On-vehicle air conditioner control method, vehicle and readable storage medium

Abstract

The present application discloses a vehicle air conditioner control method, a vehicle, and a readable storage medium, relating to the technical field of vehicles. The method includes: in response to a vehicle air conditioner control instruction, obtaining the current environment, the current vehicle operation state, and vehicle user characteristics; determining target air conditioner parameters under the current environment, the current vehicle operation state, and vehicle user characteristics through a preset air conditioner prediction model, where the preset air conditioner prediction model is obtained by training an initial air conditioner prediction model based on a credible training sample set; and controlling the operation of the vehicle air conditioner in the vehicle according to the target air conditioner parameters. The present application aims to solve the technical problem of low intelligence level of vehicle air conditioners.

Description

Technical Field

[0001] The present application relates to the technical field of vehicles, and in particular to an in-vehicle air conditioner control method, a vehicle, and a readable storage medium. Background Art

[0002] With the development of intelligent vehicle technology, users have higher and higher requirements for the comfort of the in-vehicle environment. Traditional in-vehicle air conditioners generally adopt a manual adjustment method. However, if the manually adjusted temperature is too high, it will waste energy; if the manually adjusted temperature is too low, it will affect the riding comfort. Moreover, in actual driving scenarios, users often do not have time to set the air conditioner because they are eager to set off, which will also affect the riding comfort. Therefore, there is currently a technical problem of low intelligence of in-vehicle air conditioners.

[0003] The above content is only used to assist in understanding the technical solution of the present application, and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main purpose of the present application is to provide an in-vehicle air conditioner control method, a vehicle, and a readable storage medium, aiming to solve the technical problem of low intelligence of in-vehicle air conditioners.

[0005] To achieve the above purpose, the present application provides an in-vehicle air conditioner control method, and the method includes:

[0006] Respond to an in-vehicle air conditioner control instruction, and obtain the current environment, the current vehicle operating state, and the vehicle user characteristics;

[0007] Determine the target air conditioner parameters under the current environment, the current vehicle operating state, and the vehicle user characteristics through a preset air conditioner prediction model, where the preset air conditioner prediction model is obtained by training an initial air conditioner prediction model based on a credible training sample set;

[0008] Control the operation of the in-vehicle air conditioner in the vehicle according to the target air conditioner parameters.

[0009] In an embodiment, the method includes:

[0010] Obtain the air conditioner historical usage data of the target user group, where the target user group includes the vehicle users corresponding to the in-vehicle air conditioner and the same-type users of the vehicle users, the air conditioner historical usage data includes multiple sub-usage data, and the sub-usage data includes at least one of the following: historical vehicle user characteristics, historical air conditioner feedback adjustment data, historical air conditioner setting parameters, historical environment, and historical vehicle operating state;

[0011] Screen out a credible training sample set from the air conditioner historical usage data according to multiple preset large models;

[0012] Determine a preset air conditioner prediction model based on the trusted training sample set and the initial air conditioner prediction model.

[0013] In one embodiment, the step of screening out a trusted training sample set from the historical air conditioner usage data based on multiple preset large models includes:

[0014] For each preset large model, input the historical air conditioner usage data into the preset large model, and remove the abnormal sub-usage data in the historical air conditioner usage data through the preset large model to obtain a normal usage data set screened out by the preset large model from the historical air conditioner usage data;

[0015] Extract the target normal data jointly included in all the normal usage data sets to obtain a trusted training sample set.

[0016] In one embodiment, the step of determining a preset air conditioner prediction model based on the trusted training sample set and the initial air conditioner prediction model includes;

[0017] Perform scene feature extraction on the trusted training sample set to obtain multiple scene parameter training features. Each scene parameter training feature includes a scene training feature and an air conditioner parameter setting feature associated with the scene training feature. The scene training feature includes at least one of the following: training vehicle user features, training air conditioner feedback adjustment data, training environment, and training vehicle operating state. The air conditioner parameter setting feature includes at least one of the following: air conditioner switch setting parameter, seat heating setting parameter, blowing setting parameter, air conditioner temperature setting parameter, air conditioner air volume setting parameter, temperature adjustment mode setting parameter, and circulation mode setting parameter;

[0018] Train the initial air conditioner prediction model based on the multiple scene parameter training features to obtain a trained teacher model;

[0019] Obtain a student model with a parameter magnitude smaller than that of the teacher model, update the student model through the teacher model, and use the updated student model as the preset air conditioner prediction model.

[0020] In one embodiment, the target air conditioner parameters include at least one of a target air conditioner switch parameter, a target seat heating parameter, a target blowing mode, a target air conditioner temperature, a target air conditioner air volume, a target circulation mode, and a target temperature adjustment mode; the step of determining the target air conditioner parameters in the current environment, the current vehicle operating state, and the vehicle user characteristics through the preset air conditioner prediction model includes:

[0021] Input the current environment, the current vehicle operating state, and the vehicle user characteristics into a preset air conditioner prediction model, and determine the target air conditioner switch parameter, the target seat heating parameter, the target blowing mode, the target air conditioner temperature, the target air conditioner air volume, the target circulation mode, and / or the target temperature adjustment mode under the current environment, the current vehicle operating state, and the vehicle user characteristics through the preset air conditioner prediction model;

[0022] Among them, the current environment includes at least one of the current weather, the current outside temperature, the current inside temperature, the current inside humidity, the current outside rainfall, the current outside environment image, and the current inside environment image. The current vehicle operating state includes the current vehicle battery power. The vehicle user characteristics include at least one of the age, gender, current usage month, current vehicle latitude, and user behavior of the vehicle user.

[0023] In one embodiment, the method further includes:

[0024] In a preset driving scenario, if the current outside temperature is greater than a preset high temperature threshold, the output of the preset air conditioner prediction model includes: the target air conditioner switch parameter is on, the target air conditioner air volume includes an increased air volume, and / or the target blowing mode includes blowing downward and / or blowing cold air;

[0025] In a preset driving scenario, if the current outside temperature is less than a preset low temperature threshold, the output of the preset air conditioner prediction model includes: the target air conditioner switch parameter is on, the target air conditioner air volume includes an increased air volume, and / or the target blowing mode includes blowing downward and / or blowing hot air;

[0026] Among them, the preset driving scenario includes: the current weather is rainy, the current inside humidity is greater than a preset humidity threshold, the current outside rainfall is greater than a preset rainfall threshold, and / or the detected water accumulation in the current outside environment image is greater than a preset water accumulation threshold.

[0027] In one embodiment, the method includes:

[0028] When it is detected that the outside air quality in the current outside environment image is less than a preset quality standard, the output of the preset air conditioner prediction model includes that the target circulation mode is the internal circulation.

[0029] In one embodiment, the method further includes:

[0030] When the absolute value of the difference between the target comfort temperature corresponding to the current clothing thickness of the vehicle occupant detected in the current in-vehicle environment image and the current in-vehicle temperature is greater than a preset temperature difference threshold, and the vehicle is in a preset rapid temperature adjustment mode, and the temperature adjustment duration from the current in-vehicle temperature to the target comfort temperature is less than the expected driving duration of the vehicle, it is determined that the output of the preset air conditioner prediction model includes: the target temperature adjustment mode includes the preset rapid temperature adjustment mode, and the target air conditioner temperature includes adjusting the current in-vehicle temperature to the target comfort temperature;

[0031] Among them, the target comfort temperature is obtained by adjusting the preset standard somatosensory comfort temperature based on the current clothing thickness to obtain an initial comfort temperature, and then adjusting the initial comfort temperature according to the obtained temperature preference of the vehicle occupant.

[0032] In addition, to achieve the above object, the present application also provides an in-vehicle air conditioner control device, and the device includes:

[0033] An acquisition module, configured to acquire the current environment, the current vehicle operating state, and the vehicle user characteristics in response to an in-vehicle air conditioner control instruction;

[0034] A determination module, configured to determine the target air conditioner parameters under the current environment, the current vehicle operating state, and the vehicle user characteristics through a preset air conditioner prediction model, where the preset air conditioner prediction model is obtained by training an initial air conditioner prediction model based on a credible training sample set;

[0035] An operation module, configured to control the operation of the in-vehicle air conditioner in the vehicle according to the target air conditioner parameters.

[0036] In addition, to achieve the above object, the present application also provides a vehicle, and the vehicle includes a vehicle body and a controller, the controller is disposed on the vehicle body, and the controller is configured to execute the steps of implementing the in-vehicle air conditioner control method as described above.

[0037] In addition, to achieve the above object, the present application also provides a computer-readable storage medium, on which a program for implementing the in-vehicle air conditioner control method is stored, and when the program for the in-vehicle air conditioner control method is executed by a processor, the steps of the in-vehicle air conditioner control method as described above are implemented.

[0038] In addition, to achieve the above object, the present application also provides a computer program product, including a computer program, and when the computer program is executed by a processor, the steps of the in-vehicle air conditioner control method as described above are implemented.

[0039] One or more technical solutions proposed in this application have at least the following technical effects: This application can obtain the current environment, the current vehicle operating state, and the vehicle user characteristics, and determine the target air-conditioning parameters under the current environment, the current vehicle operating state, and the vehicle user characteristics through a preset air-conditioning prediction model, so that the target air-conditioning parameters can adapt to the external environment, the vehicle operating state, and meet the personalized needs of users at the same time. Further, since the preset air-conditioning prediction model is obtained by training the initial air-conditioning prediction model based on a credible training sample set, and the credible training samples are credible, the prediction accuracy of the preset air-conditioning prediction model trained by the credible training samples is higher, which is convenient for more accurately determining the target air-conditioning parameters, and controlling the operation of the in-vehicle air conditioner in the vehicle according to the target air-conditioning parameters. Without the need for user operation, the in-vehicle air conditioner can be accurately controlled, improving the intelligence level of the in-vehicle air conditioner. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings herein are incorporated into the specification and form a part of the specification, showing embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application.

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the accompanying drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0042] Figure 1 It is a schematic flowchart of the first embodiment of the in-vehicle air conditioner control method of the present application;

[0043] Figure 2 It is a schematic flowchart of the second embodiment of the in-vehicle air conditioner control method of the present application;

[0044] Figure 3 It is a schematic flowchart of determining the target air-conditioning parameters and controlling the operation of the in-vehicle air conditioner according to the target air-conditioning parameters in an example of the in-vehicle air conditioner control method of the present application;

[0045] Figure 4 It is a schematic flowchart of training to obtain a preset air-conditioning prediction model and predicting the target air-conditioning parameters through the preset air-conditioning prediction model in another example of the in-vehicle air conditioner control method of the present application;

[0046] Figure 5 It is a schematic structural diagram of the in-vehicle air conditioner control device of the present application.

[0047] The implementation, functional features, and advantages of the objectives of the present application will be further described in conjunction with the embodiments with reference to the accompanying drawings. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0048] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0049] To better understand the technical solutions of the present application, the following will be described in detail in conjunction with the accompanying drawings of the specification and specific implementation manners.

