Input system, device and method for artificial intelligence generation content of intelligent cabin and storage medium
By collecting user profiles and vehicle operating data in real time, and combining the collaborative work of vehicle-side and cloud-side devices, small and large models are used for data matching and analysis. This solves the problems of insufficient customized services and real-time data in the human-machine interaction system of the smart cockpit, thereby improving the user experience and interaction effect.
Patent Information
- Application Number
- CN202511136801.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-13
- Publication Date
- 2025-12-05
AI Technical Summary
Existing smart cockpit human-computer interaction systems lack customized services, have insufficient real-time data collection, and lack optimized prompts, thus affecting user experience.
By collecting user profiles and vehicle operating condition data in real time, combined with an optimized prompt word generation strategy, and utilizing the collaborative work of vehicle-side and cloud-side devices, data matching and analysis are performed using preset small and large models to achieve personalized prompt word generation and real-time interaction.
It improves the human-computer interaction experience of the smart cockpit, ensures real-time data and personalized services, reduces irrelevant or incorrect responses, and enhances the naturalness and fluency of user interaction with AI.
Smart Images
Figure CN121075321A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of intelligent cockpit, in particular to an intelligent cockpit artificial intelligence generated content input system, device, method and storage medium. BACKGROUND
[0002] Intelligent cockpit integrates large model artificial intelligence generated content (AIGC) technology capabilities, combined with advanced sensing, computing, communication and other technologies, to realize real-time perception, analysis and control of the vehicle interior environment and passengers automatically and intelligently, to improve the comfort, convenience and safety of the driver and passenger experience. The core of intelligent cockpit is human-computer interaction, which can understand the needs of passengers according to their multi-modal signals and provide personalized services and feedback.
[0003] In the prior art, the scheme adopted by the human-computer interaction of the intelligent cockpit is: when the user raises a demand problem in the intelligent cockpit, the intelligent cockpit will collect the user's multi-modal signals through sensors and voice recognition technology, convert them into text problems, and call the AIGC service in the cloud. The AIGC service in the cloud inputs the problem and the prompt word into the large model. The large model will output the most appropriate answer feedback to the intelligent cockpit or the user according to the problem and the prompt word.
[0004] However, the existing intelligent cockpit human-computer interaction system still has some problems, such as lack of customized services, insufficient real-time data collection, and insufficient prompt word optimization. These problems limit the further development of intelligent cockpit and reduce user experience. SUMMARY
[0005] The technical problem to be solved by the present application is that the present application provides an intelligent cockpit artificial intelligence generated content input system, device, method and storage medium. By collecting user portraits and vehicle working condition data in real time, combined with the optimized prompt word generation strategy, the human-computer interaction experience and service level of the intelligent cockpit are improved.
[0006] As one aspect of the present application, an intelligent cockpit artificial intelligence generated content input system is provided, which includes a vehicle-side device and a cloud-side device, wherein:
[0007] The vehicle-side device is configured to collect user portraits, environment, vehicle working condition data in real time, and user questions and report them to the cloud-side device, and receive the answers to the user questions fed back by the cloud-side device;
[0008] The cloud device is configured to, after receiving a user question from the vehicle terminal, call an AIGC context service to obtain corresponding user portrait, vehicle working condition, and environmental context data; match the context data and the user question by using a preset prompt word small model to obtain a prompt word template that meets a business scenario and fill the prompt word template to obtain a perfected prompt word and the user question; use a preset large model to analyze the perfected prompt word and the user question to obtain a corresponding answer and send the answer to the vehicle terminal.
[0009] The vehicle terminal device comprises at least:
[0010] The vehicle terminal data acquisition module is configured to acquire user portrait, environmental, and vehicle working condition data at a predetermined acquisition frequency.
[0011] The AI processing module is configured to receive a user question input by a user through the vehicle terminal voice module and generate an acquisition frequency change command after receiving the user question and send the acquisition frequency change command to the vehicle terminal data acquisition module to change the acquisition frequency of the vehicle terminal data acquisition module.
[0012] The TBOX device is configured to report the user portrait, environmental, and vehicle working condition data acquired by the vehicle terminal data acquisition module in real time and the user question obtained by the AI processing module to the cloud and receive an answer to the user question fed back by the cloud.
