Vehicle function recommendation method, electronic device, and computer storage medium

Abstract

The application provides a vehicle function recommendation method, an electronic device and a computer storage medium. The method comprises the following steps: obtaining a current vehicle state of a vehicle; receiving first use data of the vehicle in a historical period when the vehicle is in the vehicle state; obtaining second use data according to the first use data, wherein the second use data represents data of each target function preferred to be used by a user in the historical period; obtaining third use data according to the first use data, wherein the third use data represents data of each target function predicted to be preferred to be used by the user; calculating a use probability value of each target function according to the second use data and the third use data; determining a target probability value greater than a preset probability threshold value in the plurality of use probability values; and recommending a target function matched with the target probability value. The application can accurately recommend a target function meeting the user's preference according to the historical use preference of the target function in the historical period.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicles, and particularly relates to a vehicle function recommendation method, an electronic device and a computer storage medium. BACKGROUND

[0002] With the continuous development of social economy and the technical field of vehicles, more and more families purchase vehicles. Compared with public transport tools, vehicles can directly reach the destination from the departure place, save the waiting and transfer time, improve the travel efficiency, and provide convenience and comfort for people's life and work.

[0003] At present, vehicles are provided with various target functions, and users can use different target functions to meet the needs of entertainment, relaxation and the like. For example, a user can perform seat massage in the vehicle, can turn on the atmosphere lamp when parking, can listen to music when parking, and can wish to turn on the air conditioner before getting on the vehicle. However, there is a problem that it is relatively cumbersome for a user to start some target functions, resulting in poor use experience. SUMMARY

[0004] Therefore, the present application provides a vehicle function recommendation method, an electronic device and a computer storage medium, which can solve the problem that it is relatively cumbersome for a user to start some target functions, resulting in poor use experience.

[0005] The vehicle function recommendation method provided in the present application embodiment comprises: acquiring a vehicle state in which a vehicle currently locates; receiving first use data of the vehicle in a historical period when the vehicle locates in the vehicle state, the first use data being a number of times that each target function of the vehicle is used in the historical period; obtaining second use data according to the first use data, the second use data representing data that each target function is preferentially used by a user in the historical period; obtaining third use data according to the first use data, the third use data representing data that each target function is predicted to be preferentially used by the user; calculating a use probability value of each target function according to the second use data and the third use data; determining a target probability value greater than a preset probability threshold value in a plurality of use probability values; and recommending the target function matched with the target probability value.

[0006] Compared with the related art, the present application embodiment has at least the following advantages:

[0007] The vehicle state in which the vehicle currently is, and the first use data of the vehicle in the historical period when the vehicle is in the vehicle state are acquired. The second use data of each target function being preferred to be used by the user in the historical period and the third use data of each target function being predicted to be preferred to be used by the user are obtained from the first use data. The use probability value of each target function is calculated through the second use data and the predicted third use data, and the target probability value greater than the preset probability threshold value is determined from the plurality of use probability values. Therefore, based on the target probability value, the target function meeting the use preference of the user is accurately recommended to the user, so that the user can directly use the preferred target function without performing cumbersome operation steps, thereby improving the use experience of the user on the vehicle.

[0008] In some possible implementation manners, the first use data includes historical cross data, historical numerical data, historical classification data and function data, the historical cross data is frequency data of each target function being used in the case that the vehicle is in different historical positions and historical times, the historical numerical data is frequency data of each target function being used, the historical classification data is data obtained by classifying the historical positions and the historical times based on a preset classification rule, and the function data is name information of each target function; the second use data is obtained according to the historical cross data; and the third use data is obtained according to the historical numerical data, the historical classification data and the function data.

[0009] In some possible implementation manners, the method predicts the target function displayed in the vehicle by using a prediction model, and the prediction model includes a first model and a second model; the method further includes: acquiring a current time of the vehicle; the second use data is obtained according to the historical cross data, including: inputting the historical cross data and the current time into the first model to obtain the second use data; and the third use data is obtained according to the historical numerical data, the historical classification data and the function data, including: inputting the historical numerical data, the historical classification data, the function data and the current time into the second model to obtain the third use data.

[0010] In some possible implementation manners, the method employs a prediction model to predict the target function displayed in the vehicle, the prediction model comprising a first model and a second model; the method further comprises: obtaining a current location where the vehicle is located; the obtaining the second use data according to the historical intersection data comprises: inputting the historical intersection data and the current location into the first model to obtain the second use data; and the obtaining the third use data according to the historical numerical data, the historical classification data and the function data comprises: inputting the historical numerical data, the historical classification data, the function data and the current location into the second model to obtain the third use data.

[0011] In some possible implementation manners, the method employs a prediction model to predict the target function displayed in the vehicle, the prediction model comprising a first model and a second model; the method further comprises: obtaining a current location where the vehicle is located; the obtaining the second use data according to the historical intersection data comprises: inputting the historical intersection data and the current location into the first model to obtain the second use data; and the obtaining the third use data according to the historical numerical data, the historical classification data and the function data comprises: inputting the historical numerical data, the historical classification data, the function data and the current location into the second model to obtain the third use data.

[0012] In some possible implementation manners, the vehicle state comprises a first state, a second state and a third state; the first state is a state of the vehicle in response to an unlocking instruction, the second state is a state of the vehicle in a driving process, and the third state is a state of the vehicle in a parking process.

[0013] The second aspect of the present application discloses a vehicle function recommendation method applied to a vehicle, the vehicle comprising a plurality of target functions, the method comprising: sending a vehicle state currently taken by the vehicle to a cloud server, wherein the cloud server is configured to receive first usage data corresponding to each of the target functions, the first usage data being a number of times each of the target functions is used when the vehicle is in the vehicle state in a historical period; obtaining second usage data according to the first usage data, the second usage data representing data of each of the target functions being preferred to be used by a user in the historical period; obtaining third usage data according to the first usage data, the third usage data representing data of each of the target functions being predicted to be preferred to be used by the user; calculating a usage probability value of each of the target functions according to the second usage data and the third usage data; determining a target probability value greater than a preset probability threshold in the plurality of usage probability values; receiving the target function matching the target probability value; and displaying a recommendation list, wherein the recommendation list is at least one target function matching the target probability value determined based on the vehicle state, and the recommendation list is the target function matching the target probability value.

