A method, device, medium and program product for proactively providing vehicle services

Abstract

The present application proposes a method, device, medium and program product for proactively providing vehicle services. The method may include: acquiring the driving data of the vehicle, and determining the target driving scenario of the vehicle based on the acquired driving data; determining the target workflow corresponding to the target driving scenario, the target workflow is generated based on the vehicle usage habits of the target user and the target workflow includes at least one target vehicle service, the target user is the user currently driving the vehicle; executing the target workflow to provide the target vehicle service to the target user. According to the technical solution of the present application, it is possible to proactively provide the target vehicle service to the user based on the target driving scenario.

Description

Technical Field

[0001] The present application relates to the field of automobile technology, and in particular to a method, device, medium and program product for proactively providing vehicle services. Background Art

[0002] As the level of automobile intelligence increases, the types of services that vehicles can provide to users are also increasing. For example, vehicles can provide users with services such as intelligent navigation and voice interaction. However, in related technologies, these services usually require users to actively trigger them to start. In other words, the vehicle can only passively respond to the user's instructions to provide services. This results in users having to operate manually when using vehicle services, reducing the convenience of user operations. Especially for frequently used services, users are required to repeatedly input commands, which not only increases the user's operating burden, but also reduces the overall driving experience. Summary of the invention

[0003] The present application provides a method, device, medium and program product for proactively providing vehicle services to address deficiencies in related technologies.

[0004] According to a first aspect of one or more embodiments of the present application, a method for proactively providing vehicle services is provided, the method comprising:

[0005] Acquire driving data of the vehicle, and determine a target driving scene in which the vehicle is located according to the acquired driving data;

[0006] Determine a target workflow corresponding to the target driving scenario, wherein the target workflow is generated based on a target user's vehicle usage habits and includes at least one target vehicle service, and the target user is a user currently driving the vehicle;

[0007] The target workflow is executed to provide the target vehicle service to the target user.

[0008] Optionally, determining the target driving scene of the vehicle based on the acquired driving data includes: determining the preset conditions of each candidate driving scene in the candidate driving scenes; calculating the matching score between the driving data and each candidate driving scene based on the preset conditions of each candidate driving scene; and determining the candidate driving scene whose matching score exceeds a score threshold as the target driving scene.

[0009] Optionally, the driving data includes the current driving time, the current geographic location of the vehicle and the current operation of the target user; the preset conditions of any candidate driving scenario include preset conditions on the time dimension, preset conditions on the location dimension and preset conditions on the user operation dimension; the matching score of the driving data and any candidate driving scenario is calculated according to the preset conditions of any candidate driving scenario, including: calculating the time matching score corresponding to the current driving time according to the preset conditions of any candidate driving scenario on the time dimension; calculating the location matching score corresponding to the current geographic location according to the preset conditions of any candidate driving scenario on the location dimension; calculating the operation matching score corresponding to the current operation of the target user according to the preset conditions of any candidate driving scenario on the user operation dimension; calculating the matching score of the driving data with any candidate driving scenario based on the time matching score, the location matching score and the operation matching score.

[0010] Optionally, determining the target workflow corresponding to the target driving scenario includes: obtaining target user portrait data of the target user in the target driving scenario, and generating the target workflow based on the acquired target user portrait data; or, querying a candidate driving scenario matching the target driving scenario from a workflow database, and determining the candidate workflow corresponding to the queried candidate driving scenario as the target workflow, wherein the workflow database stores multiple candidate driving scenarios and their corresponding candidate workflows, and the candidate workflows are pre-generated based on the comprehensive user portrait data of the target user; the target user portrait data or the comprehensive user portrait data is used to characterize the target user's car usage habits.

[0011] Optionally, generating a workflow based on user portrait data includes: inputting the user portrait data and prompt words into a workflow generation model, so that the workflow generation model generates and outputs a workflow matching the user portrait data under the instruction of the prompt words.

[0012] Optionally, the workflow generation model is trained in the following manner: obtaining sample user portrait data and a sample workflow corresponding to the sample user portrait data; inputting the sample user portrait data and sample prompt words into a pre-trained workflow generation model, so that the pre-trained workflow generation model generates and outputs a sample prediction workflow matching the sample user portrait data under the instruction of the sample prompt words; and adjusting the sample prompt words and iteratively training the pre-trained workflow generation model according to the difference between the sample workflow and the sample prediction workflow.

[0013] Optionally, executing the target workflow includes: outputting the target workflow to the target user, and executing the target workflow upon receiving a confirmation instruction from the target user for the target workflow.

[0014] Optionally, the method further includes: obtaining feedback information from the target user on the target workflow, so as to optimize the target workflow based on the feedback information, wherein the feedback information includes evaluation information and / or improvement suggestions on the target workflow.

[0015] According to a second aspect of one or more embodiments of the present application, there is provided a device for actively providing vehicle services, including:

[0016] A driving scene determination unit, used to obtain driving data of a vehicle, and determine a target driving scene in which the vehicle is located according to the obtained driving data;

[0017] a workflow determination unit, configured to determine a target workflow corresponding to the target driving scenario, wherein the target workflow is generated based on a target user's vehicle usage habits and includes at least one target vehicle service, and the target user is a user currently driving the vehicle;

[0018] The workflow execution unit is used to execute the target workflow to provide the target vehicle service to the target user.

