Task processing method and device, electronic equipment and storage medium

Abstract

The invention relates to the field of vehicle intelligent control, in particular to a task processing method and device, electronic equipment and a storage medium. According to the method, the intention is determined by identifying the target voice data of the user, and the user does not need to manually install a plurality of specific application programs. The user only needs to express the requirement through voice, and the system automatically plans and executes related subtasks, so that the tedious operation of installing numerous applications and frequently updating and maintaining in a traditional mode is avoided. Secondly, vehicle-mounted resources are obtained according to the target subtask, the task is executed, and the situation that each application independently runs and occupies the resources is avoided; therefore, the resources can be dynamically allocated according to specific requirements, and unnecessary resource reutilization is avoided. And finally, information of different sources can be integrated in the process of analyzing the target sub-task and the task execution condition. When the subtasks are planned according to the target intention of the user, cooperation among different functions is realized through comprehensive planning and execution of the tasks, so that data isolation is broken.

Description

Technical Field

[0001] The present invention relates to the field of vehicle intelligent control, and in particular to a task processing method, device, electronic equipment and storage medium. Background Art

[0002] With the rapid advancement of technology, smart devices have been deeply integrated into people's lives. Smart vehicles provide rich functions such as navigation guidance and in-car entertainment for travel. For example, navigation applications can accurately plan travel routes for people, music playback applications provide a large amount of music resources for appreciation, social media applications allow people to interact and share life with friends, relatives and strangers, and health monitoring applications help people pay attention to their own physical condition.

[0003] However, the existing in-vehicle application ecosystem has exposed many drawbacks. In terms of installation and management, users have to install many applications on their devices and update and maintain them regularly, which not only takes up a lot of storage space, but also forces users to perform complex and tedious operations. In terms of resource utilization, each application runs independently, resulting in the duplication and waste of computing power and storage space. At the same time, data is isolated from each other between different applications, making it difficult to share and integrate, which greatly limits the user's ability to obtain and use information and forms a data island phenomenon. These problems combined have seriously affected the efficiency, convenience and user-friendly experience of the smart device application ecosystem. Summary of the invention

[0004] In view of this, embodiments of the present invention provide a task processing method, apparatus, electronic device and storage medium to solve the problem of how to simplify the installation and management of applications on smart devices, avoid waste of resources, and break down data silos between different applications.

[0005] In a first aspect, an embodiment of the present invention provides a task processing method, the method comprising:

[0006] Recognize the target voice data currently input by the user to obtain at least one target user intention of the user;

[0007] According to the target user intention, planning at least one target subtask associated with the target user intention;

[0008] Acquire the vehicle resources corresponding to the target subtask, and execute the target subtask based on the vehicle resources to obtain the task execution status;

[0009] The target subtask and the task execution status corresponding to the target subtask are analyzed to obtain an analysis result, and the target subtask is updated using the analysis result, and the updated target subtask is executed.

[0010] Furthermore, the step of performing recognition based on the target voice data currently input by the user to obtain at least one target user intention of the user includes:

[0011] Acquire the target voice data currently input by the user;

[0012] Parsing the target voice data to obtain text content corresponding to the target voice data;

[0013] The text content is identified, semantic information of the text content is obtained, and the semantic information is analyzed to obtain at least one target user intention of the user.

[0014] Further, the analyzing the semantic information to obtain at least one target user intention of the user includes:

[0015] Detecting the semantic information to obtain keywords contained in the semantic information;

[0016] Querying a first user intention related to the keyword, and modifying the first user intention based on the context corresponding to the keyword to obtain a second user intention;

[0017] The current behavior data of the user is detected, and if the behavior data matches the second user intention, the first user intention is used as the target user intention.

[0018] Furthermore, planning at least one target subtask associated with the target user intention according to the target user intention includes:

[0019] Obtaining a mapping relationship between a preset user intention and a preset subtask;

[0020] Based on the mapping relationship, the original subtask corresponding to the user intention is obtained, and the scene information in which the user is currently located is obtained;

[0021] A target subtask corresponding to the scene information is acquired from the original subtask.

[0022] Further, the acquiring a target subtask corresponding to the scene information from the original subtask includes:

[0023] Taking the original subtasks among the original subtasks that match the scene information as candidate subtasks;

[0024] Obtaining user preferences of the user and resource configuration of the vehicle;

[0025] The candidate subtasks are optimized using the user preferences and the resource configuration, and the optimized candidate subtasks are used as the target subtasks.

[0026] Furthermore, the analyzing the target subtask and the task execution status corresponding to the target subtask to obtain the analysis result includes:

[0027] Obtaining the user's feedback behavior and execution results based on the task execution status;

[0028] Analyze the feedback behavior to determine whether the target subtask achieves the expected effect corresponding to the target user intention, and obtain a first analysis content;

[0029] Analyze the execution result to determine the planning accuracy corresponding to the target subtask, and obtain a second analysis content;

[0030] The analysis result is generated based on the first analysis content and the second analysis content.

[0031] Further, the analyzing the feedback behavior to determine whether the target subtask achieves the expected effect corresponding to the target user intention, and obtaining the first analysis content, includes:

[0032] Acquire the action feedback data and the voice feedback data of the user based on the feedback behavior;

[0033] Analyzing the action feedback data to obtain task adjustment intention, and analyzing the voice feedback data to obtain emotional tendency;

[0034] The task adjustment intention and the emotional tendency are used as the first analysis content.

[0035] Furthermore, the analyzing the execution result to determine the planning accuracy corresponding to the target subtask, and obtaining the second analysis content, includes:

[0036] Determine the task completion status corresponding to the target subtask based on the execution result;

[0037] Analyze whether there is a deviation between the task completion status and the target user intention;

[0038] If there is a deviation between the task completion status and the target user intention, the reason for the deviation is determined and used as the second analysis content.

[0039] Further, the target subtask is updated using the analysis result, and the updated target subtask is executed, including:

[0040] Determining update content based on the first analysis content and / or the second analysis content, wherein the update content includes replacing a target subtask and / or updating a task parameter of the target subtask;

[0041] The target subtask is updated using the update content to obtain an updated target subtask, and the updated target subtask is executed.

[0042] In a second aspect, an embodiment of the present invention provides a task processing device, the device comprising:

[0043] A recognition module, configured to perform recognition based on target voice data currently input by the user to obtain at least one target user intention of the user;

[0044] A planning module, configured to plan at least one target subtask associated with the target user's intention according to the target user's intention;

[0045] An acquisition module, used to acquire the vehicle resources corresponding to the target subtask, and execute the target subtask based on the vehicle resources to obtain the task execution status;

[0046] The analysis module is used to analyze the target subtask and the task execution status corresponding to the target subtask to obtain the analysis result, and use the analysis result to update the target subtask and execute the updated target subtask.

[0047] In a third aspect, an embodiment of the present invention provides an electronic device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, computer instructions being stored in the memory, and the processor executing the method of the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0048] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the method of the first aspect or any corresponding embodiment thereof.

