Vehicle air conditioning control method, device and readable storage medium

Abstract

This application discloses a vehicle air conditioning control method, device, and readable storage medium, relating to the field of air conditioning control technology, and is used to obtain a more superior air conditioning control strategy to enhance the passenger experience. The method includes: obtaining current vehicle data at the current moment and historical vehicle data from a preset time period before the current moment, and obtaining to-be-processed vehicle data based on the current vehicle data and the historical vehicle data; determining a feature vector for the to-be-processed vehicle data; the feature vector includes a first feature vector and a second feature vector; obtaining a vehicle air conditioning control strategy based on the first feature vector, the second feature vector, and a pre-trained vehicle air conditioning control model, and controlling the vehicle air conditioning according to the to-be-processed air conditioning control strategy.

Description

Technical Field

[0001] The present application relates to the field of air-conditioning control technology, and in particular to a vehicle-mounted air-conditioning control method, device, and readable storage medium. Background Art

[0002] In recent years, with the development of machine learning technology, vehicle air conditioning systems based on machine learning algorithms have developed rapidly. Using machine learning algorithms to adjust vehicle air conditioning can reduce the driver's manual operation and reduce the probability of traffic accidents.

[0003] However, existing in-vehicle air conditioning systems based on machine learning algorithms typically require sensors to extract current in-vehicle environmental information (such as temperature and humidity). This information is then fed into a pre-trained machine learning model to generate an air conditioning control strategy. However, because the various in-vehicle environmental information is previously isolated, existing machine learning models rely solely on the inherent characteristics of each piece of environmental information for prediction, failing to fully utilize the implicit features of this information. This results in poor air conditioning control strategy performance, which in turn reduces the passenger experience. Summary of the Invention

[0004] One of the purposes of this application is to provide a vehicle-mounted air-conditioning control method, device and readable storage medium, which are committed to obtaining a more excellent air-conditioning control strategy to enhance the passenger experience.

[0005] According to the first aspect of the present application, a vehicle air-conditioning control method is provided, which includes: obtaining current vehicle data at the current moment and historical vehicle data of a preset time period before the current moment, and obtaining vehicle data to be processed based on the current vehicle data and the historical vehicle data; the vehicle data is used to reflect the vehicle operating status and the internal and external environment of the vehicle; the vehicle data to be processed has time series; determining the feature vector of the vehicle data to be processed; the feature vector includes a first feature vector and a second feature vector; the first feature vector is used to reflect the changes in the vehicle operating status over time and the changes in the internal and external environment of the vehicle over time; the second feature vector is used to reflect the correlation between the vehicle operating status and the internal and external environment of the vehicle at the current moment; based on the first feature vector, the second feature vector and a pre-trained vehicle air-conditioning control model, a vehicle air-conditioning control strategy is obtained, and the vehicle air-conditioning is controlled according to the vehicle air-conditioning control strategy.

[0006] According to the above technical means, the present application obtains the current vehicle data at the current moment and the historical vehicle data of the preset time period before the current moment, and obtains the vehicle data to be processed based on the current vehicle data and the historical vehicle data. Since the vehicle data is used to reflect the vehicle's operating status and the internal and external environment of the vehicle, the obtained vehicle data to be processed can reflect the changes in the vehicle's operating status and the internal and external environment of the vehicle over a period of time, and the vehicle data to be processed has a time series. Since the vehicle data to be processed includes not only the information of the data itself, but also the change information of the data in a time sequence. Therefore, the present application can determine the first eigenvector and the second eigenvector of the vehicle data to be processed, wherein the first eigenvector is used to reflect the changes in the vehicle's operating status over time and the changes in the internal and external environment of the vehicle over time; the second eigenvector is used to reflect the correlation between the vehicle's operating status and the internal and external environment of the vehicle at the current moment. Furthermore, the present application obtains the vehicle air-conditioning control strategy based on the first eigenvector, the second eigenvector and the pre-trained vehicle air-conditioning control model, and controls the vehicle air-conditioning according to the vehicle air-conditioning control strategy. Compared with the existing technology that only focuses on the characteristics of the data itself (i.e., numerical characteristics) when using machine learning models, this application not only focuses on the characteristics of the data itself, but also focuses on the connections between the data, and mines the implicit time series characteristics, so as to derive a better air-conditioning control strategy, thereby improving the passenger experience.

[0007] Furthermore, the vehicle data to be processed is obtained based on the current vehicle data and the historical vehicle data, including: sorting the current vehicle data and the historical vehicle data in chronological order to obtain first data; cleaning the historical vehicle data to obtain second data; the second data has continuity; and splicing the first data and the second data to obtain the vehicle data to be processed.

[0008] According to the above technical means, in order to obtain the vehicle data to be processed, the present application sorts the current vehicle data and the historical vehicle data in chronological order, so that the first data obtained has time sequence. Furthermore, the present application cleans the historical vehicle data to obtain the second data. After cleaning, the historical vehicle data can be deduplicated, filled in, etc., so that the second data obtained has continuity. In this way, the present application splices the first data with the second data, and the obtained vehicle data to be processed is not only time sequenced, but also complete and continuous, which is convenient for subsequent feature extraction.

