Interaction method and system for virtual reality software based on ai communication mode
By deploying a visual virtual assistant in a VR scene and establishing voice interaction, and using an AI processing module to analyze user intent and combine it with a knowledge base to generate real-time Q&A and evaluation reports, the problem of complex and fragmented interaction in VR automotive training software is solved, achieving an efficient immersive learning experience.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- FXB CO LTD
- Filing Date
- 2026-05-08
- Publication Date
- 2026-07-31
AI Technical Summary
Existing VR automotive training software suffers from problems such as complex interaction methods, a disconnect between operation and knowledge acquisition, a lack of real-time intelligent guidance and closed-loop evaluation, and a disconnect between AI functions and the VR environment, making it difficult to achieve both immersion and learning efficiency.
A visual virtual assistant is deployed in a VR scene. User commands or questions are collected through voice interaction. An AI processing module is used to perform intent analysis and business logic verification. Combined with a knowledge base, real-time question and answer is achieved, and personalized evaluation reports are generated. A four-layer architecture is constructed, consisting of a VR hardware layer, an AI interaction layer, a business function layer, and a data service layer.
It achieves a closed-loop, voice-driven, end-to-end interaction, reduces cognitive load, improves training efficiency and immersion, and solves problems such as interaction paradigm conflicts, separation of operation and knowledge space, and disconnect between AI and VR environments.
Smart Images
Figure CN122489704A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of virtual reality technology, and in particular to interactive methods and systems for virtual reality software based on AI communication. Background Technology
[0002] Current VR automotive training software generally suffers from the following problems: First, the interaction relies on controller buttons and virtual menus, requiring users to repeatedly switch visual focus in the immersive environment. Complex commands require multiple clicks to complete, resulting in high learning costs and low interaction efficiency. Second, when students encounter operational questions during training, they must exit the scene to consult paper or physical materials, creating a temporal and spatial disconnect between knowledge acquisition and practical operation, disrupting the continuity of the training. Third, the training process lacks proactive guidance, operational errors cannot be corrected in a timely manner, and post-task evaluation relies on manual intervention, leading to delayed and unpersonalized feedback. Fourth, existing AI question-and-answer and intelligent guidance functions are mostly implemented on physical devices, technically independent of the VR immersive environment, failing to form a closed-loop interaction driven by natural voice. Therefore, there is an urgent need for an interaction method that deeply integrates AI communication methods with the VR environment to solve the above problems. Summary of the Invention
[0003] The main objective of this invention is to provide an interactive method and system for virtual reality software based on AI communication, aiming to solve the technical problems of existing VR automotive training software, such as the conflict between graphical interface interaction and immersive environment, separation of operation and knowledge space, lack of real-time intelligent guidance and closed-loop evaluation, and the disconnect between AI functions and VR environment, which makes it difficult to achieve both immersion and learning efficiency.
[0004] To achieve the above objectives, this invention proposes an AI-based communication-assisted interaction method for virtual reality software, applied to a VR automotive training system. The VR training system deploys a visual virtual assistant within its VR scene and establishes a voice channel for voice interaction with the user. The method includes the following steps: S10, the virtual assistant is activated by a preset voice wake-up word. After being woken up, the virtual assistant generates a wake-up response voice and starts the voice recognition function to collect the user's voice commands or questions. S20: The collected speech is converted into text, and the text is input into the intent parsing model to obtain the user's intent; S30: When the intent is an operation command, a corresponding formatted business command is generated and sent to the business logic control module of the VR scene. After the business logic of the formatted business command is verified, the corresponding business operation is executed, and the execution result is fed back through the virtual assistant's voice. When the intent is a knowledge question, relevant content is retrieved from the preset knowledge base through retrieval enhancement generation technology and a reply text is generated. The reply text is input into the speech synthesis module to generate feedback speech, and the virtual assistant is controlled to broadcast the feedback speech through the voice channel. The S40 records user operation data in real time during training, and after the task is completed, it analyzes the operation data through a large model to generate a personalized evaluation report and provides voice feedback through a virtual assistant.
[0005] Furthermore, step S10 includes: activating the virtual assistant through a preset voice wake-up word; after being woken up, the virtual assistant generates a wake-up response voice and starts the voice recognition function to collect the user's voice commands or questions.
[0006] Furthermore, the formatted business instruction is a JSON data structure containing action identifiers and business parameters; the business logic verification includes: verifying whether the current VR scene's running state meets the conditions for executing the operation corresponding to the formatted business instruction.
[0007] Furthermore, the knowledge base is pre-loaded with vehicle structure knowledge data, software operation specification data, and training step text data; the retrieval enhancement generation technology combines the retrieved knowledge base content with a large language model to generate response text.
