Vehicle air conditioner control method and device and vehicle

Abstract

The invention discloses an air conditioner temperature control method and device and a vehicle, and the method comprises the steps that an environment time sequence data set of a target vehicle is obtained, and the environment time sequence data set comprises vehicle interior and exterior environment data, vehicle state data and air conditioner parameters; the in-vehicle and out-vehicle environment data at least comprises in-vehicle passenger distribution data; the vehicle state data comprises vehicle speed data, electric quantity data and driving mode data; determining a temperature control demand of the target vehicle in a preset future time period based on the environment time sequence data set; based on the temperature control requirement, an air conditioner parameter adjusting scheme of the target vehicle in a preset future time period is determined; when the air conditioner temperature adjusting scheme in the vehicle is generated, factors of passenger distribution in the vehicle and the vehicle driving state are considered, meanwhile, the generated air conditioner temperature adjusting scheme comprises the wind speed parameter and the wind direction parameter, and the air conditioner energy consumption can be reduced while the comfort degree requirement is met.

Description

Technical Field

[0001] This application relates to the field of automotive technologies, and particularly to a vehicle air-conditioning control method, apparatus, and vehicle. Background Art

[0002] In existing automotive temperature control systems, the adjustment of the air conditioner usually relies on fixed rules or a simple Proportional-Integral-Derivative Control (PID) algorithm, and it is difficult to adaptively adjust energy consumption according to dynamic environmental changes, which may lead to the following problems:

[0003] Since existing air-conditioning systems usually only consider the impact of external environmental factors (temperature, humidity, light) of the vehicle on air-conditioning control, but ignore the occupant distribution inside the vehicle and personalized occupant needs (thermal sensitivity), this may result in unnecessary energy waste and a poor temperature control experience for occupants. In addition, traditional control methods are difficult to adapt to complex dynamic scenarios (such as high-speed driving, frequent starts and stops), and cannot quickly adjust to the optimal settings.

[0004] Therefore, how to quickly generate an adaptive adjustment strategy for air-conditioning system parameters while improving comfort and reducing energy consumption has become a technical problem that needs to be urgently solved by those skilled in the art. Summary of the Invention

[0005] In view of the above problems, the present disclosure provides an air-conditioning temperature control method, apparatus, computer-readable storage medium, and vehicle that overcome or at least partially solve the above problems. The technical solutions are as follows:

[0006] A vehicle air-conditioning control method includes: obtaining an environmental time-series data set of a target vehicle, where the environmental time-series data set includes vehicle interior and exterior environmental data, vehicle state data, and air-conditioning parameters; the vehicle interior and exterior environmental data at least includes vehicle interior occupant distribution data; the vehicle state data includes vehicle speed data, power data, and driving mode data; based on the environmental time-series data set, determining the temperature control requirements of the target vehicle within a preset future time period; based on the temperature control requirements, determining an air-conditioning parameter adjustment plan for the target vehicle within a preset future time period; the air-conditioning parameters at least include a target temperature parameter, a wind direction parameter, and a wind speed parameter.

[0007] When generating an air-conditioning temperature adjustment plan inside the vehicle, factors such as the vehicle interior occupant distribution and the vehicle driving state are considered, and the generated air-conditioning temperature adjustment plan includes a wind speed parameter and a wind direction parameter, which can reduce air-conditioning energy consumption while meeting the comfort requirements.

[0008] Optionally, determining the temperature control requirement of the target vehicle within a preset future time period based on the environmental time series data set specifically includes: inputting the environmental time series data set into a pre-trained hybrid temperature control requirement prediction model to determine the temperature control requirement of the target vehicle within a preset future time period; the temperature control requirement prediction model includes a convolutional neural network layer, a long short-term memory network layer, and an improved temperature control requirement prediction layer; the convolutional neural network layer is used to extract the short-term local change features of the environmental time series data set; the long short-term memory network layer is used to capture the long-term time dependence relationship of the environmental time series data set; the temperature control requirement prediction layer is used to perform global feature modeling; in the temperature control requirement prediction layer, the multi-head attention includes at least environmental attention, occupant attention, and energy consumption attention.

[0009] By predicting the temperature control requirement of the target vehicle in the future time period through a pre-trained temperature control requirement prediction model, the air-conditioning parameters of the vehicle can be adjusted in advance to maintain a constant temperature inside the target vehicle. At the same time, by using the convolutional neural network layer and the long short-term memory network layer to extract the short-term local change features and the long-term time dependence relationship of the environmental time series data set respectively, the accuracy of the temperature control requirement prediction can be improved. By adding environmental attention, occupant attention, and energy consumption attention, different influencing factors can be adaptively focused on, improving the prediction accuracy.

[0010] Optionally, determining the air-conditioning parameter adjustment scheme of the target vehicle within a preset future time period based on the temperature control requirement specifically includes: determining the static parameter adjustment scheme corresponding to the target vehicle within the preset future time period based on the temperature control requirement of the target vehicle within the preset future time period and the environmental time series data set; determining the dynamic parameter adjustment scheme corresponding to the environmental time series data set within the preset future time period based on the temperature control requirement of the target vehicle within the preset future time period and the corresponding static parameter adjustment scheme; using the static parameter adjustment scheme and / or the dynamic parameter adjustment scheme as the air-conditioning parameter adjustment scheme of the target vehicle within the preset future time period.

[0011] When the environmental time series data of the target vehicle is relatively stable, the air-conditioning parameters of the target vehicle can be adjusted through the static parameter adjustment scheme. When the environmental time series data changes, the corresponding dynamic parameter adjustment scheme can be generated through the static parameter adjustment scheme to adapt to the change of the environmental time series data, thereby always ensuring the comfort inside the target vehicle.

[0012] Optionally, determining the static parameter adjustment plan corresponding to the target vehicle within the preset future time period based on the temperature control requirements of the target vehicle within the preset future time period and the environmental time series data set specifically includes: constructing a Bayesian optimization model by using Gaussian process regression and the upper confidence bound strategy; inputting the temperature control requirements of the target vehicle within the preset future time period and the environmental time series data set into the Bayesian optimization model; determining multiple evaluation dimensions of the Bayesian optimization model, and the types of the evaluation dimensions at least include a comfort dimension and an energy consumption dimension; and obtaining the static parameter adjustment plan output by the Bayesian optimization model.

[0013] By using the Bayesian optimization model and the predicted temperature control requirements to determine the static parameter adjustment plan corresponding to the temperature control requirements, compared with traditional grid search or random search, using Gaussian process regression can efficiently find the optimal solution with less data.