[0014] In some possible implementation manners, after the recommendation list is displayed, the method further comprises: in response to a function selection instruction for at least one target function in the recommendation list, performing a function operation corresponding to the selected target function.

[0015] The third aspect of the present application discloses an electronic device, comprising a processor and a memory, the memory being configured to store instructions, and the processor being configured to invoke the instructions in the memory, so that the electronic device performs the vehicle function recommendation method described above.

[0016] The fourth aspect of the present application discloses a computer storage medium comprising computer instructions, when the computer instructions run on an electronic device, the electronic device performs the vehicle function recommendation method described above.

[0017] It can be understood that the vehicle function recommendation method of the second aspect, the electronic device of the third aspect and the computer storage medium of the fourth aspect provided above all correspond to the method of the first aspect described above, and therefore the beneficial effects achieved thereby can refer to the beneficial effects of the corresponding method provided above, which will not be described here again. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 is a flowchart of a vehicle function recommendation method according to an embodiment of the present application.

[0019] Figure 2is a schematic diagram of an interaction between a vehicle and a cloud according to an embodiment of the present application.

[0020] Figure 3 is a schematic diagram of a structure of a Wide&Deep model according to an embodiment of the present application.

[0021] Figure 4 is a schematic diagram of an architecture of a function recommendation system according to an embodiment of the present application.

[0022] Figure 5 is another schematic diagram of an interaction between a vehicle and a cloud according to an embodiment of the present application.

[0023] Figure 6 is a schematic diagram of a hardware structure of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0024] In order to more clearly understand the above objectives, features and advantages of the present application, the following will be described in detail in connection with the attached drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict, if possible.

[0025] In the following description, a large number of specific details are set forth in order to facilitate a thorough understanding of the present application. The described embodiments are only some of the embodiments of the present application, and are not all the embodiments.

[0026] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in the description herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the present application.

[0027] Further, it should be noted that herein, the terms "comprising", "including", or any other variant thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements recited, but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without more limitations, an element defined by the phrase "comprising a" does not exclude the existence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0028] In this application, "at least one" means one or more, and "more than one" means two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and drawings of this application are used to distinguish similar objects, not to describe a specific order or sequence.

[0029] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0030] Vehicles are equipped with various target functions to cater to different application scenarios. For example, users may want to enjoy seat massage, turn on ambient lighting and listen to music while parked, turn on the air conditioning before getting in the car, or plan navigation routes in advance. However, there are issues where activating certain target functions can be cumbersome, resulting in a poor user experience.

[0031] Based on this, please refer to Figure 1 This application provides a vehicle function recommendation method. The vehicle includes multiple target functions. This method combines historical user preferences for these target functions and recommends potentially interesting functions based on the user's historical usage habits. This allows the user to directly operate the recommended functions, increasing the convenience of operation. This vehicle function recommendation method can be applied to a function recommendation system, which can be deployed in the cloud. The following description uses a cloud-based function recommendation system as an example. Specifically, the function recommendation system is deployed on a cloud server. The cloud communicates with the in-vehicle equipment installed in the vehicle. In this embodiment, the in-vehicle equipment is an in-vehicle computer. In other embodiments, the in-vehicle equipment can be other electronic devices besides an in-vehicle computer, as long as it can communicate with the cloud. The function recommendation system can also be deployed in devices other than the cloud; this application does not limit this.

[0032] Cloud is a software platform using application virtualization technology (Application Virtualization), which integrates software search, download, use, management, backup and other functions. Through the platform, various commonly used software can be packaged in an independent virtual environment, so that the application software will not be coupled with the system, and the purpose of green software use is achieved. At present, more and more vehicle manufacturers also choose to arrange part of the service in the cloud to realize seamless connection between multiple devices, devices and the cloud, and shield the differentiation of different device terminal capabilities. Thus, the user experience is improved.

[0033] The cloud can include a standalone server, or a server network or server cluster composed of servers. For example, the cloud described in the embodiments of the present application includes but is not limited to a computer, a network host, a single network server, a plurality of network server sets, or a cloud server composed of a plurality of servers. The cloud server is composed of a large number of computers or network servers based on cloud computing.

[0034] The vehicle function recommendation method includes the following steps:

[0035] Step 101, obtaining the current vehicle state of the vehicle.

[0036] In some embodiments, in response to a service request of the vehicle, the service request includes the vehicle state and the vehicle ID. The vehicle ID is used to uniquely identify the vehicle, so as to facilitate subsequent recommendation of the target function to the vehicle sending the service request. The vehicle state includes a first state, a second state and a third state; the first state is the state of the vehicle in response to the unlocking instruction, that is, when the vehicle is in the first state, the user can get on and off at any time. The second state is the state of the vehicle in the driving process, and the third state is the state of the vehicle when the vehicle is in the parking state, that is, when the gear of the vehicle is in the P gear, the state of the vehicle. In other embodiments, the vehicle state can also include only one state, for example, the vehicle state only includes the first state, and the type of the vehicle state is not limited in the present application.

[0037] In some embodiments, the current time, the current location, the environmental data and the running data of the vehicle in different vehicle states can also be obtained. The environmental data includes the temperature in the vehicle and the temperature outside the vehicle, and the running data includes the state of the window, the state of the door, etc. According to the actual design requirements, the specific content of the running data is set.

[0038] For example, when the vehicle is in the first state or the third state, the current time, the current position where the vehicle is located, the environment data and the running data are acquired, and at this time, the current time can be a time point. When the vehicle is in the second state, the current time, the current position where the vehicle is located, the environment data and the running data are acquired, and at this time, the current time can be a time period, and the current position can include a plurality of positions matched with the running track of the vehicle.