[0013] The vehicle terminal data acquisition module comprises a plurality of sensors and controllers and is configured to acquire user portrait data including user identity information, driving habits, and preferences and vehicle working condition data including vehicle speed, engine state, temperature, and humidity and environmental data.
[0014] The vehicle terminal data acquisition module is configured to acquire user portrait, environmental, and vehicle working condition data at a predetermined first acquisition frequency by default and acquire user portrait, environmental, and vehicle working condition data at a second acquisition frequency after receiving the acquisition frequency change command sent by the AI processing module; the second acquisition frequency is greater than the first acquisition frequency.
[0015] The cloud device comprises at least:
[0016] The cloud data acquisition module is configured to acquire user portrait, environmental, and vehicle working condition data reported by the vehicle terminal device in real time and a user question.
[0017] The cloud AIGC service module is configured to, after receiving a user question from the vehicle terminal, call an AIGC context service to obtain corresponding user portrait, vehicle working condition, and environmental context data; send the context data and the user question to a prompt word optimization module and send the perfected prompt word and the user question returned by the prompt word optimization module to a large model processing module; receive a corresponding answer fed back by the large model processing module and send the answer to the vehicle terminal device.
[0018] The prompt word optimization module is used for matching and perfecting the context data and the user question from the cloud AIGC service module by using a preset prompt word small model, obtaining a perfected user portrait, vehicle working condition data, and returning to the cloud AIGC service module;
[0019] The large model processing module is used for reasoning analysis according to the perfected prompt word and the user question, obtaining a corresponding answer, and transmitting to the cloud AIGC service module.
[0020] The prompt word optimization module uses a machine learning algorithm to learn and train a large amount of historical data to generate personalized prompt words.
[0021] Correspondingly, as another aspect of the present application, an input device for intelligent cockpit artificial intelligence generated content is also provided, which comprises:
[0022] The cloud data acquisition module is used for collecting the user portrait, environment, vehicle working condition data reported by the vehicle terminal device in real time, and user questions;
[0023] The cloud AIGC service module is used for obtaining the context data of the corresponding user portrait, vehicle working condition, and environment by calling the AIGC context service after receiving the user question from the vehicle terminal, sending the context data and the user question to the prompt word optimization module, sending the perfected prompt word and the user question returned by the prompt word optimization module to the large model processing module, receiving the corresponding answer fed back by the large model processing module, and issuing to the vehicle terminal device;
[0024] The prompt word optimization module is used for matching and perfecting the context data and the user question from the cloud AIGC service module by using a preset prompt word small model, obtaining a perfected prompt word and a user question;
[0025] The large model processing module is used for reasoning analysis according to the perfected prompt word and the user question, obtaining a corresponding answer, and transmitting to the cloud AIGC service module.
[0026] Correspondingly, as another aspect of the present application, an input device for intelligent cockpit artificial intelligence generated content is also provided, which comprises:
[0027] The vehicle terminal device collects the user portrait, environment, vehicle working condition data, and user questions in real time and reports to the cloud;
[0028] The cloud device obtains the context data of the corresponding user portrait, vehicle working condition, and environment by calling the AIGC context service after receiving the user question from the vehicle terminal;
[0029] The cloud device matches and perfects the context data and the user question by using a preset prompt word small model, to obtain a perfect prompt word and a user question;
[0030] The cloud device reasons and analyzes the perfect prompt word and the user question by using a preset large model, to obtain a corresponding answer, and delivers the answer to the vehicle terminal device.
[0031] The vehicle terminal device collects user portrait, environment, vehicle working condition data, and user questions in real time, and reports the data to the cloud, at least including:
[0032] Collecting user portrait, environment, vehicle working condition data at a predetermined first collection frequency;
[0033] Receiving a user question input by a user, and generating a collection frequency changing command after receiving the user question, to change the collection frequency of collecting user portrait, environment, vehicle working condition data, and collecting user portrait, environment, vehicle working condition data at a second collection frequency; the second collection frequency is greater than the first collection frequency;
[0034] Reporting the real-time collected user portrait, environment, vehicle working condition data, and user questions to the cloud device.
