Thermal comfort evaluation method and system based on AI Agent

Abstract

According to the thermal comfort evaluation method and system based on the AI Agent, multi-modal analysis is carried out by fusing the physiological data of the vehicle owner and the in-vehicle environment data, and accurate evaluation and personalized adjustment of the thermal comfort state of the vehicle owner are achieved. The multi-source data can be processed in real time, and a customized comfort adjusting scheme is generated according to physiological feedback of a car owner and environmental parameter changes. In the aspect of data processing, data processing and comfort adjusting operation are completed locally by the vehicle-mounted terminal, data are prevented from being uploaded to the cloud, and privacy security of user data is guaranteed fundamentally. Meanwhile, the method has a dynamic learning capability, can continuously optimize an adjustment strategy based on feedback data of a vehicle owner, and deeply excavates personalized preferences of the user.

Description

Technical Field

[0001] The present invention relates to the field of intelligent management technology, and in particular to a thermal comfort evaluation method and system based on AI Agent. Background Art

[0002] With the in-depth development of intelligent cockpit technology and environmental comfort research, thermal comfort evaluation systems are increasingly being used in fields such as automotive and architecture. While existing thermal comfort evaluation technologies have achieved some success, they still have many limitations and are unable to meet user demands for a comfortable, intelligent, and personalized experience.

[0003] Currently, most thermal comfort assessment technologies rely primarily on single-dimensional data collection and analysis. For example, some technologies only collect environmental data (such as temperature, humidity, and wind speed) and perform comfort assessments using pre-set standard models. However, this approach ignores the impact of individual user physiological characteristics (such as heart rate and body temperature) on thermal comfort perception and fails to accurately reflect the user's actual thermal comfort status. Different users in the same environment may experience vastly different thermal comfort levels due to differences in their physical constitutions, making it difficult to provide an accurate assessment based on analysis of single environmental data.

[0004] When it comes to personalized adjustments, existing systems lack the ability to deeply explore and dynamically respond to users' individual preferences. Most systems adjust the environment based solely on general comfort standards, failing to flexibly adapt based on historical user feedback and real-time physiological data. For example, air conditioning temperature adjustments cannot accurately adapt to individual user preferences, nor can they optimize settings in real time based on changes in user sensation. This prevents users from achieving a comfort experience that truly meets their needs.

[0005] Inadequate real-time performance and adaptability are also major shortcomings of existing technologies. Faced with complex and ever-changing driving or indoor environments, traditional thermal comfort evaluation systems are slow to respond and fail to promptly capture changes in user comfort needs. When environmental factors change rapidly, the system cannot quickly adjust environmental parameters, resulting in prolonged user discomfort, impacting both user experience and safety.

[0006] Furthermore, as users' privacy awareness grows, existing thermal comfort assessment technologies present privacy risks in data processing. Many systems upload user data to the cloud for analysis and processing, increasing the likelihood of data leakage and failing to meet users' demands for data privacy.

[0007] A search of patent literature revealed an invention patent with publication number CN119761169A, which discloses a method and system for optimizing indoor thermal comfort in buildings based on machine learning. This patent uses laser scanning and IoT sensors to establish a three-dimensional BIM information model of the building and utilizes machine learning to optimize indoor thermal comfort. By collecting, analyzing, and predicting environmental data in real time, the system automatically adjusts temperature control equipment to ensure indoor comfort. This patent relies on high-precision sensors and BIM models, requiring extremely accurate data collection. However, the system suffers from insufficient tolerance for missing data or sensor failures, and lacks in-depth analysis of individual user needs, making it impossible to flexibly adjust to suit different users and environments. Furthermore, its real-time performance is limited, and its computational complexity is high.

[0008] In summary, in response to the above-mentioned problems of the existing technology, studying a thermal comfort evaluation method and system based on AIAgent has become a key task that needs to be solved urgently. Summary of the Invention

[0009] In view of the defects in the prior art, the purpose of the present invention is to provide a thermal comfort evaluation method and system based on AIAgent.

