fusion network can minimize temperature prediction errors and optimize network parameters through backpropagation, so that the output fusion features can best characterize the temperature response pattern after the control command is executed. During the inference phase, the fusion features serve as the input to the language model prediction subnetwork, supporting its accurate modeling of the temperature evolution trajectory, thereby providing physically meaningful feature representations for the verification of preset conditions. This process breaks down the modal barriers between environmental parameters and control parameters through deep fusion at the data level, improving the model's ability to model the causal relationship between "control-environmental response."

[0101] Optionally, the environmental parameter information can be first encoded to obtain state features, and the air conditioning control information can be encoded to obtain control text features. Then, the state features and the control text features can be input into the feature fusion network to obtain the fusion features output by the feature fusion network.

[0102] Then, based on the fusion feature, the language model is used to determine whether the air-conditioning control information meets a preset condition.

[0103] The language model is also trained to determine whether the air-conditioning control information meets the preset condition.

[0104] Optionally, the fused feature has integrated the key features of environmental parameter information and air-conditioning control information, reflecting the association between the current environmental state and the air-conditioning control instructions. After the fused feature is input into the language model, the language model can perform further semantic analysis and feature mining on it.

[0105] Specifically, the multi-layered neural network structure within the language model processes the fused features layer by layer. Each layer extracts feature information at a different level, from simple low-level features to high-level abstract features. For example, the first layer might identify a simple comparison between the current temperature and the set temperature within the fused features. Subsequent layers will consider more comprehensive factors, such as humidity and human activity, that influence temperature fluctuations, thereby constructing a comprehensive representation of the environment and control scenario.

[0106] Furthermore, the preset condition indicates that the ambient temperature of the air conditioner's environment meets the target temperature after the air conditioner is controlled by the air conditioner control information. The language model infers and judges the input fused features based on its learned knowledge and patterns. This can be used to simulate the air conditioner's operation under given control information and predict ambient temperature trends. The model compares the predicted temperature with the target temperature. If the predicted temperature is close to the target temperature within an allowable error range, the air conditioner control information is considered to meet the preset condition; otherwise, it is considered not met.

[0107] During the judgment process, the language model also considers practical factors such as the air conditioner's cooling / heating capacity and the thermal inertia of the environment. For example, if the environment has high thermal inertia, even if the air conditioner control information seems reasonable, it may take some time for the temperature to reach the target value. The model comprehensively considers these factors to make a more accurate judgment.

[0108] Optionally, the decoded fusion feature may be first input into the language model to obtain control confirmation information output by the language model, where the control confirmation information is used to indicate whether the air conditioning control information satisfies the preset condition. Then, if the control confirmation information indicates that the air conditioning control information satisfies the preset condition, it is determined that the air conditioning control information satisfies the preset condition.

[0109] In other embodiments, when the control confirmation information indicates that the air-conditioning control information does not meet the preset condition, the model parameters of the language model can be updated according to the target temperature, the environmental parameter information and the air-conditioning control information output by the language model to obtain updated first target model parameters; then, according to the first target model parameters, the language model with updated parameters is obtained; the environmental parameter information is input into the language model with updated parameters to obtain the first air-conditioning control information output by the language model, until the first air-conditioning control information meets the preset condition; and the air-conditioning operation is controlled according to the first air-conditioning control information.

[0110] For example, the update process typically uses a backpropagation algorithm and an optimizer (such as stochastic gradient descent (SGD) or Adam). Specifically, the error between the ambient temperature and the target temperature under the air conditioning control information predicted by the model is calculated. This error can be measured using a loss function such as mean squared error (MSE). Based on the value of the loss function, the gradients of each model parameter are calculated through backpropagation. The optimizer then updates the model parameters based on these gradients, resulting in the updated first target model parameters.

[0111] Once the first target model parameters are obtained, the updated first target model parameters can be applied to the language model, replacing the original model parameters. This results in a language model with updated parameters. At this point, the model remains structurally the same as before, but the parameters have changed, giving it new predictive capabilities.

[0112] The environmental parameter information is then fed back into the language model with updated parameters. Based on the updated parameters, the model reprocesses and analyzes the environmental parameter information and uses the learned mapping relationships to output new air conditioning control information.

[0113] At this point, the new air conditioning control information can be checked to see if it meets the preset conditions. Specifically, it determines whether the ambient temperature can reach the target temperature after the air conditioning is controlled according to this new control information. If not, the above steps of updating the model parameters, updating the language model, and generating new control information are repeated, with continuous iterative optimization. Each iteration brings the model closer to an accurate mapping relationship until the generated new air conditioning control information meets the preset conditions.