[0009] Furthermore, determining the feature vector of the vehicle data to be processed includes: clustering the vehicle data to be processed to obtain multiple data sets; the similarity between each data in a data set is greater than or equal to a preset threshold; extracting features from each data set to obtain an initial feature vector; performing a first transformation on the initial feature vector to obtain a first feature vector; and performing a second transformation on the initial feature vector to obtain a second feature vector.

[0010] According to the above technical means, in order to facilitate feature extraction of the vehicle data to be processed, the present application clusters the vehicle data to be processed to obtain multiple data sets, wherein the similarity between the data in a data set is greater than or equal to a preset threshold. Furthermore, the present application only needs to perform feature extraction on each data set, which reduces the processing pressure. Moreover, the obtained initial feature vector is not used directly, but the initial feature vector is subjected to a first transformation to obtain a first feature vector; the initial feature vector is subjected to a second transformation to obtain a second feature vector, making the extracted features more refined.

[0011] Furthermore, performing a first transformation on the initial feature vector to obtain a first feature vector includes: inputting the initial feature vector into a preset deep cross network, performing a first transformation on the initial feature vector through the deep cross network, and outputting the first feature vector.

[0012] According to the above technical means, the present application inputs the initial feature vector into a preset deep cross network to perform a first transformation on the initial feature vector through the deep cross network and output a first feature vector. Since the deep cross network uses the idea of ​​the residual network to achieve explicit modeling of features, a more explanatory high-order feature cross can be obtained. At the same time, due to the addition of time series features, the deep cross network can pay more attention to the changes in the features at the most recent moment.

[0013] Furthermore, performing a first transformation on the initial eigenvector to obtain a first eigenvector includes: inputting the initial eigenvector into a preset multilayer perceptron, performing a second transformation on the initial eigenvector through the multilayer perceptron, and outputting a second eigenvector.

[0014] According to the above technical means, the present application inputs the initial feature vector into a preset multilayer perceptron, which performs a second transformation on the initial feature vector through the multilayer perceptron and outputs a second feature vector. Because the multilayer perceptron models the interaction between different features, it focuses on the interaction between each feature at the current moment.

[0015] In a second aspect, a vehicle air-conditioning control device is provided, which includes an acquisition unit, a determination unit and a processing unit; the acquisition unit is used to acquire current vehicle data at a current moment and historical vehicle data for a preset time period before the current moment, and obtain vehicle data to be processed based on the current vehicle data and the historical vehicle data; the vehicle data is used to reflect the vehicle operating status and the internal and external environment of the vehicle; the vehicle data to be processed has time series; the determination unit is used to determine the feature vector of the vehicle data to be processed; the feature vector includes a first feature vector and a second feature vector; the first feature vector is used to reflect the changes in the vehicle operating status over time and the changes in the internal and external environment of the vehicle over time; the second feature vector is used to reflect the correlation between the vehicle operating status and the internal and external environment of the vehicle at the current moment; the processing unit is used to obtain a vehicle air-conditioning control strategy based on the first feature vector, the second feature vector and a pre-trained vehicle air-conditioning control model, and control the vehicle air-conditioning according to the vehicle air-conditioning control strategy.

[0016] Optionally, the acquisition unit is specifically used to: sort the current vehicle data and historical vehicle data in chronological order to obtain first data; clean the historical vehicle data to obtain second data; the second data has continuity; and splice the first data and the second data to obtain the vehicle data to be processed.

[0017] Optionally, the determination unit is specifically used to: cluster the vehicle data to be processed to obtain multiple data sets; the similarity between each data in a data set is greater than or equal to a preset threshold; perform feature extraction on each data set to obtain an initial feature vector; perform a first transformation on the initial feature vector to obtain a first feature vector; perform a second transformation on the initial feature vector to obtain a second feature vector.

[0018] Optionally, the determination unit is specifically used to: input the initial feature vector into a preset deep cross network, perform a first transformation on the initial feature vector through the deep cross network, and output a first feature vector.

[0019] Optionally, the determining unit is specifically configured to: input the initial feature vector into a preset multilayer perceptron, perform a second transformation on the initial feature vector through the multilayer perceptron, and output a second feature vector.

[0020] In a third aspect, a vehicle is provided, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the vehicle air conditioning control method according to the first aspect.

[0021] In a fourth aspect, a computer-readable storage medium is provided, on which a computer program is stored. The program is executed by a processor to implement the vehicle air conditioning control method as described in the first aspect.