[0010] According to the present invention, a thermal comfort evaluation method based on AI Agent is provided, comprising the following steps:

[0011] Step S1, acquiring multimodal data including environmental data and physiological data;

[0012] Step S2, constructing a data set by multimodal data and building a thermal comfort evaluation model;

[0013] Step S3: Input the multimodal data acquired in real time into the thermal comfort evaluation model, output a thermal comfort score and environmental adjustment suggestions, and dynamically adjust the in-vehicle environment or model parameters based on the feedback data from the vehicle owner;

[0014] Step S4: encrypt and store the multimodal data and feedback data locally in the vehicle terminal;

[0015] In step S5 , the thermal comfort evaluation model interacts collaboratively with the in-vehicle environment control system.

[0016] Preferably, step S1 includes the following sub-steps:

[0017] Step S1.1, collecting environmental data in the vehicle through the vehicle temperature sensor, vehicle humidity sensor and vehicle air flow rate sensor, the environmental data including temperature data, humidity data and air flow rate data;

[0018] Step S1.2, collecting the driver's physiological data through a heart rate monitor and a skin temperature sensor, the physiological data including heart rate data, body temperature data, and skin galvanic response data;

[0019] Step S1.3, preprocessing the environmental data and the physiological data to obtain preprocessed environmental data and preprocessed physiological data, that is, multimodal data.

[0020] Preferably, in step S1.3, the preprocessing includes data denoising, normalization and outlier removal.

[0021] Preferably, step S2 includes the following sub-steps:

[0022] Step S2.1: Receive the vehicle owner's evaluation of their current thermal comfort through the vehicle interactive interface, generate a thermal comfort evaluation label, and associate and annotate the multimodal data with the thermal comfort evaluation label to form a labeled dataset for subsequent model training.

[0023] In step S2.2, based on the data set, a deep learning-based thermal comfort evaluation model is constructed to output thermal comfort scores and environmental adjustment recommendations.

[0024] Preferably, step S2.2 includes the following sub-steps:

[0025] In step S2.2.1, a convolutional neural network is used to extract the spatial features of the environmental data, and a long short-term memory network is used to process the temporal dependencies of the physiological data to obtain joint features. The training data is the dataset constructed in step S2.1.

[0026] In step S2.2.2, the joint features are input into the fully connected regression layer, and after linear transformation and activation function processing, the theoretical optimal value under the current environment is output;

[0027] Step S2.2.3: Based on the theoretical optimal values, a multi-objective optimization algorithm is integrated to generate adjustment plans for the air conditioning temperature parameters, seat heating parameters, and air conditioning wind speed parameters;

[0028] In step S2.2.4, when the car owner modifies the comfort score through the interactive interface, the model parameters are updated through the incremental learning algorithm so that the subsequent output thermal comfort score approaches the car owner's modified value.

[0029] Preferably, step S3 includes the following sub-steps:

[0030] Step S3.1: Acquire multimodal data in real time and input it into the trained thermal comfort evaluation model, and display the output thermal comfort score and environmental adjustment suggestions through the vehicle interface;

[0031] Step S3.2: If the vehicle owner provides feedback data within a preset time, the vehicle interior environment parameters are controlled or the personalized adaptation capability of the thermal comfort evaluation model is optimized;

[0032] Step S3.4: If the vehicle owner does not provide feedback data within the preset time, step S3.1 is re-executed at the preset frequency.

[0033] Preferably, step S3.2 includes the following sub-steps:

[0034] Step S3.2.1: If the feedback data is an adjustment instruction, send the adjustment instruction to the air conditioning controller, seat controller, and air flow controller via the CAN bus;

[0035] Step S3.2.1: When the feedback data is a correction value of the comfort score, the thermal comfort evaluation model parameters are updated to adapt to personalized needs.

[0036] Preferably, in step S4: an independent data partition is divided in the encrypted storage area of the vehicle-mounted MCU chip, and the multimodal data and feedback data are encrypted and stored using the AES-256 algorithm, and the encryption key is bound to the vehicle VIN code.

[0037] Preferably, in step S5, data interaction between the thermal comfort evaluation model and the intelligent navigation and vehicle-mounted intelligent seat is achieved through the vehicle-mounted Ethernet.