[0014] Optionally, based on the temperature control requirements corresponding to the target vehicle within the preset future time period and the corresponding static parameter adjustment plan to determine the dynamic parameter adjustment plan corresponding to the environmental time series data set within the preset future time period, specifically includes: obtaining the environmental change value of the environmental time series data set; using the temperature control requirements corresponding to the target vehicle within the preset future time period as the initial state space and the corresponding static parameter adjustment plan as the initial action space to construct a deep deterministic policy gradient algorithm model; and based on the deep deterministic policy gradient algorithm model, determining the dynamic parameter adjustment plan corresponding to the environmental change value.

[0015] Through the static parameter adjustment plan and the corresponding temperature control requirements, using the deep deterministic policy gradient algorithm can dynamically optimize the air-conditioning settings under different driving environments and different occupant states to achieve an adaptive temperature control strategy. At the same time, by using the experience replay strategy, the training stability can be improved, and by using the soft update strategy, the training can be made smoother.

[0016] Optionally, after determining the air-conditioning parameter adjustment plan of the target vehicle within the preset future time period based on the temperature control requirements, the method further includes: obtaining the air-conditioning parameter adjustment instruction of the occupant based on the air-conditioning parameter adjustment plan and the vehicle state data set corresponding to the air-conditioning parameter adjustment instruction; the vehicle state data set includes at least one of vehicle speed data, in-vehicle occupant distribution data, driving habit data, and destination distance data; and based on the vehicle state data set and the air-conditioning parameter adjustment instruction, updating the model parameters in the temperature control requirement prediction model, the Bayesian optimization model, and the deep deterministic policy gradient algorithm model.

[0017] By obtaining the adjustment instructions of the occupant and the corresponding vehicle state data set, the model parameters of the above model can be adjusted to make the in-vehicle model of the target vehicle more in line with the personalized needs of the occupant.

[0018] Optionally, based on the vehicle state data set and the air-conditioning parameter adjustment instruction, updating the model parameters in the temperature control demand prediction model, the Bayesian optimization model, and the deep deterministic policy gradient algorithm model specifically includes: determining the user profile of the owner corresponding to the target vehicle and the corresponding vehicle type based on the vehicle state data set; constructing a parameter training sample set based on the current environmental time series data set and the air-conditioning parameter adjustment instruction; determining target parameter training samples related to the user profile and the corresponding vehicle type in the parameter training sample set; and updating the model parameters in the temperature control demand prediction model, the Bayesian optimization model, and the deep deterministic policy gradient algorithm model of the owner and the target vehicle based on the target parameter training samples.

[0019] Optionally, the method further includes: deploying the temperature control demand prediction model, the Bayesian optimization model, and the deep deterministic policy gradient algorithm model in the target vehicle; performing real-time data stream processing through an in-memory database and a distributed message queue; receiving the optimized model parameters from the cloud at preset time intervals, and replacing the model parameters in the temperature control demand prediction model, the Bayesian optimization model, and the deep deterministic policy gradient algorithm model with the optimized model parameters.

[0020] This application also discloses an air-conditioning temperature control device, which includes: a data acquisition module that acquires the environmental time series data set of the target vehicle, and the environmental time series data set includes in-vehicle and out-of-vehicle environmental data, vehicle state data, and air-conditioning parameters; the in-vehicle and out-of-vehicle environmental data at least includes in-vehicle occupant distribution data; the vehicle state data includes vehicle speed data, power data, and driving mode data; a demand prediction module that determines the temperature control demand of the target vehicle in a preset future time period based on the environmental time series data set; a parameter adjustment module that determines the air-conditioning parameter adjustment plan of the target vehicle in a preset future time period based on the temperature control demand; and the air-conditioning parameters at least include target temperature parameters, air direction parameters, and wind speed parameters.

[0021] The present application also discloses a vehicle, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the at least one processor is enabled to: obtain an environmental time-series data set of a target vehicle, the environmental time-series data set including in-vehicle and out-of-vehicle environmental data, vehicle state data, and air-conditioning parameters; the in-vehicle and out-of-vehicle environmental data at least including in-vehicle occupant distribution data; the vehicle state data including vehicle speed data, power data, and driving mode data; determine a temperature control requirement of the target vehicle within a preset future time period based on the environmental time-series data set; determine an air-conditioning parameter adjustment scheme of the target vehicle within the preset future time period based on the temperature control requirement; the air-conditioning parameters at least including a target temperature parameter, a wind direction parameter, and a wind speed parameter.

[0022] The method proposed by the present application can bring the following beneficial effects: when generating an air-conditioning temperature adjustment scheme inside the vehicle, factors such as in-vehicle occupant distribution and vehicle driving state are considered, and at the same time, the generated air-conditioning temperature adjustment scheme includes a wind speed parameter and a wind direction parameter, which can reduce air-conditioning energy consumption while meeting the comfort requirements.

[0023] The above description is only an overview of the technical solution of the present disclosure. In order to be able to understand the technical means of the present disclosure more clearly, it can be implemented according to the content of the description. And in order to make the above and other purposes, features, and advantages of the present disclosure more obvious and understandable, the specific embodiments of the present disclosure are hereinafter specifically exemplified. Brief Description of the Drawings

[0024] By reading the detailed description of the preferred embodiments below, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present disclosure. And throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0025] Figure 1 A flowchart showing the process of a vehicle air-conditioning control method provided by an embodiment of the present application is shown;

[0026] Figure 2 A structural diagram showing a vehicle air-conditioning control device provided by an embodiment of the present application is shown;

[0027] Figure 3 A structural diagram showing a vehicle provided by an embodiment of the present application is shown. Detailed Description of the Embodiments

[0028] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully conveyed to those skilled in the art. The technical solutions provided by the embodiments of the present application will be described in detail below in conjunction with the accompanying drawings.

[0029] Existing air conditioning systems usually only consider the influence of external vehicle environment factors (temperature, humidity, light) on air conditioning control, but will ignore the occupant distribution inside the vehicle and the personalized occupant needs (thermal sensitivity), which will lead to unnecessary energy waste and poor occupant temperature control experience. In addition, traditional control methods are difficult to adapt to complex dynamic scenarios (such as high-speed driving, frequent starts and stops) and cannot quickly adjust the optimal settings.

[0030] To solve the above technical problems, Figure 1 It is a schematic flowchart of a vehicle air conditioning control method provided by one or more embodiments of this specification. This method can be applied to different types of vehicles, and this process can be executed by computing devices in the corresponding field (such as a cloud server, or a mobile terminal in the vehicle, etc.). Some input parameters or intermediate results in the process allow manual intervention and adjustment to help improve accuracy.