[0114] When the new air conditioning control information meets the preset conditions, the air conditioning operation can be controlled accordingly. For example, if the new control information sets the temperature to 25°C, the fan speed to medium, and the operating mode to cooling, these instructions will be sent to the air conditioner, which will adjust according to these instructions, bringing the ambient temperature closer to the target temperature and ultimately achieving the desired comfort effect.

[0115] This feedback-based approach of continuously optimizing the language model makes air conditioning control more intelligent and precise. It can dynamically adjust control strategies based on actual conditions, adapting to changing environments and user needs, improving air conditioning efficiency and user comfort while also helping to reduce energy consumption and achieve energy savings.

[0116] By adopting the above technical solution, the language model can be used to accurately judge whether the air-conditioning control information meets the preset conditions based on the fusion features and its powerful learning and reasoning capabilities. The accuracy and reliability of the judgment can be continuously improved through dynamic optimization, providing strong support for the realization of intelligent and efficient air-conditioning control.

[0117] In some embodiments, as Figure 3 As shown, the method can also be implemented in the following manner.

[0118] In step 201, historical target control information is obtained, and environmental parameter information of the environment in which the air conditioner is located is obtained.

[0119] In this step, the environmental parameter information of the environment in which the air conditioner is located can be obtained by the method in step 101, and the specific method will not be repeated here.

[0120] Historical target control information refers to the set of effective control instructions that have ensured the air conditioner's ambient temperature meets the target temperature over a period of time. This includes parameters such as temperature setpoint, air volume, fresh air ratio, and compressor frequency. Control instructions generated through reinforcement learning or expert systems are labeled "energy-efficiency-optimized" control information and are used to train models to balance comfort and energy consumption.

[0121] While collecting historical target control information, it is also necessary to simultaneously obtain environmental parameter information at the corresponding moment to form a dataset of "environmental state-control action" associations. This simultaneous acquisition of historical target control information and environmental parameter information constructs a training dataset in the form of "data pairs," providing the language model with a foundation for learning the mapping relationship from "environmental state to effective control action." This supervised learning mechanism, based on real-world scenario data, enables the model to capture nonlinear relationships between environmental parameters and control strategies (such as the correction factor for humidity to temperature setpoints), avoiding control biases caused by relying solely on physical model assumptions. This provides the data foundation for the model to generate accurate air conditioning control information in subsequent steps.

[0122] In step 202, the model parameters of the language model are updated according to the historical target control information, the environmental parameter information and the air-conditioning control information to obtain updated second target model parameters.

[0123] In step 203, the language model with updated parameters is obtained according to the second target model parameters.

[0124] In step 204, the environmental parameter information is input into the language model after parameter update to obtain second air-conditioning control information output by the language model until the second air-conditioning control information meets the preset condition.

[0125] In step 205, the air conditioner is controlled to operate according to the second air conditioner control information.

[0126] Using this technical solution, after parameter updates, the resulting second target model parameters enable the language model to more accurately predict air conditioning control information. When environmental parameter information is re-input, the model can use the learned mapping relationship to output air conditioning control information that better meets actual needs, thereby improving air conditioning control effectiveness and enabling the air conditioner to more accurately adjust the ambient temperature to the target temperature, enhancing user comfort and improving the air conditioner's energy efficiency.

[0127] In some embodiments, as Figure 4 As shown, the method can also be implemented in the following manner.

[0128] In step 301, an image of a preset area is acquired.

[0129] In this step, the preset area refers to a specific, pre-defined spatial range, such as a room or office area. This area is typically served by air conditioning. Images of this area can be acquired using a camera, which is installed in a suitable location to cover the preset area and ensure clear capture of occupants. The frequency of image acquisition can be set based on actual needs, for example, capturing images at intervals (e.g., 1 minute) to ensure timely capture of changes in occupants within the area.

[0130] In step 302, face recognition is performed on the image according to the target object recognition model to obtain a recognition result.

[0131] In this step, the target object recognition model is a model specifically used to recognize faces. After being trained with a large amount of face data, it has the ability to accurately recognize different facial features. The image obtained in step 301 is input into the model, and the model will detect and analyze the face in the image. Specifically, the model will locate the position of the face in the image, extract the key features of the face (such as the shape and relative position of the eyes, nose, and mouth, etc.), and compare these features with the pre-stored face template to determine whether the target user exists in the image. The recognition result is usually presented in the form of a Boolean value or identification information, such as "is the target user" or "is not the target user", or directly gives the identity of the target user.