[0019] According to a third aspect of one or more embodiments of the present application, an electronic device is provided, including:

[0020] processor;

[0021] a memory for storing processor-executable instructions;

[0022] The processor implements the method described in any one of the embodiments of the first aspect above by running the executable instructions.

[0023] According to a fourth aspect of one or more embodiments of the present application, a computer-readable storage medium is provided, on which computer instructions are stored. When the instructions are executed by a processor, the steps of the method described in any embodiment of the first aspect are implemented.

[0024] According to a fifth aspect of one or more embodiments of the present application, a computer program product is provided, comprising a computer program and / or instructions, which, when executed by a processor, implement the steps of the method described in any embodiment of the first aspect above.

[0025] It can be seen from the above technical solutions that in one or more embodiments of the present application, by obtaining the driving data of the vehicle, and determining the current target driving scene according to the driving data of the vehicle, the target workflow corresponding to the target driving scene is determined for execution. The above method realizes that the vehicle actively determines the target workflow for execution according to the target driving scene, and can automatically provide the user with each target vehicle service contained in the target workflow without the user initiating an instruction, which significantly improves the convenience of user operation and the car-using experience. In addition, the user's car-using needs in different driving scenes are usually inconsistent. Therefore, the user's current car-using needs can be more accurately determined according to the target driving scene. The target workflow corresponding to the target driving scene is generated based on the target user's car-using habits, that is, the target workflow fully considers the user habits / preferences of the target user in the target driving scene, so the execution of the target workflow can flexibly meet the target user's car-using needs in the target driving scene, and provide intelligent and personalized services to the target user.

[0026] 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

[0027] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0028] Figure 1 It is a schematic diagram of the architecture of a vehicle service system provided by an exemplary embodiment.

[0029] Figure 2 It is a flowchart of a method for actively providing vehicle services provided by an exemplary embodiment.

[0030] Figure 3 This is a schematic diagram of generating and storing candidate workflows provided by an exemplary embodiment.

[0031] Figure 4 It is a flowchart of a training method for a workflow generation model provided by an exemplary embodiment.

[0032] Figure 5 This is a schematic diagram of outputting a target workflow to a user provided by an exemplary embodiment.

[0033] Figure 6 is a block diagram of a vehicle service system provided by an exemplary embodiment.

[0034] Figure 7 It is a schematic structural diagram of an electronic device provided by an exemplary embodiment.

[0035] Figure 8 It is a block diagram of a device for proactively providing vehicle services provided by an exemplary embodiment. DETAILED DESCRIPTION

[0036] The user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0037] Next, one or more embodiments of the present application are described in detail.

[0038] Figure 1 FIG. 1 is a schematic diagram of the architecture of a vehicle service system provided by an exemplary embodiment. Figure 1 As shown, the vehicle service system may include a server 11, a network 12, and several vehicles, such as vehicles 13-14, etc. In the present application, the vehicle service system is responsible for providing vehicle services to users on the vehicle.

[0039] The server 11 may be a physical server including an independent host, or the server 11 may be a virtual server carried by a host cluster. During operation, the server 11 may run a server-side program of the vehicle service system to realize the service end of the vehicle service system.

[0040] The vehicles 13-14 may be any vehicles. The vehicles 13-14 and the server 11 may be directly or indirectly connected via wired communication or wireless communication, and this application does not impose any special restrictions.

[0041] based on Figure 1 In the vehicle service system shown, any vehicle among the vehicles 13-14 can obtain the driving data of the vehicle itself and submit the obtained driving data to the server 11. The server 11 can determine the target driving scene of the vehicle submitting the driving data based on the received driving data. Then, the server 11 determines the target workflow corresponding to the target driving scene and returns the target workflow to the vehicle submitting the driving data so that the vehicle executes the target workflow.

[0042] In another embodiment, the vehicle service system may include only the vehicle. Figure 1 Taking the vehicle 13 in the example, the vehicle 13 can obtain the driving data of the vehicle itself, and determine the target driving scene of the vehicle based on the obtained driving data. Then, the target workflow corresponding to the target driving scene is determined, and the target workflow is executed to provide the user on the vehicle with various vehicle services included in the target workflow.

[0043] Figure 2 The flowchart of a method for actively providing vehicle services provided by an exemplary embodiment is shown in FIG. 1 , and the method can be applied to a vehicle service system. Figure 2 As shown, the method may include the following steps:

[0044] S201. Acquire driving data of a vehicle, and determine a target driving scene of the vehicle according to the acquired driving data.

[0045] The driving data of a vehicle refers to various data during the driving process of the vehicle, including but not limited to: driving environment data (such as weather, traffic conditions, etc.), user operation data (such as air conditioning settings, seat position adjustment, multimedia playback settings, etc.), vehicle driving status data (such as braking distance control, vehicle speed, tire pressure, etc.). Driving data can be obtained through various sensors installed on the vehicle or vehicle networking technology. Generally, there are differences in the driving data of a vehicle in different driving scenarios. Therefore, by performing data statistics and mining on the acquired driving data, the target driving scenario represented by the driving data can be determined.