[0050] With the development of intelligent vehicle technology, users have higher and higher requirements for the comfort of the in-vehicle environment. The traditional air-conditioning system cannot be adjusted according to the personalized needs of users. Therefore, the present application provides a method for controlling an in-vehicle air conditioner. In this embodiment, the current environment, the current vehicle operating state, and the vehicle user characteristics can be obtained. Through a preset air-conditioning prediction model, the target air-conditioning parameters under the current environment, the current vehicle operating state, and the vehicle user characteristics are determined, so that the target air-conditioning parameters can adapt to the external environment (the current environment), the vehicle operating environment, and meet the personalized needs of users at the same time. Further, the preset air-conditioning prediction model can be obtained by training an initial air-conditioning prediction model based on a credible training sample set screened from multiple preset large models. Each preset large model has a powerful semantic understanding ability, so that effective screening of training samples can be carried out. In this embodiment, a credible training sample set is jointly screened by multiple preset large models, which can ensure the credibility of the credible training sample set and avoid the inaccurate screening caused by the cognitive deviation of a single preset large model. Furthermore, training the initial air-conditioning prediction model with the screened credible training sample set can improve the accuracy of the preset air-conditioning prediction model, facilitate more accurate determination of the target air-conditioning parameters, and control the operation of the in-vehicle air conditioner in the vehicle according to the target air-conditioning parameters. Without the need for user operation, the in-vehicle air conditioner can be accurately controlled, improving the intelligence level of the in-vehicle air conditioner.

[0051] An embodiment of the present application provides a method for controlling an in-vehicle air conditioner, referring to Figure 1 , Figure 1 is a schematic flowchart of the first embodiment of the method for controlling an in-vehicle air conditioner in the present application. In the first embodiment, the method for controlling an in-vehicle air conditioner includes steps S10 to step S30:

[0052] Step S10, in response to an in-vehicle air conditioner control instruction, obtain the current environment, the current vehicle operating state, and the vehicle user characteristics;

[0053] It should be noted that the vehicle can be an electric vehicle, a hybrid vehicle, etc., and this embodiment does not make specific limitations thereto. The current environment can reflect the environment where the vehicle is located. For example, the current environment can reflect the temperature, humidity, and / or weather, etc., of the place where the vehicle is located. For example, the current environment can include at least one of the current weather, the current outside temperature of the vehicle, the current inside temperature of the vehicle, the current inside humidity of the vehicle, the current outside rainfall of the vehicle, the current outside environment image, and the current inside environment image.

[0054] The current weather can be obtained through the weather API (Application Programming Interface) interface on the vehicle. The current weather can describe the weather conditions in the area where the vehicle is located. The current weather can be rainy, sunny, cloudy, etc.; rainy days can also be divided into heavy rain, heavy rain, and light rain, etc., based on the rainfall. The current outside temperature of the vehicle can include the current climate temperature detected by the meteorological bureau and / or the outside detection temperature detected by the outside temperature sensor of the vehicle. The current climate temperature is the environmental temperature in the area where the vehicle is located. The current climate temperature can be detected by the meteorological bureau, and in this embodiment, the current climate temperature can be obtained through the weather API interface of the vehicle. Different climate environments require different parameters to be set for the in-vehicle air conditioner.

[0055] The current inside temperature of the vehicle can be detected by the temperature sensor inside the vehicle, and the current outside temperature of the vehicle can be detected by the device for detecting temperature outside the vehicle. Although the current climate temperature can reflect the temperature in the area where the vehicle is located, the current climate temperature reflects the overall temperature in the area where the vehicle is located, and there may be differences in the temperatures of different places within the area. For example, affected by the environment such as buildings near the vehicle, the outside temperature of the vehicle may be higher or lower than the current climate temperature. Therefore, in this embodiment, by detecting the current outside temperature, it is also convenient to more accurately determine the corresponding parameters set for the in-vehicle air conditioner subsequently. The current outside rainfall of the vehicle can be detected by the rainfall sensor outside the vehicle.

[0056] The current inside humidity of the vehicle can be detected by the device for detecting humidity inside the vehicle, such as a humidity sensor, etc. The position where the humidity sensor is set can also be set based on the actual situation. For example, the humidity sensor inside the vehicle can be set in the foot area of the vehicle, so as to facilitate detecting whether the feet of the passengers inside the vehicle are wet, and then facilitate subsequent adaptive adjustment of the parameters set for the in-vehicle air conditioner. The current outside environment image can be obtained by the image detection device set on the outside of the vehicle body of the vehicle, and the current outside environment image can reflect the driving road conditions of the vehicle. The current inside environment image can be obtained by the image detection device set inside the vehicle. The image detection device can be a device such as a camera. The current inside environment image can reflect the state of the passengers inside the vehicle. For example, the sitting situation of the passengers inside the vehicle, the thickness of the clothes worn by the passengers inside the vehicle, etc.

[0057] The current vehicle operating state can reflect the vehicle state. The current vehicle operating state also affects the parameters set for the in-vehicle air conditioner. The current vehicle operating state may include the current vehicle power. The current vehicle operating state is the vehicle driving state detected in real time currently. The current vehicle power represents the remaining power of the vehicle currently. Since turning on the in-vehicle air conditioner consumes energy and different degrees of turning on result in different energy consumptions, it is necessary to obtain the current vehicle power, so as to facilitate determining the corresponding parameters of the in-vehicle air conditioner in combination with the current vehicle power subsequently, and thus facilitate adapting to the actual operating environment of the vehicle.

[0058] Vehicle user characteristics can reflect the personalized needs of users. Vehicle user characteristics may include at least one of age, gender, current usage month, current vehicle latitude, and user behavior. Vehicle user characteristics can be represented in the form of a vector. The current usage month represents the month when the vehicle user is currently driving the vehicle. Since the climate temperature is not necessarily the same in different months, the air conditioner parameters set in different months may also be different. Therefore, vehicle user characteristics may include the current usage month so as to be able to set the air conditioner parameters for that month in a personalized manner in combination with the usage month subsequently. The current vehicle latitude represents the latitude of the area where the vehicle user is driving the vehicle. Since the climates in different latitudes are different, the air conditioner parameters set by vehicle users in different latitudes are not necessarily the same. Therefore, vehicle user characteristics may include the current vehicle latitude so as to be able to set the air conditioner settings parameters for the vehicle user in a personalized manner in combination with the latitude subsequently.

[0059] User behavior is characterized by the behavior of the user in the vehicle. User behavior can be the current behavior of the user or the historical behavior of the user. For example, when the user currently has an operation behavior on the in-vehicle air conditioner, the user behavior is the current behavior. When the user currently does not have an operation behavior on the in-vehicle air conditioner, the user behavior can be the user's historical behavior, thus facilitating providing personalized in-vehicle air conditioner settings for the user in combination with the user's age, gender, user behavior, current usage month, and current vehicle latitude subsequently. In other embodiments, vehicle user characteristics may also include the current clothing thickness. Since the air conditioner parameters set are not necessarily the same for different clothing thicknesses, personalized air conditioner setting parameters can be provided for the user in combination with the current clothing thickness.

[0060] Exemplarily, the current weather and the current climate temperature can be obtained through the vehicle's weather API. The current humidity inside the vehicle and the current temperature inside the vehicle can be obtained through the temperature sensor and the humidity sensor inside the vehicle. The current temperature outside the vehicle can be obtained through the temperature sensor outside the vehicle. The current rainfall outside the vehicle can be detected by the rainfall sensor. The current in-vehicle environment image can be obtained through the in-vehicle image detection device. The current out-of-vehicle environment image can be obtained through the out-of-vehicle image detection device. The current vehicle battery power can be obtained through the vehicle's battery management system. Since the current weather reflects the overall weather of the region, the rainfall amounts in different places within the region may not be the same. Therefore, the rainfall outside the vehicle can be detected by the rainfall sensor to improve the accuracy of the setting of the vehicle air conditioner parameters subsequently. The age and gender of the user are obtained. If the user currently has an operation behavior on the vehicle air conditioner, then this operation behavior is taken as the user behavior. If the user currently does not have an operation behavior on the vehicle air conditioner, then the user's historical operation behavior can be taken as the user behavior. The current clothing thickness can be obtained by calling a preset vision large model to recognize the user image. The user image can be obtained through an in-vehicle camera or an infrared sensor or other devices that can be used to obtain images. The preset vision large model can be used to detect the current clothing thickness of the user. The current clothing thickness can be the clothing thickness of the vehicle driver. The preset vision large model can be VLM (Visual Language Model, large model). VLM is a multi-modal artificial intelligence model that combines visual perception and natural language processing. Thus, the current clothing thickness recognized by the preset vision large model can be used as one of the vehicle user characteristics, so that the preset air conditioner prediction model can set the corresponding air conditioner according to the current clothing thickness subsequently. In other embodiments, the current clothing thickness may not be included in the vehicle user characteristics. The current clothing thickness can also be obtained by the preset air conditioner prediction model recognizing the current in-vehicle environment image.

[0061] Step S20, determine the target air conditioner parameters under the current environment, the current vehicle operating state, and the vehicle user characteristics through the preset air conditioner prediction model, where the preset air conditioner prediction model is obtained by training the initial air conditioner prediction model based on a reliable training sample set;

[0062] It should be noted that the preset air conditioner prediction model can be used to determine the target air conditioner parameters under the current environment, the current vehicle operating state, and the vehicle user characteristics; the target air conditioner parameters are characterized as the parameters that need to be set for the vehicle air conditioner.

[0063] The trustworthy training sample set is screened out by multiple preset large models and can be used to train the initial air conditioner prediction model. For example, the data for training can be input into each preset large model, and each preset large model can screen these data, and then the trustworthy training sample set jointly screened out by all preset large models can be determined. The preset air conditioner prediction model can be obtained by training the initial air conditioner prediction model with the trustworthy training samples in advance.

[0064] The preset large models are large language models. Each preset large model is different, and each preset large model can come from different suppliers. The number of preset large models is at least 3, and the number of preset large models can also be 4, 5, etc. This embodiment does not make specific limitations. For example, the multiple preset large models can be large language models such as Wenxin Yiyan, Doubao, deepseek, Kimi, etc. This embodiment does not make specific limitations. In addition, in this embodiment, the preset air conditioner prediction model can also identify the clothing thickness of the vehicle occupants in the current in-vehicle environment image.

[0065] Exemplarily, the current environment, the current vehicle operating state, and the vehicle user characteristics can be input into the preset air conditioner prediction model, and the preset air conditioner prediction model can output the target air conditioner parameters.

[0066] Step S30, control the on-vehicle air conditioner in the vehicle to operate according to the target air conditioner parameters.

[0067] The on-vehicle air conditioner can be controlled to operate according to the target air conditioner parameters, so that the on-vehicle air conditioner can be automatically adjusted for the user and the user's needs can be met at the same time.

[0068] For example, the target air conditioner parameters can be input into the on-vehicle air conditioner, so that the on-vehicle air conditioner can operate according to the target air conditioner parameters. Among them, the data format of the target air conditioner parameters output by the preset air conditioner prediction model needs to be consistent with the data format of the vehicle interface, so as to ensure that the on-vehicle air conditioner can be controlled to operate according to the target air conditioner parameters.

[0069] Embodiments of the present application can obtain the current environment, the current vehicle operating state, and vehicle user characteristics, and determine target air-conditioning parameters under the current environment, the current vehicle operating state, and vehicle user characteristics through a preset air-conditioning prediction model, so that the target air-conditioning parameters can adapt to the external environment, the vehicle operating state, and meet the personalized needs of users at the same time. Further, since the preset air-conditioning prediction model is obtained by training an initial air-conditioning prediction model based on a credible training sample set, and the credible training samples are credible, the prediction accuracy of the preset air-conditioning prediction model trained by the credible training samples is higher, which is convenient for more accurately determining the target air-conditioning parameters. Based on the target air-conditioning parameters, the on-vehicle air conditioner in the vehicle can be controlled without the need for user operation, and the on-vehicle air conditioner can be accurately controlled, improving the intelligence level of the on-vehicle air conditioner.