[0132] In step 303, when the recognition result indicates that the target user is included in the preset area, the control information is used as user control information corresponding to the target user.

[0133] In this step, if the identification results indicate that a target user exists within the preset area, the current air conditioning control information can be designated as the user control information corresponding to the target user. This control information may include temperature settings, fan speed levels, and operating modes (cooling, heating, dehumidification, etc.). This is done to prepare for subsequent adjustments to the air conditioning control strategy based on the target user's historical control habits, ensuring that the air conditioning operation can better meet the target user's personalized needs.

[0134] In step 304, historical user control information corresponding to the target user is obtained.

[0135] In this step, historical user control information records the target user's past air conditioner control operations in similar environments. This information can be obtained from a database, which stores detailed records of each user's air conditioner control operations, including the time of control, environmental parameters (such as temperature and humidity), and the corresponding control information. By analyzing this historical information, we can understand the target user's usage habits and preferences. For example, some users like to set the temperature to 24°C and the fan speed to medium, while others prefer lower temperatures and higher fan speeds.

[0136] In step 305, the model parameters of the language model are updated according to the historical user control information, the environmental parameter information and the air-conditioning control information to obtain updated third target model parameters.

[0137] In this step, the language model has already been trained. However, to more accurately predict the air conditioning control information that meets the target user's needs, its parameters need to be updated based on the target user's historical control information, current environmental parameter information, and existing air conditioning control information. The specific update process is similar to the parameter update mentioned above. First, this information is input into the language model, and the model output is obtained through forward propagation. The loss value between the model output and the historical user control information is then calculated using the backpropagation algorithm. The gradient of the loss function with respect to the model parameters is then calculated using the backpropagation algorithm. Finally, an optimization algorithm (such as stochastic gradient descent, Adam, etc.) is used to update the model parameters based on the gradient, thereby obtaining the updated third target model parameters.

[0138] In step 306, the language model with updated parameters is obtained according to the third target model parameters.

[0139] In this step, the updated third target model parameters are applied to the language model, replacing the original model parameters. This results in a language model with updated parameters. At this point, the language model has learned the relationship between the target user's historical control habits and current environmental parameters, enabling it to more accurately predict air conditioning control information suitable for the target user.

[0140] In step 307, the environmental parameter information is input into the language model after parameter update to obtain the third air-conditioning control information output by the language model until the third air-conditioning control information meets the preset condition. The third air-conditioning control information is the control information for the air-conditioning to operate in the preset area.

[0141] In this step, the current environmental parameters are input into the updated language model. Based on its learned knowledge and patterns, the model outputs new air conditioning control information. This new control information is then checked to see if it meets the preset conditions. This condition typically means that the temperature in the preset area can reach the target temperature after the air conditioner is controlled by this control information. If not, steps 305-307 are repeated, continuously updating the model parameters and generating new control information until the new air conditioning control information meets the preset conditions. Ultimately, the new air conditioning control information that meets the preset conditions will serve as the control information for the air conditioner in the preset area.

[0142] In step 308 , the air conditioner is controlled to operate according to the third air conditioner control information.

[0143] By adopting the above technical solution, the target user is identified through face recognition technology, and the target user's historical control information and current environmental parameters are combined to continuously optimize the language model. Finally, the air-conditioning control information suitable for the target user is generated and the air-conditioning operation is controlled, realizing personalized intelligent control of the air-conditioning.

[0144] It should be noted that, since the air conditioner can be an on-board air conditioner inside a vehicle, the preset area includes multiple sub-areas. For the air conditioning control information corresponding to each sub-area, the air conditioner is controlled to operate in the sub-area using the corresponding air conditioning control information.

[0145] When the air conditioner is installed inside a vehicle, the pre-set area it serves (i.e., the interior space) typically features differentiated control requirements across multiple zones. Because passengers in different sub-zones within the vehicle (e.g., the driver's seat, front passenger seat, and left / right rear seats) may have significantly different preferences for parameters such as temperature and air volume (e.g., the driver prefers low temperatures and strong air, while children in the back seat require higher temperatures and gentle air volume), the pre-set area needs to be divided into multiple sub-zones (e.g., N sub-zones, N ≥ 2). Independently generate corresponding air conditioning control information for each sub-zone based on its environmental parameters and passenger characteristics, achieving refined zone-by-zone control.