[0038] The present invention also provides an AI Agent-based thermal comfort evaluation system, comprising:

[0039] Module M1, acquires multimodal data including environmental data and physiological data;

[0040] Module M2, constructs a thermal comfort evaluation model by constructing a data set through multimodal data;

[0041] Module M3 inputs the real-time multimodal data into the thermal comfort evaluation model, outputs a comfort score and environmental adjustment suggestions, and dynamically adjusts the in-car environment or model parameters based on the feedback data from the car owner;

[0042] Module M4, encrypts and stores multimodal data and feedback data locally on the vehicle terminal;

[0043] Module M5, collaborative interaction between thermal comfort evaluation model and in-vehicle environmental control system.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] 1. This invention innovatively integrates the driver's physiological data (such as heart rate and body temperature) with in-vehicle environmental data. Leveraging the AI Agent's powerful multi-dimensional analysis capabilities, it enables precise assessment and real-time adjustment of the driver's thermal comfort status. This invention not only efficiently processes environmental data such as in-vehicle temperature, humidity, and wind speed, but also closely integrates the driver's physiological responses and incorporates personalized preferences to dynamically adjust in-vehicle environmental parameters. This comprehensive analysis and adjustment mechanism transforms the traditional single-dimensional evaluation model, providing drivers with a more accurate and personalized thermal comfort assessment.

[0046] 2. The personalized thermal comfort adjustment mechanism introduced in this invention leverages driver feedback and real-time physiological data to dynamically adjust the in-vehicle environment. Specifically, this invention fine-tunes settings such as air conditioning temperature, air speed, and seat heating based on the driver's long-established temperature preferences and current sensory data. This ensures the driver is always in the most comfortable environment, effectively resolving the difficulty traditional systems have in meeting individual user needs.

[0047] 3. This invention possesses powerful real-time monitoring and dynamic response capabilities. By continuously monitoring the driver's physiological changes (such as heart rate fluctuations, body temperature fluctuations, etc.) as well as real-time changes in the in-vehicle environmental data, it can keenly capture subtle changes in the driver's thermal comfort needs. The system can promptly optimize the in-vehicle environment, ensuring the driver enjoys the best thermal comfort experience in all conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments with reference to the following drawings:

[0049] Figure 1 This is a flow chart of a thermal comfort evaluation method based on AI Agent in an embodiment of the present invention. DETAILED DESCRIPTION

[0050] The present invention will be described in detail below with reference to specific embodiments. The following examples will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those skilled in the art, several changes and improvements can be made without departing from the scope of the present invention. These all fall within the scope of protection of the present invention.

[0051] This patent provides a thermal comfort evaluation method and system based on AIAgent. By integrating the physiological data of the car owner (such as heart rate, body temperature, etc.) with the in-car environmental data (such as temperature, humidity, air flow rate, etc.), multimodal analysis is carried out to achieve accurate evaluation and personalized adjustment of the thermal comfort status of the car owner. It can process the above multi-source data in real time and generate customized comfort adjustment solutions based on the physiological feedback of the car owner and changes in environmental parameters. In terms of data processing, the present invention places the data processing and comfort adjustment operations locally on the vehicle terminal to avoid uploading data to the cloud, which fundamentally guarantees the privacy and security of user data. At the same time, the present invention has dynamic learning capabilities, and can continuously optimize the adjustment strategy based on the feedback data of the car owner, and deeply explore the user's personalized preferences.

[0052] Example 1:

[0053] Figure 1 This is a flow chart of a thermal comfort evaluation method based on AI Agent in an embodiment of the present invention.

[0054] like Figure 1 As shown, this embodiment provides a thermal comfort evaluation method based on AI Agent, including the following steps:

[0055] Step S1: Acquire multimodal data including environmental data and physiological data.

[0056] Specifically, step S1 includes the following sub-steps:

[0057] Step S1.1, collecting environmental data in the vehicle through the vehicle temperature sensor, vehicle humidity sensor and vehicle air flow rate sensor, the environmental data including temperature data, humidity data and air flow rate data;

[0058] Step S1.2, collecting the driver's physiological data through a heart rate monitor and a skin temperature sensor, the physiological data including heart rate data, body temperature data, and skin galvanic response data;

[0059] Step S1.3, preprocessing the environmental data and the physiological data to obtain preprocessed environmental data and preprocessed physiological data, that is, multimodal data.