[0070] Further, based on the above embodiments of the present application, in the second embodiment of the present application, the same or similar content as the above embodiments can be referred to the above introduction and will not be repeated hereinafter. On this basis, referring to Figure 2 , the control method of the on-vehicle air conditioner further includes steps A10 to A30:

[0071] Step A10, obtaining the historical air-conditioning usage data of the target user group, where the target user group includes the vehicle users corresponding to the on-vehicle air conditioner and the similar users of the vehicle users, and the historical air-conditioning usage data includes multiple sub-usage data, and the sub-usage data includes at least one of the following: historical vehicle user characteristics, historical air-conditioning feedback adjustment data, historical air-conditioning setting parameters, historical climate environment, and historical vehicle operating state;

[0072] It should be noted that the historical air-conditioning feedback adjustment data represents the adjustment operations performed by the user on the on-vehicle air conditioner after getting in the car. For example, after getting in the car, the user may default to turn on the air conditioner, but the user may adjust the on-vehicle air conditioner according to personal habits, and this adjustment is the historical air-conditioning feedback adjustment data of the user. The historical climate environment refers to the historical climate temperature and historical weather, and the historical vehicle operating state refers to the historical interior temperature, historical exterior temperature, historical interior humidity, and historical vehicle power; in other embodiments, the historical vehicle operating state may further include historical exterior rainfall. The historical air-conditioning setting parameters represent the air-conditioning setting parameters under the historical vehicle user characteristics, historical air-conditioning feedback adjustment data, historical climate environment, and historical vehicle operating state.

[0073] In this embodiment, the setting types of the air-conditioning parameters corresponding to the vehicle-mounted air conditioner can include multiple types, namely, air-conditioning switch, seat heating, blowing mode, air-conditioning temperature, air-conditioning air volume, temperature adjustment mode, and circulation mode. The air-conditioning switch indicates whether the air conditioner is turned on, and the seat heating indicates whether the seat heating is turned on and the specific heating value of the seat heating. The circulation mode can include internal circulation and external circulation, and the temperature adjustment mode can include a preset rapid temperature adjustment mode and a preset normal temperature adjustment mode. The preset rapid temperature adjustment mode can include a rapid heating mode and a rapid cooling mode. The temperature change rate of the preset normal temperature adjustment mode is less than that of the preset rapid temperature adjustment mode.

[0074] The blowing mode can include blowing downward, blowing upward, etc., and specifically can be determined based on the signals of the vehicle. Different vehicle models may include different blowing modes, and this embodiment does not make specific limitations in this regard.

[0075] The historical vehicle user characteristics refer to the characteristics of the user corresponding to the sub-usage data. The historical vehicle user characteristics include at least one of the age, gender, historical month, historical vehicle latitude, and historical user behavior of the user corresponding to the sub-usage data. The historical month refers to the month corresponding to the sub-usage data, and the historical user behavior is characterized by the operation behavior of the vehicle air conditioner when the user first enters the vehicle. The historical vehicle latitude represents the latitude corresponding to the sub-usage data. The climates of different latitudes are different, and the climates of regions with similar latitudes are also similar. Therefore, when determining similar users, similar users of vehicle users can also be matched based on latitude to improve the accuracy of determining similar users of vehicle users. In other embodiments, the historical vehicle user characteristics can also include historical clothing thickness. The historical clothing thickness represents the thickness of the clothing worn by the user corresponding to the sub-usage data, and the clothing thickness can be identified through a preset vision large model. For example, a device capable of taking images is set in the vehicle to take images of the user, and the images of the user are input into the preset vision large model, and the clothing thickness of the user is output through the preset vision large model.

[0076] The target user group includes vehicle users corresponding to vehicle air conditioners and users of the same kind as the vehicle users. Users of the same kind are those similar to vehicle users. Specifically, corresponding users of the same kind can be found based on the vehicle user characteristics corresponding to the vehicle users. Corresponding users of the same kind can be found based on the age, gender, and vehicle latitude in the vehicle user characteristics corresponding to the vehicle users. For example, users who are in the same age group as the vehicle user, of the same gender, and the absolute value of the difference in vehicle latitudes between the two vehicles is less than or equal to a preset latitude threshold can be regarded as users of the same kind as the vehicle user. Furthermore, the historical air conditioner usage data of users of the same kind can be obtained, which is also convenient for enriching the training data and helps improve the accuracy of subsequent model training. Different ages can be divided into different age groups based on the actual situation. For example, the age groups can be divided into 18 - 25, 26 - 30, 31 - 35, 36 - 40, 41 - 45, 46 - 50, etc. This embodiment does not make specific limitations in this regard. When the absolute value of the difference in vehicle latitudes between a user of the same kind and the vehicle user is less than the preset latitude threshold, it indicates that the latitudes of the user of the same kind and the vehicle user do not differ much, and the climate environments in the regions where they are located are the same. Therefore, the historical air conditioner usage data of this user of the same kind can also be used as training data.

[0077] In other embodiments, corresponding users of the same kind can also be found based on the age, gender, vehicle latitude, month, and clothing thickness in the vehicle user characteristics. Among them, the process of finding users of the same kind corresponding to age, gender, and vehicle latitude has been described above, and this embodiment will not elaborate further. Furthermore, in the case where the user is in the same age group as the vehicle user, of the same gender, and the absolute value of the difference in vehicle latitudes between the two vehicles is less than or equal to the preset latitude threshold, if the absolute value of the difference in clothing thickness between the two is less than the preset clothing difference threshold, belongs to the same month and / or the same season as the month corresponding to the vehicle user, the user can be regarded as a user of the same kind. In other embodiments, users of the same kind as the vehicle user can also be found by combining user behaviors. This embodiment does not make specific limitations in this regard. Therefore, in this embodiment, users of the same kind can be users who are of the same gender, in the same age group, and have similar vehicle latitudes as the vehicle user. In other embodiments, users of the same kind can also be users who are of the same gender, in the same age group, have similar vehicle latitudes, similar clothing thicknesses, and are in the same month, or users of the same kind can also be users who are of the same gender, in the same age group, have similar vehicle latitudes, similar clothing thicknesses, and are in the same season. It can also be users who are of the same gender, in the same age group, have similar vehicle latitudes, similar clothing thicknesses, and have the same vehicle signal as the vehicle user.

[0078] Exemplarily, obtain the historical usage data of the vehicle user, search for multiple similar users based on the vehicle user characteristics of the vehicle user, and obtain the historical usage data corresponding to each similar user. Use the historical usage data of the vehicle user and the historical usage data of all similar users as the air conditioner historical usage data. Among them, the historical usage data of the vehicle user includes multiple sub-usage data, and the historical usage data of similar users also includes multiple usage data.

[0079] Step A20, based on multiple preset large models, screen out a credible training sample set from the air conditioner historical usage data;

[0080] It should be noted that the credible training sample set is characterized by reasonable sub-usage data. Each preset large model can, based on its powerful semantic understanding ability, screen out a credible training sample set from the air conditioner usage data. To ensure the accuracy of the credible training sample set, in this embodiment, at least three different preset large models can be used to screen the air conditioner historical usage data, and the three different preset large models are characterized by different technical architectures of the large models, and each preset large model can be the latest and optimal model version under its corresponding technical architecture. For example, the different preset large models can be deepseek-R1, Doubao, and Kimi. This embodiment does not make specific limitations on this.

[0081] Exemplarily, input the air conditioner historical usage data into each preset large model, and jointly screen out a credible training sample set through all preset large models.

[0082] In a feasible embodiment, step A20 further includes steps A21 to A22:

[0083] Step A21, for each preset large model, input the air conditioner historical usage data into the preset large model, and remove the abnormal sub-usage data in the air conditioner historical usage data through the preset large model to obtain the normal usage data set screened out by the preset large model from the air conditioner historical usage data;

[0084] Step A22, extract the target normal data jointly included in all normal usage data sets to obtain a credible training sample set.

[0085] It should be noted that the air conditioner historical usage data includes multiple sub-usage data, and the abnormal usage data is the abnormal sub-usage data. The normal usage data set is a set of multiple normal data. Each preset large model can output its corresponding normal usage data set, and there may be differences in the normal usage data sets output by each preset large model.

[0086] Each piece of sub - usage data is the actual operation behavior of the vehicle user on the vehicle air conditioner. However, there are many unreasonable air - conditioner operation behaviors in the actual user behavior, and they cannot be cleaned according to the rules. If these unreasonable air - conditioner operation behaviors are also used as samples for training the model, it will cause the model to predict unreasonable air - conditioner operation behaviors, and then it will be difficult to automatically set the best parameters for the vehicle air conditioner for the user. Therefore, in this embodiment, multiple preset large models are used to clean and screen the historical air - conditioner usage data. Each preset large model can remove the abnormal usage data from the historical air - conditioner usage data, and then screen out the normal usage data set.

[0087] Although multiple preset large models are all large - language models, the semantic - understanding capabilities of different large - language models are not necessarily the same. In this embodiment, to avoid the situation of bias in the training samples caused by the screening of a single large model, multiple preset large models can be used for joint screening. For example, the optimal large - language models of multiple third - party large - model suppliers can be used as each preset large model, and the same historical air - conditioner usage data can be screened. The data that all preset large models consider credible and reasonable can be retained as the feasible training sample set.

[0088] The target normal data is characterized as the sub - usage data that all preset large models consider normal. The normal usage data sets corresponding to all preset large models respectively include the target normal data. Each piece of target normal usage data can form a credible training sample set, which can improve the credibility and reasonableness of the credible training sample set, and then facilitate improving the accuracy of the prediction of vehicle - air - conditioner parameters in the follow - up.

[0089] Exemplarily, the historical air - conditioner usage data is input into each preset large model respectively. For each preset large model, the abnormal usage data in the historical air - conditioner usage data is removed through the preset large model, and the normal usage data set screened out by the preset large model is obtained. The target normal data jointly included in all normal data sets is extracted from all normal data sets to obtain the credible training sample set. This ensures the credibility of the sample set and then facilitates improving the accuracy of subsequent model training.

[0090] Step A30: Determine the preset air - conditioner prediction model based on the credible training sample set and the initial air - conditioner prediction model.

[0091] It should be noted that the initial air - conditioner prediction model can be trained using the credible training sample set, and then the trained preset air - conditioner prediction model is determined. The initial air - conditioner prediction model can also be a large - language model, and the preset air - conditioner prediction model also belongs to the large - language model, which helps to improve the prediction accuracy of the vehicle - air - conditioner setting parameters.

[0092] Exemplarily, feature extraction can be performed on the trusted training sample set, and the initial air conditioner prediction model can be trained based on the data extracted by the feature extraction to determine the preset air conditioner prediction model.

[0093] In a feasible embodiment, step A30 further includes steps A31 to A33:

[0094] Step A31, perform scenario feature extraction on the trusted training sample set to obtain multiple scenario parameter training features. Each scenario parameter training feature includes a scenario training feature and an air conditioner parameter setting feature associated with the scenario training feature. The scenario training feature includes at least one of the following: training vehicle user feature, training air conditioner feedback adjustment data, training environment, and training vehicle operation state. The air conditioner parameter setting feature includes at least one of the following: air conditioner switch setting parameter, seat heating setting parameter, blowing setting parameter, air conditioner temperature setting parameter, air conditioner air volume setting parameter, temperature adjustment mode setting parameter, and circulation mode setting parameter;

[0095] It should be noted that the trusted training sample set includes a large amount of target normal use data. Among the target normal use data, there may be multiple target similar normal use data with the same historical air conditioner setting parameters. Therefore, scenario feature extraction can be performed on the multiple target similar normal data to extract the commonality of the multiple target similar normal data, so as to obtain the scenario parameter training features corresponding to the multiple target similar normal data. In this embodiment, multiple scenario parameter training features can be screened from the trusted training samples.