[0031] The implementation of the analysis method involved in the embodiments of the present application can be a terminal device or a server, and the present application does not make special restrictions on this. For the convenience of understanding and description, the following embodiments will be described in detail taking the in-vehicle computer as an example. It should be noted that the server can be a single device or a system composed of multiple devices, that is, a distributed server, and the present application does not make specific limitations on this.

[0032] As Figure 1 shown, the embodiments of the present application provide a vehicle air conditioning control method, including:

[0033] S101: Obtain the environmental time series data set of the target vehicle.

[0034] First, before adjusting the air conditioning parameters of the vehicle, it is necessary to determine the current environmental time series data set of the vehicle. Here, the environmental time series data set refers to the set of various types of environmental time series data inside and outside the vehicle. Among them, the types of environmental time series data include vehicle interior and exterior environmental data, vehicle state data, and air conditioning parameters. Among them, the vehicle interior and exterior environmental data at least includes the in-vehicle occupant distribution data, and the vehicle state data includes vehicle speed data, power data, and driving mode data.

[0035] Among them, the in-vehicle environmental data in the vehicle and external environmental data can include in-vehicle temperature data and in-vehicle humidity data for monitoring the in-vehicle environmental state, which can be collected by in-vehicle temperature and humidity sensors. The external environmental data can include external temperature data and external humidity data for monitoring external environmental changes, which can be collected by external temperature and humidity sensors. The vehicle and external environmental data can also include sunlight intensity data collected by an external light sensor, which is used to estimate the impact of solar radiation on the in-vehicle temperature. The in-vehicle occupant distribution data can be collected by an in-vehicle infrared sensor and is used to identify the positions of the people in the vehicle. The vehicle speed data and driving mode data are used to judge the impact of the driving state on the in-vehicle heat load, and the power data and air-conditioning parameters can be directly obtained by the vehicle computer, which are used to evaluate the impact of the air conditioner on the endurance and to feedback the air-conditioning setting parameters respectively.

[0036] It should be noted that, different from the prior art, in this technical solution, when considering factors, the vehicle speed, driving state, battery information, and in-vehicle occupant distribution information of the vehicle are added. Therefore, when generating the adjustment scheme of the air conditioner, the comfort requirements of the occupants and the energy consumption requirements of the vehicle can be taken into account simultaneously.

[0037] After obtaining the environmental time series dataset, it is necessary to preprocess the environmental time series dataset to facilitate subsequent data processing. At this time, it is first necessary to perform time synchronization on the environmental time series dataset according to the sampling frequency of the sensor. Specifically, time alignment can be performed through linear interpolation. At the same time, if there are outliers in the environmental time series dataset, the outliers need to be detected and removed. For example, the Z-score method can be used to detect and remove abnormal data. After removing the abnormal data, before inputting the environmental time series dataset into the model, the environmental time series dataset can be normalized. For example, the Min-Max normalization is used to map the data into the interval [0, 1], thereby improving the stability of model training.

[0038] S102: Based on the environmental time series dataset, determine the temperature control requirements of the target vehicle within a preset future time period.

[0039] After obtaining the environmental time series dataset of the target vehicle, the temperature control requirements of the target vehicle within a future time period can be predicted according to the environmental time series dataset. The temperature control requirements here refer to the change trends of the temperature and humidity of the target vehicle within the future time period.

[0040] In one embodiment, when predicting the temperature control demand of a target vehicle, it can be predicted by using a pre-trained temperature control demand prediction model. Here, the Transformer architecture is used as an example for illustration. Existing Transformers are mainly used for processing time-series data. However, the temperature control demand involves multiple data sources (such as the temperature and humidity inside and outside the vehicle, the occupant status, the light intensity, the vehicle driving status, etc.). Therefore, multi-modal data fusion needs to be carried out first. When performing multi-modal data fusion, it is necessary to determine the input format of sensor data. The input data X consists of multiple sensor data streams: X = [x1, x2, …… x n ,]. Among them, each x n contains multiple environmental features. For example, x n =(the temperature inside the vehicle, the temperature outside the vehicle, the humidity inside the vehicle, the humidity outside the vehicle, the sunlight intensity, the vehicle driving status, the distribution of vehicle occupants, the current air-conditioning parameters).

[0041] For time-series data such as the temperature inside the vehicle, the humidity inside the vehicle, the temperature outside the vehicle, the humidity outside the vehicle, the vehicle speed, and the sunlight intensity, they can be directly input into the Transformer architecture. If they are spatial data such as the occupant distribution and the seat temperature, feature extraction is first performed through a graph neural network, and then the extracted features are input into the Transformer architecture. The current air-conditioning parameters can be attached to the Transformer input as external variables.

[0042] To better capture the short-term and long-term changes in temperature control demand, we introduce a local time window to present the short-term changes in the environmental time-series dataset of the target vehicle. For example, the local time window can be set from 15 seconds to 10 minutes, and long-term trend prediction at the hourly level is adopted.

[0043] Specifically, when using short-term local change modeling, the local time window mechanism is used to extract short-term features within the most recent N time steps to capture the changes in occupant behavior and environmental fluctuations (such as sudden temperature changes when sunlight enters the vehicle) within a short period of time.

[0044] Long-term trends are extracted through positional encoding and sliding time window aggregation. For example: If the temperature has been rising continuously in the past 30 minutes, the system can predict that it may continue to rise in the next 10 minutes, so as to adjust the temperature control strategy in advance. The final time-series features are jointly determined by short-term local changes and long-term trends. For example, they can be determined by weighting, where the weights are controlled by adaptive attention weights.

[0045] In one embodiment, the Transformer architecture can also be optimized for multi-head attention. Based on the original Transformer multi-head attention, environmental attention focusing on the temperature and humidity outside the vehicle and light is added, as well as occupant attention focusing on the occupant distribution and seat temperature, and energy consumption attention focusing on the battery power and air-conditioning power consumption. This can adaptively focus on different influencing factors and improve the prediction accuracy.

[0046] In one embodiment, when making a prediction, the environmental time-series dataset can be input into the hybrid Transformer architecture to determine the temperature control requirements of the target vehicle within a preset future time period. Here, the hybrid Transformer architecture includes a convolutional neural network layer, a long short-term memory network layer, and an improved Transformer layer. Among them, the convolutional neural network layer is used to extract local time features from the environmental time-series dataset, the long short-term memory network layer is used to capture the long-term time dependencies of the environmental time-series dataset, and the improved Transformer layer is used to perform global feature modeling. The above multi-head attention optimization can be placed in the improved Transformer layer, and environmental attention, occupant attention, and energy consumption attention are newly added to the multi-head attention.