[0008] Furthermore, in step S40, the operational data includes the user's operational behavior, operational sequence, and compliance data during the training; the personalized evaluation report includes operational highlights, non-standard steps, error cause analysis, and improvement suggestions.
[0009] This invention also proposes an interactive system for virtual reality software based on AI communication methods, the system comprising: VR hardware modules are used to provide an immersive virtual reality environment and collect user voice input; The virtual assistant module is deployed in the VR scene and serves as a visual medium for voice interaction with users; The AI processing module is used to recognize and understand user speech, and to perform intent parsing, knowledge question and answer generation, and operation data analysis through the intent parsing model; The business execution module is used to control the VR scene engine to execute the corresponding business operations based on the formatting instructions generated by the AI processing module. The data storage module is used to store knowledge base data, user operation records, and evaluation report data.
[0010] Furthermore, the AI processing module includes: A speech recognition unit is used to convert user speech into text; Large model inference unit, used to parse user intent and generate operation instructions or question-and-answer text; A speech synthesis unit is used to convert reply text into speech; The retrieval enhancement generation unit is used to generate question-and-answer text by combining knowledge base content.
[0011] Furthermore, the business execution module includes an AI agent unit, which is used to receive formatted business instructions, verify the current VR scene status, and control the VR scene engine to automatically perform vehicle disassembly and assembly, fault diagnosis, or tool combination operations.
[0012] Furthermore, the system is integrated into a VR automotive training system, which also includes: The VR hardware layer is used to collect user voice, present 3D virtual scenes, and output voice feedback through virtual assistants. The AI interaction layer is used to parse voice commands and answer knowledge questions; The business function layer is used to connect with the automotive training business logic; The data service layer is used to store vehicle structure data, operating procedures, training operation records, and evaluation reports.
[0013] The present invention also proposes a computer-readable storage medium storing a computer program thereon, the computer program being executed by a processor to provide an interactive method for virtual reality software based on AI communication. This invention deploys a virtual assistant visually within a VR scene and establishes a voice channel. An AI agent module converts natural language commands into formatted business commands to drive the VR scene engine. Combined with a local knowledge base and enhanced retrieval generation technology, it enables real-time knowledge-based question answering. Personalized evaluation reports are generated through real-time recording of operational data and analysis of large models. Simultaneously, a four-layer architecture—VR hardware, AI interaction, business functions, and data services—is constructed, deeply integrating AI capabilities such as speech recognition, natural language understanding, large-scale model reasoning, and speech synthesis with the VR scene engine. This solves problems such as interaction paradigm conflicts, separation of operation and knowledge space, fragmented evaluation feedback, and disconnect between AI and the VR environment, forming a voice-driven, end-to-end interactive closed loop. This reduces cognitive load and improves training efficiency and immersion. Attached Figure Description
[0014] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.
[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0016] Figure 1 This is a flowchart illustrating the interaction method of virtual reality software based on AI communication, as described in this invention. Figure 2 This is a schematic diagram of the interactive system of virtual reality software assisted by AI communication method according to the present invention.
[0017] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation
[0018] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of the present invention and are not intended to limit the present invention.
[0019] To better understand the technical solution of the present invention, a detailed description will be provided below in conjunction with the accompanying drawings and specific embodiments.
[0020] like Figure 1 As shown, Figure 1 This is a flowchart illustrating the interaction method of virtual reality software based on AI communication, as described in this invention.
[0021] This invention proposes an AI-based communication-assisted interaction method for virtual reality software, applied to a VR automotive training system. The VR training system deploys a visual virtual assistant within its VR scene and establishes a voice channel for voice interaction with the user. The method includes the following steps: S10, the virtual assistant is activated by a preset voice wake-up word. After being woken up, the virtual assistant generates a wake-up response voice and starts the voice recognition function to collect the user's voice commands or questions. S20: The collected speech is converted into text, and the text is input into the intent parsing model to obtain the user's intent; S30: When the intent is an operation command, a corresponding formatted business command is generated and sent to the business logic control module of the VR scene. After the business logic of the formatted business command is verified, the corresponding business operation is executed, and the execution result is fed back through the virtual assistant's voice. When the intent is a knowledge question, relevant content is retrieved from the preset knowledge base through retrieval enhancement generation technology and a reply text is generated. The reply text is input into the speech synthesis module to generate feedback speech, and the virtual assistant is controlled to broadcast the feedback speech through the voice channel. The S40 records user operation data in real time during training, and after the task is completed, it analyzes the operation data through a large model to generate a personalized evaluation report and provides voice feedback through a virtual assistant.