[0022] Therefore, the above technical features of this application have the following beneficial effects:

[0023] (1) The present application obtains the current vehicle data at the current moment and the historical vehicle data of the preset time period before the current moment, and obtains the vehicle data to be processed based on the current vehicle data and the historical vehicle data. Since the vehicle data is used to reflect the vehicle operating status and the internal and external environment of the vehicle, the obtained vehicle data to be processed can reflect the changes in the vehicle operating status and the internal and external environment of the vehicle over a period of time, and the vehicle data to be processed has a time series. Since the vehicle data to be processed includes not only the information of the data itself, but also the change information of the data in a time sequence. Therefore, the present application can determine the first eigenvector and the second eigenvector of the vehicle data to be processed, wherein the first eigenvector is used to reflect the changes in the vehicle operating status over time and the changes in the internal and external environment of the vehicle over time; the second eigenvector is used to reflect the correlation between the vehicle operating status and the internal and external environment of the vehicle at the current moment. Furthermore, the present application obtains the vehicle air-conditioning control strategy based on the first eigenvector, the second eigenvector and the pre-trained vehicle air-conditioning control model, and controls the vehicle air-conditioning according to the vehicle air-conditioning control strategy. Compared with the existing technology that only focuses on the characteristics of the data itself (i.e., numerical characteristics) when using machine learning models, this application not only focuses on the characteristics of the data itself, but also focuses on the connections between the data, and mines the implicit time series characteristics, so as to derive a better air-conditioning control strategy, thereby improving the passenger experience.

[0024] (2) The present application sorts the current vehicle data and the historical vehicle data in chronological order, so that the first data obtained has a time sequence. Furthermore, the present application cleans the historical vehicle data to obtain the second data. After cleaning, the historical vehicle data can be deduplicated, filled in, etc., so that the second data obtained has continuity. In this way, the present application splices the first data with the second data, and the obtained vehicle data to be processed is not only time-series, but also complete and continuous, which is convenient for subsequent feature extraction.

[0025] (3) The present application clusters the vehicle data to be processed to obtain multiple data sets, wherein the similarity between each data in a data set is greater than or equal to a preset threshold. Furthermore, the present application only needs to extract features from each data set, which reduces the processing pressure. The obtained initial feature vector is not used directly, but the initial feature vector is subjected to a first transformation to obtain a first feature vector; the initial feature vector is subjected to a second transformation to obtain a second feature vector, so that the extracted features are more refined.

[0026] (4) This application inputs the initial feature vector into a preset deep cross network to perform a first transformation on the initial feature vector through the deep cross network and output a first feature vector. Since the deep cross network uses the idea of ​​the residual network to achieve explicit modeling of features, a more explanatory high-order feature cross can be obtained. At the same time, due to the addition of time series features, the deep cross network can pay more attention to the changes in the features at the most recent moment.

[0027] (5) The present application inputs the initial feature vector into a preset multilayer perceptron, which performs a second transformation on the initial feature vector and outputs a second feature vector. Since the multilayer perceptron models the interaction between different features, the multilayer perceptron focuses on the interaction between each feature at the current moment.

[0028] It should be noted that the technical effects brought about by any implementation method in the second to fourth aspects can refer to the technical effects brought about by the corresponding implementation method in the first aspect, and will not be repeated here.

[0029] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present application, and together with the specification are used to explain the principles of the present application, and do not constitute an improper limitation on the present application.

[0031] Figure 1 A schematic structural diagram of a vehicle air-conditioning control system provided in an embodiment of the present application;

[0032] Figure 2 A schematic diagram of the logic module composition of an air conditioning control device provided in an embodiment of the present application;

[0033] Figure 3 A flow chart of a vehicle air conditioning control method provided in an embodiment of the present application;

[0034] Figure 4 A schematic diagram of self-discretization of continuous features in a feature self-embedding module provided in an embodiment of the present application;

[0035] Figure 5 A schematic diagram of the structure of a deep cross network provided in an embodiment of the present application;

[0036] Figure 6 A schematic diagram of the structure of a multilayer perceptron provided in an embodiment of the present application;

[0037] Figure 7A schematic structural diagram of an air conditioning control device provided in an embodiment of the present application;

[0038] Figure 8 A schematic structural diagram of a vehicle provided in an embodiment of the present application. DETAILED DESCRIPTION

[0039] In order to enable ordinary people in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0040] It should be noted that the terms "first," "second," and the like in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or precedence. It should be understood that the numbers used in this manner are interchangeable where appropriate so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the embodiments of the present application as detailed in the appended claims.

[0041] It will also be understood that the term “comprising” indicates the presence of described features, integers, steps, operations, elements and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements and / or components.

[0042] It should be noted that the illustrations provided in the following embodiments are merely schematic illustrations of the basic concept of the present invention. Therefore, the illustrations only show components related to the present invention and are not drawn according to the number, shape, and size of components in actual implementation. In actual implementation, the type, quantity, and proportion of each component may be changed arbitrarily, and the component layout may also be more complex.

[0043] In addition, the actions and terms involved in the various embodiments of this application can refer to each other without limitation. The message names or parameter names in the messages exchanged between the various devices in the embodiments of this application are only examples, and other names can also be used in specific implementations without limitation.

[0044] To facilitate the clear description of the technical solutions of the embodiments of the present application, in the embodiments of the present application, the words "first" and "second" are used to distinguish between identical or similar items with substantially the same functions and effects. Those skilled in the art will understand that the words "first" and "second" do not limit the quantity or execution order, and the words "first" and "second" do not necessarily mean different.

[0045] It should be noted that, in this application, words such as "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described in this application as "exemplary" or "for example" should not be construed as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0046] In this application, "at least one" means one or more, and "more" means two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following items" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can mean: a, b, c, ab, ac, bc, or abc, where a, b, c can be single or multiple.