[0046] In one embodiment, candidate driving scenes and preset conditions for each candidate driving scene may be set, so as to determine the target driving scene according to the preset conditions. Generally, the user may set the candidate driving scenes and their preset conditions, or the vehicle service system may set them according to the user's daily vehicle usage data. This application does not limit the specific setting method. Table 1 shows four candidate driving scenes and their preset conditions provided by an exemplary embodiment. Please refer to Table 1:

[0047] Table 1

[0048]

[0049] When determining the target driving scene of the vehicle based on the acquired driving data, first determine the candidate driving scenes and their preset conditions, then calculate the matching scores of the acquired driving data and each candidate driving scene based on the preset conditions of each candidate driving scene, and determine the candidate driving scene whose matching score exceeds the score threshold as the target driving scene. The matching score is used to characterize the similarity between the driving data and the preset conditions. The higher the matching score, the more similar the driving data is to the preset conditions, that is, the more the driving data can meet the preset conditions. Combined with Table 1, assuming that the score threshold is 0.8, the acquired driving data includes: the current time is 17:00 on Friday afternoon, and the user's navigation destination is school. The matching scores of the driving data and the preset conditions of the four candidate driving scenes in Table 1 are calculated respectively, and the matching score of the commuting scene is 0, the matching score of the long-distance travel scene is 0, the matching score of the shopping scene is 0.5, and the matching score of the child-picking scene is 0.85. Obviously, only the matching score of the child-pickup scenario exceeds the score threshold of 0.8, so the candidate driving scenario of child-pickup can be determined as the target driving scenario. In addition, the score threshold may not be set, but the candidate driving scenario corresponding to the maximum value of the calculated matching scores may be determined as the target driving scenario.

[0050] It should be noted that the candidate driving scenarios and their preset conditions are not static and can be updated according to user needs or adjusted according to the user's actual driving habits / preferences.

[0051] In this embodiment, by setting clear candidate driving scenes and their candidate conditions, the target driving scene is screened from the clear candidate driving scenes, which helps to improve the accuracy of the determined target driving scene. In addition, this method decomposes the determination of the target driving scene into simple condition matching, which helps to reduce the difficulty and complexity of determining the target driving scene and improve the efficiency of determining the target driving scene.

[0052] In one embodiment, driving data itself has many dimensions, so when setting the preset conditions of the candidate driving scenes, preset conditions of multiple dimensions can be set, and then the preset conditions and driving data on the same dimension can be used to calculate the matching score in a targeted manner.

[0053] Exemplarily, the preset conditions of any candidate driving scene may include preset conditions on the time dimension, preset conditions on the location dimension, and preset conditions on the user operation dimension. The driving data obtained may include the current driving time, the current geographical location of the vehicle, and the current operation of the target user. The target user refers to the user who is currently driving the vehicle. Assuming that the candidate driving scene is a working scene, the preset condition of the candidate driving scene on the time dimension is that the current time is between 7:00-9:00 in the morning, the preset condition on the location dimension is that the vehicle is located on the driving route from home to the company, and the preset condition on the user operation dimension is that the user has boarded the vehicle. Then, the time matching score can be calculated according to the current driving time of the vehicle and the preset conditions on the time dimension, the location matching score can be calculated according to the current geographical location of the vehicle and the preset conditions on the location dimension, and the operation matching score can be calculated according to the current operation of the target user and the preset conditions on the user operation dimension. The above-mentioned time matching score, location matching score, and operation matching score can all be calculated by methods such as semantic matching, keyword matching, etc., and this application does not limit the specific matching method.

[0054] Then, the matching score between the acquired driving data and the candidate driving scene "going to work" is calculated based on the time matching score, the location matching score and the operation matching score. The calculation process can adopt weighted average calculation, average calculation, standard deviation calculation and other methods, which are not limited in this application. Assuming that the weighted average calculation method is adopted to calculate the matching score between the driving data and the candidate driving scene, the calculation principle can be referred to the following formula (1):

[0055] (1)

[0056] In formula (1), Characterizes the matching score between driving data and any candidate driving scene, Characterize the time matching score, Characterizes the position matching score, Characterize the operation matching score, Representation weight, which can be set by yourself. For example, the time matching score is 0.9, the location matching score is 0.95, and the operation matching score is 1.0. is 0.4, When both are 0.3, according to formula (1), we can obtain that the matching score between the driving data and the candidate driving scene is 0.935.

[0057] It should be noted that the dimensions of the above-mentioned preset conditions are only examples. In the actual implementation of the present application, the preset conditions of any candidate driving scenario may also include preset conditions in other dimensions. Correspondingly, the acquired driving data may also include driving data in other dimensions.

[0058] In this embodiment, the matching scores of each preset dimension are calculated by using the preset conditions of the preset dimension and the driving data of the corresponding dimension, and then the matching scores of each preset dimension are combined to calculate the matching scores of the driving data and the candidate driving scene, which helps to consider the matching degree of the driving data and the candidate driving scene from multiple dimensions, improve the accuracy of the matching score finally obtained, and then improve the accuracy of the determined target driving scene. In addition, this method only needs to calculate based on the driving data of the preset dimension, without considering other invalid driving data, which greatly reduces the amount of data and effectively improves the calculation efficiency.

[0059] S202: Determine a target workflow corresponding to the target driving scenario, where the target workflow is generated based on a target user's vehicle usage habits and includes at least one target vehicle service, and the target user is a user currently driving the vehicle.