[0039] In step 102, first usage data corresponding to each target function is received.

[0040] In some embodiments, the first usage data is the number of times that each target function is used when the vehicle is in different vehicle states in a historical period. The target function can include a music function, a navigation function, an ambient light mode function, an ambient light opening function, an ambient light closing function, a seat massage opening function, a seat massage closing function, a suspension height setting function, a one-key screen-off function, a rearview mirror folding function, a child lock opening function, a child lock closing function, a window closing function, a window opening function, a trunk opening function, etc. In this embodiment, the first usage data includes the number of times that one or more target functions are operated by the user when the vehicle is in any one of the first state, the second state or the third state, and is located at different historical positions and is at different historical times in the historical period.

[0041] The first usage data of a plurality of different vehicles is stored in the cloud server. For example, the first usage data can be stored in a data warehouse. In order to accurately recommend the target function preferred by the user based on the first usage data, the frequency of use of each target function of the vehicle in the historical period needs to be acquired based on the vehicle ID and the current vehicle state of the vehicle. Specifically, in the case where the vehicle state is detected as the first state, the first usage data when the vehicle is in the first state in the historical period is acquired. Similarly, in the case where the vehicle state is detected as the second state, the first usage data when the vehicle is in the second state in the historical period is acquired. In the case where the vehicle state is detected as the third state, the first usage data when the vehicle is in the third state in the historical period is acquired.

[0042] In this embodiment, please refer to Figure 2 The cloud server can store the first usage data after data preprocessing to the data warehouse using the kafka message channel. The kafka message channel has the characteristics of high throughput, distribution, storage persistence, high availability, etc. At the same time, the first usage data is stored in the data warehouse in the cloud. In order to continuously obtain the recommended target function by using the first usage data in the later period.

[0043] Further, the cloud first performs data preprocessing operation on the first usage data and constructs a prediction model. In this embodiment, the prediction model can be a Wide&Deep model. The data preprocessing includes data analysis, data filtering, data correlation and data statistics. The data analysis, data filtering, data correlation and data statistics all belong to the prior art, and will not be described here. The specific structure of the Wide&Deep model will be described in detail below, and will not be described here.

[0044] In step 103, the second usage data is obtained according to the first usage data.

[0045] In some embodiments, the second usage data is used to represent the data that each target function is preferred to be used by the user in the historical period. The prediction model includes a first model and a second model, wherein the first model is a Wide model in the Wide&Deep model, and the second model is a Deep model in the Wide&Deep model.

[0046] In order to obtain accurate function recommendation results subsequently, the first usage data and the current time are input into the Wide model to obtain the second usage data. Alternatively, the first usage data and the current location can also be input into the Wide model to obtain the second usage data. Alternatively, the first usage data, the current time and the current location can also be input into the Wide model to obtain the second usage data. In other embodiments, the first usage data, the current time, the current location, the environment data and the running data can also be input into the Wide model to obtain the second usage data. That is, one or more of the first usage data, the current time, the current location, the environment data and the running data can be input into the Wide model to obtain the second usage data.

[0047] In this embodiment, please refer to Figure 3 In order to obtain the second usage data according to the first usage data and one or more of the current time, the current location, the environment data and the running data, it is necessary to construct the Wide&Deep model first. At the same time, the constructed Wide&Deep model can be stored in a model warehouse. When the Wide&Deep model is needed, the model can be directly called from the model warehouse. The model warehouse is essentially a database, mainly used to store a plurality of different models constructed. When a certain model is used, it can be directly called from the model warehouse, improving the efficiency of model calling.

[0048] The Wide&Deep model is a recommendation system framework combining Wide model and Deep model, aiming to balance the accuracy and diversity of the recommendation results. The model jointly trains the Wide model and the Deep model, uses the memory ability of the Wide model (directly learns the "co-occurrence frequency" in the historical data) and the generalization ability of the Deep model (prediction ability for new samples and unseen feature combinations), so as to realize the balance of the accuracy and diversity of the recommendation results.

[0049] The Wide&Deep model includes a Wide model and a Deep model. The Wide model has memory ability and can accurately recommend by using cross-features to efficiently realize memory function. It realizes certain generalization ability by adding some wide class features, but is limited to training data and cannot generalize to situations that have not appeared in the training data. The Wide model uses a sparse matrix composed of one-hot algorithm to represent the features, which can be single or cross, and needs to be designed artificially.

[0050] The Deep model has generalization ability and can recommend unseen content by learning low-dimensional dense vectors. However, when the model faces sparse data, it may over-generalize and recommend a lot of irrelevant content, thereby affecting the accuracy. The Deep model maps the sparse features of the Wide model to the embedding layer to convert them into a 1-dimensional dense matrix, realizing feature extraction and dimension reduction.

[0051] In this embodiment, for example, the Wide model wants to remember which target function has the maximum matching degree with the current time and the current location. The Wide model predicts a probability value, that is, the probability of the user finally selecting the recommended target function for the current time and the current location and the recommended target function. If the user finally selects the recommended target function, the model will remember that the probability value of the current time and the current location and the recommended target function is high.

[0052] That is, the Wide model is used to process the first use data and one or more of the current time, the current location, the environment data and the running data to obtain the second use data. The memory ability of the Wide&Deep model can be improved.

[0053] Step 104, obtaining third use data according to the first use data.

[0054] In some embodiments, the third usage data represents data that each target function is predicted to be preferred to be used by the user. Similarly, in order to obtain accurate function recommendation results subsequently, the first usage data and the current time are input into the Deep model to obtain the third usage data. Alternatively, the first usage data and the current location can also be input into the Deep model to obtain the third usage data. Alternatively, the first usage data, the current time and the current location can also be input into the Deep model to obtain the third usage data. In other embodiments, the first usage data, the current time, the current location, the environment data and the running data can also be input into the Deep model to obtain the third usage data. That is, one or more of the first usage data, and the current time, the current location, the environment data and the running data can be input into the Deep model to obtain the third usage data.