[0049] The method provided in the embodiment of the present application has the following beneficial effects:

[0050] The method provided in the embodiment of the present application first determines the intention by identifying the user's target voice data, without the need for the user to manually install multiple specific applications. The user only needs to express the needs through voice, and the system automatically plans and executes related subtasks, avoiding the tedious operations of installing many applications and frequently updating and maintaining them in the traditional way. Secondly, the vehicle resources are obtained and the tasks are executed according to the target subtasks, rather than allowing each application to run independently and occupy resources. In this way, resources can be dynamically allocated according to specific needs to avoid unnecessary resource reuse. Finally, the process of analyzing target subtasks and task execution can integrate information from different sources. When subtasks are planned according to the user's target intentions, the coordination between different functions is achieved through the comprehensive planning and execution of tasks, thereby breaking the data isolation. For example, if a user has the needs of navigation and playing music at the same time, the system can integrate the data and resources related to these two functions to better meet user needs.

[0051] The embodiment of the present application first obtains keywords by detecting semantic information, and then corrects the first user intention in combination with the context, so as to more accurately understand the user's real needs and avoid misunderstandings caused by the ambiguity of a single keyword. At the same time, the user's current behavior data is detected for further verification to ensure that the determined target user intention is more in line with the actual situation, thereby improving the accuracy and reliability of intent recognition. In the process of analyzing the user's intention to obtain the target subtask, the mapping relationship between the preset user intention and the preset subtask is obtained, which provides a basis for quickly planning subtasks. The target subtask is filtered out from the original subtask in combination with the user's current scene information, so that the planning of the subtask is more personalized and practical, which can better meet the needs of users in specific scenarios, improve user experience, and achieve more intelligent and efficient system services.

[0052] The embodiment of the present application first obtains the user's feedback behavior and execution results to fully understand the actual execution of the task. Analyzing the feedback behavior can accurately determine whether the target subtask has achieved the expected effect corresponding to the target user's intention, so as to adjust the task in time to better meet user needs and improve user experience. At the same time, analyzing the execution results can determine the planning accuracy corresponding to the target subtask, which helps to discover problems in task planning and improve them, and improve the accuracy and efficiency of task execution. The analysis results generated based on the first analysis content and the second analysis content can provide a strong basis for the continuous optimization of the system, so that the system can continuously learn and adapt to user needs and usage scenarios, and realize more intelligent services.

[0053] The embodiment of the present application first obtains the user's feedback behavior and execution results to fully understand the actual execution of the task. Analyzing the feedback behavior can accurately determine whether the target subtask has achieved the expected effect corresponding to the target user's intention, so as to adjust the task in time to better meet user needs and improve user experience. At the same time, analyzing the execution results can determine the planning accuracy corresponding to the target subtask, which helps to discover problems in task planning and improve them, and improve the accuracy and efficiency of task execution. The analysis results generated based on the first analysis content and the second analysis content can provide a strong basis for the continuous optimization of the system, so that the system can continuously learn and adapt to user needs and usage scenarios, and realize more intelligent services. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0055] Figure 1 is a flowchart of a task processing according to some embodiments of the present invention;

[0056] Figure 2 is a schematic diagram of a system framework according to some embodiments of the present invention;

[0057] Figure 3 is a schematic diagram of voice data processing according to some embodiments of the present invention;

[0058] Figure 4 is a flowchart of a task processing according to some embodiments of the present invention;

[0059] Figure 5 is a schematic diagram of vehicle-mounted resources according to some embodiments of the present invention;

[0060] Figure 6 is a complete flowchart of task processing according to some embodiments of the present invention;

[0061] Figure 7 is a structural block diagram of a task processing device according to an embodiment of the present invention;

[0062] Figure 8 It is a schematic diagram of the hardware structure of the electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0063] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0064] According to an embodiment of the present invention, a task processing method, apparatus, electronic device and storage medium are provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0065] In this embodiment, a task processing method is provided. Figure 1 is a flowchart of a task processing method according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps:

[0066] Step S101: Recognize the target voice data currently input by the user to obtain at least one target user intention of the user.

[0067] In an embodiment of the present application, the user's voice input is collected through a vehicle-mounted microphone or other equipment. Ensure the sensitivity and clarity of the microphone to accurately obtain the user's voice information. Use the ASR (automatic speech recognition) service to convert the collected audio information into text content. The ASR service can use a deep learning algorithm and improve the accuracy of speech recognition after training with a large amount of voice data. Then use natural language processing technology to analyze the text content and identify the user's intention. For example, establish an intention recognition model by training a large amount of text data with good intentions. For example, use word vector representation, part-of-speech tagging, named entity recognition and other methods to extract key information from the text, and then judge the user's intention based on this information. If words such as "tired" and "want to rest" appear in the text, it may be recognized that the user has the intention to relax.

[0068] Step S102: planning at least one target subtask associated with the target user's intention according to the target user's intention.

[0069] In an embodiment of the present application, a library of mapping relationships between user intentions and possible subtasks is first established. For example, the user intention of relaxation can correspond to subtasks such as playing soothing music, adjusting the temperature in the car, and reducing the brightness of the lights in the car. The subtasks are further refined according to the specific in-vehicle scenarios and available resources. If the vehicle supports the seat massage function, a subtask of starting the seat massage can be added when the user has the intention to relax. At the same time, the user's personalized preferences are taken into account. If it is known that the user likes a specific type of music, a music style that suits the user's preferences is selected in the subtask of playing music. For the multiple planned subtasks, they can be prioritized according to the urgency and importance of the user's intentions. For example, if the user expresses a strong sense of fatigue, then the relaxation-related subtasks can be set to a high priority.

[0070] Step S103, obtaining the vehicle resources corresponding to the target subtask, and executing the target subtask based on the vehicle resources to obtain the task execution status.

[0071] In the embodiment of the present application, first, the vehicle resources required for each subtask are determined. For example, the subtask of playing music requires obtaining music resources, which can be obtained from channels such as vehicle-mounted music storage devices and online music platforms. If it is a subtask of adjusting the temperature in the car, it is necessary to obtain the data of the vehicle temperature sensor and the permissions of the air conditioning control system. Through the interface and protocol of the vehicle system, various resources are interacted and obtained.

[0072] Execute subtasks based on the acquired resources. For the subtask of playing music, call the music player software and set the appropriate volume and play mode. For the subtask of adjusting the temperature in the car, adjust the temperature to the preset value by controlling the air conditioning system. During the execution process, record the execution status and progress of the task for subsequent analysis. Then continue to monitor the execution of the task, such as whether the music is playing normally and whether the temperature is stable at the set value. Real-time information on the execution of the task can be obtained through sensor feedback, software status monitoring, etc.

[0073] Step S104, analyzing the target subtask and the execution status of the task corresponding to the target subtask to obtain an analysis result, and using the analysis result to update the target subtask, and executing the updated target subtask.