[0047] In recent years, with the development of machine learning technology, vehicle air conditioning systems based on machine learning algorithms have developed rapidly. Using machine learning algorithms to adjust vehicle air conditioning can reduce the driver's manual operation and reduce the probability of traffic accidents.

[0048] However, existing in-vehicle air conditioning systems based on machine learning algorithms typically require sensors to extract current in-vehicle environmental information (such as temperature and humidity). This information is then fed into a pre-trained machine learning model to generate an air conditioning control strategy. However, because the various in-vehicle environmental information is previously isolated, existing machine learning models rely solely on the inherent characteristics of each piece of environmental information for prediction, failing to fully utilize the implicit features of this information. This results in poor air conditioning control strategy performance, which in turn reduces the passenger experience.

[0049] In view of this, an embodiment of the present application provides a vehicle air-conditioning control method, which obtains current vehicle data at the current moment and historical vehicle data of a preset time period before the current moment, and obtains vehicle data to be processed based on the current vehicle data and the historical vehicle data; the vehicle data is used to reflect the vehicle operating status and the internal and external environment of the vehicle; the vehicle data to be processed has time series; the feature vector of the vehicle data to be processed is determined; the feature vector includes a first feature vector and a second feature vector; the first feature vector is used to reflect the changes in the vehicle operating status over time and the changes in the internal and external environment of the vehicle over time; the second feature vector is used to reflect the correlation between the vehicle operating status and the internal and external environment of the vehicle at the current moment; based on the first feature vector, the second feature vector and a pre-trained vehicle air-conditioning control model, a vehicle air-conditioning control strategy is obtained, and the vehicle air-conditioning is controlled according to the vehicle air-conditioning control strategy.

[0050] The method provided in the embodiments of the present application is described in detail below with reference to the accompanying drawings.

[0051] Figure 1 An exemplary application scenario diagram provided by an embodiment of the present application is shown. Figure 1 As shown, the vehicle air conditioning control method provided in the embodiment of the present application can be applied to a vehicle air conditioning control system 10. The vehicle air conditioning control system 10 includes a vehicle air conditioning control device (hereinafter referred to as the air conditioning control device) 11 and a vehicle 12. The air conditioning control device 11 is connected to the vehicle 12. The air conditioning control device 11 and the vehicle 12 can be connected by wire or wirelessly, which is not limited in the embodiment of the present disclosure.

[0052] The vehicle 12 is equipped with an onboard air conditioner, and the vehicle 12 can control the onboard air conditioner through the air conditioner control device 11 to adjust the temperature inside the vehicle.

[0053] The air conditioning control device 11 is configured to obtain current vehicle data at the current moment and historical vehicle data from a preset time period before the current moment, and to obtain vehicle data to be processed based on the current vehicle data and the historical vehicle data. The air conditioning control device 11 is further configured to determine a feature vector of the vehicle data to be processed, and based on the first feature vector, the second feature vector, and a pre-trained vehicle air conditioning control model, to obtain a vehicle air conditioning control strategy, and to control the vehicle air conditioning according to the vehicle air conditioning control strategy.

[0054] Optionally, the air conditioning control device 11 includes a data preprocessing module, a time series feature module, an original sensor feature module, a feature self-embedding module, a time series feature embedding vector module, an original sensor feature embedding vector module, a deep cross network module, a multi-layer perceptron module, and an onboard air conditioning adjustment module. The collaborative processing method between the various logic modules is referred to in Figure 2 , each logic module passes Figure 2 The assistance method shown can realize the functions of the above-mentioned air-conditioning control device 11.

[0055] Figure 3 FIG. 1 is a flow chart of an address data processing method according to some exemplary embodiments. In some embodiments, the address data processing method can be applied to Figure 1 The data processing device and electronic device shown can also be applied to other similar devices.

[0056] like Figure 3 As shown, the vehicle air conditioning control method provided by the embodiment of the present disclosure includes the following S201-S203.

[0057] S201: The air conditioning control device obtains current vehicle data at a current moment and historical vehicle data of a preset time period before the current moment, and obtains vehicle data to be processed based on the current vehicle data and the historical vehicle data.

[0058] Vehicle data reflects the vehicle's operating status and the vehicle's internal and external environment. The vehicle data to be processed is time-series. For example, the vehicle data may include one or more of vehicle speed, vehicle air conditioning operating parameters, vehicle location, vehicle interior temperature, and vehicle interior humidity.

[0059] As one possible implementation, the air conditioning control device uses a sensor to obtain current vehicle data at the current moment and historical vehicle data from a preset time period before the current moment. The air conditioning control device sorts the current vehicle data and historical vehicle data in chronological order to obtain first data. The air conditioning control device then cleans the historical vehicle data to obtain second data; the second data is continuous. Furthermore, the air conditioning control device concatenates the first data and the second data to obtain vehicle data to be processed.

[0060] Exemplarily, the air conditioning control device can implement S201 through a data preprocessing module. Specifically, the data preprocessing module is responsible for processing the original sensor data. First, it ensures the uniqueness of the data. Due to transmission or other reasons, there may be repeated content in the original data, so it is necessary to deduplicate the data to ensure the uniqueness of the data. Second, it handles missing values ​​in the data. Since the sensor data comes from different channels, it cannot be guaranteed that all required sensor data can be obtained at the same time. The air conditioning control device can fill in the missing values ​​in chronological order, that is, search forward for N data, fill in the values ​​according to the average value if they are continuous values, and fill in the values ​​according to the mode if they are discrete values.