[0035] Correspondingly, as another aspect of the present application, an input method for generating content by artificial intelligence of an intelligent cockpit is also provided, which includes the following steps:
[0036] Collecting user portrait, environment, vehicle working condition data, and user questions reported by the vehicle terminal device in real time;
[0037] After receiving the user question from the vehicle terminal, calling an AIGC context service to obtain context data of the corresponding user portrait, vehicle working condition, and environment;
[0038] Matching and perfecting the context data and the user question by using a preset prompt word small model, to obtain a perfect prompt word and a user question;
[0039] Reasoning and analyzing the perfect prompt word and the user question by using a preset large model, to obtain a corresponding answer, and delivering the answer to the vehicle terminal device.
[0040] Correspondingly, as another aspect of the present application, a computer readable storage medium having a computer program stored thereon is also provided, and the computer program is executed by a processor to realize the steps of the method as described above.
[0041] The embodiment of the present application has the following beneficial effects:
[0042] The application provides an input system, device and method for intelligent cockpit artificial intelligence generated content and a storage medium.
[0043] Meanwhile, different data collection frequencies are adopted according to the requirements of different scenes. When the user initiates a question or interacts, the data collection frequency is immediately increased, thereby effectively solving the data time delay problem and ensuring that the user obtains real-time and accurate data and information.
[0044] In addition, a small model general prompt word optimization scheme is adopted in the cloud device to realize personalized customization of general prompt words and support various business scenarios. The drawbacks of prompt word template solidification are avoided, irrelevant or incorrect answers generated by the model are minimized, the interaction between the user and the AI is more natural and smooth, and the user experience and interaction effect are improved. BRIEF DESCRIPTION OF DRAWINGS
[0045] In order to more clearly illustrate the technical solutions in the embodiments of the application or the prior art, a brief introduction will be given below to the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description only constitute some embodiments of the application, and for those skilled in the art, other drawings obtained according to these drawings without creative labor are still within the scope of the application;
[0046] Figure 1 A structural schematic diagram of one embodiment of the input system for intelligent cockpit artificial intelligence generated content provided by the application;
[0047] Figure 2 A structural schematic diagram of the vehicle-side device in Figure 1
[0048] Figure 3 A structural schematic diagram of the cloud-side device in Figure 1
[0049] Figure 4 A detailed schematic diagram of the system provided by the application in actual application;
[0050] Figure 5 A main flow schematic diagram of one embodiment of the input method for intelligent cockpit artificial intelligence generated content provided by the application;
[0051] Figure 6 A principle schematic diagram of the data collection service involved in Figure 5
[0052] Figure 7 A principle schematic diagram of the data collection service involved in Figure 5 A schematic diagram of the principle of changing the collection frequency in the data collection service involved in the application;
[0053] Figure 8 For Figure 5 A schematic diagram of the principle of inputting a question by a user to obtain an answer involved in the application. DETAILED DESCRIPTION
[0054] In order to make the purpose, technical solutions and advantages of the application more clear, the application will be further described in detail below with reference to the drawings.
[0055] As Figure 1 shown, a structural schematic diagram of one embodiment of an input system for intelligent cockpit artificial intelligence generated content provided by the application is shown. In combination with Figures 2 to 4 shown, in this embodiment, the input system for intelligent cockpit artificial intelligence generated content includes a vehicle-side device and a cloud-side device, wherein:
[0056] The vehicle-side device 1 is configured to collect user portrait, environment, and vehicle working condition data in real time, input a question by a user and report to the cloud-side device, and receive an answer to the question by the user fed back by the cloud-side device. The user portrait data can include identity information, driving habits, and preferences of the user. The vehicle working condition data can include vehicle speed, acceleration, steering wheel angle, engine state, etc. The environment data can include temperature, humidity, whether there is fog, etc.
[0057] The cloud-side device 2 is configured to, after receiving a question by a user from the vehicle-side, call an AIGC context service to obtain context data of the user portrait, vehicle working condition, and environment. The context data and the question by the user are matched using a preset prompt word small model to obtain a prompt word template conforming to a business scenario and fill it to obtain a perfected prompt word and the question by the user. A preset large model is used to analyze the perfected prompt word and the question by the user to obtain a corresponding answer and send it to the vehicle-side device.