[0074] In an embodiment of the present application, the execution result of each subtask is evaluated to determine whether the expected user intention has been achieved. For example, whether the played music makes the user feel relaxed, and whether the adjusted temperature in the car meets the user's needs. The execution results can be evaluated through user feedback, physiological indicator monitoring (such as heart rate, breathing, etc.), etc. Analyze the rationality of task planning, check whether the generated subtasks fully and accurately correspond to the user's intentions, and whether there are any missing or redundant subtasks. For example, if the user still feels uncomfortable during the execution process, it may be necessary to reflect on whether the subtask planning is complete and whether other subtasks need to be added.

[0075] Update the target subtask based on the analysis results. If the execution result is not ideal, adjust the subtask parameters, such as increasing the volume of the music, adjusting the temperature range, etc. You can also delete or add new subtasks to better meet the user's intention. Execute the updated target subtask, obtain the vehicle resources again and execute the task. Continue to pay attention to the task execution and user feedback, and continue to adjust and optimize until the user's satisfaction is achieved.

[0076] In order to more clearly introduce the implementation process of the above-mentioned task processing method, the following is a system architecture diagram of the task processing method of an embodiment of the present application. Figure 2 As shown, the system framework includes: input layer, speech processing layer, model scheduling layer, infrastructure layer, storage layer, and the functions of each layer are described as follows:

[0077] Input Layer:

[0078] Users can use the microphone in the car to input their needs and intentions. The microphone should have high sensitivity and noise immunity to ensure that the user's voice can be accurately collected in various in-car environments.

[0079] The speakers and loudspeakers are used for audio output, which provide feedback to users on the system's processing results and prompt information. For example, during speech recognition, a prompt tone can be played to indicate that the system is processing the speech; when feedback is given on the task execution results, a voice broadcast can be played to inform users of the completion of the task.

[0080] Speech processing layer:

[0081] Speech recognition: Speech recognition technology is used to convert user voice input into text information. Deep learning algorithms can be used to improve the accuracy and robustness of speech recognition through training with a large amount of speech data. It adapts and optimizes to different accents, speaking speeds, and noise environments to ensure that the user's voice can be accurately recognized in various situations.

[0082] Acoustic processing: Perform acoustic processing on the collected voice signals, including noise reduction, echo cancellation, gain control, etc. These processes can improve the quality of voice signals, reduce noise interference, and make voice recognition more accurate. The acoustic processing algorithm should be optimized according to the characteristics of the in-vehicle environment to adapt to different in-vehicle noise and echo conditions.

[0083] Voice announcement: Generate voice announcement content based on the system's processing results and prompt information. Voice announcements should have clear and natural voice quality so that users can easily understand. Speech synthesis technology can be used to convert text information into voice signals, which are then played to users through the audio system. The speech synthesis algorithm should be able to generate natural and fluent speech, and can be personalized according to user needs.

[0084] Model scheduling layer:

[0085] Intent recognition: Perform intent recognition on the text information output by the speech processing layer to determine the user's target user intent. Natural language processing techniques, such as word vector representation, part-of-speech tagging, and named entity recognition, can be used to extract key information from the text, and then judge the user's intent based on this information. Establish an intent recognition model and improve the accuracy and generalization ability of intent recognition by training a large amount of text data with well-labeled intent.

[0086] Task planning: Based on the identified target user intent, at least one target subtask associated with the target user intent is planned. A library of mapping relationships between user intent and subtasks can be established, and appropriate subtasks can be selected from the library based on the user intent. Considering the vehicle scenario and available resources, the planned subtasks can be refined and optimized. For example, based on the configuration and functions of the vehicle, determine which subtasks can be executed and which subtasks need to be adjusted or replaced.

[0087] Task execution: Obtain the vehicle resources corresponding to the target subtask and execute the target subtask based on the vehicle resources. You can interact and control various vehicle resources through the interfaces and protocols of the vehicle system. Monitor and manage the task execution process to ensure the smooth execution of the task. If problems or abnormal situations occur, handle and adjust them in a timely manner.

[0088] Resource Scheduling: Responsible for managing and scheduling onboard resources to meet the needs of task execution. According to the priority of the task and the availability of resources, resources can be reasonably allocated to improve resource utilization. Resource usage is monitored and counted for resource optimization and adjustment. For example, if it is found that certain resources are often occupied, you can consider increasing the number of resources or optimizing the resource allocation strategy.

[0089] ④Infrastructure layer:

[0090] Basic data storage: basic data required for the operation of the storage system, such as speech recognition models, intent recognition models, task planning libraries, etc. These data should be updated and maintained regularly to ensure the performance and accuracy of the system. You can use a database management system or a file system for data storage, and choose the appropriate storage method based on the characteristics of the data and access requirements.

[0091] Task storage: records the task information executed by the system, including task description, execution status, execution results, etc. This information can be used for task tracking and management, as well as subsequent analysis and optimization. You can use a database or log file to store tasks in order to query and count the execution of tasks.

[0092] Historical data storage: Stores the system's historical data, such as user voice input, intent recognition results, task execution status, etc. This data can be used for user behavior analysis and system performance optimization. A database or data warehouse can be used to store historical data for data analysis and mining.

[0093] Support system operation: Provide the infrastructure and services required for system operation, such as computing resources, storage resources, network communications, etc. Ensure the stability and reliability of the system, as well as efficient operation performance.

[0094] Storage layer:

[0095] Data storage: Responsible for storing various data in the system, including user data, task data, historical data, etc. You can use a variety of storage methods such as relational databases, non-relational databases, and file systems. Choose a suitable storage solution based on the characteristics of the data and access requirements. Back up and restore data to prevent data loss and damage. Ensure data security and integrity.

[0096] System information interaction: Realize information interaction and communication between various layers of the system. Technologies such as message queues, RPC (remote procedure calls), and RESTful APIs can be used to ensure timely delivery and accurate processing of information. Monitor and manage information interaction to ensure system stability and reliability.

[0097] Cache: To improve system performance and response speed, cache technology can be used. The cache system can store frequently accessed data and calculation results so that they can be quickly retrieved the next time they are accessed.

[0098] As an example, Figure 3As shown in the figure, when the user is driving the vehicle, he says to the vehicle system: "Play some dynamic music and check if there are any gas stations nearby." First, the user's voice input audio is received by the vehicle system and converted into a text demand "Play some dynamic music and check if there are any gas stations nearby" through ASR (automatic speech recognition) technology. Then, the large model in the vehicle system begins to process this demand. For the demand of playing dynamic music, the music player software is called to search for dynamic music tracks in the music library; for the demand of checking nearby gas stations, the map navigation software is called to find the information of nearby gas stations. Then, the system prepares a response text, such as "Playing dynamic music for you, there are 3 gas stations within 5 kilometers nearby." And it is converted into a voice response through TTS (text to speech) technology and played to the user. At the same time, the system executes the action, starts playing dynamic music, and displays the location information and navigation route of nearby gas stations on the vehicle display. If the user further issues instructions, such as selecting one of the gas stations for navigation, the system will perform corresponding actions according to the instructions, plan the route to the gas station and provide voice navigation prompts.