[0060] Furthermore, in step S1.3, preprocessing includes data denoising, normalization, and outlier removal, converting all data into the same standardized range to ensure data consistency and accuracy.

[0061] Step S2: constructing a data set using multimodal data and building a thermal comfort evaluation model.

[0062] Specifically, step S2 includes the following sub-steps:

[0063] Step S2.1: Receive the vehicle owner's evaluation of their current thermal comfort through the vehicle interactive interface, generate a thermal comfort evaluation label, and associate and annotate the multimodal data with the thermal comfort evaluation label to form a labeled dataset for subsequent model training.

[0064] In step S2.2, based on the data set, a deep learning-based thermal comfort evaluation model is constructed to output thermal comfort scores and environmental adjustment recommendations.

[0065] Specifically, the deep learning model processes environmental data and physiological data, identifies patterns and features in the data, evaluates thermal comfort status in real time, and provides environmental adjustment recommendations.

[0066] Furthermore, step S2.2 includes the following sub-steps:

[0067] In step S2.2.1, a convolutional neural network (CNN) is used to extract the spatial features of the environmental data, and a long short-term memory network (LSTM) is used to process the temporal dependencies of the physiological data to obtain the joint feature F(x). The training data is the dataset constructed in step S2.1.

[0068] Step S2.2.2: Input the joint feature F(x) into the fully connected regression layer, and after linear transformation W1, W2 and ReLU activation function processing, output the theoretical optimal value (T opt , H opt , V opt ), the formula is as follows:

[0069] (T opt , H opt , V opt )=W2·σ(W1·F(x)+b1)+b2,

[0070] Among them, W2, W1 and b1, b2 are the weight matrix and bias of the regression layer respectively; σ(·) is the activation function ReLu; this process is a forward propagation, without back propagation, the model output is the same as the label of the car owner's feedback For comparison, the joint loss function is:

[0071] LOSS=λ1·MSE(S,S label )+λ2·((TT label ) 2 +(HH label ) 2 +(VV 1abel ) 2 )

[0072] Among them, the first part λ1·MSE(S,S label) is the comfort score regression error, which measures the comfort score S predicted by the model and the true label S label The mean square error of the second part λ2·((TT label ) 2 +(HH label ) 2 +(VV label ) 2 ) is the regression error of each environmental adjustment parameter, and the square error between it and the corresponding label is calculated respectively. Among them, λ1 and λ2 are weight coefficients, which are used to balance the importance ratio of comfort score regression and environmental parameter regression.

[0073] In step S2.2.3, based on the theoretical optimal values, a multi-objective optimization algorithm is integrated to generate adjustment plans for the air conditioning temperature parameters, seat heating parameters, and air conditioning wind speed parameters.

[0074] The multi-objective optimization algorithm realizes parameter adjustment by establishing an objective function. The objective function J is expressed as:

[0075] J=ω1(TT ipt ) 2 +ω2(Hh opt ) 2 +ω3(VV opt ) 2

[0076] Among them, T, H, and V are the current setting values of air conditioning temperature, seat heating power, and air conditioning wind speed, respectively. opt , H opt , V opt is the theoretical optimal value output by the model, ω1, ω2, and ω3 are the weight coefficients of the adjustment target, which are used to balance the adjustment priority of different parameters.

[0077] With the goal of minimizing the objective function J, the gradient descent method is applied to iterate the current parameters T, H, V until they converge to the optimal solution T * , H * , V * They are respectively used as recommended parameters for air conditioning temperature, seat heating power and air conditioning wind speed, as environmental adjustment suggestions.

[0078] In step S2.2.4, when the car owner modifies the comfort score through the interactive interface, the model parameters are updated through the incremental learning algorithm so that the subsequent output thermal comfort score approaches the car owner's modified value.