[0146] In this way, through the above technology, the multi-sub-zone independent control solution of the vehicle air conditioner transforms the traditional "device-centric" control mode into a "passenger-centric" personalized service mode, while improving the driving experience, achieving comprehensive optimization of energy efficiency and safety, which is in line with the development trend of intelligent connected vehicles.

[0147] Device embodiment

[0148] Figure 5 This is a block diagram of an air conditioning control device provided in an embodiment of the present application. Figure 5 As shown, the apparatus 400 may include:

[0149] The acquisition module 401 is used to obtain environmental parameter information of the environment in which the air conditioner is located;

[0150] An output module 402 is configured to input the environmental parameter information into a language model to obtain air conditioning control information output by the language model;

[0151] a determination module 403 for determining, based on the environmental parameter information and the air conditioning control information, using the language model, that the air conditioning control information satisfies a preset condition, wherein the preset condition indicates that after the air conditioning is controlled by the air conditioning control information, the temperature of the environment in which the air conditioning is located satisfies a target temperature;

[0152] The control module 404 is configured to control the operation of the air conditioner according to the air conditioner control information when the preset condition is met.

[0153] Optionally, the judgment module is used to obtain fusion features output by the feature fusion network based on the environmental parameter information and the air-conditioning control information through a feature fusion network; and based on the fusion features, determine through the language model whether the air-conditioning control information meets preset conditions.

[0154] Optionally, the judgment module is used to encode the environmental parameter information to obtain state features, and to encode the air-conditioning control information to obtain control text features; the state features and the control text features are input into the feature fusion network to obtain fusion features output by the feature fusion network.

[0155] Optionally, the judgment module is used to input the decoded fusion feature into the language model to obtain control confirmation information output by the language model, and the control confirmation information is used to indicate whether the air-conditioning control information meets the preset condition; when the control confirmation information indicates that the air-conditioning control information meets the preset condition, it is determined that the air-conditioning control information meets the preset condition.

[0156] Optionally, the device further includes:

[0157] When the control confirmation information indicates that the air-conditioning control information does not meet the preset condition, the model parameters of the language model are updated according to the target temperature, the environmental parameter information and the air-conditioning control information output by the language model to obtain updated first target model parameters; according to the first target model parameters, the language model with updated parameters is obtained; the environmental parameter information is input into the language model with updated parameters to obtain first air-conditioning control information output by the language model, until the first air-conditioning control information meets the preset condition; and the air-conditioning operation is controlled according to the first air-conditioning control information.

[0158] Optionally, the device further includes:

[0159] Acquire historical target control information; update the model parameters of the language model according to the historical target control information, the environmental parameter information and the air-conditioning control information to obtain updated second target model parameters; obtain the language model after parameter update according to the second target model parameters; input the environmental parameter information into the language model after parameter update to obtain second air-conditioning control information output by the language model, until the second air-conditioning control information meets the preset condition; and control the operation of the air conditioner according to the second air-conditioning control information.

[0160] Optionally, the device further includes:

[0161] An image of a preset area is obtained; face recognition is performed on the image according to a target object recognition model to obtain a recognition result; if the recognition result indicates that the preset area includes a target user, the control information is used as user control information corresponding to the target user.

[0162] Optionally, the device further includes:

[0163] Obtain historical user control information corresponding to the target user; update the model parameters of the language model based on the historical user control information, the environmental parameter information and the air-conditioning control information to obtain updated third target model parameters; obtain the language model after parameter update based on the third target model parameters; input the environmental parameter information into the language model after parameter update to obtain third air-conditioning control information output by the language model, until the third air-conditioning control information meets the preset condition, the third air-conditioning control information being control information for the operation of the air-conditioning in the preset area; and control the operation of the air-conditioning according to the third air-conditioning control information.

[0164] Optionally, the preset area includes multiple sub-areas, and the air conditioner can be controlled to operate in the sub-area using the corresponding air conditioning control information according to each sub-area.