[0099] The method provided in the embodiment of the present application first determines the intention by identifying the user's target voice data, without the need for the user to manually install multiple specific applications. The user only needs to express the needs through voice, and the system automatically plans and executes related subtasks, avoiding the tedious operations of installing many applications and frequently updating and maintaining them in the traditional way. Secondly, the vehicle resources are obtained and the tasks are executed according to the target subtasks, rather than allowing each application to run independently and occupy resources. In this way, resources can be dynamically allocated according to specific needs to avoid unnecessary resource reuse. Finally, the process of analyzing target subtasks and task execution can integrate information from different sources. When subtasks are planned according to the user's target intentions, the coordination between different functions is achieved through the comprehensive planning and execution of tasks, thereby breaking the data isolation. For example, if a user has the needs of navigation and playing music at the same time, the system can integrate the data and resources related to these two functions to better meet user needs.

[0100] Figure 4 is a flowchart of a task processing method according to an embodiment of the present invention. Figure 4 As shown, the process includes the following steps:

[0101] Step S201: Recognize the target voice data currently input by the user to obtain at least one target user intention of the user.

[0102] In the embodiment of the present application, the target voice data currently input by the user is recognized to obtain at least one target user intention of the user, including the following steps A1-A3:

[0103] Step A1, obtaining the target voice data currently input by the user.

[0104] Specifically, the car computer is usually equipped with a microphone. When the user interacts with the car computer through voice in the car, the microphone of the car computer starts to collect sound signals. The microphone should have good sensitivity and anti-interference ability to ensure that the user's voice can be accurately collected in various environments in the car (such as noise during driving, passengers' conversations, etc.). The car computer system continuously monitors the microphone input, and once it detects that the sound signal exceeds a certain threshold, it starts to collect voice data. The collected voice data can be stored and processed in a specific audio format (such as WAV, MP3, etc.).

[0105] In addition, users can input voice commands through the voice assistant application on their mobile phones. The microphone of the mobile phone also needs to have high-quality acquisition capabilities to adapt to different usage scenarios (such as noisy outdoor environments, quiet indoor environments, etc.). When the user activates the voice assistant on the mobile phone and starts speaking, the mobile phone application collects voice data through the microphone. The collected voice data can be transmitted to the car system or other related devices connected to the mobile phone via Bluetooth, Wi-Fi or mobile network for further processing.

[0106] Step A2: parse the target voice data to obtain text content corresponding to the target voice data.

[0107] Specifically, the collected target voice data is preprocessed, including noise reduction, echo removal, audio enhancement and other operations. These processes can improve the quality of the voice signal and reduce errors in the subsequent voice recognition process. The voice recognition model is called to convert the preprocessed voice data into text content. The voice recognition model can be a deep learning-based model that can accurately convert voice into text after being trained with a large amount of voice data. The voice recognition model will analyze the voice based on the acoustic characteristics and language model, recognize the words in the voice signal frame by frame, and combine them into complete text content.

[0108] Step A3, identifying text content, obtaining semantic information of the text content, and analyzing the semantic information to obtain at least one target user intention of the user.

[0109] Specifically, the semantic information is analyzed to obtain at least one target user intention of the user, including: detecting the semantic information to obtain the keywords contained in the semantic information; querying the first user intention related to the keywords, and modifying the first user intention based on the contextual context corresponding to the keywords to obtain the second user intention; detecting the user's current behavior data, and if the behavior data matches the second user intention, taking the first user intention as the target user intention.

[0110] First, the semantic information of the acquired text content is deeply analyzed. Methods such as lexical analysis, syntactic analysis, and semantic role labeling in natural language processing technology can be used. For example, lexical analysis can be used to determine the part of speech and meaning of each word, syntactic analysis can be used to understand the structure of the sentence, and semantic role labeling can be used to determine the semantic role of each component in the sentence.

[0111] Secondly, keyword extraction algorithms are used to extract key words from semantic information. Common keyword extraction algorithms include TF-IDF (term frequency-inverse document frequency) algorithm, TextRank algorithm, etc. These algorithms determine keywords based on the frequency of occurrence of the words in the text and their importance in the entire corpus. For example, if words such as "music", "play", and "song" appear frequently in the text, these words may be extracted as keywords.

[0112] Then, obtain a pre-established mapping database of keywords and user intentions. After the keywords are extracted, query the database for possible user intentions related to these keywords. For example, if the keywords are "music" and "play", the possible first user intention is to listen to music. Consider the context of the keywords to further modify the first user intention. The context may include the context before and after the sentence, the user's historical interaction records, etc. For example, if the text mentions "I want to listen to music and relax while driving", then combined with the context, it can be determined that the user's intention is not only to listen to music, but also includes the need to relax in the driving scenario, so as to modify the first user intention and obtain a more accurate second user intention.

[0113] Next, the sensors and monitoring equipment of the vehicle system are used to detect the user's current behavior data. For example, the user's facial expressions and body postures can be detected through the in-vehicle camera, the vehicle's driving status (such as speed, acceleration, etc.) can be detected through the vehicle sensor, and the user's operating behavior can be detected through the interaction record between the user and the vehicle system. Match the detected behavior data with the second user's intention. If the behavior data is consistent with the second user's intention, it means that the user's actual behavior supports the inferred user intention. For example, if the second user's intention is to relax by listening to music, and it is detected that the user is adjusting the volume, switching music tracks, etc., these behaviors match the intention of relaxing by listening to music.

[0114] Finally, if the behavior data matches the second user intent, the first user intent is used as the target user intent. This means that after keyword analysis, contextual correction, and behavior data verification, the user's final intent is determined. For example, if the above process determines that the user's intent is to listen to music and relax, then this intent becomes the target user intent, and the system can use it to plan and execute subsequent tasks.

[0115] Step S202: According to the target user intention, at least one target subtask associated with the target user intention is planned.

[0116] In an embodiment of the present application, at least one target subtask associated with the target user's intention is planned according to the target user's intention, including the following steps B1-B3:

[0117] Step B1, obtaining a mapping relationship between a preset user intention and a preset subtask.

[0118] Specifically, first, create a database to store the mapping relationship between preset user intents and preset subtasks. This database can be a relational database or a non-relational database, depending on the needs and performance requirements of the system. For each preset user intent, determine a series of preset subtasks associated with it. For example, if the user intent is "relax", the preset subtasks may include playing soothing music, adjusting the temperature in the car, lowering the seat back, etc.

[0119] Step B2: Based on the mapping relationship, the original subtask corresponding to the user's intention is obtained, and the scene information in which the user is currently located is obtained.

[0120] Specifically, the target user intent of the user is determined according to the previous steps. The user intent can be extracted from the user input through speech recognition, semantic analysis and other technologies. Once the user intent is determined, the corresponding preset subtasks are searched in the mapping database, and these subtasks are the original subtasks.

[0121] Use the sensors and devices of the vehicle system to obtain the scene information of the user. For example, the vehicle location information can be obtained through the vehicle GPS to determine whether the user is driving on a city road, highway or rural road; the temperature inside the car can be obtained through the temperature sensor inside the car; the light sensor can be used to obtain the light intensity inside the car, etc. It can also be combined with time information to determine whether it is daytime or nighttime, weekday or weekend, etc. This scene information will be used to further filter subtasks.