[0096] The scenarios corresponding to different scenario parameter training features are different, and the air conditioner parameter setting features corresponding to different scenarios are different. The air conditioner parameter setting feature in each scenario parameter training feature can be characterized as the parameter set for the in-vehicle air conditioner in the corresponding scenario. For example, the air conditioner switch setting parameter, seat heating setting parameter, blowing setting parameter, air conditioner temperature setting parameter, air conditioner air volume setting parameter, temperature adjustment mode setting parameter, and / or circulation mode setting parameter, etc. The air conditioner parameter setting feature can be determined based on the historical air conditioner setting parameters in multiple similar target normal use data. The historical air conditioner setting parameters of multiple similar target normal use data are also similar. For example, the historical air conditioner setting parameters can be the same, or the average value of each historical air conditioner setting parameter can be taken as the air conditioner setting parameter. The air conditioner switch setting parameter represents setting the in-vehicle air conditioner switch, the seat heating setting parameter represents setting the seat heating switch, the blowing setting parameter represents setting the blowing direction, the air conditioner temperature setting parameter represents setting the in-vehicle air conditioner temperature, the air conditioner air volume setting parameter represents setting the air volume of the in-vehicle air conditioner, the temperature adjustment mode setting parameter represents setting the temperature adjustment mode of the in-vehicle air conditioner, such as rapid heating, cooling, etc., and the circulation mode setting parameter represents setting the circulation mode of the in-vehicle air conditioner, such as internal circulation and external circulation, etc.

[0097] The scenario training features in each scenario parameter training feature can reflect the scenario corresponding to the air conditioner parameter setting features. The scenario training features include the training vehicle user features corresponding to the air conditioner parameter setting features, the training air conditioner feedback adjustment data, the training climate environment, and the training vehicle operation environment. Since the scenario training features are extracted from multiple target similar normal data, the training vehicle user features in the scenario training features are the commonalities of the historical vehicle user features corresponding to the multiple target similar normal data. Generally, the historical vehicle user features of the multiple target similar normal data are the same, or the age, gender, and vehicle latitude in the historical vehicle user features are the same. The training vehicle user features can also include age, gender, and vehicle latitude, etc. The training air conditioner feedback adjustment data, the training climate environment, and the training vehicle operation environment are all the same data extracted from each target similar normal data. For example, the historical air conditioner feedback adjustment data, the historical climate environment, and the historical vehicle operation state in the multiple target similar normal data are the same, so that the corresponding training air conditioner feedback adjustment data, the training climate environment, and the training vehicle operation environment can be extracted.

[0098] For example, in this embodiment, the environmental scenario features corresponding to rapid temperature rise can be extracted from the trusted training sample set, and the air conditioner setting parameter features corresponding to rapid temperature rise and the environmental scenario features can be used as a scenario parameter training feature. The environmental scenario features corresponding to rapid temperature drop can also be extracted from the trusted training sample set, and the air conditioner setting parameter features corresponding to rapid temperature drop and the environmental scenario features can be used as another scenario parameter training feature. The environmental scenario features corresponding to other different air conditioner setting parameter features can also be extracted to construct the corresponding scenario parameter training features, and this embodiment does not make specific limitations on this.

[0099] Exemplarily, the extraction data sets corresponding to different historical air conditioner setting parameters can be extracted from the trusted training samples. For each different historical air conditioner setting parameter, there is a corresponding extraction data set. Each extraction data set can include multiple target normal usage data. The historical air conditioner setting parameters of the target normal usage data in the same extraction data set are the same. For the same extraction data set, the historical air conditioner setting parameter of any target normal usage data in the extraction data set can be used as the air conditioner parameter setting feature, and then the commonalities of the historical vehicle user features, the historical air conditioner feedback adjustment data, the historical climate environment, and the historical vehicle operation state in each target normal usage data in the extraction data set can be extracted respectively, so that the training vehicle user features, the training air conditioner feedback adjustment data, the training climate environment, and the training vehicle operation environment corresponding to the air conditioner parameter setting feature can be obtained.

[0100] Multiple scene parameter training features can each correspond to different scenes, and the air conditioner setting parameters corresponding to different scenes are also different. In this embodiment, multiple scene parameter training features are extracted from the trusted training sample set, so as to facilitate the subsequent training of the initial air conditioner prediction model to learn the air conditioner setting parameters in different scenes. Furthermore, the final preset air conditioner prediction model can handle the air conditioner setting parameters in different scenes, thereby facilitating the improvement of the intelligence of the vehicle-mounted air conditioner and ensuring the comfort of users at the same time.

[0101] Step A32: Train the initial air conditioner prediction model according to multiple scene parameter training features to obtain a trained teacher model.

[0102] Step A33: Obtain a student model with a parameter order of magnitude smaller than that of the teacher model, update the student model through the teacher model, and use the updated student model as the preset air conditioner prediction model.

[0103] It should be noted that the initial air conditioner prediction model can be a large language model. The parameter order of magnitude of the teacher model is greater than that of the student model. The parameter order of magnitude is the order of magnitude of the model training parameters (that is, the range of the size of the quantity). The parameter order of magnitude can reflect the scale of the model. The larger the parameter order of magnitude, the larger the scale of the model, and the smaller the parameter order of magnitude, the smaller the scale of the model. The scale of the student model is smaller than that of the teacher model, and the student model also belongs to the large language model. For example, the teacher model can be a 70B (Billion) large language model, and the student model can be a 1.5B large language model. The model output representation is the output distribution of the teacher model.

[0104] The initial air conditioner prediction model can be trained according to a large number of scene parameter training features. Exemplarily, the scene training features in the scene parameter training features can be input into the initial air conditioner prediction model. The initial air conditioner prediction model outputs the air conditioner parameter results. The model error between the air conditioner parameter results and the air conditioner parameter setting features in the scene training features can be calculated. If the model error is less than the preset loss threshold, the trained teacher model can be obtained. If the model error is greater than or equal to the preset loss threshold, new scene parameter training features can be input into the model to be trained until the model error is less than the preset loss threshold. The preset loss threshold can be set based on the actual situation, and this embodiment does not make specific limitations on this.

[0105] In this embodiment, a student model with a parameter order of magnitude smaller than that of the teacher model is obtained, so that accurate prediction of air-conditioning parameters can still be achieved while reducing the parameter order of magnitude. This embodiment is based on the principle of model distillation. Model Distillation is a technique for transferring the knowledge of a complex model (teacher model) to a simple model (student model). Its core principle is to make the student model mimic the output distribution of the teacher model, so that the student model can maintain high performance while significantly reducing the computational complexity. Therefore, in this embodiment, a student model with a parameter order of magnitude smaller than that of the teacher model is obtained, and the student model learns the model output of the teacher model, so that the student model can mimic the output of the teacher model and reduce the complexity at the same time. Furthermore, the learned model can be used as a preset air-conditioning prediction model.

[0106] Thus, the preset air-conditioning prediction model can be deployed in the vehicle. Since the preset air-conditioning prediction model can reduce the computational complexity while maintaining a high prediction accuracy, it can also support the operation of the preset air-conditioning prediction model on the vehicle side, without imposing a large operating pressure on the vehicle side. Furthermore, it is convenient to improve the prediction efficiency of vehicle air-conditioning parameters and the user experience.

[0107] For a better understanding of this embodiment, please refer to Figure 3 , Figure 3It is: a schematic flow chart for determining target air conditioner parameters and controlling the operation of the vehicle-mounted air conditioner according to the target air conditioner parameters; a brief description of the process from model training to model prediction in this embodiment is as follows. Step N1: Obtain historical air conditioner usage data. Step N2: Multiple preset large models screen the historical air conditioner usage data to obtain a credible training sample set. Step N3: Extract features from the credible training sample set to obtain a large number of scene parameter training features. Step N4: Use the large number of scene parameter training features to train the initial air conditioner prediction model to determine the preset air conditioner prediction model. Step N5: Deploy the preset air conditioner prediction model in the vehicle and periodically update the preset air conditioner prediction model. Periodically updating the preset air conditioner prediction model can be to obtain new usage data of the vehicle-mounted air conditioner and then update the preset air conditioner prediction model with the new usage data, which helps to improve the prediction accuracy of the preset air conditioner prediction model. The period for updating the preset air conditioner prediction model is not limited in this embodiment and can be set based on the actual situation. Step N6: Obtain the current environment, the current vehicle operating state, and the vehicle user characteristics. Combine step N6 with the preset air conditioner prediction model in step N5, and determine the target air conditioner parameters under the current environment, the current vehicle operating state, and the vehicle user characteristics through the preset air conditioner prediction model. For example, step N7: Determine the target air conditioner parameters. Step N8: Control the operation of the vehicle-mounted air conditioner according to the target air conditioner parameters. In addition, it should be noted that if the preset air conditioner prediction model cannot determine what the corresponding target air conditioner parameters are under the current environment, the current vehicle operating state, and the vehicle user characteristics, the teacher model in the cloud can be called to determine the target air conditioner parameters under the current environment, the current vehicle operating state, and the vehicle user characteristics. Since the preset air conditioner prediction model is deployed offline on the vehicle side, the scenarios for which the preset air conditioner prediction model can predict the target air conditioner parameters are the scenarios that exist in the credible training sample set. And because the preset air conditioner prediction model belongs to the student model and the number of parameters of the student model is relatively small, there may be a situation where the target air conditioner parameters in individual scenarios cannot be accurately determined. Therefore, the teacher model in the cloud can be called for prediction. Although there may be individual scenarios in which the preset air conditioner prediction model is difficult to accurately predict the target air conditioner parameters, this does not mean that the number of scenarios that the preset air conditioner prediction model can predict is small. Because in this embodiment, a large number of scene parameter training features can be extracted from the credible training sample set, so that the preset air conditioner prediction model can predict the corresponding target air conditioner parameters in a large number of scenarios.

[0108] In a feasible embodiment, the target air-conditioning parameters include at least one of a target air-conditioning switch parameter, a target seat heating parameter, a target blowing mode, a target air-conditioning temperature, a target air-conditioning air volume, a target circulation mode, and a target temperature adjustment mode; step S30 further includes step S31: inputting the current environment, the current vehicle operating state, and the vehicle user characteristics into a preset air-conditioning prediction model, and determining the target air-conditioning switch parameter, the target seat heating parameter, the target blowing mode, the target air-conditioning temperature, the target air-conditioning air volume, the target circulation mode, and / or the target temperature adjustment mode under the current environment, the current vehicle operating state, and the vehicle user characteristics through the preset air-conditioning prediction model;

[0109] Among them, the current environment includes at least one of the current weather, the current outside temperature, the current inside temperature, the current inside humidity, the current outside rainfall, the current outside environment image, and the current inside environment image; the current vehicle operating state includes the current vehicle battery level; the vehicle user characteristics include at least one of the age, gender, current usage month, current vehicle latitude, and user behavior of the vehicle user.