[0047] Taking the following scenario as an example, the scenario description is as follows: the outside temperature is 35°C, the inside temperature is 28°C, the light intensity is 800 W / m2, there is 1 person in the front row and 2 people in the back row of the occupants, the current air-conditioning setting is 24°C, and the wind speed is level 3. At this time, the temperature control requirement prediction task is to predict the temperature control requirements in the next 10 minutes. At this time, the sensor collects data, uses the convolutional neural network layer to extract the temperature mutation characteristics, extracts the long-term trend through the long short-term memory network layer, predicts the future temperature control requirements through the improved Transformer layer, and outputs the predicted future temperature control requirements. For example, it is predicted that the inside temperature of the vehicle will rise to 29.2°C in the next 10 minutes, and the future humidity is predicted to be 58%.

[0048] S103: Based on the temperature control requirements, determine the air-conditioning parameter adjustment plan for the target vehicle within a preset future time period.

[0049] After obtaining the temperature control requirements in the future time period, the air-conditioning parameter adjustment plan for the target vehicle within a preset future time period can be determined based on the temperature control requirements, so as to adjust the air-conditioning parameters in advance to prevent the temperature of the target vehicle from changing frequently or requiring the occupants to adjust. Among them, in addition to the target temperature parameter in the prior art, the present application additionally adds a wind direction parameter and a wind speed parameter to improve the comfort of the occupants in the target vehicle.

[0050] In one embodiment, when determining the parameter adjustment scheme according to the temperature control requirement, it is necessary to determine the static parameter adjustment scheme corresponding to the target vehicle in the preset future time period based on the temperature control requirement of the target vehicle in the preset future time period and the environmental time series data set. It should be noted that the static parameter adjustment scheme here refers to the optimal static setting in a certain determined environment, that is, if the environmental time series data is always maintained, the static parameter adjustment scheme is the optimal adjustment scheme that takes into account both comfort and economy.

[0051] After obtaining the static parameter adjustment scheme, based on the temperature control requirement corresponding to the target vehicle in the preset future time period and the corresponding static parameter adjustment scheme, the dynamic parameter adjustment scheme corresponding to the environmental time series data set in the preset future time period can be determined. After obtaining the static parameter adjustment scheme and the dynamic parameter adjustment scheme, the static parameter adjustment scheme and / or the dynamic parameter adjustment scheme can be used as the air-conditioning parameter adjustment scheme of the target vehicle in the preset future time period.

[0052] If the environmental time series data of the target vehicle is relatively stable, the static parameter adjustment scheme is the optimal adjustment scheme that takes into account both comfort and economy. If the environmental time series data changes in a short time, the air-conditioning parameters can be adjusted through the dynamic parameter adjustment scheme after using the static parameter adjustment scheme. If the frequency of change of the environmental time series data is relatively high in a long time, the air-conditioning parameters can be adjusted only through the dynamic parameter adjustment scheme. At this time, the previous dynamic parameter adjustment scheme and the temperature control requirement corresponding to the target vehicle in the preset future time period are used as inputs to generate the dynamic parameter adjustment scheme. It should be noted that when adjusting the air-conditioning parameters, the set temperature of the air-conditioning may increase, but the body sensation of the occupants can be made unaffected by adjusting the wind speed and direction.

[0053] In one embodiment, when generating the static parameter adjustment scheme, the predicted temperature control requirement can be optimized by Bayesian optimization to obtain the static parameter adjustment scheme. Specifically, the Gaussian process regression strategy and the upper confidence bound strategy can be used to construct a Bayesian optimization model, and the temperature control requirement of the target vehicle in the preset future time period and the environmental time series data set are input into the Bayesian optimization model. And determine multiple evaluation dimensions of the Bayesian optimization model. The types of the evaluation dimensions here at least include the comfort dimension and the energy consumption dimension. Finally, the static parameter adjustment scheme output by the Bayesian optimization model can be obtained. It should be noted that when the evaluation dimensions are the comfort dimension and the energy consumption dimension, the optimization goal of the Bayesian optimization at this time is to maximize the comfort and minimize the air-conditioning power consumption at the same time. The static parameter adjustment scheme output by the Bayesian optimization should include the temperature setting, the wind speed setting, and the wind direction setting.

[0054] The core of Bayesian optimization is to find an objective function to measure the comfort and energy consumption of different air-conditioning setting schemes and find the optimal setting.

[0055] Therefore, in the static parameter adjustment scheme output at this time, the magnitude of the air-conditioning temperature setting value, the wind speed, and the wind direction are related to the comfort function, the energy consumption function, and the weights corresponding to the two functions. Among them, comfort is related to the air-conditioning temperature setting value, and energy consumption is related to the wind speed and the wind direction. The optimization goal of the air-conditioning temperature setting value is to maintain comfort without wasting energy. The optimization goal of the wind speed is to minimize the wind speed to reduce energy consumption. The optimization goal of the wind direction is to intelligently supply air according to the occupant's position and try to blow towards, so that the occupant can directly feel the air-conditioning temperature. When calculating comfort, the Predicted Mean Vote (PMV) can be used as an indicator to calculate human comfort. Specifically, the predicted mean thermal sensation is related to data such as air temperature, humidity, wind speed, occupant clothing, and metabolic rate. The lower the predicted mean thermal sensation value, the higher the comfort level. When calculating energy consumption, the energy consumption value is related to the wind speed and the temperature difference between the inside and outside of the vehicle. The air-conditioning temperature setting value, wind speed, and wind direction provided by Bayesian optimization are the optimal static settings in the current environment. However, in a complex environment (such as vehicle speed changes, occupant seat adjustment), reinforcement learning (such as DDPG) is required to further optimize the adaptive adjustment strategy.

[0056] Specifically, after obtaining the static parameter adjustment scheme, if the environmental time series data set changes, obtain the environmental change value of the environmental time series data set, use the temperature control demand corresponding to the target vehicle in the preset future time period as the initial state space, use the corresponding static parameter adjustment scheme as the initial action space, construct a deep deterministic policy gradient algorithm model, and based on the deep deterministic policy gradient algorithm model, determine the dynamic parameter adjustment scheme corresponding to the environmental change value.

[0057] Specifically, the deep deterministic policy gradient algorithm model includes a policy network Actor and a value network Critic. The policy network is used to receive the environmental state and output the optimized air-conditioning control strategy. The value network is used to evaluate the benefits of the current policy and optimize the policy.