[0058] As Figure 2 shown, in a specific example, the vehicle-side device 1 at least includes:
[0059] The vehicle-side data collection module 10 is configured to collect user portrait, environment, and vehicle working condition data at a predetermined collection frequency. The vehicle-side data collection module includes a plurality of sensors and controllers configured to collect user portrait data including identity information, driving habits, and preferences of the user, vehicle working condition data including vehicle speed, engine state, temperature, and humidity, and environment data.
[0060] The vehicle-end data acquisition module collects user portrait, environment, and vehicle working condition data at a predetermined first acquisition frequency by default. After receiving an acquisition frequency change command sent by the AI processing module, the vehicle-end data acquisition module collects user portrait, environment, and vehicle working condition data at a second acquisition frequency. The second acquisition frequency is greater than the first acquisition frequency.
[0061] The AI processing module 11 receives user questions input by the user through the vehicle-end voice module, and generates an acquisition frequency change command after receiving the user questions, and sends the acquisition frequency change command to the vehicle-end data acquisition module to change the acquisition frequency thereof.
[0062] The TBOX device 12 reports user portrait, environment, and vehicle working condition data collected by the vehicle-end data acquisition module in real time, and user questions obtained by the AI processing module to the cloud end, and receives answers to the user questions fed back by the cloud end.
[0063] As shown in Figure 3 In specific examples, the cloud-end device 2 at least includes:
[0064] The cloud-end data acquisition module 20 collects user portrait, environment, and vehicle working condition data reported by the vehicle-end device in real time, and user questions, etc. These data are stored in the cloud-end device;
[0065] The cloud-end AIGC service module 21 obtains context data of user portrait, vehicle working condition, and environment after receiving user questions from the vehicle end, and sends the context data and the user questions to the prompt word optimization module, and obtains improved prompt words and user questions and sends them to the large model processing module, receives corresponding answers fed back by the large model processing module, and issues them to the vehicle end;
[0066] The prompt word optimization module 22 matches the context data and the user questions from the cloud-end AIGC service module using a preset prompt word small model, obtains a prompt word template that meets the business scenario and fills it in, improves user portrait and vehicle working condition data, and returns the improved prompt words and user questions to the cloud-end AIGC service module. In specific examples, the prompt word optimization module 22 uses machine learning algorithms to learn and train a large amount of historical data to generate personalized prompt words. It can be understood that this module uses a small model to learn and train a large amount of historical data, which can generate appropriate prompt words according to different scenarios and needs, and improve the naturalness and fluency of human-computer interaction.
[0067] The large model processing module 23 performs reasoning analysis according to the improved prompt words and user questions sent by the cloud-end AIGC service module, obtains corresponding answers, and transmits them to the cloud-end AIGC service module.
[0068] It can be understood that the large model processing module 23 here can select advanced large model architectures such as GPT series (such as GPT-3, GPT-4), BERT, T5, etc. as the basis. These models perform well in natural language processing, image generation, etc., and have strong language understanding and generation capabilities. And pre-use large-scale, diverse data sets to pre-train the model. This helps the model learn rich language knowledge and general features, laying a solid foundation for subsequent fine-tuning and application. The pre-training process can be carried out in the cloud, using distributed computing resources to speed up the training process.
[0069] In combination Figure 4 As shown in the figure, in the embodiment of the application, the whole vehicle cloud calling link is composed of two parts: vehicle end and cloud end. The vehicle end collects data by the data collection module, coordinates user problems and collects frequency change commands by the AI module; requests the cloud gateway (MQTT) through the TBox; the cloud gateway distributes to the specific cloud AIGC application (APP) service after receiving the request, and then the cloud AIGC application service coordinates the calling of the AIGC context service to obtain user context information, then calls the prompt word small model to obtain the prompt word, and finally calls the large model to ask questions, and the obtained answer original link is returned.
[0070] It can be understood that the scheme provided by the application can be widely applied to user use of AI services in the cockpit, such as voice dialogue, intelligent recommendation, image generation, etc. Large model business scenarios.