[0022] This embodiment uses VR automotive power battery disassembly and assembly training as an example to provide a detailed description of the interaction method and system of virtual reality software based on AI communication proposed in this invention.
[0023] When the VR automotive training system starts, it first performs initial configuration. The VR hardware module (including the VR headset and controllers) loads the virtual training scene and deploys a visualized virtual assistant model within the scene. This virtual assistant is presented in a human-like form (such as "Xiao Feng"), serving as a visual medium for user-system voice interaction. Simultaneously, the system establishes a voice channel for user interaction. This channel uses the microphone array built into the VR headset to collect user voice input, and outputs system voice feedback through the headset's headphones, forming a two-way voice interaction link. After the user enters the "Power Battery Disassembly and Assembly Laboratory" virtual scene while wearing the VR headset, the visualized virtual assistant "Xiao Feng" appears at a fixed location within the scene, and the voice channel is in a ready-to-awaken state.
[0024] S10, the virtual assistant is activated by a preset voice wake-up word. After being woken up, the virtual assistant generates a wake-up response voice and starts the voice recognition function to collect the user's voice commands or questions.
[0025] In this embodiment, the user activates the virtual assistant by using the voice wake-up keyword "Xiaofeng, Xiaofeng". After the VR headset's microphone picks up the wake-up word, the system recognizes it and triggers the virtual assistant to switch from standby mode to interactive mode. Once activated, the virtual assistant generates a wake-up response voice message, "Hello, how can I help you?", through the voice channel, and initiates the voice recognition function to collect subsequent voice commands or questions from the user. During the training process, the user may generate two types of voice input: operational commands, such as "Help me lift the vehicle"; and knowledge-based questions, such as "How to properly disassemble the power battery module". The system collects these voice inputs through the voice channel and proceeds to the subsequent processing flow.
[0026] S20 converts the collected speech into text and inputs the text into the intent parsing model to obtain the user's intent.
[0027] In this embodiment, the system converts the collected user speech into text using a speech recognition unit. For example, the user speech "Help me lift the vehicle" is converted into the text string "Help me lift the vehicle"; the user speech "How to properly disassemble the power battery module" is converted into the text string "How to properly disassemble the power battery module". Subsequently, the system inputs this text into an intent parsing model (built based on a large model) for semantic understanding to obtain the user's intent type. The intent parsing model outputs classification results: for "Help me lift the vehicle", it is identified as an "operation command" intent; for "How to properly disassemble the power battery module", it is identified as a "knowledge question" intent.
[0028] S30: When the intent is an operation command, a corresponding formatted business command is generated and sent to the business logic control module of the VR scene. After the business logic of the formatted business command is verified, the corresponding business operation is executed, and the execution result is fed back through the virtual assistant's voice. When the intent is a knowledge question, relevant content is retrieved from the preset knowledge base through retrieval enhancement generation technology and a reply text is generated. The reply text is input into the speech synthesis module to generate feedback speech, and the virtual assistant is controlled to broadcast the feedback speech through the voice channel. In this embodiment, when the intent parsing result is an "operation instruction," the system performs the following operations. First, the large model inference unit generates a formatted business instruction based on the parsed intent. For example, for the instruction "Help me lift the vehicle," the system generates a JSON-formatted business instruction: {"action":"lift_vehicle", "params":{"height":"safe_height"}}. This JSON data structure contains the action identifier "lift_vehicle" and the business parameter "height," clearly defining the type and parameters of the operation to be performed, corresponding to the further scheme's "formatted business instruction is a JSON data structure containing the action identifier and business parameters." Then, the system sends the formatted business instruction to the VR scene's business logic control module. After receiving the instruction, the business logic control module first performs business logic verification: verifying whether the current VR scene's operating state meets the conditions for executing the vehicle lifting operation, such as determining whether the vehicle has stopped, whether the lift is available, and whether there are other operations currently conflicting, corresponding to the further scheme's "business logic verification includes: verifying whether the current VR scene's operating state meets the conditions for executing the operation corresponding to the formatted business instruction." If the verification passes (e.g., the vehicle is in a liftable state), the business logic control module calls the VR scene engine interface to control the VR scene engine to execute the vehicle lifting animation, raising the vehicle to a safe height. After the operation is completed, the business logic control module returns the execution status to the AI processing module. The AI processing module generates a feedback voice message, "The vehicle has been raised to a safe height for you," through the speech synthesis unit, and controls the virtual assistant to broadcast this feedback message through the voice channel. The user hears the feedback through the VR headset, completing the closed loop of the operation command.