[0061] S202: The air conditioning control device determines a feature vector of the vehicle data to be processed.

[0062] Among them, the eigenvector includes a first eigenvector and a second eigenvector; the first eigenvector is used to reflect the changes in the vehicle's operating status over time and the changes in the vehicle's internal and external environment over time; the second eigenvector is used to reflect the correlation between the vehicle's operating status and the vehicle's internal and external environment at the current moment.

[0063] As one possible implementation, the air conditioning control device clusters the vehicle data to be processed to obtain multiple data sets, wherein the similarity between each data set is greater than or equal to a preset threshold. The air conditioning control device then performs feature extraction on each data set to obtain an initial feature vector. Furthermore, the air conditioning control device performs a first transformation on the initial feature vector to obtain a first feature vector, and then performs a second transformation on the initial feature vector to obtain a second feature vector.

[0064] Optionally, the air conditioning control device inputs the initial feature vector into a preset deep cross network to perform a first transformation on the initial feature vector through the deep cross network and output a first feature vector.

[0065] Optionally, the air conditioning control device inputs the initial feature vector into a preset multi-layer perceptron, so as to perform a second transformation on the initial feature vector through the multi-layer perceptron and output a second feature vector.

[0066] It should be noted that since the vehicle data to be processed contains both time series features and original sensor features, the vehicle data to be processed passes through the time series feature module and the original sensor feature module respectively, and is transmitted to the feature self-embedding module for feature extraction. Furthermore, the feature self-embedding module is responsible for mapping the features accepted by the machine learning model and generating the feature embedding vector (Feature Embedding Vector) required by the subsequent modules. The original discrete features are often represented by one-hot encoding and then converted into feature embedding vectors through the feature embedding matrix. Continuous features need to be discretized. Discretization of continuous features can speed up the iteration speed of the model, improve the robustness of the model, and improve the nonlinear expression ability of the model. The feature self-embedding module can capture the relationship between continuous features and embedding vectors. Through automatic multi-box discretization, the module can automatically model continuous features and obtain the final feature embedding vector.

[0067] For example, Figure 4 As shown in the figure, a schematic diagram of the self-discretization of continuous features in the feature self-embedding module is shown. Where x is a continuous feature (i.e., the feature corresponding to the vehicle data to be processed), which can be expressed as: x = [x1, x2, ..., x n , Δx1, Δx2, ..., Δx m ]

[0068] where x i is the original continuous feature, Δx i It is the time series feature constructed by the data preprocessing module.

[0069] The air conditioning control device has two layers of neural networks with short circuits so that the features are dispersed to H i In a bucket, this process can be described as:

[0070] h j =LeakyReLU(w j x j )

[0071]

[0072] in and is the learnable parameter of the self-discrete continuous feature j, and the final result Representative features in H j The projection on the buckets is performed, and finally the final feature embedding vector is obtained through a unified Embedding matrix.

[0073] The original discrete features can be directly converted into feature vectors through the embedding matrix, e i∈R d is the embedding vector of the i-th feature after transformation, and d is the length of the feature embedding vector. Therefore, the feature embedding vector output by the final feature embedding module can be expressed as:

[0074] e=[e1,e2,...,e m+k ]

[0075] Where e indicates that the original input has m continuous features and k discrete features, and finally m+k feature embedding vectors are obtained.

[0076] Figure 5 This is a schematic diagram of the structure of the deep cross network. The feature embedding vector e output by the feature embedding module is fed into the deep cross network. Then the calculation process of the l+1th layer deep cross network can be described as:

[0077] x l+1 =x0⊙(W l x l +b l )+x l

[0078] Where W l It is the learnable parameter of the lth layer, where X0 is the feature embedding vector e. After multiple calculations, the final output vector is obtained.

[0079] Figure 6 This is a schematic diagram of the structure of a multi-layer perceptron. Assuming that the network has a total of l layers, the forward calculation process can be described as:

[0080] z (1) =σ(W (1) e+b (1) )

[0081] …

[0082] z （l) =σ(W (l) z (l-1) +b (l) )

[0083] where z (l) , W (l) , b (l) are the output, weight matrix and bias value of the lth layer respectively, and σ is the activation function.

[0084] As can be understood, feature extraction of the vehicle data to be processed mainly consists of two parts: explicit feature interaction and implicit feature interaction. The explicit feature interaction part uses a deep cross network (DCN). DCN uses the idea of ​​residual network to achieve explicit modeling of features, thereby obtaining more interpretable high-order feature interactions. The implicit feature interaction part uses a tower-like multilayer perceptron (MLP) to model the interactions between different features, thereby implicitly capturing high-order feature interactions. At the same time, due to the addition of time series features, DCN can pay more attention to the changes in the features at the most recent moment, allowing the MLP to focus on the interaction of each feature at the current moment.