[0060] Since the target users have different car-using habits in different driving scenarios, a workflow corresponding to each driving scenario can be generated based on the target users' car-using habits. The target users' car-using habits (or car-using preferences) may include driving habits, riding habits, etc., such as the air-conditioning setting temperature, songs that are usually played, braking distance, etc. Any workflow includes at least one vehicle service. Vehicle services refer to services provided by the vehicle to users, such as automatic adjustment of seat position, automatic adjustment of air-conditioning temperature, and other services. When a workflow contains two or more vehicle services, each vehicle service must be implemented in a specific order, which may include a parallel order and / or a serial order. Table 2 shows five driving scenarios and their corresponding workflows provided by an exemplary embodiment. Please refer to Table 2:

[0061] Table 2

[0062]

[0063] Table 2 is only an example, and the implementation order of each vehicle service in the workflow can be determined according to the actual vehicle usage habits of the target users.

[0064] In one embodiment, the daily car use data of the target user can be collected through the vehicle system and the smart cockpit interface, and the messy daily car use data can be cleaned and normalized to obtain structured user portrait data, which is used to characterize the user's car use habits / preferences. User portrait data includes but is not limited to: user personal data, security domain data, entertainment domain data, and economic domain data. Among them, user personal data can include user name, age, occupation, etc. and user personal information. Security domain data can be understood as data related to the user's driving safety habits, such as braking distance, seat position, vehicle speed, etc. Entertainment domain data can be understood as data related to entertainment habits, such as radio stations, playing songs and other multimedia data settings, air conditioning temperature settings, etc. Economic domain data can be understood as data related to vehicle expenses, such as power saving / fuel saving mode, power consumption / fuel consumption mode, etc. User portrait data can be stored in a portrait database so that the user portrait data of the target user can be obtained from the portrait database during the implementation of the present application solution. The portrait database can be deployed locally in the vehicle or on a server, and this application does not limit this. User portrait data is not immutable and can be updated in time as the user's car use habits change. In addition, in the process of collecting user vehicle usage data and generating and storing user portrait data, it is necessary to strengthen the privacy protection of various data to avoid data leakage that may lead to the abuse of user data.

[0065] In this embodiment, the user portrait data of the target user in all driving scenarios is called "comprehensive user portrait data", and the user portrait data of the target user in the target driving scenario is called "target user portrait data".

[0066] Candidate workflows corresponding to multiple candidate driving scenarios can be pre-generated based on the comprehensive user portrait data stored in the portrait database, and multiple candidate driving scenarios and their corresponding candidate workflows can be associated and stored in the workflow database. Among them, the comprehensive user portrait data is used to characterize the target user's car usage habits in each candidate driving scenario. Exemplarily, during the storage process, a relational database or a NoSQL database can be used as a workflow database, and the candidate workflows can be stored in JSON format (JSON format facilitates subsequent parsing and management of candidate workflows). In addition, the candidate driving scenarios, candidate workflows, and the car usage requirements corresponding to the candidate driving scenarios can also be associated and stored. For example, the candidate driving scenario is the commuting scenario, the car usage requirements are 7:00-9:00 in the morning and from home to the company, and the candidate workflow is [adjust the air conditioning temperature to 22°C, play podcasts, navigate to the company, and query the weather forecast], then the storage format can refer to the following example:

[0067] {

[0068] "workflow_id": "morning_commute_001",

[0069] "scenario": "morning_commute",

[0070] "conditions": {

[0071] "time_range": "07:00-09:00",

[0072] "location_range": {

[0073] "start": "home",

[0074] "end": "company"

[0075] }

[0076] },

[0077] "actions": [

[0078] {"type": "set_ac_temperature", "value": 22},

[0079] {"type": "play_audio", "content": "podcast"},

[0080] {"type": "navigate", "destination": "company"},

[0081] {"type": "send_weather_report"} ]

[0083] }

[0084] In the above example, scenario represents the candidate driving scenario, conditions represent the vehicle usage requirements corresponding to the candidate driving scenario, and actions represent the candidate workflow.

[0085] The workflow database stores multiple candidate driving scenarios and their corresponding candidate workflows. In the process of implementing the present application, after determining the target driving scenario, the workflow database can be queried for candidate driving scenarios that match the target driving scenario, and the candidate workflow corresponding to the queried candidate driving scenario can be determined as the target workflow. This method helps to improve the efficiency of determining the target workflow and reduce errors or delays that may occur during the real-time generation process.

[0086] In addition, after determining the target driving scenario, the target workflow corresponding to the target driving scenario can be temporarily generated according to the target user portrait data. Exemplarily, the target user portrait data is obtained from the portrait database, and the target user portrait data is used to characterize the target user's car usage habits / preferences in the target driving scenario. For example, the target user portrait data includes: the target user is accustomed to setting the air conditioner to 22°C in the working scenario, and listens to podcasts while driving and listens to the singer Xiaobai's songs while waiting for traffic lights. Then, based on the target user portrait data, a target workflow can be generated: the air conditioning temperature is set to 22°C → play podcasts while driving → play singer Xiaobai's songs while waiting for traffic lights. This method of temporarily generating a target workflow realizes the real-time generation of the target workflow based on the latest collected target user portrait data, which can further improve the accuracy of the target workflow, so that the target workflow can flexibly meet the current car needs of the target user.