[0055] In the present embodiment, for example, the Deep model first converts the current time, the current location and the first usage data into low-dimensional vectors, and finds data close to the current time and the current location from the first usage data in the vector space. Thus, the third usage data is obtained.

[0056] That is, the current time, the current location and the first usage data are processed by the Deep model to obtain the third usage data. The generalization ability of the Wide & Deep model can be improved, so that the Wide & Deep model can realize recommendation of unseen content.

[0057] Further, the first usage data includes historical numerical data, historical classification data, function data and historical cross data. The historical cross data is frequency data of each target function being used when the vehicle is in different historical locations and historical times. The historical numerical data is frequency data of each target function of the vehicle being used. The historical classification data is data classified based on a preset classification rule for historical locations and historical times, and the function data is name information of each target function.

[0058] In the present embodiment, the historical numerical data can include frequency data of the user operating one or more target functions at historical times or frequency data of the user operating one or more target functions at historical locations.

[0059] In order to obtain the historical classification data, preset classification rules need to be set. The preset classification rules include a time classification rule and a location classification rule. The time classification rule is used to classify the historical time when the vehicle is in the vehicle state. The location classification rule is used to classify the historical location when the vehicle is in the vehicle state. For example, 24 hours of a day are divided into six stages of early morning, morning, noon, afternoon, evening and night, and the early morning, morning, noon, afternoon, evening and night can be mutually exclusive and cover 24 hours of a day. It is determined that the historical time when the vehicle is in the vehicle state belongs to which one of the early morning, morning, noon, afternoon, evening and night. The location is divided into six types of shopping mall, school, hospital, bank, bookstore and hotel, and it is determined that the historical location when the vehicle is in the vehicle state belongs to which one of the shopping mall, school, hospital, bank, bookstore and hotel. In other embodiments, 24 hours of a day can be divided into 4 stages, 8 stages, etc. according to actual needs, or the location can be divided into other types, which are not limited in the present application.

[0060] The function data can include list information of target functions when the vehicle is in the first state, list information of target functions when the vehicle is in the second state, and list information of target functions when the vehicle is in the third state. When the vehicle is in the first state, the target functions can include a music function, a navigation function, an ambient light mode function, an ambient light opening function, an ambient light closing function, a seat massage opening function, a seat massage closing function, a suspension height setting function, a one-key screen-off function, a rearview mirror folding function, a reading light opening function, a reading light closing function, a window opening function, and a window closing function.

[0061] When the vehicle is in the second state, the target functions can include a music function, a navigation function, an ambient light mode function, an ambient light opening function, an ambient light closing function, a seat massage opening function, a seat massage closing function, a suspension height setting function, a one-key screen-off function, a rearview mirror folding function, a reading light opening function, a reading light closing function, a window opening function, and a window closing function.

[0062] When the vehicle is in the third state, the target functions can include a music function, a navigation function, an ambient light mode function, an ambient light opening function, an ambient light closing function, a seat massage opening function, a seat massage closing function, a suspension height setting function, a one-key screen-off function, a rearview mirror folding function, a child lock opening function, a child lock closing function, a window full closing function, a window full opening function, and a trunk opening function. In other embodiments, the function data can also include other target functions, for example, a rest mode switch, an energy recovery intensity, and a driving mode selection. The scope of the function data is not limited in the present application.

[0063] In this embodiment, the historical cross data can be obtained based on the historical numerical data, the historical classification data and the function data. The historical cross data is a combination of different aspects. Taking the combination of time and location as an example, the historical cross data can include frequency data of using one or more target functions in the morning when the vehicle is in a shopping mall, frequency data of using one or more target functions in the morning when the vehicle is in a school, frequency data of using one or more target functions at noon when the vehicle is in a shopping mall, and the like. That is, the data obtained by crossing the data in the historical numerical data, the historical classification data and the function data with each other is the historical cross data. In this embodiment, the data obtained by crossing the data in the historical numerical data, the historical classification data and the function data with each other is the historical cross data. As long as the frequency information of using one or more target functions at different historical times and historical locations in the historical period can be reflected from the historical cross data.

[0064] In some embodiments, the historical cross data and the current time are input into the Wide model to obtain the second use data. Alternatively, the historical cross data and the current location are input into the Wide model to obtain the second use data. Alternatively, the historical cross data, the current time and the current location are input into the Wide model to obtain the second use data. Alternatively, the historical cross data, the current time, the current location, the environment data and the running data are input into the Wide model to obtain the second use data. That is, one or more of the historical cross data and the current time, the current location, the environment data and the running data are input into the Wide model to obtain the second use data.

[0065] Please continue to refer to Figure 3 , the Wide & Deep model includes a feature input layer, an embedding layer, a hidden layer and an output layer. One or more of the historical cross data and the current time, the current location, the environment data and the running data are input into the feature input layer, and the Wide model processes the data input into the feature input layer to obtain the second use data. Then the second use data is directly input into the output layer. In this way, the Wide model can remember the cross information between the historical location and the historical time in the historical cross data and whether each target function is used. For example, when the frequency data of the seat massage in the historical cross data is 5 times in the morning when the vehicle is in a shopping mall (assuming that the vehicle is in the third state at this time). The Wide model can remember that the user prefers to use the seat massage function in the morning when the vehicle is in a shopping mall (the vehicle is in the third state).

[0066] In the embodiment, the Wide model processes one or more of the historical cross data, the current time, the current location, the environment data and the operation data by using the following formula to obtain the second use data. The formula is as follows:

[0067]

[0068] wherein, x represents the data input to the Wide model by the feature input layer, y represents the data of the output layer, θ represents the feature coefficient, and β represents the constant term.

