KR-20260065484-A - SYSTEM AND METHOD OF BUILDING CUSTOMIZED SCENARIO FOR VEHICLE WITH LARGE LANGUAGE MODEL BASED ON SERVICE-ORIENTED ARCHITECTURE

Abstract

The present invention relates to a system and method for constructing customized scenarios for a vehicle using Service Oriented Architecture (SOA)-based Large Language Models (LLMs). The system may include: a voice recognition device configured to acquire user voice information, convert the acquired user voice information into user request information in text form, classify the converted user request information, determine whether the type of user request information is a customized scenario, and, if the type of user request information is determined to be a customized scenario, generate user request information for the customized scenario; an SOA atomic function library configured to provide state information of sensors and actuators of the vehicle; and a control device configured to analyze user request information based on LLMs and generate a customized scenario plan suitable for user requirements using the user request information of the customized scenario and the SOA atomic function library.

Inventors

• 니우동
• 양팽
• 황루쟁

Assignees

• 현대자동차주식회사
• 기아 주식회사

Dates

Publication Date

20260508

Application Date

20250707

Priority Date

20241101

Claims (18)

1. In a system for building customized scenarios for vehicles using Service Oriented Architecture (SOA)-based Large Language Models (LLMs), A voice recognition device configured to acquire user voice information, convert the acquired user voice information into user request information in text form, classify the converted user request information, determine whether the type of user request information is a customized scenario, and, if the type of user request information is determined to be a customized scenario, generate user request information for a customized scenario; An SOA atomic function library configured to provide status information of the vehicle's sensors and actuators; and A control device configured to be communically connected to a voice recognition device and an SOA atomic function library, and to analyze user requirement information based on LLMs and generate a customized scenario plan suitable for user requirements using user requirement information of a customized scenario and the SOA atomic function library; A system for building customized vehicle scenarios using SOA-based LLMs including
2. In paragraph 1, A system for building customized scenarios for a vehicle using SOA-based LLMs, further comprising a display device configured to communicate with a control unit and to display the generated customized scenario plan to the user.
3. In paragraph 2, A facial recognition device configured to be communicably connected to a control device, recognize a user's facial information, determine whether the user is a registered user, and send the determined user identity information to the control device; and A seat seating detection device configured to be communicably connected to a control device, determine a user's seating position, and send the determined user seating position information to the control device; Includes more, A system for constructing a customized scenario for a vehicle using SOA-based LLMs, further configured to analyze user requirement information based on LLMs and generate a customized scenario suitable for user requirements by utilizing user requirement information of a customized scenario and an SOA atomic function library when the above-mentioned control device determines that the user is a registered user or that the user is sitting in the driver's seat.
4. In paragraph 2 or 3, The above display device A confirmation button indicating satisfaction with the customized scenario plan generated through the user's selection; and A rebuild button that indicates dissatisfaction with the customized scenario solution generated through the user's selection and prompts the user to regenerate the customized scenario solution; A system for building customized vehicle scenarios with SOA-based LLMs that further include
5. In paragraph 4, The above control device If the user is not satisfied with the generated customized scenario solution, the customized scenario solution is regenerated based on the user's feedback information until the user is satisfied; A system for constructing customized scenarios for a vehicle using SOA-based LLMs, further configured to store the generated customized scenario plan or store the generated customized scenario plan and corresponding user identity information or seating position information when the user is satisfied with the generated customized scenario plan.
6. In paragraph 5, The above control device A system for building customized scenarios for a vehicle using SOA-based LLMs configured to execute the stored customized scenario plan when the conditions of the stored customized scenario plan are satisfied.
7. In paragraph 1, Perform training and prompt tuning processing for LLMs using pre-configured customized scenario information; The above LLMs are SOA-based LLMs installed in the vehicle's control unit or in the cloud, and are systems for building customized scenarios for the vehicle.
8. In paragraph 1, The voice recognition device above A microphone configured to capture the user's voice signal; Automatic speech recognition (ASR) module configured to convert a user's voice signal into text information; A natural language processing (NLP) module configured to process and analyze text information to understand the structure and meaning of phrases; and A natural language understanding (NLU) module configured to understand text information and make corresponding decisions as needed; A system for building customized vehicle scenarios using SOA-based LLMs including
9. In paragraph 3, The above facial recognition device A camera configured to acquire an image of the user's face; and A processor configured to convert an acquired facial image into facial information, extract features from the facial information, and compare the extracted features with facial features stored in a database to determine whether the user is a registered user; A system for building customized vehicle scenarios using SOA-based LLMs including
10. Regarding the method of building customized vehicle scenarios with SOA-based LLMs, By a voice recognition device, Step of acquiring user voice information; A step of converting acquired user voice information into user request information in text form; A step of classifying the types of converted user requirement information; A step to determine whether the type of user requirement information is a customized scenario; and If it is determined that the type of user requirement information is a customized scenario, a step of generating user requirement information for the customized scenario; A step of providing state information of the vehicle's sensors and actuators by the SOA atomic function library; and By the control device, A step of analyzing user requirement information based on LLMs using user requirement information of customized scenarios and SOA atomic function libraries; and Step of generating a customized scenario plan suitable for user requirements; A method for building customized vehicle scenarios with SOA-based LLMs including
11. In Paragraph 10, A method for building customized scenarios for a vehicle using SOA-based LLMs, which further includes the step of displaying to the user a customized scenario plan generated by a display device.
12. In Paragraph 11, By facial recognition device Step of recognizing the user's facial information; A step to determine whether the user is a registered user; and A step of sending determined user identity information to a control device; By a seat seating detection device Step of determining the user's seating position; and A step of sending determined user seating position information to a control device; and By the control device If it is determined that the user is a registered user or that the user is sitting in the driver's seat, a step of analyzing user requirement information based on LLMs using user requirement information of a customized scenario and the SOA atomic function library; and Step of generating a customized scenario plan suitable for user requirements; A method for building customized vehicle scenarios with SOA-based LLMs that include more.
13. In Article 11 or Article 12, A step of receiving an indication from the user of satisfaction regarding the customized scenario plan generated by selecting a confirmation button on the display device; A step of indicating that the user is not satisfied with the customized scenario plan generated by the user selecting a rebuild button on the display device; and Step of regenerating a customized scenario plan; A method for building customized vehicle scenarios with SOA-based LLMs that include more.
14. In Paragraph 13, If the user is not satisfied with the generated customized scenario plan, the control device regenerates the customized scenario plan according to the user's feedback information until the user is satisfied; and If the user is satisfied with the solution of the generated customized scenario, the step of saving the generated customized scenario solution or saving the generated customized scenario solution and the corresponding user identity information or seating location information; A method for building customized vehicle scenarios with SOA-based LLMs that include more.
15. In Paragraph 14, A method for constructing a vehicle's customized scenario using SOA-based LLMs that execute the stored customized scenario plan when the conditions of the stored customized scenario plan are satisfied.
16. In Paragraph 11, It further includes a step of performing training and prompt tuning processing on LLMs using pre-configured custom scenario information, and The above LLMs are SOA-based LLMs provided to control units within the vehicle or the cloud, a method for building customized scenarios for the vehicle.
17. In Paragraph 10, Capturing the user's voice signal by the microphone; Converts the user's voice signal into text information by an Automatic Voice Recognition (ARS) module; Text information is processed and analyzed by a Natural Language Processing (NLP) module to understand the structure and meaning of phrases; A method for building customized vehicle scenarios using SOA-based LLMs that understand text information and make corresponding decisions via a Natural Language Understanding (NLU) module.
18. In Paragraph 12, Acquire the user's facial image using a camera; A method for constructing a customized scenario for a vehicle using an SOA-based LLMS that converts a facial image acquired by a processor into facial information, extracts features from the facial information, compares the extracted features with facial features stored in a database, and determines whether the user is a registered user.