[0087] In one embodiment, a workflow generation model may be used to generate a workflow. There are many workflow generation models, such as BERT (Bidirectional Encoder Representations from Transformers), Large Language Model (LLM), etc. This application does not limit the specific workflow generation model.

[0088] The user portrait data and prompt words are input into the workflow generation model. The prompt words can help the model better understand the user's needs and guide the model to generate results that match the input. In this embodiment, the workflow generation model can generate and output a workflow that matches the input user portrait data under the instruction of the prompt words.

[0089] Since the workflow generation model has powerful language understanding and data analysis capabilities, it can generate reasonable and accurate workflows using the workflow generation model, so that the generated workflows can be well adapted to the user's car use needs in various driving scenarios, thereby providing users with precise and high-quality services / functions by executing accurate workflows, which helps to improve user satisfaction.

[0090] Figure 3 FIG. 1 is a schematic diagram of generating and storing candidate workflows provided by an exemplary embodiment. Figure 3 As shown, the user portrait data and prompt words are input into the workflow generation model, so that the workflow generation model, under the instruction of the prompt words, summarizes multiple candidate driving scenarios according to the input user portrait data, and generates candidate workflows corresponding to each candidate driving scenario for output. Figure 3In the example, the workflow generation model outputs candidate driving scene 1 to candidate driving scene n, and candidate workflow 1 to candidate workflow n. There is a one-to-one correspondence between candidate workflows and candidate driving scenes. The generated candidate driving scenes and their corresponding candidate workflows can be associated and stored in the workflow database. It should be noted that the user profile data can be updated regularly to generate new candidate scenes and their corresponding new candidate workflows based on the updated user profile data, or to update the candidate workflows corresponding to the original candidate scenes.

[0091] In one embodiment, Figure 4 FIG. 1 is a flowchart of a training method for a workflow generation model shown in an exemplary embodiment. Figure 4 As shown, the method may include the following steps:

[0092] S401. Obtain sample user portrait data and a sample workflow corresponding to the sample user portrait data.

[0093] The sample user profile data and sample workflow are known and determined. In order to improve the prediction accuracy of the workflow generation model, multiple sample user profile data and their corresponding sample workflows can be obtained, each sample workflow contains a preset number of sample vehicle services and each sample vehicle service is executed in a preset order.

[0094] For example, sample user portrait data can be found in Table 3:

[0095] Table 3

[0096]

[0097] The sample workflows that match the above sample user portrait data include sample workflows corresponding to the working scene and sample workflows corresponding to the leaving work scene. Among them, the sample workflow corresponding to the working scene is: set the air conditioner temperature to 22℃→navigate to the company→play a podcast. The sample workflow corresponding to the leaving work scene is: set the air conditioner temperature to 24℃→navigate to home→play soothing music.

[0098] S402. Input the sample user portrait data and sample prompt words into a pre-trained workflow generation model, so that the pre-trained workflow generation model generates and outputs a sample prediction workflow matching the sample user portrait data under the instruction of the sample prompt words.

[0099] In this embodiment, the pre-trained workflow generation model refers to a model that has completed preliminary training, and the preliminary training is based on a large amount of unlabeled general data. In order to improve the performance of the pre-trained workflow generation model in the vertical task of generating workflows, the pre-trained workflow generation model can be fine-tuned using workflow-related data to obtain a model suitable for the workflow generation task. For example, the pre-trained workflow generation model is fine-tuned using sample user portrait data and sample workflows corresponding to the sample user portrait data. During the fine-tuning process, the pre-trained workflow generation model generates a sample prediction workflow.

[0100] The sample prediction workflow includes at least one sample prediction vehicle service. When the sample prediction workflow includes two or more sample prediction vehicle services, each sample prediction vehicle service is executed in the sample prediction order. The sample prompt words can be obtained by initialization, and then the sample prompt words are adjusted during the training process so that the adjusted sample prompt words can provide clear and accurate instructions for the workflow generation model. The specific process of generating the sample prediction workflow can refer to the embodiment shown in S202, which will not be repeated here.

[0101] S403: According to the difference between the sample workflow and the sample prediction workflow, adjust the sample prompt words and iteratively train the pre-trained workflow generation model.

[0102] The differences between the sample workflow and the sample prediction workflow include: the difference between the sample vehicle services and the sample prediction vehicle services, and the difference between the preset order of each sample vehicle service and the sample prediction order of each sample prediction vehicle service.

[0103] After multiple iterations of training, when the obtained workflow generation model meets the predefined training target or reaches the predefined number of iterations, the training of the workflow generation model is completed. And through multiple adjustments, the prompt words for the workflow generation model are determined. The trained workflow generation model can be used to generate corresponding workflows according to user portrait data.

[0104] Through the above training method, not only can accurate and reasonable prompt words be obtained, but also the workflow generation model can perform very well in the vertical task of generating workflows, thereby achieving a workflow with higher accuracy by using the workflow generation model.

[0105] S203: Execute the target workflow to provide the target vehicle service to the target user.

[0106] After the target workflow is determined, the corresponding functional component can be called to execute the target workflow to provide the target user with each target vehicle service included in the target workflow.