[0069] In some embodiments, the historical numerical data, the historical classification data, the function data and the current time are input into the Deep model to obtain the third use data. Alternatively, the historical numerical data, the historical classification data, the function data and the current location are input into the Deep model to obtain the third use data. Alternatively, the historical numerical data, the historical classification data, the function data, the current time and the current location are input into the Deep model to obtain the third use data. Alternatively, the historical numerical data, the historical classification data, the function data, the current time, the current location, the environment data and the operation data are input into the Deep model to obtain the third use data. Similarly, that is, one or more of the historical numerical data, the historical classification data, the function data, the current time, the current location, the environment data and the operation data can be input into the Deep model to obtain the third use data.

[0070] In some embodiments, inputting one or more of the historical numerical data, the historical classification data, the function data, the current time, the current location, the environment data and the operation data into the Deep model to obtain the third use data comprises: performing vectorization processing on the historical classification data and one or more of the current time, the current location, the environment data and the operation data to obtain a behavior classification vector. Performing vectorization processing on the function data to obtain a function vector. Obtaining the third use data according to the historical numerical data, the behavior classification vector and the function vector.

[0071] In the embodiment, one or more of the historical numerical data, the historical classification data, the function data, the current time, the current location, the environment data and the operation data are input into the feature input layer. Since the historical classification data and the function data are discrete data, and the historical numerical data is continuous data, in order to enable the subsequent Deep model to convert the continuous data into a low-dimensional vector, and then based on the processing of the low-dimensional vector, thereby improving the generalization ability of the model, the embedding layer converts one or more of the historical classification data, the function data and the current time, the current location, the environment data and the operation data transmitted by the feature input layer into continuous data.

[0072] Further, the historical classification data is converted into continuous data using a one-hot algorithm. The one-hot algorithm is also known as one-bit effective coding, which mainly uses an N-bit state register to encode N states, each state is represented by an independent register bit, and only one bit is effective at any time. The one-hot algorithm is a representation of a categorical variable as a binary vector. The classification value is required to be mapped to an integer value. Then, each integer value is represented as a binary vector. For example, if the time when the vehicle is in the parking state in the historical classification data is "night", this value is not a number and cannot be directly input into the hidden layer, so the historical classification data needs to be one-hot processed. Based on the time classification rule described above, the value of night is set to 1, and the values of the other five features are set to 0, so that the non-numeric to numeric conversion is completed, and can be input into the hidden layer. When the running data also needs to be converted, the data type conversion process of the historical classification data can be referred to. To avoid repetition, it will not be described here.

[0073] Since there are many names of target functions in the function data, and there is a relationship between each target function. For example, there is a relationship between the two target functions of "turning on the atmosphere lamp" and "turning off the atmosphere lamp". Therefore, the data type conversion rule of the one-hot algorithm is not applicable. The Word2Vec model needs to be used to convert the function data to reduce the number of features after vectorization and describe the relationship between functions. Specifically, first, the name of each target function is segmented, for example, "turn on the atmosphere lamp" is divided into "turn on" and "atmosphere lamp". After segmentation, a preset word library is established, that is, the segmented words are collected together, and then the Word2Vec model is used in combination with the preset word library to output the word vector corresponding to each function, and the output dimension of the model can be determined. For example, "turn on the atmosphere lamp", the output dimension is 5, then the Word2Vec model outputs the word vector of "turn on the atmosphere lamp" as ['0.12', '0.94', '0.58', '0.39', '0.22']. In this way, the data type conversion is completed, and the converted data is input into the hidden layer.

[0074] In some embodiments, since the historical numerical data is continuous data, it does not need to be processed by the embedding layer, and can be directly transmitted to the hidden layer. The hidden layer includes multiple layers of first activation functions, and each layer of the first activation function is a ReLu activation function. After the hidden layer receives the historical numerical data, the behavior classification vector and the function vector transmitted by the embedding layer, the historical numerical data, the behavior classification vector and the function vector are processed by multiple layers of first activation functions, and the hidden layer can deeply learn the complex relationship between the data to obtain third use data.

[0075] Step 105, according to the second use data and the third use data, the use probability value of each target function is calculated.

[0076] Specifically, the output layer is provided with a second activation function, and the second activation function is a Sigmoid function. The second use data and the third use data are input into the Sigmoid function to calculate a plurality of use probability values of each target function.

[0077] The output of the Wide&Deep model combines the outputs of the Wide model and the Deep model, and a plurality of use probability values are obtained by using the following formula.

[0078]

[0079] wherein, is a Sigmoid function, represents historical cross data, is a parameter of the Wide model, is a parameter corresponding to the first activation function of the last layer in the Deep model, is a bias parameter.

[0080] In some embodiments, during the training process of the Deep model, the model parameters of the Deep model are optimized using a regularization and batch normalization mechanism, so that the influence of model overfitting can be reduced. The formula of the regularization and batch normalization mechanism is as follows:

[0081]

[0082] wherein, , ,

[0083] x represents data input into the Deep model, y represents data output by the Deep model, represents that the Deep model uses a neural network function, θ represents a parameter in the function, λ is a hyperparameter that needs to be adjusted according to the model effect, and the smaller the hyperparameter is, the simpler the neural network is, and the less likely overfitting occurs. m is the batch sample size.

[0084] Step 106, determining a target probability value greater than a preset probability threshold value in the plurality of use probability values.

[0085] wherein, the preset probability threshold value can be set to 90%, 92% or 95%, and can be set according to actual needs. The application does not limit the specific value of the preset probability threshold value.

[0086] Step 107, recommending a target function matched with the target probability value.

[0087] In this embodiment, target functions matching the target probability value are recommended to the vehicle. After receiving a target function matching the target probability value, the vehicle can reasonably judge the recommended target function based on its own vehicle status, thereby obtaining a recommendation list, and then recommending the predicted user preference list to the user. The vehicle status also includes the operational status of multiple vehicle components, such as the trunk, windows, rearview mirrors, and seats.

[0088] For example, a vehicle's ability to reasonably determine recommended target functions based on its own vehicle status means that: when the vehicle is in a parking space and there are obstacles behind it preventing the trunk from being opened, the target function of opening the trunk should not be recommended. Similarly, when the vehicle is parked and the trunk is piled with many items, preventing the rear seats from activating the massage function, the target function of activating the rear seat massage should not be recommended.