[0122] Step B3, obtaining the target subtask corresponding to the scene information from the original subtask.

[0123] Specifically, a target subtask corresponding to the scene information is obtained from the original subtask, including: taking the original subtask that matches the scene information in the original subtask as a candidate subtask; obtaining the user's user preference and the vehicle's resource configuration; optimizing the candidate subtask using the user preference and resource configuration, and taking the optimized candidate subtask as the target subtask.

[0124] First, the original subtasks are analyzed one by one to determine the degree of match between each subtask and the current scene information. For example, if the scene information shows that the vehicle is driving at high speed, then subtasks related to safe driving (such as adjusting the cruising speed, reminding to pay attention to road conditions, etc.) may be more suitable; if it is a night driving scene, subtasks such as turning on the interior ambient light may be more appropriate.

[0125] Secondly, obtain user preference information from multiple channels. For example, by analyzing the user's historical operation records, understand the user's frequently used functions and setting preferences; read the user's preferences for music type, temperature range, seat angle, etc. from the user's personal settings; and actively collect user preference information through questionnaires or user feedback. Organize this user preference information into a structured data format for subsequent subtask optimization. At the same time, use the diagnostic tools or sensors of the vehicle system to obtain the vehicle's resource configuration information. This includes the vehicle's hardware equipment configuration, such as the brand and performance of the audio system, whether the seat has a massage function, the cooling and heating capacity of the air-conditioning system, etc.; it also includes software resource configuration, such as the media format supported by the vehicle entertainment system, the functions of the navigation software, etc.

[0126] Then, for each candidate subtask, make adjustments based on user preferences. For example, if the user prefers soft music, in the candidate subtask of playing music, select soft music tracks or music playlists; if the user likes lower temperatures, adjust the temperature setting values ​​in the candidate subtasks related to temperature. Consider the user's frequency of use and importance preferences for different functions. If the user often uses the seat heating function, in cold weather scenarios, prioritize the candidate subtasks related to seat heating. At the same time, determine the feasibility and optimization direction of the candidate subtasks based on the vehicle's resource configuration. If the vehicle's audio system is of high quality, in the candidate subtask of playing music, you can improve the music playback quality parameters, such as adding sound effect settings, adjusting the volume range, etc.; if the vehicle has advanced seat adjustment functions, in the candidate subtask of adjusting seats, you can make full use of these functions to provide more comfortable seat settings. For resource-constrained situations, make corresponding adjustments. For example, if the vehicle has limited storage space, in the candidate subtasks of downloading music or updating software, choose a more space-saving method or prioritize downloading the content that users use most frequently.

[0127] Finally, after optimizing user preferences and vehicle resource configuration, the final set of candidate subtasks is obtained. These optimized candidate subtasks are the target subtasks, which will be used in the subsequent task execution phase. Ensure that the target subtasks not only meet the user's personalized needs, but also make full use of the vehicle's resource configuration to provide a better user experience.

[0128] Step S203, obtaining the vehicle resources corresponding to the target subtask, and executing the target subtask based on the vehicle resources to obtain the task execution status.

[0129] In the embodiment of the present application, the vehicle resources are as follows: Figure 5 As shown, including music resources, video resources, navigation resources, radio resources and so on.

[0130] Analyze the target subtasks For each target subtask, carefully analyze its specific requirements. For example, if the target subtask is to play music, then the required in-vehicle resources are music resources; if it is to watch videos, video resources are needed; if it is to navigate, navigation resources are relied on, etc. Based on the analysis results of the target subtasks, list the types of in-vehicle resources required for each subtask. In this way, you can clearly understand what specific resources need to be obtained in order to perform these subtasks.

[0131] If the target subtask requires music resources, they can be obtained from the following channels: Onboard storage devices: Check the vehicle's built-in storage devices, such as hard drives or memory cards, to see if there are any music files pre-stored by the user. Online music platforms: If the vehicle supports Internet connection, online music platforms can be accessed through the vehicle system to obtain a large amount of music resources. This may require users to log in to their music platform accounts in order to obtain personalized music playlists and recommendations. Bluetooth connection: If the user has music resources on their mobile phone or other devices, these devices can be connected to the vehicle system via Bluetooth connection to obtain music resources.

[0132] If the target subtask requires video resources, they can be obtained from the following channels: In-vehicle multimedia system storage: The multimedia systems of some high-end vehicles may have a certain number of video resources built in for users to watch when parking or in specific situations. USB device: Users can insert a USB device with video files stored in it into the USB port of the in-vehicle system to obtain video resources.

[0133] If the target subtask requires navigation resources, they can be obtained from the following channels: In-vehicle navigation system: The vehicle's built-in navigation system usually stores map data and navigation software, and these resources can be used directly for navigation. Map data can be updated regularly to ensure accuracy. Mobile navigation synchronization: Synchronize the navigation application on the mobile phone with the vehicle system through Bluetooth or other connection methods to take advantage of the richer navigation resources and real-time traffic information on the mobile phone.

[0134] If the target subtask requires radio resources, they can be obtained in the following ways: Car radio: The vehicle's radio can receive various radio station signals and provide a variety of radio resources such as music, news, talk shows, etc. Internet radio: If the car system supports Internet connection, you can access the Internet radio platform to obtain more types of radio programs.

[0135] Once the music resources are acquired, music is played according to the requirements of the target subtask. Music can be played through the car audio system, and the volume, playback mode (single loop, random play, etc.) and sound effect settings can be adjusted. At the same time, the stability and smoothness of music playback are monitored to ensure that there are no freezes or interruptions.

[0136] For video playback tasks, use the vehicle display to play video resources under appropriate circumstances (such as when the vehicle is stationary). You can adjust parameters such as video playback progress, volume, and picture quality. Ensure the clarity and smoothness of video playback, as well as compatibility with the vehicle display.

[0137] When performing navigation tasks, start the in-vehicle navigation system or synchronized mobile navigation application. Enter the destination information, the system calculates the best route based on the navigation resources, and guides the user to the destination through voice prompts and map display. Monitor the vehicle's location and driving status in real time, and adjust the route suggestions in time according to traffic conditions.

[0138] For radio broadcast tasks, select the appropriate radio frequency and listen to radio programs through the car radio or online radio platform. You can adjust the volume and switch radio channels to meet the user's listening needs.

[0139] During the execution of the target subtask, the task execution is continuously monitored. For example, for the music playing task, monitor whether the music is playing normally and whether the volume is stable; for the navigation task, monitor whether the vehicle position is accurate and whether the route guidance is clear. Record the execution status of each target subtask, including the start time, end time, whether there are any problems in the middle, etc. These records can be used for subsequent analysis and optimization, as well as to provide the task execution history when the user needs it. At the same time, collect user feedback information to understand the user's satisfaction with the task execution. Users can provide feedback through voice commands, touch screen operations or other interactive methods, such as evaluation of music playback, accuracy of navigation routes, etc.