[0071] As Figure 5 As shown in the figure, a main flow diagram of one embodiment of an input method for intelligent cockpit artificial intelligence generated content provided by the application is shown. It is implemented by the system described in Figures 1 to 4 As shown in the figure, a main flow diagram of one embodiment of an input method for intelligent cockpit artificial intelligence generated content provided by the application is shown. It is implemented by the system described in Figures 6 to 8 In this embodiment, the method includes at least the following steps:
[0072] Step S10, the vehicle end device collects user portrait, environment, vehicle working condition data in real time, and user questions and reports to the cloud end;
[0073] In a specific example, the step S10 includes at least:
[0074] Step S100, collecting user portrait, environment, vehicle working condition data at a predetermined first collection frequency;
[0075] Step S101, receiving a user question input by a user, and generating a collection frequency changing command after receiving the user question, changing the collection frequency of collecting user portrait, environment, and vehicle working condition data, and collecting user portrait, environment, and vehicle working condition data at a second collection frequency; the second collection frequency is greater than the first collection frequency, and in a specific example, the first collection frequency is to collect once every 10 seconds, and the second collection frequency is to collect once every 1 second;
[0076] Step S102, reporting the real-time collected user portrait, environment, and vehicle working condition data to a cloud device.
[0077] Specifically, as shown in Figure 6 The vehicle-end data collection module collects information including but not limited to vehicle information (vin) in the car computer, user identification (userId), and vehicle temperature, speed in the sensing system, and environment information (current latitude and longitude position), and reports the information to the cloud-end data collection module;
[0078] Meanwhile, in order to obtain real-time data and optimize user experience, in the example of the present application, the vehicle-end collection module reports data every 10 seconds when idle; when the user initiates a question, the vehicle voice module is woken up; the voice module issues an instruction to the vehicle-end collection module to immediately sample and report, and the collection frequency is 1 second / time, and the data is updated in time by means of 4G / 5G network; when the user finishes the question, the problem and the vehicle-end data are updated at the same time, and specific reference can be made to Figure 7 .
[0079] Step S11, after the cloud device receives the user question from the vehicle-end, the AIGC context service is called to obtain the context data of the user portrait, vehicle working condition, and environment;
[0080] As shown in Figure 8 When the user initiates a question, the corresponding cloud AIGC APP service receives the vehicle-end request, and the AIGC context service is called to obtain the context data of the user portrait, vehicle working condition, and environment. Specifically, the AIGC context service calls the required services to obtain the user portrait (age, gender, hobby, etc.), vehicle working condition (vin, vehicle model, version, battery capacity, etc.), and environment (regional location) data according to the reported data. In the example of Figure 8 , the AIGC context service is called to obtain the following information:
[0081] The user information (userInfo) is:
[0082] User identification (userId): uid123, age: 28, gender: male, hobby: football;
[0083] The vehicle information (vehicleInfo) is:
[0084] Vin: vin1, vehicle type: gac-m3-max, version: 2024.01.100, battery capacity: 10AH;
[0085] The environment information (environment) is:
[0086] The area (area) is Guangzhou, Guangdong Province.
[0087] In step S12, the cloud device matches the context data and the user question with a preset prompt word small model to obtain a prompt word template that conforms to the business scenario and is filled in to obtain a perfect prompt word and user question.
[0088] Then, the AIGC APP service uses the current context data and the user question to request the prompt word small model. The prompt word small model infers a prompt word template that matches the business scenario and is filled in to perfect the user portrait, vehicle working condition data, etc., to form a perfect prompt word and user question.
[0089] As shown in Figure 8 , the perfect prompt word and user question are:
[0090] "You are now an intelligent cockpit assistant. The user is driving a {model} type car at a speed of {speed} on the roads of {area}. The temperature in the car is {temperature}. The user is {age} years old, {gender}, and the hobby is {hobby}. Based on the above background, answer the following question {question} of the user."
[0091] In step S13, the cloud device uses a preset large model to analyze the perfect prompt word and user question to obtain a corresponding answer and deliver it to the vehicle terminal device for corresponding processing or display.
[0092] Specifically, the AIGC application service provides the returned perfect prompt word and question to a pre-trained large model for questioning to obtain an answer.
[0093] For more details, reference can be made to the foregoing description of Figures 1 to 4 , which will not be repeated here.