[0058] The deep deterministic policy gradient algorithm model includes a state space, an action space, and a corresponding reward function. Here, the state space contains multiple states, and each state corresponds to different environmental time series data, such as the temperature inside the vehicle, the temperature outside the vehicle, humidity, light, the status of the occupants, the remaining battery power, etc. The action space contains multiple actions, and each action corresponds to a parameter adjustment plan, that is, corresponding to setting the temperature, wind speed, and wind direction. The reward function is a weighted operation of comfort and energy consumption, where comfort can be calculated by PMV, and energy consumption is calculated by the air conditioner power model. When performing reinforcement learning training, the deep deterministic policy gradient algorithm model uses experience replay to improve training stability and uses a soft update strategy to make the training smoother. Finally, the Actor policy network trained by the deep deterministic policy gradient algorithm model can dynamically optimize the air conditioner settings under different driving environments and different occupant statuses, realizing an adaptive temperature control strategy.

[0059] In one embodiment, after determining the static parameter adjustment plan or the dynamic parameter adjustment plan, when adjusting the wind direction of the air conditioning system, in order to avoid the wind direction blowing directly on the vehicle occupants, the body data of the occupants, such as height, arm length, torso length, etc., can be determined according to the infrared data collected by the infrared sensor inside the vehicle. At the same time, obtain the wind direction preferences from the occupants, such as whether they can accept the direct blowing of the air, whether they can accept the direct blowing on the skin, etc., and then fine-tune the wind direction parameters in the parameter adjustment plan according to the body data and the wind direction preferences to prevent the air blown by the air conditioning system from blowing directly on the skin or torso of the occupants. When performing fine-tuning, the air outlet direction of the air conditioning outlet can be determined based on the angle between the area where the occupants can accept the blowing and the air conditioning outlet. If the occupants cannot accept the direct blowing on the skin, the wind direction can be adjusted to blow on the clothes of the occupant's torso. When making adjustments, the influence of factors such as wind speed, outlet air temperature, and vehicle interior temperature can also be considered. If the wind speed is large or the temperature difference between the outlet air temperature and the vehicle interior temperature is large, it should be avoided that the air blown by the air conditioning system blows on the skin or torso of the occupants, improving the occupant experience.

[0060] In one embodiment, since different occupants have their own personalized needs, after adjusting the air conditioning system in the target vehicle through a static parameter adjustment scheme or a dynamic parameter adjustment scheme, if the occupants in the target vehicle (especially the main driver) adjust the air conditioning parameters by themselves within a preset time period (such as one minute), such as issuing a voice command, or manipulating the temperature and wind speed on the in-vehicle computer, or adjusting the air outlet direction at the air outlet, then at this time, obtain the air conditioning parameter adjustment instruction of the occupant based on the air conditioning parameter adjustment scheme, and the vehicle state data set corresponding to the air conditioning parameter adjustment instruction. The vehicle state data set here includes information such as vehicle speed data, in-vehicle occupant distribution data, driving habit data, and destination distance data. Then, based on the vehicle state data set and the air conditioning parameter adjustment instruction, update the model parameters in the hybrid Transformer architecture, Bayesian optimization model, and deep deterministic policy gradient algorithm model, so that the output of the above architecture and model is closer to the expectations of the occupants.

[0061] In one embodiment, when obtaining the air conditioning parameter adjustment instruction, it can be set to obtain only the air conditioning parameter adjustment instruction of the target occupant, for example, only obtain the air conditioning parameter adjustment instruction of the main driver, or only obtain the air conditioning parameter adjustment instruction of the co-driver, or not obtain the air conditioning parameter adjustment instruction.

[0062] Furthermore, when collecting the vehicle state dataset and the air-conditioning parameter adjustment instructions and adjusting the model parameters in the hybrid Transformer architecture, the Bayesian optimization model, and the deep deterministic policy gradient algorithm model, for different types of target vehicles and different types of drivers, the vehicle state dataset of the same or similar type of target vehicle and the air-conditioning parameter adjustment instructions of the drivers with the same or similar driving styles can be selected to train the hybrid Transformer architecture, the Bayesian optimization model, and the deep deterministic policy gradient algorithm model of the corresponding owner of the target vehicle. Specifically, based on the vehicle state dataset, the user portrait of the owner and the corresponding vehicle type can be determined. Based on the current environmental time-series dataset and the air-conditioning parameter adjustment instructions, a parameter training sample set is constructed. The target parameter training samples related to the user portrait and the corresponding vehicle type are determined in the parameter training sample set, and based on the target parameter training samples, the model parameters in the hybrid Transformer architecture, the Bayesian optimization model, and the deep deterministic policy gradient algorithm model of the corresponding owner of the target vehicle are updated. Among them, when determining the user portrait of the owner, the driving style of the owner, such as a fuel-consuming driving style and a fuel-saving driving style, can be determined based on the historical driving data of the owner. When determining the vehicle type of the target vehicle, the first vehicle type of the same vehicle model can be selected based on the vehicle model, or the second vehicle type of the reference vehicle with a price close to that of the target vehicle can be determined according to the current vehicle price of the target vehicle. The parameter training samples corresponding to the vehicles of the first vehicle type and the second vehicle type are used as the target parameter training samples. When determining the user portrait, the average fuel consumption difference between the owner and the vehicles of the same type as the target vehicle can be determined according to the historical driving data of the owner, and the user portrait of the owner can be determined according to the magnitude and sign of the average fuel consumption difference.

[0063] In one embodiment, the model parameters corresponding to different occupants can be stored in the cloud. When the driver of the vehicle is changed, the model parameters corresponding to the current driver can be obtained from the cloud and enabled. The historical driving data and the model parameters can be stored in the storage device of the computer device in advance. When it is necessary to determine the user portrait of the owner, the computer device can select the historical driving data from the storage device. Of course, the computer device can also obtain the historical driving data from other external devices. For example, the historical driving data is stored in the cloud. When it is necessary to determine the user portrait of the owner, the computer device can obtain the historical driving data from the cloud. The acquisition method of the historical driving data in this embodiment is not limited.

[0064] Among them, the hybrid Transformer architecture, Bayesian optimization model, and deep deterministic policy gradient algorithm model are mathematical models constructed based on machine learning algorithms. The constructed hybrid Transformer architecture, Bayesian optimization model, and deep deterministic policy gradient algorithm model are pre-trained through a training data set. When the set training accuracy and accuracy are reached, it is determined that the hybrid Transformer architecture, Bayesian optimization model, and deep deterministic policy gradient algorithm model trained this time are completed, so as to be used for the generation of air-conditioning parameter adjustment schemes.