[0079] In step S3, the multimodal data acquired in real time is input into the thermal comfort evaluation model, a thermal comfort score and environmental adjustment suggestions are output, and the in-vehicle environment or model parameters are dynamically adjusted according to the feedback data of the vehicle owner.

[0080] Specifically, step S3 includes the following sub-steps:

[0081] Step S3.1: Acquire multimodal data in real time and input it into the trained thermal comfort evaluation model, and display the output thermal comfort score and environmental adjustment suggestions through the vehicle interface;

[0082] Step S3.2: If the vehicle owner provides feedback data (including settings for environmental parameters such as temperature, humidity, and wind speed) within a preset time, the in-vehicle environmental parameters are controlled or the personalized adaptation capability of the thermal comfort evaluation model is optimized;

[0083] Furthermore, step S3.2 includes the following sub-steps:

[0084] Step S3.2.1: If the feedback data is an adjustment instruction, the adjustment instruction is sent to the air conditioning controller, seat controller and air flow rate controller via the CAN bus.

[0085] In this embodiment, the interior environment of the car is adjusted in real time according to the adjustment instructions given by the car owner through the car touch screen or voice assistant.

[0086] Step S3.2.1: When the feedback data is a correction value of the comfort score, the thermal comfort evaluation model parameters are updated to adapt to personalized needs.

[0087] Step S3.4: If the vehicle owner does not provide feedback data within the preset time, step S3.1 is re-executed at the preset frequency.

[0088] Step S4: encrypt and store the multimodal data and feedback data locally in the vehicle terminal.

[0089] Specifically, in step S4: an independent data partition is divided in the encrypted storage area of the vehicle-mounted MCU chip, and the multimodal data and feedback data are encrypted and stored using the AES-256 algorithm, and the encryption key is bound to the vehicle VIN code.

[0090] In this embodiment, all collected environmental and physiological data is processed locally on the vehicle system, ensuring user data privacy. The driver's physiological data, comfort feedback data, and other information are not uploaded to the cloud or external servers but are stored and used only on the vehicle terminal. All data stored on the vehicle terminal is encrypted, especially sensitive physiological data. Data can only be accessed and used with the vehicle owner's authorization. Vehicle owners can review, manage, and delete their personal data at any time, ensuring data controllability and transparency.

[0091] In step S5 , the thermal comfort evaluation model interacts collaboratively with the in-vehicle environment control system.

[0092] Specifically, in step S5, data interaction between the thermal comfort evaluation model and the intelligent navigation and vehicle-mounted intelligent seat is achieved through the vehicle-mounted Ethernet.

[0093] In this embodiment, the thermal comfort evaluation model works in conjunction with other systems such as the in-vehicle smart seat and smart navigation. For example, when the in-vehicle smart seat reports a temperature adjustment signal, the thermal comfort evaluation model is triggered to re-evaluate the thermal comfort score and environmental adjustment recommendations.

[0094] Example 2:

[0095] The present invention also provides a thermal comfort evaluation system based on AI Agent. The thermal comfort evaluation system based on AI Agent can be implemented by executing the process steps of the thermal comfort evaluation method based on AI Agent. That is, those skilled in the art can understand the thermal comfort evaluation method based on AI Agent as a preferred implementation of the thermal comfort evaluation system based on AI Agent.

[0096] Specifically, the AI Agent-based thermal comfort evaluation system includes:

[0097] Module M1, acquires multimodal data including environmental data and physiological data;

[0098] Module M2 constructs and trains a thermal comfort evaluation model using multimodal data to form a data set, and obtains a trained thermal comfort evaluation model.

[0099] Module M3 inputs the real-time multimodal data into the trained thermal comfort evaluation model, outputs a comfort score and environmental adjustment suggestions, and dynamically adjusts the in-car environment or model parameters based on the driver's feedback data;

[0100] Module M4, encrypts and stores multimodal data and feedback data locally on the vehicle terminal;

[0101] Module M5, collaborative interaction between thermal comfort evaluation model and in-vehicle environmental control system.