[0094] Correspondingly, as another aspect of the present application, an input method for intelligent cockpit artificial intelligence generated content is also provided. In specific embodiments, it is applied in a cloud device, which includes the following steps:
[0095] Real-time collection of user portrait, environment, vehicle working condition data reported by the vehicle terminal device, and user questions;
[0096] After receiving the user question from the vehicle terminal, the AIGC context service is called to obtain the corresponding user portrait, vehicle working condition, and environment context data;
[0097] The context data and user question are matched using a preset prompt word small model to obtain a prompt word template that meets the business scenario and fill in the user portrait and vehicle working condition data to obtain a perfected prompt word and user question;
[0098] The perfected prompt word and user question are analyzed using a preset large model to obtain a corresponding answer and send it to the vehicle terminal.
[0099] For more details, refer to the foregoing description of Figures 1 to 8 , which will not be repeated here.
[0100] Correspondingly, as another aspect of the present application, a computer readable storage medium is also provided, which stores a computer program that, when executed by a processor, implements the steps of the method described above. Figures 5 to 8 For more details, refer to the foregoing description of Figures 5 to 8 , which will not be repeated here.
[0101] Implementing the embodiments of the present application has the following advantages:
[0102] The present application provides an intelligent cockpit artificial intelligence content generation input system, device, method and storage medium. By utilizing the existing data acquisition channel in the vehicle, user portrait and vehicle working condition data are automatically obtained, and real-time analysis is performed in the cloud to improve data utilization and analysis efficiency;
[0103] At the same time, different data acquisition frequencies are adopted according to the needs of different scenarios. When the user initiates a question or interacts, the data acquisition frequency is immediately increased, effectively solving the data time delay problem and ensuring that the user obtains real-time and accurate data and information;
[0104] In addition, a small model general prompt word optimization scheme is adopted in the cloud device to realize personalized customization of general prompt words and support various business scenarios. This avoids the disadvantages of fixed prompt word templates and minimizes the generation of irrelevant or incorrect answers by the model, making the user-AI interaction more natural and smooth and improving user experience and interaction effect.
[0105] The computer program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other processing devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other processing devices to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions specified in the flowchart block or blocks. Figure 1 The flowchart and / or block diagram in the variation can describe any device by any combination of software and / or hardware. Figure 1 The flowchart and / or block diagram in the variation can describe any device by any combination of software and / or hardware.
[0106] The above-described embodiments are merely possible implementations of the present application, and thus are not intended to limit the scope of the present application. It is therefore intended that the scope of the present application be determined by the following claims, and that equivalents be included therein as well.
Claims
1. An input system for intelligent cockpit artificial intelligence generated content, comprising a vehicle-side device and a cloud-side device, characterized in that, Wherein: The vehicle terminal device is configured to collect user portrait, environment, and vehicle working condition data in real time, and report user questions to the cloud terminal device, and receive answers to the user questions fed back by the cloud terminal device; The cloud terminal device is configured to call an AIGC context service to obtain context data of the user portrait, vehicle working condition, and environment after receiving the user questions from the vehicle terminal device; match and perfect the context data and the user questions using a preset prompt word small model to obtain perfect prompt words and user questions; and use a preset large model to perform reasoning analysis on the perfect prompt words and user questions to obtain corresponding answers and send the answers to the vehicle terminal device.
2. The system of claim 1, wherein, The vehicle terminal device at least includes: A vehicle terminal data collection module configured to collect user portrait, environment, and vehicle working condition data at a predetermined collection frequency; An AI processing module configured to receive user questions input by a user through a vehicle terminal voice module, and generate a collection frequency change command after receiving the user questions and send the command to the vehicle terminal data collection module to change the collection frequency thereof; A TBOX device configured to report user portrait, environment, and vehicle working condition data collected by the vehicle terminal data collection module in real time and user questions obtained by the AI processing module to the cloud, and receive answers to the user questions fed back by the cloud.
3. The system of claim 2, wherein, The vehicle terminal data collection module includes a plurality of sensors and controllers configured to collect user portrait data including user identity information, driving habits, and preferences, vehicle working condition data including vehicle speed, engine state, temperature, and humidity, and environment data; The vehicle terminal data collection module collects user portrait, environment, and vehicle working condition data at a predetermined first collection frequency by default, and collects user portrait, environment, and vehicle working condition data at a second collection frequency after receiving the collection frequency change command sent by the AI processing module; the second collection frequency is greater than the first collection frequency.