[0140] Step S204, analyzing the target subtask and the execution status of the task corresponding to the target subtask to obtain an analysis result, and using the analysis result to update the target subtask, and executing the updated target subtask.

[0141] In the embodiment of the present application, analyzing the target subtask and the task execution status corresponding to the target subtask to obtain the analysis result includes the following steps C1-C4:

[0142] Step C1, obtaining user feedback behavior and execution results based on task execution status.

[0143] Specifically, during the execution of the target subtask, the various sensors and monitoring tools of the vehicle system are continuously used to track the execution of the task. For example, for the subtask of playing music, the smoothness of the music playing, the volume, the current track being played, etc. can be monitored; for the subtask of navigation, the real-time location of the vehicle, the accuracy of the navigation route, the clarity of the voice prompts, etc. can be monitored.

[0144] Collect user feedback through various methods, including the following:

[0145] Voice feedback: The vehicle's voice recognition system listens to the user's voice comments and instructions. For example, the user may say "the music is too loud" or "the navigation is inaccurate".

[0146] Touch screen operation: If the in-vehicle system has a touch screen, observe whether the user makes adjustments through the touch screen, such as adjusting the volume, switching tracks, re-planning the navigation route, etc. These operations can reflect the user's dissatisfaction with the current task execution or further needs.

[0147] Physical button operation: Physical buttons on the vehicle can also provide clues to user feedback. For example, frequent pressing of the volume adjustment button by the user may indicate dissatisfaction with the current volume.

[0148] Step C2, analyzing the feedback behavior to determine whether the target subtask achieves the expected effect corresponding to the target user's intention, and obtaining the first analysis content.

[0149] Specifically, the feedback behavior is analyzed to determine whether the target subtask has achieved the expected effect corresponding to the target user's intention, and the first analysis content is obtained, including: obtaining the user's action feedback data and voice feedback data based on the feedback behavior; analyzing the action feedback data to obtain the task adjustment intention, and analyzing the voice feedback data to obtain the emotional tendency; taking the task adjustment intention and the emotional tendency as the first analysis content.

[0150] Specifically, the collected action feedback data is classified. For example, the operations on the touch screen are divided into different categories, such as volume adjustment, track switching, navigation operation, etc. The meaning of each action is interpreted to determine the user's task adjustment intention. For example, if a user frequently clicks the volume up button, it may mean that he wants to increase the volume of the music; if a user drags on the navigation interface, it may mean that he wants to view different route options. Create a mapping relationship between action feedback data and task adjustment intention. By analyzing and summarizing a large number of user actions, the task adjustment intentions corresponding to different actions are determined. For example, a specific combination of touch screen operations may mean that the user wants to switch to a specific music playback mode. Using this mapping relationship, the user's action feedback data is converted into a clear task adjustment intention.

[0151] Then, the sentiment analysis algorithm in natural language processing technology is used to determine whether the sentiment expressed in the user's voice is positive, negative, or neutral. For example, if the user says "This music is nice", the sentiment is positive; if the user says "The navigation is terrible", the sentiment is negative. In addition to determining the sentiment, the intensity of the sentiment can also be evaluated. For example, the sentiment intensity expressed by the user when they say "The music is a bit loud" and "The music is too loud and I can't stand it" is different. By evaluating the intensity of the sentiment, we can better understand the user's satisfaction with the task execution and the degree of urgency to adjust.

[0152] The task adjustment intention and emotional tendency obtained by analysis are integrated to form the first analysis content. The first analysis content can be presented in a structured manner, for example, including a specific description of the task adjustment intention, the type and intensity of the emotional tendency, etc. The first analysis content can serve as an important basis for subsequent task optimization. According to the task adjustment intention, the system can make corresponding adjustments to the currently executed task, such as adjusting the volume, changing the track, replanning the navigation route, etc. At the same time, according to the emotional tendency, the system can understand the user's satisfaction with the task execution, so as to further improve and optimize the task execution effect and enhance the user experience.

[0153] Step C3, analyzing the execution result to determine the planning accuracy corresponding to the target subtask, and obtaining the second analysis content.

[0154] In an embodiment of the present application, the execution result is analyzed to determine the planning accuracy corresponding to the target subtask, and a second analysis content is obtained, including: determining the task completion status corresponding to the target subtask based on the execution result; analyzing whether there is a deviation between the task completion status and the target user intention; if there is a deviation between the task completion status and the target user intention, determining the reason for the deviation, and using the reason as the second analysis content.

[0155] First, the collected execution result data is analyzed to evaluate the completion of the target subtask. The successful completion of the task can be judged based on the pre-set standards or indicators. For example, for the subtask of playing music, if the music can be played normally and the volume meets the user's requirements, the task can be considered to be completed well; for the navigation subtask, if the vehicle can accurately drive according to the planned route and arrive at the destination within a reasonable time, the task is completed well.

[0156] Review the target user intent identified earlier to ensure a clear understanding of the user's needs. For example, if the user's target user intent is to relax while driving, then the relevant subtasks may include playing soothing music, adjusting the temperature in the car, etc. Compare the execution results of the target subtasks with the target user intent. Determine whether the execution results meet the user's expectations and are consistent with the user's intent. For example, if the user wants to play soothing music to relax, but the music actually played is too fast or too loud, then there is a deviation.

[0157] When it is found that there is a deviation between the task completion and the target user's intention, it is necessary to deeply analyze the reasons for the deviation. Possible reasons include but are not limited to the following aspects: ① Inaccurate understanding of user intention: errors may occur in the process of identifying user intention, resulting in the planned subtasks not meeting the user's real needs. ② Resource limitations: Insufficient or faulty on-board resources affect the execution of subtasks. For example, the music resource library does not have the type of music expected by the user, or there is a signal problem with the navigation system. ③ Interference from external factors: For example, traffic congestion affects the execution of navigation tasks, or noise in the car interferes with the effect of music playback. ④ Unreasonable subtask planning: The actual situation may not be taken into account when planning subtasks, resulting in unsatisfactory execution results.

[0158] Through further investigation and analysis, determine the specific cause of the deviation. The cause can be determined by combining user feedback, system log records, and review of the execution process. For example, check the user's operation records during the task execution process to analyze whether some user operations caused the deviation; check the system's error log to see if there are resource failures or software problems. Once the cause of the deviation is determined, record it as the second analysis content. The second analysis content can include the specific manifestations of the deviation, the determined causes, and possible solution suggestions. For example, if the cause of the deviation is an inaccurate understanding of the user's intent, you can record the erroneous understanding and the correct interpretation of the user's intent; if it is a resource limitation, you can explain the specific resource problem and possible solutions, such as updating the music resource library or fixing the navigation system signal problem.

[0159] Step C4, generating an analysis result based on the first analysis content and the second analysis content.