[0102] Specifically, module M1 includes the following sub-steps:

[0103] Module M1.1 collects environmental data inside the vehicle through the vehicle temperature sensor, vehicle humidity sensor, and vehicle air flow rate sensor. The environmental data includes temperature data, humidity data, and air flow rate data;

[0104] Module M1.2 collects the driver's physiological data through a heart rate monitor and skin temperature sensor. The physiological data includes heart rate data, body temperature data, and skin galvanic response data;

[0105] Module M1.3 preprocesses the environmental data and the physiological data to obtain preprocessed environmental data and preprocessed physiological data, namely, multimodal data.

[0106] Furthermore, in module M1.3, preprocessing includes data denoising, normalization, and outlier removal, which uniformly converts all data into the same standardized range to ensure data consistency and accuracy.

[0107] Specifically, module M2 includes the following sub-steps:

[0108] Module M2.1 receives the driver's evaluation of their current thermal comfort through the vehicle's interactive interface, generates a thermal comfort evaluation label, and associates and annotates the multimodal data with the thermal comfort evaluation label to form a labeled dataset for subsequent model training.

[0109] Module M2.2, based on the data set, builds a deep learning-based thermal comfort evaluation model to output thermal comfort scores and environmental adjustment recommendations.

[0110] Specifically, the deep learning model processes environmental data and physiological data, identifies patterns and features in the data, evaluates thermal comfort status in real time, and provides environmental adjustment recommendations.

[0111] Furthermore, module M2.2 includes the following sub-steps:

[0112] Module M2.2.1 uses a convolutional neural network (CNN) to extract the spatial features of environmental data and a long short-term memory network (LSTM) to process the temporal dependencies of physiological data to obtain the joint feature F(x). The training data is the dataset constructed by module M2.1.

[0113] Module M2.2.2 inputs the joint feature F(x) into the fully connected regression layer, and after linear transformation W1, W2 and ReLU activation function processing, outputs the theoretical optimal value (T opt , H opt , V opt ), the formula is as follows:

[0114] (T opt , H opt , V opt )=W2·σ(W1·F(x)+b1)+b2,

[0115] Among them, W2, W1 and b1, b2 are the weight matrix and bias of the regression layer respectively; σ(·) is the activation function ReLu; this process is a forward propagation, without back propagation, the model output is the same as the label of the car owner's feedback For comparison, the joint loss function is:

[0116] LOSS=λ1·MSE(S,S label )+λ2·((TT label ) 2+(HH label ) 2 +(VV label ) 2 )

[0117] Among them, the first part λ1·MSE(S,S label ) is the comfort score regression error, which measures the comfort score S predicted by the model and the true label S label The mean square error of the second part λ2·((TT label ) 2 +(HH label ) 2 +(VV label ) 2 ) is the regression error of each environmental adjustment parameter, and the square error between it and the corresponding label is calculated respectively. Among them, λ1 and λ2 are weight coefficients, which are used to balance the importance ratio of comfort score regression and environmental parameter regression.

[0118] Module M2.2.3, based on the theoretical optimal values, integrates a multi-objective optimization algorithm to generate adjustment plans for air conditioning temperature parameters, seat heating parameters, and air conditioning wind speed parameters.

[0119] The multi-objective optimization algorithm realizes parameter adjustment by establishing an objective function. The objective function J is expressed as:

[0120] J=ω1(TT opt ) 2 +ω2(HH opt ) 2 +ω3(VV opt ) 2

[0121] Among them, T, H, and V are the current setting values of air conditioning temperature, seat heating power, and air conditioning wind speed, respectively. opt , H opt , V opt is the theoretical optimal value output by the model, ω1, ω2, and ω3 are the weight coefficients of the adjustment target, which are used to balance the adjustment priority of different parameters.

[0122] With the goal of minimizing the objective function J, the gradient descent method is applied to iterate the current parameters T, H, V until they converge to the optimal solution T * , H * , V * They are respectively used as recommended parameters for air conditioning temperature, seat heating power and air conditioning wind speed, as environmental adjustment suggestions.

[0123] In module M2.2.4, when the car owner modifies the comfort score through the interactive interface, the model parameters are updated through the incremental learning algorithm so that the subsequent output thermal comfort score approaches the car owner's modified value.