[0165] In summary, an embodiment of the present invention provides an air-conditioning control device, which is used to obtain environmental parameter information of the environment in which the air-conditioning is located, and input the environmental parameter information into a language model for predicting air-conditioning control information, and output the air-conditioning control information by analyzing the environmental parameter information using the language model. At the same time, the environmental parameter information and the air-conditioning control information are combined to verify whether the air-conditioning control information meets a preset condition through the language model. The preset condition indicates that the ambient temperature meets the target temperature after the air-conditioning is controlled by the air-conditioning control information. When the preset condition is met, the air-conditioning control information is executed, thereby predicting the trend of human body temperature changes in advance through the prediction ability of the language model and generating a control signal. Control instructions, combined with the preset condition verification mechanism to compensate for lag and shorten control delay; and by inputting both environmental parameter information and air-conditioning control information into the language model, using the preset conditions to verify whether the ambient temperature meets expectations after the control information is executed, forming a "prediction-verification-correction" closed-loop control mechanism, effectively responding to uncertain factors such as sudden environmental changes, reducing the risk of misadjustment, and improving system robustness; upgrading air-conditioning control from passive adjustment based on current errors to active prediction based on future needs, accurately bridging the time difference between body perception lag and system control, and at the same time achieving personalized and dynamic adaptation through data-driven, fundamentally improving user temperature comfort experience and system energy efficiency.

[0166] The air conditioning control device in the embodiment of the present application can be an electronic device or a component in an electronic device, such as an integrated circuit or a chip. The electronic device can be a terminal or other devices other than a terminal. For example, the electronic device can be a GPU BOX, a mobile phone, a tablet computer, a laptop computer, a PDA, a vehicle-mounted electronic device, a mobile Internet device (Mobile Internet Device, MID), an augmented reality (augmented reality, AR) / virtual reality (virtual reality, VR) device, a robot, a wearable device, an ultra-mobile personal computer (Ultra-mobile personal computer, UMPC), a netbook or a personal digital assistant (Personal Digital Assistant, PDA), etc. It can also be a server, a network attached storage (Network Attached Storage, NAS), a personal computer (Personal Computer, PC), a television (Television, TV), an ATM or a self-service machine, etc., and the embodiment of the present application does not specifically limit it.

[0167] The air conditioning control device provided in the embodiment of the present application can implement each process implemented in the above-mentioned method embodiment. To avoid repetition, they will not be described here.

[0168] Alternatively, as Figure 6 As shown, an embodiment of the present application also provides an electronic device, including a processor and a memory, wherein the memory stores a program or instruction that can be run on the processor, and when the program or instruction is executed by the processor, the various steps of the above-mentioned air-conditioning control method embodiment are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0169] Alternatively, as Figure 7 As shown, the embodiment of the present application also provides an air conditioner, including Figure 6 Electronic devices shown.

[0170] In an embodiment of the present application, a memory may be used to store software programs and various data. The memory may mainly include a first storage area for storing programs or instructions and a second storage area for storing data, wherein the first storage area may store an operating system, an application program or instruction required for at least one function (such as a sound playback function, an image playback function, etc.), etc. In addition, the memory may include a volatile memory or a non-volatile memory, or the memory may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. Volatile memory can be random access memory (RAM), static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM) and direct RAM bus random access memory (DRRAM). The memory in the embodiments of the present application includes but is not limited to these and any other suitable types of memory.

[0171] The processor may include one or more processing units; optionally, the processor may integrate an application processor and a modem processor, wherein the application processor primarily handles operations related to the operating system, user interface, and application programs, and the modem processor primarily processes wireless communication signals, such as a baseband processor. It is understood that the modem processor may not be integrated into the processor.

[0172] An embodiment of the present application also provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the various processes of the above-mentioned air-conditioning control method embodiment are implemented and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0173] The processor is the processor in the electronic device described in the above embodiment. The readable storage medium includes a computer readable storage medium, such as a computer read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0174] An embodiment of the present application provides a computer program product, which is stored in a storage medium. The program product is executed by at least one processor to implement the various processes of the above-mentioned air-conditioning control method embodiment and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0175] It should be noted that, in this article, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the statement "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device comprising the element. In addition, it should be noted that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in the opposite order according to the functions involved. For example, the described method may be performed in an order different from that described, and various steps may also be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0176] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the relevant technology, can be embodied in the form of a computer software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0177] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of this application, ordinary technicians in this field can also make many forms without departing from the purpose of this application and the scope of protection of the claims, all of which are within the protection of this application.

Claims

1. An air conditioning control method, characterized in that: The method comprises: Obtain environmental parameter information of the environment in which the air conditioner is located; Inputting the environmental parameter information into a language model to obtain air conditioning control information output by the language model; determining, based on the environmental parameter information and the air conditioning control information, through the language model, that the air conditioning control information satisfies a preset condition, wherein the preset condition indicates that after the air conditioning is controlled by the air conditioning control information, the temperature of the environment in which the air conditioning is located satisfies a target temperature; When the preset condition is met, the air conditioner is controlled to operate according to the air conditioner control information.