[0089] In other embodiments, data on audio and video playback by the user over a historical period can be processed using the Wide&Deep model to recommend audio and video content that the user may be interested in. Alternatively, navigation data by the user over a historical period can be processed using the Wide&Deep model to recommend navigation routes, etc.

[0090] Please combine Figure 4 , Figure 4 It includes three aspects: interactive data, offline training, and online push. Interactive data includes initial usage data, a specific vehicle control pool, and a prediction model. The specific vehicle control pool is a data pool matching the vehicle and its current vehicle state. The specific vehicle control pool includes a first vehicle control pool, a second vehicle control pool, and a third vehicle control pool. The first vehicle control pool stores the initial usage data when the vehicle is in its first state; the second vehicle control pool stores the initial usage data when the vehicle is in its second state; and the third vehicle control pool stores the initial usage data when the vehicle is in its third state. The prediction model includes the pre-constructed Wide&Deep model.

[0091] Offline training comprises a big data platform and a training platform. The big data platform includes functions for collecting event tracking data, data ETL (Extract, Transform, and Load) of frequently used target functions, and feature data mining. Event tracking data is primarily used to collect initial vehicle usage data. Data ETL is used for preprocessing this initial usage data. The historical frequently used target functions section stores information about target functions used within a historical period. Feature data mining is used to extract features from the initial usage data. The prediction model is also housed within the training platform. The training platform is primarily used to train the prediction model based on data from the big data platform to improve its recommendation capabilities.

[0092] The main function of the online push includes: when the vehicle is in different vehicle states, the vehicle sends the current time, the current location, the environment data and the running data to the cloud server. The cloud server receives the first use data when the vehicle is in the vehicle state in the historical period based on the vehicle state of the vehicle. Then, the first use data is input into the Wide&Deep model, and the Wide&Deep model calculates a plurality of use probability values. Then, the plurality of use probability values are screened, for example, the plurality of use probability values can be sorted, and the part of use probability values meeting the requirements are taken as target probability values. Finally, according to the target probability values, the recommended target function is determined.

[0093] Compared with the related art, the embodiments of the present application have at least the following advantages:

[0094] By obtaining the vehicle state in which the vehicle is currently located, and receiving the first use data when the vehicle is in the vehicle state in the historical period, the frequency information of each target function used by the vehicle in the historical period is obtained. On the one hand, the first use data and one or more of the current time, the current location, the environment data and the running data are input into the Wide model, so that the Wide model can memorize the data that each target function is preferred to be used by the user in the historical period, so as to improve the memory ability of the Wide model to the interaction relationship between the data; on the other hand, the first use data and one or more of the current time, the current location, the environment data and the running data are input into the Deep model, so that the Deep model can predict the data that each target function is preferred to be used by the user, so as to improve the generalization ability of the Deep model to the interaction relationship between the data. Finally, according to the second use data and the third use data, the use probability value of each target function is calculated. Based on the part of use probability values greater than the preset probability threshold, the target function meeting the use preference of the user is accurately recommended to the user, so that the user can directly use the preferred target function without performing tedious operation steps, thereby improving the use experience of the user to the vehicle.

[0095] Please refer to Figure 5 The interaction schematic diagram between the cloud and the vehicle provided by the embodiments of the present application is shown in FIG. 1. The specific steps are as follows:

[0096] Step S1: When the vehicle is in different vehicle states, the vehicle sends a service request to the cloud server.

[0097] In some embodiments, the vehicle state and the vehicle ID are included in the service request. The vehicle ID is used to uniquely identify the vehicle. The vehicle state includes a first state, a second state and a third state. The first state is a state of the vehicle in response to an unlock instruction, that is, when the vehicle is in the first state, the user can get on and off at any time. The second state is a state of the vehicle during driving, and the third state is a state of the vehicle when the vehicle is parked, that is, when the gear of the vehicle is in P.

[0098] In some embodiments, the vehicle can send one or more of the current time, the current location, the environmental data and the running data of the vehicle to the cloud server.

[0099] In this embodiment, the environmental data includes the temperature inside the vehicle and the temperature outside the vehicle, and the running data can include the state of the window and the state of the door, etc. The content of the running data is set according to the actual design requirements.

[0100] Step S2: The cloud server receives the first usage data corresponding to each target function, and the first usage data is the number of times each target function is used when the vehicle is in the vehicle state in the historical period.

[0101] Specifically, the first usage data when the vehicle is in the vehicle state in the historical period is found from the data warehouse. The other specific contents of this step are the same as those of step 102, and are not repeated here.

[0102] Step S3: The cloud server uses the built-in Word&Deep model to obtain the second usage data and the third usage data according to the first usage data.

[0103] In some embodiments, the cloud server can also use the built-in Word&Deep model to obtain the second usage data and the third usage data according to the first usage data, and one or more of the current time, the current location, the environmental data and the running data. The cloud server calls the built-in Word&Deep model from the model warehouse. The specific content of the cloud server using the Word&Deep model to process data is the same as that of steps 103 and 104, and is not repeated here.

[0104] Step S4: The cloud server calculates the usage probability value of each target function according to the second usage data and the third usage data.

[0105] The specific content of this step is the same as that of step 105, and is not repeated here.

[0106] Step S5: The cloud server determines the target probability value greater than the preset probability threshold value in the plurality of usage probability values.

[0107] The specific content of this step is the same as step 106, and to avoid repetition, it will not be described here.

[0108] Step S6: The cloud server sends the target function matching the target probability value to the vehicle.

[0109] Step S7, the vehicle determines the recommendation list according to the vehicle state, and displays the recommendation list.

[0110] In this embodiment, after the vehicle receives the target function matching the target probability value sent by the cloud server, the vehicle can determine the recommendation list according to the vehicle state. The recommendation list is a list of target function information. Based on the recommendation list, the user is recommended. The vehicle state also includes the working state of the vehicle components, including the trunk, the window, the rearview mirror, the seat, etc.