[0110] It should be noted that the target air-conditioning switch parameter can be on or off, etc. When the target air-conditioning switch parameter is on, it indicates that the vehicle-mounted air conditioner is turned on; when the target air-conditioning switch parameter is off, it indicates that the vehicle-mounted air conditioner is turned off. The target seat heating parameter can be to turn on the seat heating or turn off the seat heating, etc. The target blowing mode can be blowing upward or blowing downward, etc. The air-conditioning blowing control parameters at different positions in the vehicle can be different or the same.

[0111] The target air-conditioning temperature can be the specific set temperature of the vehicle-mounted air conditioner; the target air-conditioning air volume can be the specific set air volume of the vehicle-mounted air conditioner. The set air volumes at different positions in the vehicle can be different or the same. The target circulation mode can be to set the vehicle-mounted air conditioner to the internal circulation or the external circulation, etc. The target temperature adjustment mode can be to set the vehicle-mounted air conditioner to perform preset rapid temperature adjustment or preset normal temperature adjustment, etc.

[0112] The current environment, the current vehicle operating state, and the vehicle user characteristics will all affect the target air-conditioning parameters. The vehicle user characteristics can refer to the characteristics of the driver. For example, when the current climate temperature is less than the preset low temperature threshold and the current clothing thickness is relatively thick, the vehicle-mounted air conditioner can be set to the heating mode. And the thicker the current clothing thickness, the lower the set temperature can be, and the smaller the set air volume can also be. The thinner the current clothing thickness, the higher the set temperature of the vehicle-mounted air conditioner can be, and the larger the set air volume can also be. When the current climate temperature is greater than the preset high temperature threshold, the vehicle-mounted air conditioner can be set to the cooling mode. If the current inside humidity is higher, the air volume can be increased. The set temperature may also be different in different months. In addition, even for the same age, gender, month, clothing thickness, and user behavior, if the current vehicle latitude is different, the corresponding target air-conditioning parameters may also be different.

[0113] Exemplarily, the current environment, the current vehicle operating state, and the vehicle user characteristics are all input into a preset air conditioner prediction model, and the target air conditioner switch parameter, the target seat heating parameter, the target blowing mode, the target air conditioner temperature, and the target air conditioner air volume in the scenario of the current environment, the current vehicle operating state, and the vehicle user characteristics are determined through the preset air conditioner prediction model.

[0114] In this embodiment, various air conditioner setting parameters of the vehicle-mounted air conditioner can be determined through the preset air conditioner large model, so that manual setting is not required, the intelligence level of the vehicle-mounted air conditioner is improved, and the user comfort can also be improved.

[0115] For a better understanding of this embodiment, please refer to Figure 4 , Figure 4 which is a schematic flow chart of training a preset air conditioner prediction model and predicting target air conditioner parameters through the preset air conditioner prediction model. A brief description of the process of determining the target air conditioner parameters in this embodiment is as follows: The historical air conditioner usage data may include historical vehicle user characteristics, historical air conditioner feedback adjustment data, historical air conditioner setting parameters, historical climate environment, and historical vehicle operating state. The historical air conditioner feedback adjustment data may be the user's historical air conditioner feedback adjustment behavior, and the historical vehicle user characteristics may also include the user's historical behavior, and the user's historical behavior may refer to the behavior of the user adjusting the air conditioner for the first time after getting on the vehicle. A credible training sample set of the historical air conditioner usage data can be determined. Feature extraction refers to extracting a large number of scene parameter training features from the credible training sample set. The initial air conditioner prediction model can be trained using the scene parameter training features extracted from the credible training sample set to obtain a trained teacher model, and the student model is used to learn from the teacher model to obtain the preset air conditioner prediction model. When training the initial air conditioner prediction model, the data format of the data input into the initial air conditioner prediction model needs to be consistent with the data format of the data input into the preset air conditioner prediction model, and the data format needs to conform to the specification of the vehicle-side interface, so that the vehicle can receive the target air conditioner parameters output by the preset air conditioner prediction model. The real-time data may include vehicle user characteristics, the current environment, and the current vehicle operating state. Among them, the vehicle user characteristics include the user's behavior, and the user's behavior may be the user's current behavior or historical behavior. For example, when the user has no current behavior, the user's behavior is the historical behavior, and the historical behavior may include the user's historical air conditioner feedback adjustment behavior. Inputting the real-time data into the preset air conditioner prediction model can obtain the target air conditioner parameters, and the target air conditioner parameters may include: the target air conditioner switch parameter, the target seat heating parameter, the target blowing mode, the target air conditioner temperature, the target air conditioner air volume, the target temperature adjustment mode, and the target circulation mode.

[0116] In a feasible embodiment, the vehicle air conditioner control method further includes step X10: in a preset driving scenario, if the current outside temperature is greater than a preset high temperature threshold, the output of the preset air conditioner prediction model includes: the target air conditioner switch parameter is on, the target air conditioner air volume includes an increased air volume, and / or the target blowing mode includes blowing downward and / or blowing cold air;

[0117] In a preset driving scenario, if the current outside temperature is less than a preset low temperature threshold, the output of the preset air conditioner prediction model includes: the target air conditioner switch parameter is on, the target air conditioner air volume includes an increased air volume, and / or the target blowing mode includes blowing downward and / or blowing hot air;

[0118] Wherein, the preset driving scenario includes: the current weather is rainy, the current humidity inside the vehicle is greater than a preset humidity threshold, the current rainfall outside the vehicle is greater than a preset rainfall threshold, and / or the detected water accumulation in the current outside environment image of the vehicle is greater than a preset water accumulation threshold.

[0119] It should be noted that when the current weather is rainy, it means that it is raining in the area where the vehicle is located. Therefore, when the passengers enter the vehicle, body parts such as shoes or trouser legs may be wet by the rain. For example, it may be detected that the current humidity inside the vehicle is greater than a preset humidity threshold. At this time, the air volume can be increased and blown downward, which can help the user dry body parts such as shoes and / or trouser legs, improving the comfort of the passengers in the vehicle. When the humidity inside the vehicle is greater than the preset humidity threshold, it means that the humidity inside the vehicle is relatively high and the passengers in the vehicle may be wet by the rain. For example, it may be the parts such as shoes and / or trouser legs.

[0120] It is also possible to turn on the vehicle air conditioner, increase the air volume, and the vehicle air conditioner blows downward when it is detected that the rainfall outside the vehicle is greater than the preset rainfall threshold, so as to facilitate drying the body parts such as shoes and / or trouser legs of the passengers in the vehicle in time. For example, when it is detected that the current weather in the area where the vehicle is located is cloudy and the probability of rain is 50%, it may not be possible to determine whether it is raining at the current location of the vehicle only based on the current weather. Therefore, at this time, it is possible to determine whether it is raining at the location of the vehicle based on the rainfall outside the vehicle, which is also convenient for improving the accuracy of the prediction of the preset air conditioner prediction model, and is also convenient for determining the specific preset target value of the air conditioner air volume and the specific setting parameters of the vehicle air conditioner based on the vehicle rainfall. Thus, both the accuracy of setting the vehicle air conditioner parameters and the comfort of the passengers in the vehicle are improved.

[0121] When it is detected that the accumulated water volume in the current external vehicle environment image is greater than the preset accumulated water threshold, the vehicle-mounted air conditioner can be turned on, the air volume can be increased, and the vehicle-mounted air conditioner blows downward, so as to facilitate drying the shoes and / or body parts such as trouser legs of the vehicle occupants in time. The preset accumulated water threshold and the preset rainfall threshold can be set based on the actual situation, and this embodiment does not make specific limitations on this. The preset air conditioner prediction model can be used to detect the water accumulation area in the current external vehicle environment image, so as to detect the accumulated water volume in the preset air conditioner prediction model through the water accumulation area. Furthermore, when the accumulated water volume is greater than the preset accumulated water threshold, the vehicle-mounted air conditioner can be turned on in time, the air volume can be increased, and the vehicle-mounted air conditioner blows downward. For example, the water accumulation area in the current external vehicle environment image can be detected, and the water accumulation area can be used as the accumulated water volume. In other embodiments, the water accumulation depth in the water accumulation area can also be detected, and based on the water accumulation depth and the water accumulation area, the accumulated water volume can be calculated. This embodiment does not make specific limitations on this. In other embodiments, the preset visual large model can also be called to detect the current external vehicle environment image to detect the accumulated water volume in the current external vehicle environment image.

[0122] Increasing the air volume can be characterized as continuously increasing the air volume. In other embodiments, increasing the air volume can include increasing the air conditioner air volume to a preset target value, and the preset target value is also output by the preset air conditioner prediction model; blowing downward helps to dry parts such as trouser legs or shoes.

[0123] Exemplarily, in the case where the current weather is rainy, the current humidity inside the vehicle is greater than the preset humidity threshold, the current rainfall outside the vehicle is greater than the preset rainfall threshold, and / or it is detected that the accumulated water volume in the current external vehicle environment image is greater than the preset accumulated water threshold, if the current external temperature is greater than the preset high temperature threshold, it indicates that the current external temperature is relatively high, and there may be a situation where wet shoes get into the vehicle. Therefore, the output of the preset air conditioner prediction model includes: the target air conditioner switch parameter is on, the target air conditioner air volume includes increasing the air volume, the target blowing mode includes blowing downward, and / or the target blowing mode includes blowing downward and / or blowing cold air; blowing cold air is convenient for providing a more comfortable environment for the user and avoiding the situation that the vehicle occupants are hot and wet, resulting in a poor riding experience for the vehicle occupants. In other embodiments, the output of the preset air conditioner prediction model can also include: the target air conditioner temperature includes decreasing the temperature to create a more comfortable riding environment for the vehicle occupants.

[0124] When the current weather is rainy, the current humidity inside the vehicle is greater than the preset humidity threshold, the current rainfall outside the vehicle is greater than the preset rainfall threshold, and / or the detected accumulated water volume in the current external environment image of the vehicle is greater than the preset accumulated water threshold, if the current external temperature is less than the preset low temperature threshold, it indicates that the current climate temperature is relatively low and there may be a situation where wet shoes get into the vehicle. Therefore, the output of the preset air-conditioning prediction model includes: the target air-conditioning switch parameter is on, the target air volume of the air-conditioning includes increasing the air volume, and / or the target blowing mode includes blowing downward and / or blowing hot air; thus, it is convenient to blow hot air in the case of relatively low temperature to help the passengers inside the vehicle quickly dry the wet body parts and improve the comfort of the human body. In other embodiments, the output of the preset air-conditioning prediction model may further include: the target air-conditioning temperature includes increasing the temperature to create a more comfortable riding environment for the passengers inside the vehicle.

[0125] In this embodiment, in scenarios such as rainy days and a large amount of accumulated water outside the vehicle, services such as drying the passengers inside the vehicle who are wet by water are automatically provided, thereby facilitating the improvement of the comfort of the passengers inside the vehicle. And since this application can also determine the specific setting parameters of the vehicle-mounted air-conditioning in combination with the rainfall outside the vehicle, it can also improve the accuracy of predicting the parameters of the vehicle-mounted air-conditioning. Because the current weather is obtained through the weather API interface, and the data source of the weather API interface is the meteorological bureau, and the meteorological bureau generally estimates the overall weather conditions of the region, and there may be different weathers in the same region at the same time. For example, there may be cloudy and rainy conditions at the same time. Therefore, in this embodiment, combining the rainfall outside the vehicle can improve the accuracy of predicting the parameters of the vehicle-mounted air-conditioning. This embodiment can provide a more comfortable experience for users getting into the vehicle on rainy days.