Description

System and Method of Building Customized Scenarios for Vehicle with Large Language Model Based on Service-Oriented Architecture The present invention relates to smart vehicles and software architectures, and more specifically, to a system and method for constructing customized scenarios for a vehicle using Service-Oriented Architecture (SOA)-based Large Language Models (LLMs). Currently, vehicle developers are developing smart scenario applications that run on a Service-Oriented Architecture (SOA) and allow users to define various functions themselves through the vehicle's sensors and actuators. SOA is a design mode that organizes applications into various independent services that can be updated and extended independently without affecting the overall system. In this context, various vehicle functions (e.g., driver assistance, entertainment systems, navigation, etc.) can be designed as independent "atomic" services and combined and customized according to user requirements. However, this customization process may present some challenges. Users require a certain level of technical knowledge to understand and operate such complex systems, and manual editing and customization are relatively difficult and time-consuming. Therefore, there is a need for a system and method that can accurately understand user requirements and resolve the difficulty users face in setting up customized scenarios themselves. The matters described in this background technology section are written to enhance understanding of the background of the invention and may include matters that are not prior art already known to those skilled in the art to which this technology belongs. Hereinafter, exemplary embodiments of the present invention will be described in more detail by combining the drawings. For clarity, the same components in different drawings are indicated by the same reference numerals. It should be noted that the drawings are illustrative only and are not necessarily drawn in proportion. FIG. 1 is a block diagram of a system for constructing customized scenarios for a vehicle using SOA-based LLMs according to an embodiment of the present invention. FIG. 2 is a block diagram of a system for building customized scenarios for a vehicle using SOA-based LLMs according to another embodiment of the present invention. Figure 3 is a flowchart of a method for constructing customized scenarios for a vehicle using SOA-based LLMs according to an embodiment of the present invention. FIG. 4 is a flowchart of a method for constructing customized scenarios for a vehicle using SOA-based LLMs according to another embodiment of the present invention. Hereinafter, embodiments of the present invention are described in detail. These embodiments are implemented based on the technical methods of the present invention, and detailed implementation methods and specific operation processes are disclosed; however, the scope of protection of the present invention is not limited to the following embodiments. Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings. FIG. 1 is a block diagram of a system for constructing customized scenarios for a vehicle using an SOA-based LLMS according to an embodiment of the present invention. Referring to FIG. 1, a system for constructing a customized scenario for a vehicle using an SOA-based LLMS according to an embodiment of the present invention may include a control device (100), a voice recognition device (200), an SOA atomic function library (300), and a display device (400). The control device (100) is connected to communicate with the voice recognition device (200), the SOA atomic function library (300), and the display device (400). The voice recognition device (200) acquires user voice information, converts the acquired user voice information into user request information in text form, and classifies the converted user request information into multiple types. Additionally, the voice recognition device (200) can be configured to determine whether the type of user request information is a customized scenario, and if it is determined that the type of user request information is a customized scenario, to generate user request information for the customized scenario. The SOA atomic function library (300) can be configured to provide status information of the vehicle's sensors and actuators. The control device (100) can be configured to use the user request information of the customized scenario and the SOA atomic function library (300) to analyze the user request information from the voice recognition device (200) based on LLMs (Large Language Models) and to generate a customized scenario plan suitable for the user's request. The display device (400) can be configured to display the generated customized scenario plan to the user. According to an embodiment of the present invention, a communicable connection may include a connection via wired communication a