2. The method according to claim 1, characterized in that The determining, based on the environmental parameter information and the air conditioning control information, by using the language model, that the air conditioning control information satisfies a preset condition includes: According to the environmental parameter information and the air conditioning control information, a feature fusion network is used to obtain fusion features output by the feature fusion network; According to the fusion feature, it is determined through the language model whether the air-conditioning control information meets a preset condition.

3. The method according to claim 2, characterized in that The step of obtaining the fusion features output by the feature fusion network according to the environmental parameter information and the air conditioning control information includes: Encoding the environmental parameter information to obtain state features, and encoding the air conditioning control information to obtain control text features; The state feature and the control text feature are input into the feature fusion network to obtain the fusion feature output by the feature fusion network.

4. The method according to claim 2, characterized in that Determining, based on the fusion feature and using the language model, that the air conditioning control information meets a preset condition includes: Inputting the decoded fusion feature into the language model to obtain control confirmation information output by the language model, wherein the control confirmation information is used to indicate whether the air conditioning control information meets the preset condition; In a case where the control confirmation information indicates that the air-conditioning control information satisfies the preset condition, it is determined that the air-conditioning control information satisfies the preset condition.

5. The method according to claim 4, characterized in that The method further comprises: If the control confirmation information indicates that the air-conditioning control information does not meet the preset condition, updating the model parameters of the language model according to the target temperature, the environmental parameter information, and the air-conditioning control information output by the language model to obtain updated first target model parameters; Obtaining the language model after parameter update according to the first target model parameters; Inputting the environmental parameter information into the language model after parameter update to obtain first air-conditioning control information output by the language model, until the first air-conditioning control information meets the preset condition; The operation of the air conditioner is controlled according to the first air conditioner control information.

6. The method according to claim 1, characterized in that The method further comprises: Obtain historical target control information; updating the model parameters of the language model according to the historical target control information, the environmental parameter information, and the air-conditioning control information to obtain updated second target model parameters; Obtaining the language model after parameter update according to the second target model parameters; Inputting the environmental parameter information into the language model after parameter update to obtain second air-conditioning control information output by the language model, until the second air-conditioning control information meets the preset condition; The operation of the air conditioner is controlled according to the second air conditioner control information.

7. The method according to claim 1, characterized in that The method further comprises: Get the image of the preset area; Performing face recognition on the image to obtain a recognition result according to a target object recognition model; In a case where the recognition result indicates that the preset area includes the target user, the control information is used as user control information corresponding to the target user.

8. The method according to claim 7, characterized in that The method further comprises: Obtaining historical user control information corresponding to the target user; updating the model parameters of the language model according to the historical user control information, the environmental parameter information, and the air-conditioning control information to obtain updated third target model parameters; Obtaining the language model after parameter update according to the third target model parameters; Inputting the environmental parameter information into the language model after parameter update to obtain third air-conditioning control information output by the language model until the third air-conditioning control information satisfies the preset condition, the third air-conditioning control information being control information for the air-conditioning to operate in the preset area; The operation of the air conditioner is controlled according to the third air conditioner control information.

9. The method according to claim 8, characterized in that The preset area includes a plurality of sub-areas, and the method further includes: For the air conditioning control information corresponding to each sub-area, the air conditioner is controlled to operate in the sub-area using the corresponding air conditioning control information.

10. An air conditioning control device, characterized in that: The device comprises: An acquisition module is used to obtain environmental parameter information of the environment in which the air conditioner is located; an output module, configured to input the environmental parameter information into a language model and obtain air-conditioning control information output by the language model; a judgment module, configured to determine, based on the environmental parameter information and the air conditioning control information, by using the language model, whether the air conditioning control information satisfies a preset condition, wherein the preset condition indicates that after the air conditioning is controlled by the air conditioning control information, the temperature of the environment in which the air conditioning is located satisfies a target temperature; A control module is used to control the operation of the air conditioner according to the air conditioner control information when the preset conditions are met.

11. An electronic device, characterized in that: include: A processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method according to any one of claims 1 to 9 when executing the program.

12. A readable storage medium, characterized in that: When the instructions or transactions in the storage medium are executed by a processor of an electronic device, the electronic device is enabled to perform the method according to any one of claims 1 to 9.

13. A computer program product, characterized in that The method comprises instructions or transactions, which, when executed by a processor in an electronic device, cause the electronic device to perform the method according to any one of claims 1 to 10.

14. An air conditioner, characterized in that: The air conditioner includes the electronic device according to claim 11.

Images