[0107] In the above embodiment, the vehicle actively determines the target workflow to be executed according to the target driving scenario, without the need for the user to initiate an instruction, so that the various target vehicle services contained in the target workflow are automatically provided to the user at the appropriate time (i.e., in the target driving scenario), which significantly improves the convenience of user operation and the vehicle use experience. In addition, the user's vehicle use needs in different driving scenarios are usually inconsistent. Therefore, the user's current vehicle use needs can be determined more accurately according to the target driving scenario. The target workflow corresponding to the target driving scenario is generated based on the target user's vehicle use habits, that is, the target workflow fully considers the target user's user habits / preferences in the target driving scenario. Therefore, executing the target workflow can flexibly meet the target user's vehicle use needs in the target driving scenario, and provide intelligent and personalized services to the target user.

[0108] In one embodiment, after the target workflow is determined, the target workflow can be output to the target user to determine whether to execute the target workflow based on the target user's instructions. There are many ways to output. For example, the target workflow and prompt information on whether to execute the target workflow can be displayed on the human-computer interaction interface, or a voice broadcast can be made to the target user on whether to execute the target workflow. This application does not limit the specific output method. In combination with the foregoing embodiment, a prompt message "The target workflow is to set the air conditioning temperature to 24°C → plan a route home → play soothing music. Do you want to execute the above target workflow?" can be displayed on the human-computer interaction interface (see Figure 5 ), or announce to the target user by voice "whether to execute the target workflow of setting the air conditioning temperature to 24°C, planning a route home, and playing soothing music".

[0109] If the target user confirms that the target vehicle services included in the target workflow are correct and the execution order between the target vehicle services is also correct, then the target user will issue a confirmation instruction (such as voice confirmation or triggering the consent control in the human-computer interaction interface). After receiving the target user's confirmation instruction for the target workflow, the vehicle service system can control the target functional components corresponding to each target vehicle service to execute the corresponding target vehicle service. For example, for the target workflow "set the air conditioning temperature to 24°C → plan the route home → play soothing music", if the target user confirms to execute the target workflow, the vehicle service system will send a control instruction to set the temperature to 24°C to the air conditioning controller, so that the air conditioning controller responds to the control instruction and sets the air conditioning temperature in the car to 24°C. And send a control instruction to plan the optimal route from the company to home to the navigation system control unit, so that the navigation system control unit responds to the control instruction and plans the optimal driving route from the company to home. And send a control instruction to play soothing music to the multimedia control unit, so that the multimedia control unit responds to the control instruction and plays soothing music through the speaker.

[0110] If the target user believes that there is an error in the target workflow, the target vehicle tasks or the execution order between the target vehicle tasks contained in the target workflow can be manually adjusted, so that the vehicle service system executes the adjusted target workflow.

[0111] In this embodiment, the target user checks and confirms the generated target workflow to ensure that the vehicle service system can execute the correct target workflow, thereby providing the target user with the correct vehicle service. Moreover, for the correct target workflow, the target user only needs to perform a confirmation operation, and the vehicle can automatically execute all vehicle services included in the target workflow, effectively improving the convenience of user driving.

[0112] In one embodiment, feedback information of the target user for the target workflow can be obtained, and the feedback information may include evaluation information and / or improvement suggestions for the target workflow, so as to improve the target workflow according to the feedback information. The feedback information of the target user for the target workflow can be collected in real time during the execution of the target workflow, or the feedback information of the target user for the target workflow can be collected after the execution of the target workflow is completed. There are many ways to collect, such as displaying prompt information for collecting feedback information on the human-computer interaction interface on the vehicle, or sending prompt information for collecting feedback information to the client bound to the target user, and this application does not limit this.

[0113] Figure 6 FIG. 1 is a block diagram of a vehicle service system provided by an exemplary embodiment. Figure 6As shown, the vehicle service system includes: a data collection module 601, a data processing module 602, a workflow generation module 603, a workflow storage module 604, a scene recognition module 605, a target workflow determination module 606 and a workflow execution module 607.

[0114] The data collection module 601 is used to collect the user's daily car use data in various driving scenarios, and send the collected daily car use data to the data processing module 602. The data processing module 602 is used to clean and normalize the daily car use data to remove noise and redundant information in the daily car use data and obtain structured user portrait data. The data processing module 602 sends the user portrait data to the workflow generation module 603. The workflow generation module 603 is used to generate workflows corresponding to various driving scenarios based on the user portrait data, and associate the workflows and their corresponding driving scenarios and send them to the workflow storage module 604. The workflow storage module 604 is used to store, update and manage various driving scenarios and their corresponding workflows.

[0115] The scene recognition module 605 is used to obtain the driving data of the vehicle and determine the target driving scene of the vehicle based on the obtained driving data. The scene recognition module 605 sends the determined target driving scene to the target workflow determination module 606. The target workflow determination module 606 can obtain the target workflow corresponding to the target driving scene from the workflow storage module 604 or obtain the temporarily generated target workflow from the workflow generation module 603, and send the obtained target workflow to the workflow execution module 607. The workflow execution module 607 is used to call the vehicle computer / vehicle control interface to control the corresponding functional components to execute the target workflow.