[0160] The embodiment of the present application first obtains the user's feedback behavior and execution results to fully understand the actual execution of the task. Analyzing the feedback behavior can accurately determine whether the target subtask has achieved the expected effect corresponding to the target user's intention, so as to adjust the task in time to better meet user needs and improve user experience. At the same time, analyzing the execution results can determine the planning accuracy corresponding to the target subtask, which helps to discover problems in task planning and improve them, and improve the accuracy and efficiency of task execution. The analysis results generated based on the first analysis content and the second analysis content can provide a strong basis for the continuous optimization of the system, so that the system can continuously learn and adapt to user needs and usage scenarios, and realize more intelligent services.

[0161] In an embodiment of the present application, the target subtask is updated using the analysis results, and the updated target subtask is executed, including: determining the update content based on the first analysis content and / or the second analysis content, wherein the update content includes replacing the target subtask and / or updating the task parameters of the target subtask; updating the target subtask using the update content to obtain an updated target subtask, and executing the updated target subtask.

[0162] Specifically, review the first analysis content, that is, the analysis results of whether the target subtask achieves the expected effect corresponding to the target user's intent. If it is found that the target subtask does not achieve the expected effect, it is necessary to determine the specific problem. For example, if the user reports that the music playback is not smooth or the music type does not meet expectations, this indicates that the user intent is not met in the music playback subtask. Determine possible update directions based on the problem. If the music playback is not smooth, consider changing the music playback source or optimizing the network connection; if the music type does not meet expectations, you can adjust the music playlist or reselect the music style according to user preferences.

[0163] At the same time, check the second analysis content, that is, the analysis results of the planning accuracy corresponding to the target subtask. If the planning accuracy is not high, it may be that the task parameter settings are unreasonable or there is a problem with the task planning itself. For example, if the route planned by the navigation subtask is congested or takes a detour, it may be necessary to adjust the parameters of the navigation algorithm or replan the route. Determine the update content based on the planning accuracy problem. If it is a task parameter problem, you can adjust the parameter value, such as adjusting the weight parameter for avoiding congestion in the navigation subtask; if it is a task planning problem, you can consider replacing the entire subtask, such as replacing the original navigation subtask with a different navigation software for planning.

[0164] Combine the first analysis content with the second analysis content, and comprehensively consider and determine the update content. It may be necessary to replace the target subtask and update the task parameters at the same time. For example, if the user is not satisfied with the music playback and has opinions on the navigation route, on the one hand, the music playback software can be replaced (replace the target subtask), and on the other hand, the navigation parameters can be adjusted to improve the planning accuracy (update the task parameters).

[0165] Update the target subtask according to the determined update content. If the target subtask is to be replaced, the new subtask needs to be integrated into the system and ensure coordination and compatibility with other subtasks. For example, if the music playback subtask is replaced with a new music service, it is necessary to ensure that the new music service can work well with the control interface and voice interaction system of the vehicle system. If the task parameters are updated, the corresponding parameter setting points need to be found in the system and adjusted. For example, in the navigation subtask, the parameters of the navigation algorithm are adjusted through the system settings interface.

[0166] After the update is completed, execute the updated target subtask. Reacquire the vehicle resources and execute the task according to the new task flow. For example, after updating the music playback subtask, reacquire the music resources and play the music; after updating the navigation subtask, replan the route and navigate. During the execution process, continuously monitor the task execution and user feedback to ensure that the updated target subtask can effectively meet user needs. If there are still problems, you can analyze and update again to continuously optimize the system performance and user experience.

[0167] The embodiment of the present application first obtains the user's feedback behavior and execution results to fully understand the actual execution of the task. Analyzing the feedback behavior can accurately determine whether the target subtask has achieved the expected effect corresponding to the target user's intention, so as to adjust the task in time to better meet user needs and improve user experience. At the same time, analyzing the execution results can determine the planning accuracy corresponding to the target subtask, which helps to discover problems in task planning and improve them, and improve the accuracy and efficiency of task execution. The analysis results generated based on the first analysis content and the second analysis content can provide a strong basis for the continuous optimization of the system, so that the system can continuously learn and adapt to user needs and usage scenarios, and realize more intelligent services.

[0168] Figure 6 This is a complete task processing flow provided by the embodiment of the present application, such as Figure 6 As shown, the user inputs voice based on the in-vehicle system during driving: "I want to listen to relaxing music and plan a route to the beach." The system converts the user's voice input into text content through the ASR (automatic speech recognition) tool.

[0169] First, the system identifies the user's target user intent, which is to listen to relaxing music and navigate to the beach. Based on this intent, two target subtasks are planned: subtask 1 is to play relaxing music, and subtask 2 is to plan a route to the beach. For subtask 1, the system obtains music resources and plays relaxing music tracks. At the same time, the execution of tasks is continuously monitored, such as whether the music is played normally and whether the volume is appropriate. If the user indicates through feedback that the music is not relaxing enough or the volume is too loud, the system will analyze these feedback behaviors and execution results. For example, if the analysis determines that the user's expected effect on the music is not achieved, it may be that the music style selection is inaccurate, and this is used as the first analysis content. Then, based on the first analysis content, the update content is determined, such as changing the music playlist or adjusting the volume parameters. The subtask is updated with the updated content, and the relaxing music that meets the user's needs is played again.

[0170] For subtask 2, the system calls navigation resources to plan a route to the beach. During the execution process, if traffic congestion or other situations occur, causing the planning accuracy to decrease, the system will analyze the execution results, determine the planning accuracy problem and use it as the second analysis content. Then, based on the second analysis content, determine the update content, such as replanning a route to avoid congestion. After updating the subtask, execute the updated navigation task. Throughout the process, the system continuously reflects and optimizes based on user feedback and task execution to ensure that each subtask can better meet user intentions and enhance the user experience during driving.

[0171] In this embodiment, a task processing device is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware of a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0172] The present application embodiment provides a task processing device, such as Figure 7 As shown, the device comprises:

[0173] Recognition module 701, used to perform recognition based on target voice data currently input by the user to obtain at least one target user intention of the user;

[0174] A planning module 702, configured to plan at least one target subtask associated with the target user's intention according to the target user's intention;

[0175] The acquisition module 703 is used to acquire the vehicle resources corresponding to the target subtask, and execute the target subtask based on the vehicle resources to obtain the task execution status;

[0176] The analysis module 704 is used to analyze the target subtask and the execution status of the task corresponding to the target subtask, obtain the analysis result, update the target subtask using the analysis result, and execute the updated target subtask.

[0177] In an embodiment of the present application, the recognition module 701 is used to obtain the target voice data currently input by the user; parse the target voice data to obtain the text content corresponding to the target voice data; recognize the text content to obtain the semantic information of the text content, and analyze the semantic information to obtain at least one target user intention of the user.

[0178] In an embodiment of the present application, the identification module 701 is specifically used to detect semantic information and obtain keywords contained in the semantic information; query the first user intention related to the keyword, and modify the first user intention based on the context corresponding to the keyword to obtain the second user intention; detect the user's current behavior data, and if the behavior data matches the second user intention, the first user intention is used as the target user intention.