[0065] In one embodiment, in order to ensure that the intelligent temperature control system can run in real time and is applicable to different vehicle models and driving environments, the above architectures and models, optimization algorithms, and control systems need to be integrated into the in-vehicle computing platform and optimized for deployment. For example, the hybrid Transformer architecture, Bayesian optimization model, and deep deterministic policy gradient algorithm model can be deployed in the target vehicle through TensorRT, and real-time data stream processing can be performed through an in-memory database and a distributed message queue to improve the response speed of the hybrid Transformer architecture, Bayesian optimization model, and deep deterministic policy gradient algorithm model; at preset time intervals, receive the optimized model parameters from the cloud and use the optimized model parameters to replace the model parameters in the hybrid Transformer architecture, Bayesian optimization model, and deep deterministic policy gradient algorithm model.

[0066] Verify the system performance in laboratory simulation + actual road tests. The test environment is shown in the following table:

[0067] Table of test scenarios and test conditions

[0068] Test scenario Test conditions Summer sunny day (high temperature and strong light) Outdoor temperature 36°C, light intensity 900 W / m2 Winter low temperature (cold environment) Outdoor temperature -5°C, humidity 40% High-speed driving (energy-saving optimization) Vehicle speed 120 km / h, SOC (state of charge) 40% Urban low-speed (occupant dynamic adjustment) Vehicle speed 30 km / h, number of occupants varies Idle stop (energy-saving optimization) Vehicle stationary, SOC 20% (low battery mode)

[0069] During the test, we measure and compare the following core indicators:

[0070] Table of core indicator comparisons

[0071]

[0072] As can be seen from the above table of core indicator comparisons, compared with the traditional fixed air-conditioning setting, this solution calculates the optimal temperature control strategy through AI, reducing the air-conditioning energy consumption by 15% and increasing the electric vehicle's cruising range by 5-8%. Based on the occupant distribution and environmental prediction to adjust the temperature control, the occupant comfort (PMV score) is improved by 10%. The traditional temperature control adjustment lags behind, while this solution adaptively adjusts the wind speed and direction through reinforcement learning, increasing the adjustment response speed by 30%. This solution can be intelligently optimized in different driving modes such as high-speed driving, low-speed urban driving, and parking idling, effectively improving the temperature control effect.

[0073] As shown Figure 2 in the figure, an embodiment of the present application further provides an air conditioner temperature control device, including:

[0074] A data acquisition module 201, which acquires an environmental time series data set of a target vehicle, and the environmental time series data set includes in-vehicle and out-of-vehicle environmental data, vehicle state data, and air conditioner parameters; the in-vehicle and out-of-vehicle environmental data at least includes in-vehicle occupant distribution data; the vehicle state data includes vehicle speed data, power data, and driving mode data.

[0075] A demand prediction module 202, which determines the temperature control demand of the target vehicle within a preset future time period based on the environmental time series data set.

[0076] A parameter adjustment module 203, which determines an air conditioner parameter adjustment plan for the target vehicle within a preset future time period based on the temperature control demand; the air conditioner parameters at least include a target temperature parameter, a wind direction parameter, and a wind speed parameter.

[0077] In a specific embodiment, the demand prediction module 202 includes:

[0078] Input the environmental time series data set into a hybrid pre-trained temperature control demand prediction model to determine the temperature control demand of the target vehicle within a preset future time period; the temperature control demand prediction model includes a convolutional neural network layer, a long short-term memory network layer, and an improved temperature control demand prediction layer; the convolutional neural network layer is used to extract short-term local change features of the environmental time series data set; the long short-term memory network layer is used to capture long-term time dependence relationships of the environmental time series data set; the temperature control demand prediction layer is used for global feature modeling; in the temperature control demand prediction layer, the multi-head attention at least includes environmental attention, occupant attention, and energy consumption attention.

[0079] In a specific embodiment, the parameter adjustment module 203 includes:

[0080] Based on the temperature control demand of the target vehicle within the preset future time period and the environmental time series data set, determine a static parameter adjustment plan for the target vehicle within the preset future time period; based on the temperature control demand of the target vehicle within the preset future time period and the corresponding static parameter adjustment plan, determine a dynamic parameter adjustment plan for the environmental time series data set within the preset future time period; use the static parameter adjustment plan and / or the dynamic parameter adjustment plan as the air conditioner parameter adjustment plan for the target vehicle within the preset future time period.

[0081] In a specific embodiment, the parameter adjustment module 203 includes:

[0082] A Bayesian optimization model is constructed by using Gaussian process regression and an upper confidence bound strategy; the temperature control requirements of the target vehicle within the preset future time period and the environmental time series data set are input into the Bayesian optimization model; multiple evaluation dimensions of the Bayesian optimization model are determined, and the types of the evaluation dimensions at least include a comfort dimension and an energy consumption dimension; a static parameter adjustment scheme output by the Bayesian optimization model is obtained.

[0083] In a specific embodiment, the parameter adjustment module 203 includes:

[0084] An environmental change value of the environmental time series data set is obtained; the temperature control requirements corresponding to the target vehicle within the preset future time period are used as an initial state space, and the corresponding static parameter adjustment scheme is used as an initial action space to construct a deep deterministic policy gradient algorithm model; based on the deep deterministic policy gradient algorithm model, a dynamic parameter adjustment scheme corresponding to the environmental change value is determined.

[0085] In a specific embodiment, the parameter adjustment module 203 includes:

[0086] An air-conditioning parameter adjustment instruction of the occupant based on the air-conditioning parameter adjustment scheme and a vehicle state data set corresponding to the air-conditioning parameter adjustment instruction are obtained; the vehicle state data set includes at least one of vehicle speed data, in-vehicle occupant distribution data, driving habit data, and destination distance data; based on the vehicle state data set and the air-conditioning parameter adjustment instruction, the model parameters in the temperature control requirement prediction model, the Bayesian optimization model, and the deep deterministic policy gradient algorithm model are updated.

[0087] In a specific embodiment, the parameter adjustment module 203 includes:

[0088] Based on the vehicle state data set, a user portrait of the owner corresponding to the target vehicle and the corresponding vehicle type are determined; based on the current environmental time series data set and the air-conditioning parameter adjustment instruction, a parameter training sample set is constructed; target parameter training samples related to the user portrait and the corresponding vehicle type are determined in the parameter training sample set; learning and training are performed based on the target parameter training samples, and the model parameters in the temperature control requirement prediction model, the Bayesian optimization model, and the deep deterministic policy gradient algorithm model of the target vehicle are updated.