[0116] Figure 7 is a schematic diagram of the structure of an electronic device according to an exemplary embodiment of the present application. Figure 7 At the hardware level, the electronic device includes a processor 702, an internal bus 704, a network interface 706, a memory 708, and a non-volatile memory 710, and may also include hardware required for other services. The processor 702 reads the corresponding computer program from the non-volatile memory 710 into the memory 708 and then runs it. Of course, in addition to the software implementation, this application does not exclude other implementations, such as logic devices or a combination of software and hardware, etc., that is, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0117] Corresponding to the above method embodiment, the present application also provides an embodiment of a device for actively providing vehicle services.

[0118] Figure 8is a block diagram of a device for actively providing vehicle services according to an exemplary embodiment of the present application. The device can be applied to a vehicle service system. Figure 8 The device includes a driving scene determination unit 801, a workflow determination unit 802 and a workflow execution unit 803, wherein:

[0119] The driving scene determination unit 801 is configured to obtain driving data of the vehicle and determine a target driving scene of the vehicle according to the obtained driving data;

[0120] A workflow determination unit 802 is configured to determine a target workflow corresponding to the target driving scenario, wherein the target workflow is generated based on the vehicle usage habits of a target user and the target workflow includes at least one target vehicle service, and the target user is a user currently driving the vehicle;

[0121] The workflow execution unit 803 is configured to execute the target workflow to provide the target vehicle service to the target user.

[0122] Optionally, the driving scene determination unit 801 is specifically used to: determine the preset conditions of each candidate driving scene in the candidate driving scenes; calculate the matching score of the driving data and each candidate driving scene according to the preset conditions of each candidate driving scene; and determine the candidate driving scene whose matching score exceeds the score threshold as the target driving scene.

[0123] Optionally, the driving data includes the current driving time, the current geographic location of the vehicle and the current operation of the target user; the preset conditions of any candidate driving scenario include preset conditions on the time dimension, preset conditions on the location dimension and preset conditions on the user operation dimension; the matching score of the driving data and any candidate driving scenario is calculated according to the preset conditions of any candidate driving scenario, including: calculating the time matching score corresponding to the current driving time according to the preset conditions of any candidate driving scenario on the time dimension; calculating the location matching score corresponding to the current geographic location according to the preset conditions of any candidate driving scenario on the location dimension; calculating the operation matching score corresponding to the current operation of the target user according to the preset conditions of any candidate driving scenario on the user operation dimension; calculating the matching score of the driving data with any candidate driving scenario based on the time matching score, the location matching score and the operation matching score.

[0124] Optionally, the workflow determination unit 802 is specifically used to: obtain target user portrait data of the target user in the target driving scenario, and generate the target workflow based on the acquired target user portrait data; or, query candidate driving scenarios matching the target driving scenarios from a workflow database, and determine the candidate workflow corresponding to the queried candidate driving scenarios as the target workflow, wherein the workflow database stores multiple candidate driving scenarios and their corresponding candidate workflows, and the candidate workflows are pre-generated based on the comprehensive user portrait data of the target user; the target user portrait data or the comprehensive user portrait data is used to characterize the target user's car usage habits.

[0125] Optionally, generating a workflow based on user portrait data includes: inputting the user portrait data and prompt words into a workflow generation model, so that the workflow generation model generates and outputs a workflow matching the user portrait data under the instruction of the prompt words.

[0126] Optionally, the workflow generation model is trained in the following manner: obtaining sample user portrait data and a sample workflow corresponding to the sample user portrait data; inputting the sample user portrait data and sample prompt words into a pre-trained workflow generation model, so that the pre-trained workflow generation model generates and outputs a sample prediction workflow matching the sample user portrait data under the instruction of the sample prompt words; and adjusting the sample prompt words and iteratively training the pre-trained workflow generation model according to the difference between the sample workflow and the sample prediction workflow.

[0127] Optionally, the workflow execution unit 803 is specifically configured to: output the target workflow to the target user, and execute the target workflow upon receiving a confirmation instruction from the target user for the target workflow.

[0128] Optionally, the device further comprises:

[0129] The feedback unit 804 is configured to obtain feedback information from the target user on the target workflow to optimize the target workflow based on the feedback information, wherein the feedback information includes evaluation information and / or improvement suggestions on the target workflow.

[0130] The implementation process of the functions and effects of each module in the above-mentioned device is specifically described in the implementation process of the corresponding steps in the above-mentioned method, which will not be repeated here.

[0131] The devices or modules described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer, which may be in the form of a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email transceiver, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0132] In a typical configuration, a computer includes one or more processors, including a central processing unit (CPU) and a graphics processing unit (GPU), an input / output interface, a network interface, and a memory. The CPU is used for computational simulation, and the GPU is used for outputting high-quality 3D images.

[0133] Memory may include non-permanent storage in a computer-readable medium, in the form of random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0134] Computer-readable media include permanent and non-permanent, removable and non-removable media that can be used to store information by any method or technology. Information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, disk storage, quantum memory, graphene-based storage media or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include temporary computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0135] Corresponding to the embodiments of the aforementioned method, the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the steps of any embodiment of the aforementioned method are implemented.

[0136] Corresponding to the embodiments of the aforementioned method, the present application also provides a computer program product, including a computer program and / or instructions, which implement the steps of any embodiment of the aforementioned method when executed by a processor.