[0126] In another feasible embodiment, the vehicle-mounted air-conditioning control method further includes step Y10: when it is detected that the air quality outside the vehicle in the current external environment image is less than the preset quality standard, the output of the preset air-conditioning prediction model includes that the target circulation mode is the internal circulation mode.

[0127] It should be noted that the current external environment image can be obtained in real time, so it is convenient to detect the air quality outside the vehicle in real time during the vehicle driving process. The air quality outside the vehicle can reflect the air conditions outside the vehicle during the vehicle driving process. When the air quality outside the vehicle is relatively poor, the circulation mode of the vehicle-mounted air-conditioning can be adjusted to the internal circulation mode to prevent the air inside the vehicle from being polluted by the outside, thereby facilitating the improvement of the comfort of the passengers inside the vehicle.

[0128] By means of a preset air conditioner prediction model, the outdoor air quality of the current outdoor environment image can be detected. For example, the type of facility pollution of the building facilities in the current outdoor environment image can be detected, the road dust concentration in the current outdoor environment image can be detected, and / or the cloud color in the current outdoor environment image can be detected, etc. The outdoor air quality can be determined whether it is less than the preset quality standard according to the detected facility pollution degree, road dust concentration and / or cloud color. For example, if the detected type of facility pollution is an air pollution-related building, it can be determined that the outdoor air quality is less than the preset quality standard; if the detected road dust concentration is greater than the preset concentration threshold, it can be determined that the outdoor air quality is less than the preset quality standard; if the detected cloud color is gray or yellow, it can also be determined that the outdoor air quality is less than the preset quality standard. The preset quality standard, preset concentration threshold, etc. can all be determined based on the actual situation, and this embodiment does not make specific settings for this.

[0129] The road dust concentration can reflect the environmental quality status of the road and the degree of air pollution, and the road dust concentration can also be reflected in the outdoor environment image, which can be specifically manifested as visible dust or haze in the image. Therefore, the preset air conditioner prediction model can identify the road dust concentration in the current outdoor environment image by detecting the distribution of dust in the outdoor environment image, or detecting the density of dust or the density of haze.

[0130] The type of facility pollution of building facilities can reflect the pollution sources or environmental impact characteristics existing in the driving environment. The type of facility pollution can include facilities such as chemical plants, garbage stations, civil engineering, high-rise buildings, etc. Civil engineering indicates that the road is under construction, so there will be air pollution.

[0131] The circulation mode of the vehicle-mounted air conditioner can include the internal circulation and the external circulation. The internal circulation means that only the air inside the vehicle is circulated; the external circulation is to introduce new air from outside the vehicle, filter and adjust it, and then send it into the vehicle to achieve the renewal and ventilation of the air inside the vehicle. The internal circulation is applicable to scenarios where the external air quality is poor or the temperature inside the vehicle needs to be adjusted quickly, while the external circulation is applicable to scenarios where the air quality inside the vehicle needs to be improved or the external environment is relatively fresh. Therefore, in this embodiment, when the outdoor air quality is less than the preset quality standard, the circulation mode of the vehicle-mounted air conditioner is controlled to be the internal circulation, which can prevent the air inside the vehicle from being polluted by the air outside the vehicle.

[0132] An air pollution-related building is characterized as a building that will reduce the air quality. The preset air conditioner prediction model can identify whether the type of facility pollution of the building facilities belongs to an air pollution-related building. For example, chemical plants, garbage stations, and civil engineering belong to air pollution-related buildings. The preset concentration threshold can be set based on the actual situation, and this embodiment does not make specific limitations for this.

[0133] Exemplarily, the current external environment image of the vehicle during driving is obtained in real time, and the facility pollution type of the building facilities in the current external environment image is detected through a preset air conditioner prediction model. If the facility pollution type is an air pollution type building, it can be determined that the external air quality is less than the preset quality standard; and / or, the road dust concentration in the current external environment image is detected through the preset air conditioner prediction model. If the road dust concentration is greater than the preset concentration threshold, it can be determined that the external air quality is less than the preset quality standard; and / or, the cloud color in the current external environment image is detected through the preset air conditioner prediction model. If the cloud color is gray or yellow, it can be determined that the external air quality is less than the preset quality standard. When the external air quality is less than the preset quality standard, the output of the preset air conditioner prediction model includes that the target circulation mode is the internal circulation mode.

[0134] In this embodiment, by detecting the external air quality of the vehicle in real time during driving, the circulation mode of the vehicle can be dynamically switched to the internal circulation mode when the external air quality is poor, thereby facilitating to ensure the comfort of the passengers in the vehicle. At the same time, there is no need for manual identification of whether to switch to the internal circulation mode, nor manual operation, thus improving the intelligence of the vehicle air conditioner.

[0135] In another feasible embodiment, the vehicle air conditioner control method further includes step Z10: when the absolute value of the difference between the target comfort temperature corresponding to the current clothing thickness of the vehicle occupants detected in the current internal environment image and the current internal temperature of the vehicle is greater than the preset temperature difference threshold, and the vehicle is in the preset rapid temperature adjustment mode, and the temperature adjustment duration from the current internal temperature to the target comfort temperature is less than the estimated driving duration of the vehicle, it is determined that the output of the preset air conditioner prediction model includes: the target temperature adjustment mode includes the preset rapid temperature adjustment mode, and the target air conditioner temperature includes adjusting the current internal temperature of the vehicle to the target comfort temperature;

[0136] Among them, the target comfort temperature is obtained by adjusting the preset standard body feeling comfort temperature to obtain an initial comfort temperature based on the current clothing thickness, and then adjusting the initial comfort temperature according to the temperature preference of the vehicle occupants obtained.

[0137] It should be noted that the current in-vehicle environment image is detected by an in-vehicle image detection device. The current in-vehicle environment image can reflect the clothing situation of the in-vehicle occupants, and the current clothing thickness can be determined by detecting the clothing of the in-vehicle occupants in the current in-vehicle environment image. The preset air-conditioning prediction model can identify the clothing of the in-vehicle occupants in the current in-vehicle environment image. When there are multiple passengers in the vehicle, the current in-vehicle environment image including all the passengers in the vehicle can be obtained. The current in-vehicle environment image including all the passengers can be obtained by splicing multiple seat images, and the seat images are also obtained by the in-vehicle image detection device. An image detection device can be set for each seat in the vehicle so as to obtain the seat image corresponding to each seat. When there is an occupant in the seat, the seat image corresponding to the seat includes the occupant. When there are multiple in-vehicle occupants in the vehicle, the current clothing thickness of the driver in the current in-vehicle environment image can be identified, the current clothing thickness of any in-vehicle passenger can also be identified, the current clothing thickness of the in-vehicle passengers whose age is greater than the preset high-age threshold can also be identified, and / or the current clothing thickness of the in-vehicle passengers whose age is less than the preset low-age threshold can be identified. The preset high-age threshold and the preset low-age threshold can be set based on the actual situation, and the present embodiment does not make specific limitations thereto.

[0138] When there are multiple passengers in the vehicle, the target comfort temperature can correspond to the current clothing thickness of any in-vehicle occupant. For example, it can be the current clothing thickness of the driver, the current clothing thickness of the in-vehicle passengers whose age is greater than the preset high-age threshold, or the current clothing thickness of the in-vehicle passengers whose age is less than the preset low-age threshold. Specifically, it can be determined based on the actual situation, and the present embodiment does not make specific limitations thereto. The preset air-conditioning prediction model can identify the current clothing thickness of the in-vehicle occupants by identifying the coats worn by the in-vehicle occupants.

[0139] The target comfort temperature represents the comfort temperature that the human body can perceive under the current clothing thickness. The initial comfort temperature can be obtained by adjusting the preset standard somatosensory comfort temperature based on the current clothing thickness, and the target comfort temperature can be obtained by adjusting the initial comfort temperature according to the temperature preference of the in-vehicle occupants obtained. Specifically, the difference between the preset standard somatosensory comfort temperature and the clothing heat value corresponding to the current clothing thickness can be calculated based on the clothing heat resistance value corresponding to the current clothing thickness to obtain the initial comfort temperature.

[0140] The preset standard somatosensory comfort temperature is the suitable temperature that the human body perceives when it is in a static state and not wearing clothes. For example, the preset standard somatosensory comfort temperature can be 26 °C, 27 °C or 28 °C, etc. Specifically, it can be selected based on the actual situation, and the present embodiment does not make specific limitations thereto.

[0141] The thermal resistance value of clothing is a physical quantity that characterizes the heat insulation performance of clothing. Different clothing has different corresponding thermal resistance values due to different thicknesses or materials. Each piece of clothing has its own corresponding thermal resistance value, and a preset air conditioner prediction model can determine the thermal resistance value of the clothing corresponding to the clothing by identifying the material and thickness of the clothing. For example, thick down jackets or woolen sweaters have higher thermal resistance values, while light and thin short-sleeved clothing has lower thermal resistance values. Clothing with different thermal resistance values will affect the human body's perception of the ambient temperature. Therefore, when setting the parameters of the vehicle-mounted air conditioner, the thermal resistance characteristics of the clothing need to be considered.

[0142] Temperature preferences can include being heat-preferred, cold-preferred, and standard. Being heat-preferred indicates that the driver may be afraid of cold and tends to set the temperature of the vehicle-mounted air conditioner to a higher value; being cold-preferred indicates that the driver may be afraid of heat and tends to set the temperature of the vehicle-mounted air conditioner to a lower value; standard indicates that the temperature requirement of the vehicle occupants is normal. When the temperature preference is standard, the target comfort temperature is the initial comfort temperature.

[0143] The temperature preference can be pre-set by the vehicle occupants, and options such as cold-preferred, heat-preferred, and standard can be provided for the vehicle occupants to choose. It can also be determined whether the temperature preference of the vehicle occupants is cold-preferred, heat-preferred, or standard based on the historical air conditioner usage data of the vehicle occupants.

[0144] Exemplarily, the difference between the preset standard somatosensory comfort temperature and the thermal resistance value of the clothing corresponding to the current clothing thickness is used as the initial comfort temperature. Different people have different corresponding temperature preferences. Therefore, the initial comfort temperature can be adjusted according to the temperature preference. When the temperature preference is cold-preferred, the initial comfort temperature can be reduced to obtain the target comfort temperature. For example, the initial comfort temperature can be reduced by 1 to 3 °C, etc., and it can be specifically set based on the actual situation. For example, the degree of reduction of the initial comfort temperature can be determined based on the cold-preferred degree of the vehicle occupants (such as the driver). The higher the cold-preferred degree, the more the degree of reduction, and the lower the cold-preferred degree, the less the degree of reduction. The cold-preferred degree can also be input by the vehicle occupants. In other embodiments, the preset default temperature can also be directly reduced. The preset default temperature can be custom-set. For example, it can be 1, or 2, or 3, etc. The difference between the initial comfort temperature and the preset default temperature is the target comfort temperature.