[0124] Specifically, module M3 includes the following sub-steps:

[0125] Module M3.1 acquires multimodal data in real time and inputs it into a trained thermal comfort evaluation model, displaying the output thermal comfort score and environmental adjustment suggestions through the vehicle interface;

[0126] Module M3.2: If the vehicle owner provides feedback data (including settings for environmental parameters such as temperature, humidity, and wind speed) within a preset time, it controls the in-vehicle environmental parameters or optimizes the personalized adaptation capabilities of the thermal comfort evaluation model;

[0127] Furthermore, module M3.2 includes the following sub-steps:

[0128] Module M3.2.1, if the feedback data is an adjustment instruction, sends the adjustment instruction to the air conditioning controller, seat controller and air flow rate controller through the CAN bus.

[0129] In this embodiment, the interior environment of the car is adjusted in real time according to the adjustment instructions given by the car owner through the car touch screen or voice assistant.

[0130] Module M3.2.1: When the feedback data is a correction value of the comfort score, the thermal comfort evaluation model parameters are updated to adapt to personalized needs.

[0131] Module M3.4: If the vehicle owner does not provide feedback data within the preset time, module M3.1 will be re-executed at the preset frequency.

[0132] Specifically, in module M4: independent data partitions are divided in the encrypted storage area of the on-board MCU chip, and the AES-256 algorithm is used to encrypt and store multimodal data and feedback data. The encryption key is bound to the vehicle VIN code.

[0133] In this embodiment, all collected environmental and physiological data is processed locally on the vehicle system, ensuring user data privacy. The driver's physiological data, comfort feedback data, and other information are not uploaded to the cloud or external servers but are stored and used only on the vehicle terminal. All data stored on the vehicle terminal is encrypted, especially sensitive physiological data. Data can only be accessed and used with the vehicle owner's authorization. Vehicle owners can review, manage, and delete their personal data at any time, ensuring data controllability and transparency.

[0134] Specifically, in module M5, data interaction between the thermal comfort evaluation model and intelligent navigation and vehicle-mounted intelligent seats is realized through the vehicle Ethernet.

[0135] In this embodiment, the thermal comfort evaluation model works in conjunction with other systems such as the in-vehicle smart seat and smart navigation. For example, when the in-vehicle smart seat reports a temperature adjustment signal, the thermal comfort evaluation model is triggered to re-evaluate the thermal comfort score and environmental adjustment recommendations.

[0136] Those skilled in the art will appreciate that, in addition to implementing the system and its various devices, modules, and units provided by the present invention in purely computer-readable program code, it is entirely possible to implement the same functions of the system and its various devices, modules, and units provided by the present invention in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system and its various devices, modules, and units provided by the present invention can be considered a hardware component, and the devices, modules, and units included therein for implementing various functions can also be considered as structures within the hardware component; the devices, modules, and units for implementing various functions can also be considered as both software modules implementing the method and structures within the hardware component.

[0137] The above describes specific embodiments of the present invention. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art may make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. The embodiments of this application and the features in the embodiments may be combined with each other in any manner unless there is a conflict.

Claims

1. A thermal comfort evaluation method based on AI Agent, characterized in that: The steps include: Step S1, acquiring multimodal data including environmental data and physiological data; Step S2, constructing a thermal comfort evaluation model by forming a data set using the multimodal data; Step S3: Inputting the multimodal data acquired in real time into the thermal comfort evaluation model, outputting a thermal comfort score and environmental adjustment suggestions, and dynamically adjusting the in-vehicle environment or model parameters based on the feedback data from the vehicle owner; Step S4, encrypting and storing the multimodal data and the feedback data locally in the vehicle terminal; In step S5, the thermal comfort evaluation model interacts collaboratively with the in-vehicle environment control system.

2. The thermal comfort evaluation method based on AIAgent according to claim 1, characterized in that: The step S1 includes the following sub-steps: Step S1.1, collecting environmental data in the vehicle through the vehicle temperature sensor, the vehicle humidity sensor, and the vehicle air flow rate sensor, wherein the environmental data includes temperature data, humidity data, and air flow rate data; Step S1.2, collecting physiological data of the vehicle owner through a heart rate monitor and a skin temperature sensor, wherein the physiological data includes heart rate data, body temperature data, and skin galvanic response data; Step S1.3, preprocessing the environmental data and the physiological data to obtain preprocessed environmental data and preprocessed physiological data, that is, multimodal data.