[0085] S203 : The air-conditioning control device obtains a vehicle air-conditioning control strategy based on the first eigenvector, the second eigenvector, and a pre-trained vehicle air-conditioning control model, and controls the vehicle air-conditioning according to the vehicle air-conditioning control strategy.

[0086] As a possible implementation, the air conditioning control device inputs the first eigenvector and the second eigenvector into a pre-trained vehicle air conditioning control model to obtain a vehicle air conditioning control strategy. Further, the air conditioning control device controls the vehicle air conditioning according to the vehicle air conditioning control strategy.

[0087] Exemplarily, the air conditioning control device can implement S203 via an onboard air conditioning adjustment module. Specifically, the onboard air conditioning adjustment module is responsible for converting the abstract prediction results of the explicit feature interaction and implicit feature interaction components in the feature interaction module into actual air conditioning operation. For classification problems such as air conditioning airflow mode, the onboard air conditioning adjustment module outputs the probabilities of different classification results. For regression problems such as air conditioning temperature, the onboard air conditioning adjustment module outputs the specific numerical value predicted by the model.

[0088] In practical applications, the vehicle air conditioning adjustment module is responsible for converting the results of the final deep cross network and multi-layer perceptron into specific operations. For classification problems such as blowing mode, this process can be described as follows:

[0089]

[0090] Where m represents the dimension of the output of DCN and MLP after the connection, and n represents the final number of classifications. It is the probability value of each category currently output by the model.

[0091] For regression problems, additional processing is required, which can be described as:

[0092]

[0093] where yi Indicates the value of the category. Represents the probability of the category, and the final result can be obtained by multiplying and accumulating This is the result predicted by the model.

[0094] The technical solution provided by the embodiment of the present application brings at least the following beneficial effects: (1) The present application obtains the current vehicle data at the current moment and the historical vehicle data of the preset time period before the current moment, and obtains the vehicle data to be processed based on the current vehicle data and the historical vehicle data. Since the vehicle data is used to reflect the vehicle operating status and the internal and external environment of the vehicle, the obtained vehicle data to be processed can reflect the changes in the vehicle operating status and the internal and external environment of the vehicle over a period of time, and the vehicle data to be processed has a time sequence. Since the vehicle data to be processed includes not only the information of the data itself, but also the change information of the data in the time sequence. Therefore, the present application can determine the first eigenvector and the second eigenvector of the vehicle data to be processed, wherein the first eigenvector is used to reflect the changes in the vehicle operating status over time and the changes in the internal and external environment of the vehicle over time; the second eigenvector is used to reflect the correlation between the vehicle operating status and the internal and external environment of the vehicle at the current moment. Furthermore, the present application obtains the vehicle air-conditioning control strategy based on the first eigenvector, the second eigenvector and the pre-trained vehicle air-conditioning control model, and controls the vehicle air-conditioning according to the vehicle air-conditioning control strategy. Compared with the existing technology that only focuses on the characteristics of the data itself (i.e., numerical characteristics) when using machine learning models, this application not only focuses on the characteristics of the data itself, but also focuses on the connections between the data, and mines the implicit time series characteristics, so as to derive a better air-conditioning control strategy, thereby improving the passenger experience.

[0095] (2) The present application sorts the current vehicle data and the historical vehicle data in chronological order, so that the first data obtained has a time sequence. Furthermore, the present application cleans the historical vehicle data to obtain the second data. After cleaning, the historical vehicle data can be deduplicated, filled in, etc., so that the second data obtained has continuity. In this way, the present application splices the first data with the second data, and the obtained vehicle data to be processed is not only time-series, but also complete and continuous, which is convenient for subsequent feature extraction.

[0096] (3) The present application clusters the vehicle data to be processed to obtain multiple data sets, wherein the similarity between each data in a data set is greater than or equal to a preset threshold. Furthermore, the present application only needs to extract features from each data set, which reduces the processing pressure. The obtained initial feature vector is not used directly, but the initial feature vector is subjected to a first transformation to obtain a first feature vector; the initial feature vector is subjected to a second transformation to obtain a second feature vector, so that the extracted features are more refined.

[0097] (4) This application inputs the initial feature vector into a preset deep cross network to perform a first transformation on the initial feature vector through the deep cross network and output a first feature vector. Since the deep cross network uses the idea of ​​the residual network to achieve explicit modeling of features, a more explanatory high-order feature cross can be obtained. At the same time, due to the addition of time series features, the deep cross network can pay more attention to the changes in the features at the most recent moment.

[0098] (5) The present application inputs the initial feature vector into a preset multilayer perceptron, which performs a second transformation on the initial feature vector and outputs a second feature vector. Since the multilayer perceptron models the interaction between different features, the multilayer perceptron focuses on the interaction between each feature at the current moment.

[0099] The above embodiments mainly introduce the solutions provided by the embodiments of the present application from the perspective of the device (equipment). It is understandable that in order to implement the above method, the device or equipment includes hardware structures and / or software modules corresponding to the execution of each method flow, and these hardware structures and / or software modules corresponding to the execution of each method flow can constitute a material information determination device. Those skilled in the art should easily realize that, in combination with the algorithm steps of each example described in the embodiments invented herein, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0100] In the embodiment of the present application, the functional modules of the device or equipment can be divided according to the above method examples. For example, the device or equipment can be divided into functional modules corresponding to each function, or two or more functions can be integrated into one processing module. The above-mentioned integrated modules can be implemented in the form of hardware or in the form of software functional modules. It should be noted that the division of modules in the embodiment of the present application is schematic and is only a logical functional division. There may be other division methods in actual implementation.