[0145] When the temperature preference is towards the hot side, the initial comfortable temperature can be increased to obtain the target comfortable temperature. The initial comfortable temperature can be increased by 1 to 3 °C, etc., and can be specifically set based on the actual situation. For example, the degree of increase in the initial comfortable temperature can be determined based on the degree of heat preference of the vehicle occupants (e.g., the driver). The higher the degree of heat preference, the more the degree of reduction; the lower the degree of heat preference, the less the degree of reduction. The degree of heat preference can also be input by the vehicle occupants. In other embodiments, the preset default temperature can also be directly reduced. The preset default temperature can be custom-set. For example, it can be 1, or 2, or 3, etc. The sum of the initial comfortable temperature and the preset default temperature is the target comfortable temperature. When the driver's temperature preference is standard, the initial comfortable temperature can be directly used as the target comfortable temperature. If the clothing thickness is less than or equal to the preset thickness threshold, there is no need to determine the target comfortable temperature, and the current state of the vehicle air conditioner can be directly maintained.

[0146] When the difference between the target comfortable temperature and the current vehicle interior temperature is greater than the preset temperature difference threshold, it indicates that the current vehicle interior temperature has not reached the target comfortable temperature, so it may be necessary to adjust the vehicle interior temperature. The preset temperature difference threshold can be set based on the actual situation, and this embodiment does not make specific limitations on it. The preset rapid temperature adjustment mode is one of the temperature adjustment modes of the vehicle air conditioner. The preset rapid temperature adjustment mode represents rapidly adjusting the vehicle interior temperature. The preset rapid temperature adjustment mode includes a rapid heating mode and a rapid cooling mode.

[0147] The estimated driving duration can be the estimated duration of the vehicle's current trip, and the estimated driving duration can be determined based on the vehicle's navigation route. The rapid heating mode represents rapid heating to quickly increase the vehicle interior temperature. The rapid cooling mode represents rapid cooling to quickly increase the vehicle interior temperature. The shortest temperature adjustment duration represents the duration required to adjust from the current vehicle interior temperature to the target comfortable temperature in the preset rapid temperature adjustment mode.

[0148] If the absolute value of the difference between the target comfortable temperature and the current vehicle interior temperature is greater than the preset temperature difference threshold, and when the vehicle is in the preset rapid temperature adjustment mode, the temperature adjustment duration from the current vehicle interior temperature to the target comfortable temperature is less than the vehicle's estimated driving duration, it indicates that the current vehicle interior temperature has not reached the comfortable temperature, and if the preset rapid temperature adjustment mode is used, the current vehicle interior temperature can be adjusted to the target comfortable temperature before the vehicle stops driving. Therefore, in this scenario, the preset rapid temperature adjustment mode can be used for temperature adjustment.

[0149] Exemplarily, when there are multiple passengers in the vehicle, the target comfort temperature corresponding to the current clothing thickness of any in-vehicle passenger can be determined. When the number of in-vehicle passengers in the vehicle is 1, the target comfort temperature corresponding to the current clothing thickness of this in-vehicle passenger can be directly determined. When the absolute value of the difference between the target comfort temperature and the current in-vehicle temperature is greater than the preset temperature difference threshold, and the vehicle is in the preset rapid temperature adjustment mode, and the temperature adjustment duration from the current in-vehicle temperature to the target comfort temperature is less than the expected driving duration of the vehicle, it is determined that the output of the preset air-conditioning prediction model includes: the target temperature adjustment mode includes the preset rapid temperature adjustment mode, and the target air-conditioning temperature includes adjusting the current in-vehicle temperature to the target comfort temperature. Thus, the current in-vehicle temperature can be adjusted to a comfortable temperature for in-vehicle passengers before the vehicle stops driving, thereby facilitating the improvement of the user's riding experience. Specifically, when the vehicle-mounted air conditioner is currently in the heating mode, the preset rapid temperature adjustment mode can be the rapid heating mode, and when the vehicle-mounted air conditioner is currently in the cooling mode, the preset rapid temperature adjustment mode can be the rapid cooling mode.

[0150] Further, this embodiment may further include steps a1 to a4:

[0151] Step a1, when the vehicle turns on the light clothing mode, determine the driver's clothing thickness of the driver from the current in-vehicle environment image through the preset air-conditioning prediction model, where the driver's clothing thickness includes the driver's outer coat thickness and the inner clothing thickness corresponding to the driver's outer coat thickness and the current climate temperature;

[0152] It should be noted that the light clothing mode means that the in-vehicle temperature can support users to wear thin clothes. The light clothing mode can be manually turned on by the user. After being turned on once, there is no need to set it again when the vehicle restarts driving later, and the light clothing mode will be automatically turned on. The preset air-conditioning prediction model can identify the type of outer coat worn by the human body in the clothing image, and thus can identify the thickness of the outer coat in the clothing image. The driver's clothing thickness includes the driver's outer coat thickness and the inner clothing thickness. Since the user wears an outer coat, it may be impossible to detect the inner clothing inside the outer coat, and thus it is difficult to identify the thickness of the inner clothing. Therefore, at this time, the corresponding inner clothing thickness can be determined based on the current climate temperature, because the current climate temperature will affect the thickness of the clothes worn by the human body. Therefore, after determining the driver's outer coat thickness, the corresponding inner clothing thickness can also be determined based on the current climate temperature. Different current climate temperatures and different driver's outer coat thicknesses may correspond to different inner clothing thicknesses.

[0153] Step a2, if the clothing thickness is greater than the preset thickness threshold, calculate the difference between the preset standard body feeling comfort temperature and the clothing thermal resistance value corresponding to the inner clothing thickness to obtain the first comfort temperature, and adjust the first comfort temperature according to the obtained driving temperature preference of the driver to obtain the second comfort temperature of the driver after the driver's clothing thickness is reduced;

[0154] It should be noted that the preset thickness threshold can be set based on the actual situation, and this embodiment does not make specific limitations on this. When the clothing thickness is greater than the preset thickness threshold, it indicates that the driver may be wearing thick clothes. Wearing thick clothes may affect the movement of the human body and may also be less comfortable. Therefore, in this embodiment, when it is detected that the clothing thickness is greater than the preset thickness threshold, the second comfortable temperature after the driver's clothing thickness is reduced is determined, so as to ensure that the driver will not feel cold after taking off the coat.

[0155] The driving temperature preference can include being heat-preferred, cold-preferred, and standard. Being heat-preferred means that the driver may be afraid of cold and tends to set the temperature of the vehicle air conditioner to a higher value; being cold-preferred means that the driver may be afraid of heat and tends to set the temperature of the vehicle air conditioner to a lower value; standard indicates that the driver's demand for temperature is normal. When the driving temperature preference is standard, the second comfortable temperature is the first comfortable temperature.

[0156] The driver's driving temperature preference can be pre-set by the driver. Options such as being cold-preferred, heat-preferred, and standard can be provided for the driver to choose. It can also be determined whether the driver's driving temperature preference is cold-preferred, heat-preferred, or standard by analyzing the driver's historical air conditioner usage data.

[0157] The second comfortable temperature is the temperature at which the driver will not feel cold after reducing the clothing. Reducing the driver's clothing thickness generally means that the driver takes off the coat. Therefore, the second comfortable temperature can be determined based on the driving temperature preference and the first comfortable temperature, and the first comfortable temperature can be determined based on the thickness of the inner clothing and the preset standard body-sensation comfortable temperature.

[0158] Exemplarily, if the clothing thickness is greater than the preset thickness threshold, the difference between the preset standard body-sensation comfortable temperature and the clothing thermal resistance value corresponding to the inner clothing thickness is used as the first comfortable temperature. Since different drivers have different corresponding driving temperature preferences, the first comfortable temperature can be adjusted according to the driving temperature preference to obtain the second comfortable temperature after the driver takes off the coat. When the driver's driving temperature preference is cold-preferred, the first comfortable temperature can be reduced to obtain the second comfortable temperature. For example, the first comfortable temperature can be reduced by 1 - 3 °C, etc., and it can be specifically set based on the actual situation. For example, the degree of reduction of the first comfortable temperature can be determined based on the driver's cold-preference degree. The higher the cold-preference degree, the greater the degree of reduction; the lower the cold-preference degree, the smaller the degree of reduction. The cold-preference degree can also be input by the driver. In other embodiments, the preset default temperature can also be directly reduced. The preset default temperature can be custom-set. For example, it can be 1, or 2, or 3, etc. The difference between the first comfortable temperature and the preset default temperature is the second comfortable temperature.

[0159] When the driver's driving temperature preference is on the hot side, the first comfort temperature can be increased to obtain the second comfort temperature. The first comfort temperature can be increased by 1 - 3 °C, etc., and can be specifically set based on the actual situation. For example, the degree of increase in the first comfort temperature can be determined based on the driver's hot preference. The higher the hot preference, the more the degree of increase; the lower the hot preference, the less the degree of increase. The hot preference can also be input by the driver. In other embodiments, the preset default temperature can also be directly reduced. The preset default temperature can be custom-set. For example, it can be 1, or 2, or 3, etc. The sum of the first comfort temperature and the preset default temperature is the second comfort temperature. When the driver's driving temperature preference is standard, the first comfort temperature can be directly used as the second comfort temperature. If the clothing thickness is less than or equal to the preset thickness threshold, there is no need to determine the second comfort temperature, and the current state of the vehicle air conditioner can be directly maintained.

[0160] Step a3: Obtain the current vehicle interior temperature and the estimated driving duration of the vehicle, and estimate the shortest heating duration for the vehicle air conditioner to heat up from the current vehicle interior temperature to the second comfort temperature in the rapid heating mode.

[0161] It should be noted that the estimated driving duration can be the estimated duration of the vehicle's current trip, and the estimated driving duration can be determined based on the vehicle's navigation route. The shortest heating duration represents the duration required to heat up from the current vehicle interior temperature to the second comfort temperature in the rapid heating mode.

[0162] Step a4: If the shortest heating duration is less than the estimated driving duration, control the vehicle air conditioner to operate in the rapid heating mode and output a prompt to remind the driver to reduce the clothing thickness after the shortest heating duration.

[0163] It should be noted that the output prompt can be a voice prompt. For example, when the second comfort temperature is 28 °C and the shortest heating duration is 8 minutes, the voice prompt can be: It is detected that you are wearing a thick coat, and the rapid heating to 28 °C has been started. You can adjust to the light clothing mode after 8 minutes. Thus, the driver can be reminded that taking off the coat after 8 minutes will not make them feel cold. Additionally, in this embodiment, it can be that after the vehicle air conditioner responds to the target air conditioner parameters and operates, it then responds to the corresponding control mode of the vehicle air conditioner in the light clothing mode.

[0164] Exemplarily, if the shortest heating-up duration is less than the expected driving duration, it indicates that the vehicle can rapidly heat up to the second comfortable temperature before the end of the vehicle's driving. At this time, the vehicle-mounted air conditioner can be controlled to operate in the rapid heating-up mode, and a prompt can be output to improve the driver's comfort. After the temperature inside the vehicle reaches the second comfortable temperature, the rapid heating-up mode can be turned off, and the vehicle-mounted air conditioner can be controlled to continue operating according to the currently set temperature. If the shortest heating-up duration is greater than or equal to the expected driving duration, it indicates that the vehicle cannot rapidly heat up to the second comfortable temperature before the end of the vehicle's driving. At this time, rapid heating-up can be not carried out, and the current operating state of the vehicle-mounted air conditioner can be maintained. Meanwhile, a prompt can also be output to remind the driver that taking off the coat currently may make the driver feel cold. Thereby, energy consumption can be saved.