[0111] For example, the vehicle can reasonably judge the target function matching the target probability value in combination with its own vehicle state, which means that when the vehicle is in a certain parking space and there is an obstacle behind the vehicle that cannot open the trunk, at this time, the target function of opening the trunk should not be recommended. Or, when the vehicle is parked, there are many items in the trunk, which makes the rear seat unable to open the seat massage target function, at this time, the rear seat should not be recommended to open the seat massage target function.

[0112] Compared with the related art, the embodiments of the present application have at least the following advantages:

[0113] By obtaining the current vehicle state of the vehicle, and receiving the first use data when the vehicle is in the vehicle state in the historical period, the frequency information of each target function used in the historical period is obtained. On the one hand, the first use data and one or more of the current time, the current location, the environmental data and the running data are input into the Wide model, so that the Wide model can remember the data that each target function is preferred by the user in the historical period, to improve the memory ability of the Wide model to the interaction between data; on the other hand, the first use data and one or more of the current time, the current location, the environmental data and the running data are input into the Deep model, so that the Deep model can predict the data that each target function is preferred by the user, to improve the generalization ability of the Deep model to the interaction between data. Finally, according to the second use data and the third use data, the use probability value of each target function is calculated. Based on the part of the use probability value greater than the preset probability threshold, the target function meeting the user's use preference is accurately recommended to the user, so that the user can directly use the preferred target function without performing tedious operation steps, thereby improving the user's experience of using the vehicle.

[0114] The embodiment of the application further provides a vehicle function recommendation method, which is applied to a vehicle and includes: sending a vehicle state in which the vehicle currently locates to a cloud server. The cloud server is configured to receive first usage data corresponding to each target function, the first usage data being a number of times each target function is used when the vehicle is in the vehicle state in a historical period.

[0115] Further, the cloud server is further configured to obtain second usage data and third usage data according to the first usage data, the second usage data representing data of the target functions being preferred by a user in the historical period, and the third usage data representing data of the target functions being predicted to be preferred by the user.

[0116] The cloud server is further configured to calculate usage probability values of the target functions according to the second usage data and the third usage data, and determine a target probability value greater than a preset probability threshold value in the usage probability values.

[0117] The vehicle is configured to receive a target function matched with the target probability value and display a recommendation list, wherein the recommendation list is at least one target function matched with the target probability value and determined based on the vehicle state. In the embodiment, after the vehicle receives the target function matched with the target probability value sent by the cloud server, the vehicle can determine the recommendation list according to the vehicle state.

[0118] Further, after the vehicle displays the recommendation list, the vehicle performs a function operation corresponding to the selected target function in response to a function selection instruction for the recommendation list. For example, the recommendation list includes an opening reading lamp function, a closing ambient light function, and a closing window function, and the user can select one or more of the opening reading lamp function, the closing ambient light function, and the closing window function. If the user selects the opening reading lamp function and the closing ambient light function, the vehicle controls the reading lamp to be in an opening state and the ambient light to be in a closing state.

[0119] Compared with the related art, the embodiment of the application has at least the following advantages:

[0120] By obtaining the vehicle state in which the vehicle currently locates, receiving the first usage data when the vehicle is in the vehicle state in the historical period, and obtaining the frequency information of the target functions used in the historical period, the first usage data is input into a prediction model to obtain the second usage data and the third usage data. According to the second usage data and the third usage data, the usage probability values of the target functions are calculated. Based on the part of the usage probability values greater than the preset probability threshold value, the target function preferred by the user is accurately recommended to the user, so that the user can directly use the preferred target function without performing complicated operation steps, thereby improving the user experience of the vehicle.

[0121] Please refer toFigure 6 A hardware structure schematic diagram of an electronic device 1000 is provided for embodiments of the present application. As shown, the electronic device 1000 can include a processor 1001, a memory 1002. The memory 1002 is configured to store one or more computer programs 1003. The one or more computer programs 1003 are configured to be executed by the processor 1001. The one or more computer programs 1003 include instructions that can be used to implement the above-described method in the electronic device 1000. Figure 6 It can be understood that the structure illustrated in the embodiments does not constitute a specific limitation on the electronic device 1000. In other embodiments, the electronic device 1000 can include more or fewer components than those shown, or combine some components, or split some components, or different arrangement of components.

[0122] The processor 1001 can include one or more processing units. For example, the processor 1001 can include an application processor (AP), a modem, a graphics processing unit (GPU), an image signal processor (ISP), a controller, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural-network processing unit (NPU), etc. Different processing units can be independent devices or integrated in one or more processors.

[0123] The processor 1001 can also be provided with a memory for storing instructions and data. In some embodiments, the memory in the processor 1001 is a cache memory. The memory can save instructions or data that the processor 1001 has just used or repeatedly uses. If the processor 1001 needs to use the instructions or data again, it can directly call from the memory. This avoids repeated access and reduces the waiting time of the processor 1001, thereby improving the efficiency of the system.

[0124]

[0125] ​In some embodiments, the processor 1001 can include one or more interfaces. The interfaces can include an inter-integrated circuit (I2C) interface, an inter-integrated circuit sound (I2S) interface, a pulse code modulation (PCM) interface, a universal asynchronous receiver / transmitter (UART) interface, a mobile industry processor interface (MIPI), a general-purpose input / output (GPIO) interface, a SIM interface, and / or a USB interface, etc.

[0126] In some embodiments, the processor 1001 is configured to execute single instruction multiple data (SIMD), very long instruction word (VLIW), and / or other acceleration schemes.

[0127] In some embodiments, the memory 1002 can include a high-speed random access memory, and can further include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state memory device.

[0128] The embodiments also provide a computer-readable storage medium, which stores computer instructions, and when the instructions run on an electronic device, the electronic device executes the above-mentioned related method steps to realize the vehicle function recommendation method in the above-mentioned embodiments.