[0137] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A method for proactively providing vehicle services, characterized in that: include: Acquire driving data of the vehicle, and determine the target driving scene of the vehicle according to the acquired driving data, wherein the driving data includes the current driving time, the current geographical location of the vehicle and the current operation of the target user; Determine a target workflow corresponding to the target driving scenario, wherein the target workflow is generated based on a target user's vehicle usage habits and includes at least one target vehicle service, and the target user is a user currently driving the vehicle; Executing the target workflow to provide the target vehicle service to the target user; The determining the target driving scene of the vehicle according to the acquired driving data includes: Determine the preset conditions of each candidate driving scene in the candidate driving scenes; the preset conditions of any candidate driving scene include preset conditions in the time dimension, preset conditions in the location dimension, and preset conditions in the user operation dimension; Calculating the matching score of the driving data with each candidate driving scene according to the preset conditions of each candidate driving scene respectively; wherein, calculating the matching score of the driving data with any candidate driving scene according to the preset conditions of the candidate driving scene includes: calculating the time matching score corresponding to the current driving moment according to the preset conditions of the candidate driving scene in the time dimension; calculating the location matching score corresponding to the current geographical location according to the preset conditions of the candidate driving scene in the location dimension; calculating the operation matching score corresponding to the current operation of the target user according to the preset conditions of the candidate driving scene in the user operation dimension; calculating the matching score of the driving data with any candidate driving scene based on the time matching score, the location matching score and the operation matching score; The candidate driving scene whose matching score exceeds the score threshold is determined as the target driving scene.

2. The method according to claim 1, characterized in that The determining of the target workflow corresponding to the target driving scenario includes: Acquire target user portrait data of the target user in the target driving scenario, and generate the target workflow according to the acquired target user portrait data; or, Querying a candidate driving scene matching the target driving scene from a workflow database, and determining a candidate workflow corresponding to the queried candidate driving scene as the target workflow, wherein the workflow database stores a plurality of candidate driving scenes and their corresponding candidate workflows, and the candidate workflows are pre-generated based on the comprehensive user portrait data of the target user; The target user portrait data or the comprehensive user portrait data is used to characterize the vehicle usage habits of the target user.

3. The method according to claim 2, characterized in that Generate workflows based on user portrait data, including: The user portrait data and prompt words are input into a workflow generation model, so that the workflow generation model generates and outputs a workflow matching the user portrait data under the instruction of the prompt words.

4. The method according to claim 3, characterized in that The workflow generation model is trained in the following way: Obtain sample user portrait data and a sample workflow corresponding to the sample user portrait data; Inputting the sample user portrait data and the sample prompt word into a pre-trained workflow generation model, so that the pre-trained workflow generation model generates and outputs a sample prediction workflow matching the sample user portrait data under the instruction of the sample prompt word; According to the difference between the sample workflow and the sample prediction workflow, the sample prompt words are adjusted and the pre-trained workflow generation model is iteratively trained.

5. The method according to claim 1, characterized in that The executing the target workflow includes: The target workflow is output to the target user, and upon receiving a confirmation instruction from the target user for the target workflow, the target workflow is executed.

6. The method according to claim 1, characterized in that Also includes: Acquire feedback information of the target user on the target workflow to optimize the target workflow based on the feedback information, wherein the feedback information includes evaluation information and / or improvement suggestions on the target workflow.

7. A device for actively providing vehicle services, characterized in that: include: A driving scene determination unit, used to obtain driving data of a vehicle, and determine a target driving scene of the vehicle according to the obtained driving data, wherein the driving data includes a current driving time, a current geographical location of the vehicle, and a current operation of a target user; a workflow determination unit, configured to determine a target workflow corresponding to the target driving scenario, wherein the target workflow is generated based on a target user's vehicle usage habits and includes at least one target vehicle service, and the target user is a user currently driving the vehicle; A workflow execution unit, configured to execute the target workflow to provide the target vehicle service to the target user; The determining the target driving scene of the vehicle according to the acquired driving data includes: Determine the preset conditions of each candidate driving scene in the candidate driving scenes; the preset conditions of any candidate driving scene include preset conditions in the time dimension, preset conditions in the location dimension, and preset conditions in the user operation dimension; Calculating the matching score of the driving data with each candidate driving scene according to the preset conditions of each candidate driving scene respectively; wherein, calculating the matching score of the driving data with any candidate driving scene according to the preset conditions of the candidate driving scene includes: calculating the time matching score corresponding to the current driving moment according to the preset conditions of the candidate driving scene in the time dimension; calculating the location matching score corresponding to the current geographical location according to the preset conditions of the candidate driving scene in the location dimension; calculating the operation matching score corresponding to the current operation of the target user according to the preset conditions of the candidate driving scene in the user operation dimension; calculating the matching score of the driving data with any candidate driving scene based on the time matching score, the location matching score and the operation matching score; The candidate driving scene whose matching score exceeds the score threshold is determined as the target driving scene.

8. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor implements the method according to any one of claims 1 to 6 by running the executable instructions.

9. A computer-readable storage medium having computer instructions stored thereon, characterized in that: When the instruction is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that The method comprises a computer program and / or instructions, which, when executed by a processor, implement the steps of the method according to any one of claims 1 to 6.

Images