[0179] In an embodiment of the present application, the planning module 702 is used to obtain a mapping relationship between a preset user intention and a preset subtask; obtain the original subtask corresponding to the user intention based on the mapping relationship, and obtain the scene information in which the user is currently located; and obtain the target subtask corresponding to the scene information from the original subtask.

[0180] In an embodiment of the present application, the planning module 702 is specifically used to select the original subtasks in the original subtasks that match the scene information as candidate subtasks; obtain the user preferences of the user and the resource configuration of the vehicle; optimize the candidate subtasks using the user preferences and resource configuration, and use the optimized candidate subtasks as target subtasks.

[0181] In the embodiment of the present application, the analysis module 704 includes:

[0182] The acquisition submodule is used to obtain the user's feedback behavior and execution results based on the task execution status;

[0183] The first analysis submodule is used to analyze the feedback behavior to determine whether the target subtask achieves the expected effect corresponding to the target user's intention, and obtain the first analysis content;

[0184] The second analysis submodule is used to analyze the execution result to determine the planning accuracy corresponding to the target subtask and obtain the second analysis content;

[0185] The generating submodule is used to generate an analysis result based on the first analysis content and the second analysis content.

[0186] In an embodiment of the present application, the first analysis submodule is used to obtain the user's action feedback data and voice feedback data based on the feedback behavior; analyze the action feedback data to obtain the task adjustment intention, and analyze the voice feedback data to obtain the emotional tendency; and use the task adjustment intention and emotional tendency as the first analysis content.

[0187] In an embodiment of the present application, the second analysis submodule is used to determine the task completion status corresponding to the target subtask based on the execution result; analyze whether there is a deviation between the task completion status and the target user intention; if there is a deviation between the task completion status and the target user intention, determine the reason for the deviation and use the reason as the second analysis content.

[0188] In an embodiment of the present application, the analysis module 704 is used to determine the update content based on the first analysis content and / or the second analysis content, wherein the update content includes replacing the target subtask and / or updating the task parameters of the target subtask; using the update content to update the target subtask to obtain an updated target subtask, and executing the updated target subtask.

[0189] See also Figure 8 , Figure 8 is a schematic diagram of the structure of an electronic device provided by an optional embodiment of the present invention, such as Figure 7 As shown, the electronic device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. The various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process instructions executed in the electronic device, including instructions stored in or on the memory to display graphical information of the GUI on an external input / output device (such as a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple electronic devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system).

[0190] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

[0191] The memory 20 stores instructions executable by at least one processor 10, so that at least one processor 10 executes the method shown in the above embodiment.

[0192] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created by the use of an electronic device based on the presentation of a small program landing page, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the electronic device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0193] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.

[0194] The electronic device further comprises a communication interface 30 for the electronic device to communicate with other devices or a communication network.

[0195] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

[0196] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A task processing method, characterized in that: The method comprises: Recognize the target voice data currently input by the user to obtain at least one target user intention of the user; According to the target user intention, planning at least one target subtask associated with the target user intention; Acquire the vehicle resources corresponding to the target subtask, and execute the target subtask based on the vehicle resources to obtain the task execution status; The target subtask and the task execution status corresponding to the target subtask are analyzed to obtain an analysis result, and the target subtask is updated using the analysis result, and the updated target subtask is executed.

2. The method according to claim 1, characterized in that: The identifying based on the target voice data currently input by the user to obtain at least one target user intention of the user includes: Acquire the target voice data currently input by the user; Parsing the target voice data to obtain text content corresponding to the target voice data; The text content is identified, semantic information of the text content is obtained, and the semantic information is analyzed to obtain at least one target user intention of the user.

3. The method according to claim 2, characterized in that The analyzing the semantic information to obtain at least one target user intention of the user includes: Detecting the semantic information to obtain keywords contained in the semantic information; Querying a first user intention related to the keyword, and modifying the first user intention based on the context corresponding to the keyword to obtain a second user intention; The current behavior data of the user is detected, and if the behavior data matches the second user intention, the first user intention is used as the target user intention.

4. The method according to claim 1, characterized in that: The step of planning at least one target subtask associated with the target user intention according to the target user intention includes: Obtaining a mapping relationship between a preset user intention and a preset subtask; Based on the mapping relationship, the original subtask corresponding to the user intention is obtained, and the scene information in which the user is currently located is obtained; A target subtask corresponding to the scene information is acquired from the original subtask.

5. The method according to claim 4, characterized in that The acquiring a target subtask corresponding to the scene information from the original subtask includes: Taking the original subtasks among the original subtasks that match the scene information as candidate subtasks; Obtaining user preferences of the user and resource configuration of the vehicle; The candidate subtasks are optimized using the user preferences and the resource configuration, and the optimized candidate subtasks are used as the target subtasks.

6. The method according to claim 1, characterized in that The analyzing the target subtask and the task execution status corresponding to the target subtask to obtain the analysis result includes: Obtaining the user's feedback behavior and execution results based on the task execution status; Analyze the feedback behavior to determine whether the target subtask achieves the expected effect corresponding to the target user intention, and obtain a first analysis content; Analyze the execution result to determine the planning accuracy corresponding to the target subtask, and obtain a second analysis content; The analysis result is generated based on the first analysis content and the second analysis content.

7. The method according to claim 6, characterized in that The analyzing the feedback behavior to determine whether the target subtask achieves the expected effect corresponding to the target user intention, and obtaining the first analysis content, includes: Acquire the action feedback data and the voice feedback data of the user based on the feedback behavior; Analyzing the action feedback data to obtain task adjustment intention, and analyzing the voice feedback data to obtain emotional tendency; The task adjustment intention and the emotional tendency are used as the first analysis content.

8. The method according to claim 6, characterized in that The analyzing the execution result to determine the planning accuracy corresponding to the target subtask, and obtaining the second analysis content, includes: Determine the task completion status corresponding to the target subtask based on the execution result; Analyze whether there is a deviation between the task completion status and the target user intention; If there is a deviation between the task completion status and the target user intention, the reason for the deviation is determined and used as the second analysis content.

9. The method according to claim 6, characterized in that Updating the target subtask using the analysis result and executing the updated target subtask includes: Determining update content based on the first analysis content and / or the second analysis content, wherein the update content includes replacing a target subtask and / or updating a task parameter of the target subtask; The target subtask is updated using the update content to obtain an updated target subtask, and the updated target subtask is executed.

10. A task processing device, characterized in that: The device comprises: A recognition module, configured to perform recognition based on target voice data currently input by the user to obtain at least one target user intention of the user; A planning module, configured to plan at least one target subtask associated with the target user's intention according to the target user's intention; An acquisition module, used to acquire the vehicle resources corresponding to the target subtask, and execute the target subtask based on the vehicle resources to obtain the task execution status; The analysis module is used to analyze the target subtask and the task execution status corresponding to the target subtask to obtain the analysis result, and use the analysis result to update the target subtask and execute the updated target subtask.

11. An electronic device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the method according to any one of claims 1 to 9 by executing the computer instructions.

12. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the method according to any one of claims 1 to 9.

Images