3. The thermal comfort evaluation method based on AIAgent according to claim 2, characterized in that: In step S1.3, the preprocessing includes data denoising, normalization and outlier removal.

4. The thermal comfort evaluation method based on AIAgent according to claim 1, characterized in that: The step S2 includes the following sub-steps: Step S2.1: Receive the vehicle owner's evaluation of their current thermal comfort through the vehicle interactive interface, generate a thermal comfort evaluation label, and associate and annotate the multimodal data with the thermal comfort evaluation label to form a labeled dataset for subsequent model training. Step S2.2: Based on the data set, a deep learning-based thermal comfort evaluation model is constructed to output a thermal comfort score and environmental adjustment suggestions.

5. The thermal comfort evaluation method based on AIAgent according to claim 4 is characterized in that: The step S2.2 includes the following sub-steps: Step S2.2.1, using a convolutional neural network to extract spatial features of environmental data and a long short-term memory network to process temporal dependencies of physiological data to obtain joint features, with the training data being the dataset constructed in step S2.1; Step S2.2.2: Input the joint features into the fully connected regression layer, and after linear transformation and activation function processing, output the theoretical optimal value under the current environment; Step S2.2.3: Based on the theoretical optimal values, a multi-objective optimization algorithm is integrated to generate adjustment plans for the air conditioning temperature parameters, seat heating parameters, and air conditioning wind speed parameters; In step S2.2.4, when the car owner modifies the comfort score through the interactive interface, the model parameters are updated through the incremental learning algorithm so that the subsequent output thermal comfort score approaches the car owner's modified value.

6. The thermal comfort evaluation method based on AIAgent according to claim 1, characterized in that: The step S3 includes the following sub-steps: Step S3.1, acquiring multimodal data in real time and inputting it into the trained thermal comfort evaluation model, and displaying the output thermal comfort score and environmental adjustment suggestions through the vehicle interface; Step S3.2: If the vehicle owner provides feedback data within a preset time, the vehicle interior environment parameters are controlled or the personalized adaptation capability of the thermal comfort evaluation model is optimized; Step S3.4: If the vehicle owner does not provide feedback data within the preset time, step S3.1 is re-executed at the preset frequency.

7. The thermal comfort evaluation method based on AIAgent according to claim 6, characterized in that: The step S3.2 includes the following sub-steps: Step S3.2.1: If the feedback data is an adjustment instruction, send the adjustment instruction to the air conditioning controller, seat controller, and air flow controller via the CAN bus; Step S3.2.1: When the feedback data is a correction value of the comfort score, the thermal comfort evaluation model parameters are updated to adapt to personalized needs.

8. The thermal comfort evaluation method based on AIAgent according to claim 1, characterized in that: In the step S4: an independent data partition is divided in the encrypted storage area of the vehicle-mounted MCU chip, and the multimodal data and the feedback data are encrypted and stored using the AES-256 algorithm, and the encryption key is bound to the vehicle VIN code.

9. The thermal comfort evaluation method based on AIAgent according to claim 1, characterized in that: In step S5, data interaction between the thermal comfort evaluation model and the intelligent navigation and vehicle-mounted intelligent seat is achieved through vehicle-mounted Ethernet.

10. A thermal comfort evaluation system based on AI Agent, characterized by: include: Module M1, acquires multimodal data including environmental data and physiological data; Module M2, constructs a thermal comfort evaluation model by constructing a data set through multimodal data; Module M3 inputs the real-time multimodal data into the thermal comfort evaluation model, outputs a thermal comfort score and environmental adjustment suggestions, and dynamically adjusts the in-car environment or model parameters based on the feedback data from the car owner; Module M4, encrypting and storing the multimodal data and the feedback data locally in the vehicle terminal; Module M5, the thermal comfort evaluation model interacts collaboratively with the in-vehicle environment control system.

Images