[0101] Figure 7 FIG. 1 is a structural diagram of an air conditioning control device according to an exemplary embodiment. Figure 7 As shown, the air conditioning control device 30 provided in the embodiment of the present application includes an acquisition unit 301 , a determination unit 302 and a processing unit 303 .

[0102] The acquisition unit 301 is used to obtain the current vehicle data at the current moment and the historical vehicle data of the preset time period before the current moment, and obtain the vehicle data to be processed based on the current vehicle data and the historical vehicle data; the vehicle data is used to reflect the vehicle operating status and the internal and external environment of the vehicle; the vehicle data to be processed has time series; the determination unit 302 is used to determine the feature vector of the vehicle data to be processed; the feature vector includes a first feature vector and a second feature vector; the first feature vector is used to reflect the changes in the vehicle operating status over time and the changes in the internal and external environment of the vehicle over time; the second feature vector is used to reflect the correlation between the vehicle operating status and the internal and external environment of the vehicle at the current moment; the processing unit 303 is used to obtain the vehicle air-conditioning control strategy based on the first feature vector, the second feature vector and a pre-trained vehicle air-conditioning control model, and control the vehicle air-conditioning according to the vehicle air-conditioning control strategy.

[0103] Optionally, the acquisition unit 301 is specifically used to: sort the current vehicle data and historical vehicle data in chronological order to obtain first data; clean the historical vehicle data to obtain second data; the second data has continuity; and splice the first data and the second data to obtain the vehicle data to be processed.

[0104] Optionally, the determination unit 302 is specifically used to: cluster the vehicle data to be processed to obtain multiple data sets; the similarity between each data in a data set is greater than or equal to a preset threshold; perform feature extraction on each data set to obtain an initial feature vector; perform a first transformation on the initial feature vector to obtain a first feature vector; perform a second transformation on the initial feature vector to obtain a second feature vector.

[0105] Optionally, the determining unit 302 is specifically configured to: input the initial feature vector into a preset deep cross network, perform a first transformation on the initial feature vector through the deep cross network, and output a first feature vector.

[0106] Optionally, the determining unit 302 is specifically configured to: input the initial feature vector into a preset multilayer perceptron, perform a second transformation on the initial feature vector through the multilayer perceptron, and output a second feature vector.

[0107] Figure 8 This is a schematic diagram of the structure of a vehicle provided by this application. Figure 8 The vehicle 40 may include at least one processor 401 and a memory 402 for storing processor-executable instructions, wherein the processor 401 is configured to execute instructions in the memory 402 to implement the vehicle air conditioning control method in the above embodiment.

[0108] In addition, the vehicle 40 may further include a communication bus 403 and at least one communication interface 404 .

[0109] The processor 401 may be a central processing unit (CPU), a microprocessor unit, an ASIC, or one or more integrated circuits for controlling the execution of the program of the present application.

[0110] The communication bus 403 may include a pathway for transmitting information between the aforementioned components.

[0111] The communication interface 404 uses any transceiver or other device for communicating with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area network (WLAN), etc.

[0112] The memory 402 may be a read-only memory (ROM) or other type of static storage device that can store static information and instructions, a random access memory (RAM) or other type of dynamic storage device that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, an optical disc storage (including a compact disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory may exist independently and be connected to the processor 401 via a bus. The memory may also be integrated with the processor 401.

[0113] The memory 402 is used to store instructions for executing the solution of the present application, and the execution is controlled by the processor 401. The processor 401 is used to execute the instructions stored in the memory 402, thereby realizing the functions of the method of the present application.

[0114] As an example, combining Figure 7 The functions implemented by the acquisition unit 301, the determination unit 302 and the processing unit 303 in the air conditioning control device 30 are the same as those implemented by the Figure 8 The function of the processor 401 in is the same.

[0115] In a specific implementation, as an embodiment, the processor 401 may include one or more CPUs, such as Figure 8CPU0 and CPU1 in.

[0116] In a specific implementation, as an embodiment, the vehicle 40 may include multiple processors, such as Figure 8 4 and 5. The processors 401 and 407 are shown in FIG. Each of these processors may be a single-CPU processor or a multi-CPU processor. A processor herein may refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions).

[0117] In a specific implementation, as an embodiment, the vehicle 40 may further include an output device 405 and an input device 406. The output device 405 communicates with the processor 401 and can display information in a variety of ways. For example, the output device 405 can be a liquid crystal display (LCD), a light emitting diode (LED) display device, a cathode ray tube (CRT) display device, or a projector. The input device 406 communicates with the processor 401 and can accept user input in a variety of ways. For example, the input device 406 can be a mouse, a keyboard, a touch screen device, or a sensor device.

[0118] Those skilled in the art will understand that Figure 8 The structure shown in the figure does not constitute a limitation of the vehicle 40, and the vehicle 40 may include more or fewer components than shown, or combine certain components, or adopt a different arrangement of components.