[0129] In the embodiments, the electronic device and the computer-readable storage medium are used to execute the corresponding methods provided above, and thus the beneficial effects of the electronic device and the computer-readable storage medium can refer to the beneficial effects of the corresponding methods provided above, which will not be described here.

[0130] In practical applications, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above.

[0131] In several embodiments provided in the present application, the disclosed apparatus and method can be implemented in other manners. For example, the division of the apparatus embodiments is merely illustrative, and the division of the modules or units can be changed according to actual conditions, such as a plurality of units or components being combined or integrated into another apparatus, or some features being ignored or not executed. In addition, the display or discussion of a coupling or direct coupling or communication connection between the modules or units can be indirect coupling or communication connection through some interfaces, and can be electrical, mechanical or other forms.

[0132] The units described as separate components can or can not be physically separate, and the components shown as units can be one physical unit or multiple physical units, i.e., can be located in one place or distributed in multiple different places. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0133] In addition, each functional unit in the various embodiments of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0134] The integrated unit, if realized in the form of a software functional unit and sold or used as an independent product, can be stored in a readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present application essentially or the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions to make a device (which can be a single chip, a chip, etc.) or a processor (processor) execute all or part of the steps of the method described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (Read-Only Memory, ROM), a random access memory (Random Access Memory, RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0135] Finally, it should be noted that the above embodiments are merely used to illustrate the technical solutions of the present application and not to limit the present application. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present application.

Claims

1. A vehicle function recommendation method characterized by comprising: The vehicle comprises a plurality of target functions, and the method comprises: obtaining a vehicle state in which the vehicle currently is; receiving first use data corresponding to each of the target functions, the first use data being a number of times each of the target functions is used when the vehicle is in the vehicle state in a historical period; obtaining second use data from the first use data, the second use data representing data of each of the target functions being preferentially used by a user in the historical period; obtaining third use data from the first use data, the third use data representing data of each of the target functions being predicted to be preferentially used by a user; calculating a use probability value of each of the target functions according to the second use data and the third use data; determining a target probability value greater than a preset probability threshold in the plurality of use probability values; and recommending the target function matching the target probability value.

2. The vehicle function recommendation method according to claim 1, characterized by, The first use data comprises historical cross data, historical numerical data, historical classification data and function data, the historical cross data being frequency data of each of the target functions being used in different historical positions and historical times of the vehicle, the historical numerical data being frequency data of each of the target functions being used, the historical classification data being data classified based on a preset classification rule on the historical positions and the historical times, and the function data being name information of each of the target functions. The obtaining of the second use data from the first use data comprises: obtaining the second use data according to the historical cross data. The obtaining of the third use data from the first use data comprises: obtaining the third use data according to the historical numerical data, the historical classification data and the function data.

3. The vehicle function recommendation method according to claim 2, characterized by, The method uses a prediction model to predict the target functions displayed in the vehicle, the prediction model comprising a first model and a second model; and the method further comprises: obtaining a current time of the vehicle; The obtaining of the second use data according to the historical cross data comprises: inputting the historical cross data and the current time into the first model to obtain the second use data. The obtaining of the third use data according to the historical numerical data, the historical classification data and the function data comprises: inputting the historical numerical data, the historical classification data, the function data and the current time into the second model to obtain the third use data.

4. The vehicle function recommendation method according to claim 2, characterized by, The method uses a prediction model to predict the target functions displayed in the vehicle, the prediction model comprising a first model and a second model; and the method further comprises: obtaining a current position of the vehicle; The obtaining of the second use data according to the historical cross data comprises: inputting the historical cross data and the current position into the first model to obtain the second use data. The obtaining of the third use data according to the historical numerical data, the historical classification data and the function data comprises: inputting the historical numerical data, the historical classification data, the function data and the current position into a second model to obtain the third use data.

5. The vehicle function recommendation method according to claim 2, characterized by, The method adopts a prediction model to predict the target function displayed in the vehicle, the prediction model comprising a first model and a second model; the method further comprises: obtaining a current time and a current position of the vehicle; the second use data is obtained according to the historical cross data, comprising: inputting the historical cross data, the current time and the current position into the first model to obtain the second use data; the third use data is obtained according to the historical numerical data, the historical classification data and the function data, comprising: inputting the historical numerical data, the historical classification data, the function data, the current time and the current position into the second model to obtain the third use data.

6. The vehicle function recommendation method according to claim 1, characterized by, The vehicle state comprises a first state, a second state and a third state; the first state is a state of the vehicle in response to an unlocking instruction, the second state is a state of the vehicle in a driving process, and the third state is a state of the vehicle in a parking process.

7. A vehicle function recommendation method characterized by The method comprises: sending a current vehicle state of the vehicle to a cloud server, wherein the cloud server is configured to receive first use data corresponding to each target function, the first use data being a number of times each target function is used when the vehicle is in the vehicle state in a historical period; obtain second use data from the first use data, the second use data representing data of each target function being preferred by a user in the historical period; obtain third use data from the first use data, the third use data representing data of each target function being predicted to be preferred by a user; calculate a use probability value of each target function according to the second use data and the third use data; and determine a target probability value greater than a preset probability threshold from a plurality of use probability values; receive the target function matching the target probability value; display a recommendation list, wherein the recommendation list is at least one target function matching the target probability value determined based on the vehicle state.

8. The vehicle function recommendation method according to claim 7, characterized by, After displaying the recommendation list, the method further comprises: in response to a function selection instruction for at least one target function in the recommendation list, performing a function operation corresponding to the selected target function.

9. An electronic device, comprising: The electronic device comprises a processor and a memory, the memory being configured to store instructions, and the processor being configured to invoke the instructions in the memory to enable the electronic device to perform the vehicle function recommendation method of any one of claims 1 to 8.

10. A computer storage medium, characterized in that The computer instructions, when executed on an electronic device, enable the electronic device to perform the vehicle function recommendation method of any one of claims 1 to 8.

Images