4. The system of claim 3, wherein, The cloud terminal device at least includes: A cloud terminal data collection module configured to collect user portrait, environment, and vehicle working condition data reported by the vehicle terminal device and user questions in real time; A cloud AIGC service module configured to call an AIGC context service to obtain context data of the user portrait, vehicle working condition, and environment after receiving the user questions from the vehicle terminal device; send the context data and the user questions to a prompt word optimization module, and send perfect prompt words and user questions returned by the prompt word optimization module to a large model processing module; receive corresponding answers fed back by the large model processing module, and send the answers to the vehicle terminal device; The prompt word optimization module is configured to match and perfect the context data and the user questions from the cloud AIGC service module using a preset prompt word small model to obtain perfect user portrait, vehicle working condition data, and return the data to the cloud AIGC service module; The large model processing module is configured to perform reasoning analysis on the perfect prompt words and user questions to obtain corresponding answers and transmit the answers to the cloud AIGC service module.
5. The system of claim 4, wherein, The prompt word optimization module uses a machine learning algorithm to learn and train a large amount of historical data to generate personalized prompt words.
6. An input device for intelligent cockpit artificial intelligence generated content, characterized in that, Including: The cloud data collection module is configured to collect user portrait, environment, and vehicle working condition data reported by the vehicle terminal device in real time, and collect user questions; The cloud AIGC service module is configured to, after receiving the user questions from the vehicle terminal, call an AIGC context service to obtain context data of the user portrait, vehicle working condition, and environment; The context data and the user questions are sent to the prompt word optimization module, and the improved prompt words and the user questions returned by the prompt word optimization module are sent to the large model processing module; The corresponding answers fed back by the large model processing module are received and sent to the vehicle terminal device; The prompt word optimization module is configured to improve the context data and the user questions from the cloud AIGC service module by matching using a preset prompt word small model, to obtain improved prompt words and user questions. The large model processing module is configured to perform reasoning analysis according to the improved prompt words and the user questions, to obtain corresponding answers and transmit the answers to the cloud AIGC service module.
7. An input method of intelligent cabin artificial intelligence generated content, characterized in that, The method comprises the following steps: The vehicle terminal device collects user portrait, environment, and vehicle working condition data in real time, and collects user questions and reports the questions to the cloud; The cloud device receives the user questions from the vehicle terminal, calls an AIGC context service to obtain context data of the user portrait, vehicle working condition, and environment; The cloud device improves the context data and the user questions by matching using a preset prompt word small model, to obtain improved prompt words and user questions; The cloud device performs reasoning analysis on the improved prompt words and the user questions using a preset large model, to obtain corresponding answers and sends the answers to the vehicle terminal device.
8. The method of claim 7, wherein, The vehicle terminal device collects user portrait, environment, and vehicle working condition data in real time, and collects user questions and reports the questions to the cloud, at least comprising: collecting user portrait, environment, and vehicle working condition data at a predetermined first collection frequency; receiving user input of user questions, and generating a collection frequency change command after receiving the user questions, to change the collection frequency of collecting user portrait, environment, and vehicle working condition data, and collecting user portrait, environment, and vehicle working condition data at a second collection frequency; the second collection frequency is greater than the first collection frequency; reporting the collected user portrait, environment, and vehicle working condition data, and the user questions to the cloud device.
9. An input method of intelligent cabin artificial intelligence generated content, characterized in that, The method comprises the following steps: collecting user portrait, environment, and vehicle working condition data reported by the vehicle terminal device in real time, and collecting user questions; receiving the user questions from the vehicle terminal, calling an AIGC context service to obtain context data of the user portrait, vehicle working condition, and environment; improving the context data and the user questions by matching using a preset prompt word small model, to obtain improved prompt words and user questions; performing reasoning analysis on the improved prompt words and the user questions using a preset large model, to obtain corresponding answers and send the answers to the vehicle terminal device.
10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by a processor to implement the steps of the method of any one of claims 7 to 9.