[0089] In a specific embodiment, the parameter adjustment module 203 includes:

[0090] Deploy the temperature control demand prediction model, the Bayesian optimization model, and the deep deterministic policy gradient algorithm model in the target vehicle; perform real-time data stream processing through an in-memory database and a distributed message queue; at preset time intervals, receive the optimized model parameters from the cloud, and use the optimized model parameters to replace the model parameters in the temperature control demand prediction model, the Bayesian optimization model, and the deep deterministic policy gradient algorithm model.

[0091] Figure 3 It is a schematic structural diagram of a vehicle provided by an embodiment of the present application.

[0092] Exemplarily, as Figure 3 shown, the vehicle includes: a memory 301 and a processor 302. Among them, an executable program code 3011 is stored in the memory 301, and the processor 302 is used to call and execute the executable program code 3011 to execute a vehicle torque control method.

[0093] In this embodiment, the vehicle can be divided into functional modules according to the above method examples. For example, it can correspond to each functional module, or two or more functions can be integrated into one processing module. The above integrated module can be implemented in the form of hardware. It should be noted that the division of modules in this embodiment is illustrative, only a logical function division, and there can be other division methods in actual implementation.

[0094] In the case of dividing each functional module according to each function, the vehicle can include:

[0095] A data acquisition module, which acquires an environmental time series data set of the target vehicle. The environmental time series data set includes vehicle interior and exterior environmental data, vehicle state data, and air conditioning parameters; the vehicle interior and exterior environmental data at least includes vehicle interior occupant distribution data; the vehicle state data includes vehicle speed data, power data, and driving mode data.

[0096] A demand prediction module, which determines the temperature control demand of the target vehicle within a preset future time period based on the environmental time series data set.

[0097] A parameter adjustment module, which determines an air conditioning parameter adjustment plan for the target vehicle within a preset future time period based on the temperature control demand; the air conditioning parameters at least include a target temperature parameter, a wind direction parameter, and a wind speed parameter.

[0098] It should be noted that all relevant contents of each step involved in the above method embodiment can be cited in the function description of the corresponding functional module, and will not be elaborated here.

[0099] The vehicle provided in this embodiment is used to execute the above air conditioner temperature control method, and thus can achieve the same effect as the above implementation method.

[0100] In the case of adopting an integrated unit, the vehicle may include a processing module and a storage module. Among them, the processing module can be used to control and manage the actions of the vehicle. The storage module can be used to support the vehicle to execute mutual program codes, data, etc.

[0101] Among them, the processing module can be a processor or a controller, which can implement or execute various exemplary logic blocks, modules, and circuits described in combination with the disclosure of this application. The processor can also be a combination of computing functions, such as including a combination of one or more microprocessors, a combination of digital signal processing (DSP) and a microprocessor, etc. The storage module can be a memory.

[0102] The embodiment of this application also provides a non-volatile computer storage medium, storing computer-executable instructions, and the computer-executable instructions are set as:

[0103] Obtain the environmental time series data set of the target vehicle, where the environmental time series data set includes in-vehicle and out-of-vehicle environmental data, vehicle state data, and air conditioner parameters; the in-vehicle and out-of-vehicle environmental data at least includes in-vehicle occupant distribution data; the vehicle state data includes vehicle speed data, power data, and driving mode data; based on the environmental time series data set, determine the temperature control requirement of the target vehicle within a preset future time period; based on the temperature control requirement, determine the air conditioner parameter adjustment plan of the target vehicle within a preset future time period; the air conditioner parameters at least include target temperature parameters, wind direction parameters, and wind speed parameters.

[0104] This embodiment can divide the functions of the vehicle according to the above method examples. For example, it can correspond to each functional module, or integrate two or more functions into one processing module. The above integrated module can be implemented in the form of hardware. It should be noted that the division of modules in this embodiment is illustrative, only a logical function division, and there may be other division methods in actual implementation.

[0105] In the case of dividing each functional module according to each function, the vehicle may include: a data acquisition module, a demand prediction module, etc. It should be noted that all relevant contents of each step involved in the above method embodiment can be cited in the function description of the corresponding functional module, and will not be repeated here.

[0106] The vehicle provided in this embodiment is used to execute the above-mentioned vehicle air conditioner control method, so the same effects as those of the above implementation method can be achieved. In the case of adopting an integrated unit, the vehicle may include a processing module and a storage module. Among them, the processing module can be used to control and manage the actions of the vehicle. The storage module can be used to support the vehicle to execute mutual program codes, data, etc.

[0107] Among them, the processing module can be a processor or a controller, which can implement or execute various exemplary logic blocks, modules and circuits described in connection with the disclosure of the present application. The processor can also be a combination that realizes computing functions, such as including a combination of one or more microprocessors, a combination of digital signal processing (DSP) and a microprocessor, etc. The storage module can be a memory.

[0108] This embodiment also provides a computer-readable storage medium. Computer program codes are stored in the computer-readable storage medium (including but not limited to disk memories, CD-ROMs, optical memories, etc.). When the computer program codes run on a computer, the computer is enabled to execute the above-mentioned related method steps to implement the air conditioner temperature control method provided in the above embodiment.

[0109] This embodiment also provides a computer program product. When the computer program product runs on a computer, the computer is enabled to execute the above-mentioned related steps to implement an air conditioner temperature control method provided in the above embodiment. Among them, the beneficial effects of the above embodiment can be referred to the beneficial effects in the corresponding method provided above, and will not be elaborated here.

[0110] Through the description of the above embodiments, those skilled in the art can understand that for the convenience and brevity of description, only the above-mentioned division of each functional module is used as an example. In actual applications, the above functions can be allocated to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above.

[0111] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of modules or units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces. The indirect coupling or communication connection of devices or units can be in electrical, mechanical or other forms. In the description of the present disclosure, it should be understood that if terms such as "upper", "lower", "front", "rear", "left" and "right" are used to indicate the orientation or positional relationship, it is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the indicated position or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present disclosure.

[0112] It should be noted that in this text, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or further includes elements inherent to such process, method, commodity or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, method, commodity or device including the element.

[0113] The above are only the embodiments of the present disclosure and are not used to limit the present disclosure. For those skilled in the art, the present disclosure can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present disclosure should be included within the scope of the claims of the present disclosure.