[0119] In addition, the present application also provides a computer-readable storage medium. When the instructions in the computer-readable storage medium are executed by the processor of the vehicle, the vehicle can execute the vehicle air conditioning control method provided in the above embodiment.

[0120] In addition, the present application also provides a computer program product, including computer instructions. When the computer instructions are executed on a vehicle, the vehicle executes the vehicle air conditioning control method provided in the above embodiment.

[0121] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the inventions herein. This application is intended to cover any variations, uses, or adaptations of the present application that follow the general principles of the present application and include common knowledge or customary techniques in the art not invented herein. The description and examples are to be considered merely as exemplary, and the true scope and spirit of the present application are indicated by the claims.

Claims

1. A vehicle air conditioning control method, characterized in that: The method comprises: Acquire current vehicle data at the current moment and historical vehicle data for a preset time period before the current moment, and obtain vehicle data to be processed based on the current vehicle data and the historical vehicle data; the vehicle data is used to reflect the vehicle's operating status and the vehicle's internal and external environment; the vehicle data to be processed has a time series nature; Determine the feature vector of the vehicle data to be processed; the feature vector includes a first feature vector and a second feature vector; the first feature vector is used to reflect the change of the vehicle operating state over time and the change of the internal and external environment of the vehicle over time; the second feature vector is used to reflect the correlation between the vehicle operating state and the internal and external environment of the vehicle at the current moment; the first feature vector is obtained by inputting an initial feature vector into a preset deep cross network to perform a first transformation on the initial feature vector through the deep cross network, and the second feature vector is obtained by inputting the initial feature vector into a preset multilayer perceptron to perform a second transformation on the initial feature vector through the multilayer perceptron, and the initial feature vector is obtained based on the vehicle data to be processed; Based on the first eigenvector, the second eigenvector and a pre-trained vehicle air-conditioning control model, a vehicle air-conditioning control strategy is obtained, and the vehicle air-conditioning is controlled according to the vehicle air-conditioning control strategy.

2. The vehicle air conditioning control method according to claim 1, characterized in that: The obtaining the vehicle data to be processed according to the current vehicle data and the historical vehicle data includes: Sorting the current vehicle data and the historical vehicle data in chronological order to obtain first data; Cleaning the historical vehicle data to obtain second data; the second data has continuity; The first data and the second data are concatenated to obtain the vehicle data to be processed.

3. The vehicle air conditioning control method according to claim 1, characterized in that: The determining of the feature vector of the vehicle data to be processed includes: Clustering the vehicle data to be processed to obtain multiple data sets; the similarity between each data in a data set is greater than or equal to a preset threshold; Performing feature extraction on each of the data sets to obtain an initial feature vector; Performing a first transformation on the initial eigenvector to obtain the first eigenvector; Perform a second transformation on the initial eigenvector to obtain the second eigenvector.

4. A vehicle air conditioning control device, characterized in that: The device includes an acquisition unit, a determination unit, and a processing unit; The acquisition unit is configured to acquire current vehicle data at a current moment and historical vehicle data for a preset time period before the current moment, and obtain vehicle data to be processed based on the current vehicle data and the historical vehicle data; the vehicle data is configured to reflect the vehicle's operating status and the vehicle's internal and external environment; and the vehicle data to be processed is time-series. The determining unit is used to determine the feature vector of the vehicle data to be processed; the feature vector includes a first feature vector and a second feature vector; the first feature vector is used to reflect the change of the vehicle operation state and the change of the internal and external environment of the vehicle over time; the second feature vector is used to reflect the correlation between the vehicle operation state and the internal and external environment of the vehicle at the current moment; the first feature vector is obtained by inputting an initial feature vector into a preset deep cross network to perform a first transformation on the initial feature vector through the deep cross network, and the second feature vector is obtained by inputting the initial feature vector into a preset multilayer perceptron to perform a second transformation on the initial feature vector through the multilayer perceptron, and the initial feature vector is obtained based on the vehicle data to be processed; The processing unit is used to obtain a vehicle air conditioning control strategy based on the first eigenvector, the second eigenvector and a pre-trained vehicle air conditioning control model, and control the vehicle air conditioning according to the vehicle air conditioning control strategy.

5. The vehicle air conditioning control device according to claim 4, characterized in that: The acquisition unit is specifically configured to: Sorting the current vehicle data and the historical vehicle data in chronological order to obtain first data; Cleaning the historical vehicle data to obtain second data; the second data has continuity; The first data and the second data are concatenated to obtain the vehicle data to be processed.

6. The vehicle air conditioning control device according to claim 4, characterized in that: The determining unit is specifically configured to: Clustering the vehicle data to be processed to obtain multiple data sets; the similarity between each data in a data set is greater than or equal to a preset threshold; Performing feature extraction on each of the data sets to obtain an initial feature vector; Performing a first transformation on the initial eigenvector to obtain the first eigenvector; Perform a second transformation on the initial eigenvector to obtain the second eigenvector.

7. A vehicle, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the vehicle air conditioning control method according to any one of claims 1 to 3.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the vehicle air conditioning control method according to any one of claims 1 to 3.

Images