[0029] In another embodiment, when the intent parsing result is "knowledge question", the system performs the following operations: The system inputs the user's question text "How to properly disassemble the power battery module" into the retrieval enhancement generation unit. This unit first searches a preset knowledge base. The knowledge base pre-stores vehicle structure knowledge data (such as the structural composition and connection method of the power battery module), software operation specification data (such as the standard procedure for disassembly in VR training software), and training step text data (such as disassembly sequence and safety precautions), corresponding to the further scheme "the knowledge base pre-stores vehicle structure knowledge data, software operation specification data, and training step text data". Based on the question content, the retrieval enhancement generation unit retrieves document fragments related to "power battery module disassembly" from the knowledge base, such as: "Step 1, disconnect the high-voltage harness; Step 2, remove the fixing bolts; Step 3, use a special tool to remove the module; Step 4, check the connector status". After retrieving the relevant content, the system inputs the retrieval results along with the user's question into the large model inference unit. The large-scale model inference unit combines the retrieved knowledge base content to generate a natural language response text: "When disassembling the power battery module, please follow these steps: First, confirm that the vehicle is powered off, disconnect the high-voltage wiring harness and wait 5 minutes; second, use a 13mm socket wrench to remove the 8 fixing bolts around the module; then, use a special extraction tool to vertically lift the module upwards; finally, check the connector at the bottom of the module for damage. Please wear insulated gloves during the operation." This process corresponds to the further solution's "retrieval-enhanced generation technology, which combines the retrieved knowledge base content with the large language model to generate response text." This response text is input into the speech synthesis unit to generate a natural and fluent feedback voice. The system controls the virtual assistant to broadcast this feedback voice through the voice channel, allowing users to receive professional and accurate operation guidance without leaving the VR scene, achieving real-time integration of the operation space and the knowledge space.
[0030] The S40 records user operation data in real time during training, and after the task is completed, it analyzes the operation data through a large model to generate a personalized evaluation report and provides voice feedback through a virtual assistant.
[0031] In this embodiment, during the user's practical training, the system records all user actions in real time through the operation analysis module. For example, in the power battery module disassembly training, the system records the user's operation sequence: whether the high-voltage harness was disconnected first, whether the correct tools were used, and whether the bolt disassembly sequence was correct, etc.; it records the operation standardization: whether there were any violations (such as not wearing insulated gloves), whether the operation time was reasonable, etc.; and it records the operation compliance: whether the operation conforms to the preset training step standards. This operation data, including operation behavior, operation sequence, and compliance data, is stored in real time to the data storage module, corresponding to the further scheme where "operation data includes the user's operation behavior, operation sequence, and compliance data in the training." After the user completes the power battery module disassembly operation, they click the "Submit Task" button on the handle, triggering the operation analysis and evaluation process. The system uploads all operation data recorded during this training to the AI processing module, where the large model inference unit performs in-depth analysis of this operation data. The analysis process includes comparing the user's operation sequence with the standard operating procedure to identify operational highlights (such as correct tool use and reasonable operation speed), non-standard steps (such as not waiting 5 minutes after disconnecting the wiring harness), and reasons for errors (such as neglecting to remove a fixing bolt). Based on the analysis results, the system generates a personalized evaluation report, which includes operational highlights, non-standard steps, error cause analysis, and improvement suggestions. This corresponds to the further solution where "the personalized evaluation report includes operational highlights, non-standard steps, error cause analysis, and improvement suggestions." For example, the evaluation report shows: "The standardization of this operation is 85 points. Highlights: The tools were used correctly, and the bolt disassembly sequence was standardized; Shortcomings: Failure to wait 5 minutes after disconnecting the high-voltage harness poses a safety hazard; Improvement suggestion: Please use a multimeter to confirm that there is no residual voltage after disconnecting the high-voltage harness before continuing the subsequent operation." The text of this evaluation report is converted into speech by a speech synthesis unit, and the virtual assistant reads it to the user point by point through the voice channel: "Your standardization of this operation is 85 points. The highlight is the correct use of tools. The shortcoming is that you did not wait 5 minutes after disconnecting the high-voltage harness. It is recommended that you use a multimeter to confirm that there is no residual voltage before continuing next time." The user listens to the evaluation report through the VR headset, forming a complete closed loop of "operation, recording, analysis, and feedback."
[0032] Furthermore, in step S40, the operational data includes the user's operational behavior, operational sequence, and compliance data during the training; the personalized evaluation report includes operational highlights, non-standard steps, error cause analysis, and improvement suggestions.
[0033] like Figure 2 As shown, Figure 2 This is a schematic diagram of the interactive system of virtual reality software assisted by AI communication method according to the present invention.