Claims

1. A vehicle air conditioner control method, characterized in that, Including: Obtain an environmental time-series data set of the target vehicle, where the environmental time-series data set includes in-vehicle and out-of-vehicle environmental data, vehicle state data, and air-conditioning parameters; The in-vehicle and out-of-vehicle environmental data at least includes in-vehicle occupant distribution data; The vehicle state data includes vehicle speed data, power data, and driving mode data; Based on the environmental time-series data set, determine the temperature control demand of the target vehicle within a preset future time period; Based on the temperature control demand, determine an air-conditioning parameter adjustment plan for the target vehicle within a preset future time period; the air-conditioning parameters at least include a target temperature parameter, a wind direction parameter, and a wind speed parameter.

2. The method according to claim 1, characterized in that, The determining the temperature control demand of the target vehicle within a preset future time period based on the environmental time-series data set specifically includes: Input the environmental time-series data set into a hybrid pre-trained temperature control demand prediction model to determine the temperature control demand of the target vehicle within a preset future time period; The temperature control demand prediction model includes a convolutional neural network layer, a long short-term memory network layer, and an improved temperature control demand prediction layer; The convolutional neural network layer is used to extract short-term local change features of the environmental time-series data set; The long short-term memory network layer is used to capture long-term time-dependent relationships of the environmental time-series data set; The temperature control demand prediction layer is used for global feature modeling; in the temperature control demand prediction layer, at least environmental attention, occupant attention, and energy consumption attention are included in the multi-head attention.

3. The method according to claim 2, characterized in that, The determining the air-conditioning parameter adjustment plan for the target vehicle within a preset future time period based on the temperature control demand specifically includes: Based on the temperature control demand of the target vehicle within the preset future time period and the environmental time-series data set, determine a static parameter adjustment plan corresponding to the target vehicle within the preset future time period; Based on the temperature control demand corresponding to the target vehicle within the preset future time period and the corresponding static parameter adjustment plan, determine a dynamic parameter adjustment plan corresponding to the environmental time-series data set within the preset future time period; Use the static parameter adjustment plan and / or the dynamic parameter adjustment plan as the air-conditioning parameter adjustment plan for the target vehicle within a preset future time period.

4. The method according to claim 3, wherein The determining the static parameter adjustment plan corresponding to the target vehicle within the preset future time period based on the temperature control demand of the target vehicle within the preset future time period and the environmental time-series data set specifically includes: Adopt Gaussian process regression and upper confidence bound strategy to construct a Bayesian optimization model; Input the temperature control demand of the target vehicle within the preset future time period and the environmental time-series data set into the Bayesian optimization model; Determine multiple evaluation dimensions of the Bayesian optimization model, and the types of the evaluation dimensions at least include a comfort dimension and an energy consumption dimension; Obtain the static parameter adjustment plan output by the Bayesian optimization model.

5. The method according to claim 4, characterized in that, The determining the dynamic parameter adjustment plan corresponding to the environmental time-series data set within the preset future time period based on the temperature control demand corresponding to the target vehicle within the preset future time period and the corresponding static parameter adjustment plan specifically includes: Obtain the environmental change value of the environmental time series dataset; Take the temperature control requirements of the target vehicle corresponding to the preset future time period as the initial state space, and take the corresponding static parameter adjustment scheme as the initial action space to construct a deep deterministic policy gradient algorithm model; Based on the deep deterministic policy gradient algorithm model, determine the dynamic parameter adjustment scheme corresponding to the environmental change value.

6. The method according to claim 5, characterized in that, After determining the air-conditioning parameter adjustment scheme of the target vehicle in the preset future time period based on the temperature control requirements, the method further includes: Obtain the air-conditioning parameter adjustment instruction of the occupant based on the air-conditioning parameter adjustment scheme, and the vehicle state dataset corresponding to the air-conditioning parameter adjustment instruction; The vehicle state dataset includes at least one of vehicle speed data, in-vehicle occupant distribution data, driving habit data, and destination distance data; Based on the vehicle state dataset and the air-conditioning parameter adjustment instruction, update the model parameters in the temperature control requirement prediction model, the Bayesian optimization model, and the deep deterministic policy gradient algorithm model.

7. The method according to claim 6, characterized in that The updating the model parameters in the temperature control requirement prediction model, the Bayesian optimization model, and the deep deterministic policy gradient algorithm model based on the vehicle state dataset and the air-conditioning parameter adjustment instruction specifically includes: Based on the vehicle state dataset, determine the user profile of the owner corresponding to the target vehicle and the corresponding vehicle type; Based on the current environmental time series dataset and the air-conditioning parameter adjustment instruction, construct a parameter training sample set; Determine the target parameter training samples related to the user profile and the corresponding vehicle type in the parameter training sample set; Based on the target parameter training samples, perform learning and training to update the model parameters in the temperature control requirement prediction model, the Bayesian optimization model, and the deep deterministic policy gradient algorithm model of the target vehicle.

8. The method according to claim 6, characterized in that, The method further includes: Deploy the temperature control requirement prediction model, the Bayesian optimization model, and the deep deterministic policy gradient algorithm model in the target vehicle; Perform real-time data stream processing through an in-memory database and a distributed message queue; At preset intervals, receive the optimized model parameters from the cloud, and use the optimized model parameters to replace the model parameters in the temperature control requirement prediction model, the Bayesian optimization model, and the deep deterministic policy gradient algorithm model.

9. An air conditioner temperature control device, characterized in that, The device includes: A data acquisition module that acquires the environmental time series dataset of the target vehicle, where the environmental time series dataset includes in-vehicle and out-of-vehicle environmental data, vehicle state data, and air-conditioning parameters; the in-vehicle and out-of-vehicle environmental data includes at least in-vehicle occupant distribution data; the vehicle state data includes vehicle speed data, power data, and driving mode data; A demand prediction module that determines the temperature control requirements of the target vehicle in a preset future time period based on the environmental time series dataset; A parameter adjustment module that determines the air-conditioning parameter adjustment scheme of the target vehicle in a preset future time period based on the temperature control requirements; the air-conditioning parameters include at least target temperature parameters, wind direction parameters, and wind speed parameters.

10. A vehicle, characterized in that, Includes: At least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the at least one processor is enabled to execute: obtain an environmental time series data set of a target vehicle, the environmental time series data set including in-vehicle and out-of-vehicle environmental data, vehicle state data, and air conditioning parameters; the in-vehicle and out-of-vehicle environmental data at least including in-vehicle occupant distribution data; the vehicle state data including vehicle speed data, power data, and driving mode data; determine a temperature control requirement of the target vehicle within a preset future time period based on the environmental time series data set; determine an air conditioning parameter adjustment scheme of the target vehicle within a preset future time period based on the temperature control requirement; the air conditioning parameters at least including a target temperature parameter, a wind direction parameter, and a wind speed parameter.

Images