[0165] In other embodiments, after the driver takes off the coat, the type of the inner garment actually worn by the driver can be recognized. If the clothing thermal resistance value corresponding to the actual inner garment is greater than the clothing thermal resistance value of the expected inner garment, the temperature inside the vehicle can be reduced. For example, the set temperature of the vehicle-mounted air conditioner can be reduced. If the clothing thermal resistance value corresponding to the actual inner garment is less than the clothing thermal resistance value of the expected inner garment, the temperature inside the vehicle can be appropriately increased. For example, the set temperature of the vehicle-mounted air conditioner can be increased. The clothing thermal resistance value of the expected inner garment can be the clothing thermal resistance value of the inner garment thickness determined by the preset visual large model based on the driver's coat thickness and the current climate temperature when the driver is wearing the coat.

[0166] In this embodiment, by determining the second comfortable temperature after the driver takes off the coat, it is further convenient to raise the temperature inside the vehicle to the second comfortable temperature, so that the driver can drive while wearing light clothing, thereby improving the driving comfort.

[0167] The embodiment of the present application further provides a vehicle-mounted air conditioner control device. Please refer to Figure 5 , the device includes:

[0168] An acquisition module 10, configured to acquire the current environment, the current vehicle operating state, and the vehicle user characteristics in response to a vehicle-mounted air conditioner control instruction;

[0169] A determination module 20, configured to determine the target air conditioner parameters under the current environment, the current vehicle operating state, and the vehicle user characteristics through a preset air conditioner prediction model, where the preset air conditioner prediction model is obtained by training an initial air conditioner prediction model based on a credible training sample set;

[0170] An operation module 30, configured to control the operation of the vehicle-mounted air conditioner in the vehicle according to the target air conditioner parameters. The present application provides a vehicle including a vehicle body and a controller. The controller is arranged in the vehicle body, and the controller is used to execute to implement the vehicle-mounted air conditioner control method.

[0171] The vehicle provided by this application adopts the vehicle-mounted air conditioner control method in the above-mentioned embodiment, which can solve the technical problem of low intelligence of the vehicle-mounted air conditioner. Compared with the prior art, the beneficial effects of the vehicle provided by this application are the same as those of the vehicle-mounted air conditioner control method provided by the above-mentioned embodiment, and other technical features in this vehicle are the same as those disclosed in the method of the previous embodiment, which will not be elaborated here.

[0172] It should be understood that each part disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in a suitable manner in any one or more embodiments or examples.

[0173] The above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed in this application can easily think of changes or substitutions, which should all be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

[0174] This embodiment provides a computer-readable storage medium with computer-readable program instructions stored thereon. The computer-readable program instructions are used to execute the vehicle air conditioner control method in the first embodiment above. The computer-readable storage medium provided by the embodiments of the present application can be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor devices, apparatuses, or components, or any combination of the above. More specific examples of the computer-readable storage medium may include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable EPROM (Electrical Programmable Read Only Memory), or flash memory, optical fibers, portable compact disc CD-ROM (compact disc read-only memory), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this embodiment, the computer-readable storage medium can be any tangible medium that contains or stores a program, and this program can be used by or combined with an instruction execution device, apparatus, or component. The program code contained on the computer-readable storage medium can be transmitted by any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above. The above computer-readable storage medium can be included in an electronic device; it can also exist separately without being assembled into the electronic device. The above computer-readable storage medium carries one or more programs. When the above one or more programs are executed by an electronic device, the electronic device is caused to: obtain the current environment, the current vehicle operating state, and the vehicle user characteristics; determine the target air conditioner parameters under the current environment, the current vehicle operating state, and the vehicle user characteristics through a preset air conditioner prediction model, where the preset air conditioner prediction model is obtained by training an initial air conditioner prediction model based on a credible training sample set jointly selected from multiple preset large models; and control the operation of the vehicle air conditioner in the vehicle according to the target air conditioner parameters.

[0175] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages or combinations thereof. The above-mentioned programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a LAN (local area network) or a WAN (Wide Area Network), or may be connected to an external computer (e.g., by connecting through the Internet using an Internet service provider).

[0176] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of devices, methods, and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a portion of code that contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, may be implemented by a dedicated hardware-based device that performs the specified functions or operations, or may be implemented by a combination of dedicated hardware and computer instructions.

[0177] The modules described in the embodiments of the present disclosure may be implemented in software or in hardware. Among them, the name of the module does not constitute a limitation on the unit itself in some cases. The computer-readable storage medium provided in the present application stores computer-readable program instructions for performing the above-mentioned vehicle air conditioner control method, aiming to solve the technical problem of low intelligence of vehicle air conditioners. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in the embodiments of the present application are the same as those of the vehicle air conditioner control method provided in the above embodiments, and will not be elaborated herein.

[0178] The present application also provides a computer program product, including a computer program which, when executed by a processor, implements the steps of the vehicle air conditioner control method as described above. The computer program product provided by the present application aims to solve the technical problem of the low intelligence level of vehicle air conditioners. Compared with the prior art, the beneficial effects of the computer program product provided by the embodiments of the present application are the same as those of the vehicle air conditioner control method provided by the above embodiments, and will not be elaborated herein. The above are only the preferred embodiments of the present application, and do not limit the patent scope of the present application accordingly. Any equivalent structural or equivalent process transformation made by using the content of the specification and drawings of the present application, or directly or indirectly applied to other related technical fields, shall be equally included in the patent scope of the present application.

Claims

1. A vehicle air conditioner control method, characterized in that, The method described above includes: In response to an in-vehicle air conditioner control instruction, obtain the current environment, the current vehicle operating state, and vehicle user characteristics; Determine target air conditioner parameters under the current environment, the current vehicle operating state, and vehicle user characteristics through a preset air conditioner prediction model, where the preset air conditioner prediction model is obtained by training an initial air conditioner prediction model based on a credible training sample set; Control the operation of the in-vehicle air conditioner in the vehicle according to the target air conditioner parameters; Among them, the steps of determining the preset air conditioner prediction model include: obtaining the historical air conditioner usage data of a target user group, where the target user group includes the vehicle users corresponding to the in-vehicle air conditioner and the same-type users of the vehicle users, the historical air conditioner usage data includes multiple sub-usage data, and the sub-usage data includes at least one of the following: historical vehicle user characteristics, historical air conditioner feedback adjustment data, historical air conditioner setting parameters, historical environment, and historical vehicle operating state; Screen out a credible training sample set from the historical air conditioner usage data according to multiple preset large models; Extract scene feature parameters from the credible training sample set to obtain multiple scene parameter training features, where each scene parameter training feature includes a scene training feature and an air conditioner parameter setting feature associated with the scene training feature, and the scene training feature includes at least one of the following: training vehicle user characteristics, training air conditioner feedback adjustment data, training environment, and training vehicle operating state, and the air conditioner parameter setting feature includes at least one of the following: air conditioner switch setting parameter, seat heating setting parameter, blowing setting parameter, air conditioner temperature setting parameter, air conditioner air volume setting parameter, temperature adjustment mode setting parameter, and circulation mode setting parameter; Train the initial air conditioner prediction model according to the multiple scene parameter training features to obtain a trained teacher model; Obtain a student model with a parameter order of magnitude smaller than that of the teacher model, update the student model through the teacher model, and use the updated student model as the preset air conditioner prediction model.

2. The vehicle air conditioner control method according to claim 1, wherein The step of screening out a credible training sample set from the historical air conditioner usage data according to multiple preset large models includes: For each preset large model, input the historical air conditioner usage data into the preset large model, and remove the abnormal sub-usage data in the historical air conditioner usage data through the preset large model to obtain a normal usage data set screened out by the preset large model from the historical air conditioner usage data; Extract the target normal data jointly included in all the normal usage data sets to obtain a credible training sample set.

3. The vehicle air conditioner control method according to claim 1, wherein, The target air conditioner parameters include at least one of a target air conditioner switch parameter, a target seat heating parameter, a target blowing mode, a target air conditioner temperature, a target air conditioner air volume, a target circulation mode, and a target temperature adjustment mode; The step of determining target air conditioner parameters under the current environment, the current vehicle operating state, and vehicle user characteristics through a preset air conditioner prediction model includes: Input the current environment, the current vehicle operating state, and the vehicle user characteristics into a preset air conditioner prediction model, and determine the target air conditioner switch parameter, the target seat heating parameter, the target blowing mode, the target air conditioner temperature, the target air conditioner air volume, the target circulation mode, and / or the target temperature adjustment mode under the current environment, the current vehicle operating state, and the vehicle user characteristics through the preset air conditioner prediction model; Among them, the current environment includes at least one of the current weather, the current outside vehicle temperature, the current inside vehicle temperature, the current inside vehicle humidity, the current outside vehicle rainfall, the current outside vehicle environment image, and the current inside vehicle environment image. The current vehicle operating state includes the current vehicle battery power, and the vehicle user characteristics include at least one of the age, gender, current usage month, current vehicle latitude, and user behavior of the vehicle user.

4. The vehicle air conditioner control method according to claim 3, wherein, The method further includes: Under a preset driving scenario, if the current outside vehicle temperature is greater than a preset high temperature threshold, the output of the preset air conditioner prediction model includes: the target air conditioner switch parameter is on, the target air conditioner air volume includes increasing the air volume, and / or the target blowing mode includes blowing downward and / or blowing cold air; Under a preset driving scenario, if the current outside vehicle temperature is less than a preset low temperature threshold, the output of the preset air conditioner prediction model includes: the target air conditioner switch parameter is on, the target air conditioner air volume includes increasing the air volume, and / or the target blowing mode includes blowing downward and / or blowing hot air; Among them, the preset driving scenario includes: the current weather is rainy, the current inside vehicle humidity is greater than a preset humidity threshold, the current outside vehicle rainfall is greater than a preset rainfall threshold, and / or it is detected that the accumulated water volume in the current outside vehicle environment image is greater than a preset accumulated water threshold.

5. The vehicle air conditioner control method according to claim 3, wherein, The method includes: In the case where the outside vehicle air quality in the current outside vehicle environment image is detected to be less than a preset quality standard, the output of the preset air conditioner prediction model includes that the target circulation mode is the internal circulation mode.

6. The vehicle air conditioner control method according to claim 3, characterized in that, The method further includes: When the absolute value of the difference between the target comfort temperature corresponding to the current clothing thickness of the vehicle occupants detected in the current inside vehicle environment image and the current inside vehicle temperature is greater than a preset temperature difference threshold, and the vehicle is in a preset rapid temperature adjustment mode, and the temperature adjustment duration from the current inside vehicle temperature to the target comfort temperature is less than the expected driving duration of the vehicle, it is determined that the output of the preset air conditioner prediction model includes: the target temperature adjustment mode includes the preset rapid temperature adjustment mode, and the target air conditioner temperature includes adjusting the current inside vehicle temperature to the target comfort temperature; Among them, the target comfort temperature is obtained by adjusting the preset standard body sensation comfort temperature to obtain an initial comfort temperature based on the current clothing thickness, and then adjusting the initial comfort temperature according to the obtained temperature preference of the vehicle occupants.

7. A vehicle, characterized in that, The vehicle includes a vehicle body and a controller. The controller is provided in the vehicle body, and the controller is used to execute the steps of implementing the on-vehicle air conditioner control method according to any one of claims 1 to 6.

8. A readable storage medium, characterized in that, The readable storage medium is a computer-readable storage medium, and a program for implementing the vehicle air conditioner control method is stored on the computer-readable storage medium. The program for implementing the vehicle air conditioner control method is executed by a processor to implement the steps of the vehicle air conditioner control method according to any one of claims 1 to 6.

Images