[0034] Reference Figure 2 The present invention also proposes an interactive system for virtual reality software based on AI communication, the system comprising: VR hardware module 10 is used to provide an immersive virtual reality environment and collect the user's voice input; Virtual assistant module 20, deployed in the VR scene, serves as a visual medium for voice interaction with users; AI processing module 30 is used to recognize and understand user speech, and to perform intent parsing, knowledge question and answer generation and operation data analysis through intent parsing model; The business execution module 40 is used to control the VR scene engine to execute the corresponding business operations based on the formatting instructions generated by the AI processing module 30. Data storage module 50 is used to store knowledge base data, user operation records, and evaluation report data.
[0035] Furthermore, the AI processing module 30 includes: A speech recognition unit is used to convert user speech into text; Large model inference unit, used to parse user intent and generate operation instructions or question-and-answer text; A speech synthesis unit is used to convert reply text into speech; The retrieval enhancement generation unit is used to generate question-and-answer text by combining knowledge base content.
[0036] Furthermore, the business execution module 40 includes an AI agent unit, which is used to receive formatted business instructions, verify the current VR scene status, and control the VR scene engine to automatically perform vehicle disassembly and assembly, fault diagnosis, or tool combination operations.
[0037] Furthermore, the system is integrated into a VR automotive training system, which also includes: The VR hardware layer is used to collect user voice, present 3D virtual scenes, and output voice feedback through virtual assistants. The AI interaction layer is used to parse voice commands and answer knowledge questions; The business function layer is used to connect with the automotive training business logic; The data service layer is used to store vehicle structure data, operating procedures, training operation records, and evaluation reports.
[0038] The present invention also proposes a computer-readable storage medium storing a computer program thereon, the computer program being executed by a processor to provide an interactive method for virtual reality software based on AI communication. This invention deploys a virtual assistant visually within a VR scene and establishes a voice channel. An AI agent module converts natural language commands into formatted business commands to drive the VR scene engine. Combined with a local knowledge base and enhanced retrieval generation technology, it enables real-time knowledge-based question answering. Personalized evaluation reports are generated through real-time recording of operational data and analysis of large models. Simultaneously, a four-layer architecture—VR hardware, AI interaction, business functions, and data services—is constructed, deeply integrating AI capabilities such as speech recognition, natural language understanding, large-scale model reasoning, and speech synthesis with the VR scene engine. This solves problems such as interaction paradigm conflicts, separation of operation and knowledge space, fragmented evaluation feedback, and disconnect between AI and the VR environment, forming a voice-driven, end-to-end interactive closed loop. This reduces cognitive load and improves training efficiency and immersion.
[0039] The above description is only a part of the embodiments of the present invention and does not limit the patent scope of the present invention. All equivalent structural transformations made under the technical concept of the present invention using the contents of the present invention specification and drawings, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.
Claims
1. An AI-based communication mode assisted virtual reality software interaction method, characterized in that, The method, applied to a VR automotive training system, includes the following steps: A visual virtual assistant is deployed within the VR scene of the VR automotive training system, and a voice channel is established for voice interaction with the user. S10, the virtual assistant is activated by a preset voice wake-up word. After being woken up, the virtual assistant generates a wake-up response voice and starts the voice recognition function to collect the user's voice commands or questions. S20, the collected speech is converted into text, and the text is input into the intent parsing model to obtain the user intent; S30, when the intent is an operation command, a corresponding formatted business command is generated and sent to the business logic control module of the VR scene to perform business logic verification on the formatted business command and then execute the corresponding business operation, and the execution result is fed back through the virtual assistant's voice; when the intent is a knowledge question, relevant content is retrieved from the preset knowledge base through retrieval enhancement generation technology and a reply text is generated, the reply text is input to the speech synthesis module to generate feedback voice, and the virtual assistant is controlled to broadcast the feedback voice through the voice channel; S40 records the user's operation data in real time during the training process, and after the task is completed, analyzes the operation data through a large model to generate a personalized evaluation report, and provides voice feedback through the virtual assistant. 2.The AI communication mode based interaction method of virtual reality software according to claim 1, wherein, Step S10 includes: activating the virtual assistant by a preset voice wake-up word; after being woken up, the virtual assistant generates a wake-up response voice and starts the voice recognition function to collect the user's voice commands or questions. 3.The AI communication mode based interaction method of virtual reality software according to claim 1, wherein, The formatted business instruction is a JSON data structure containing action identifiers and business parameters; The business logic verification includes: verifying whether the current VR scene's running state meets the conditions for executing the operation corresponding to the formatted business instruction.
4. The interaction method for virtual reality software based on AI communication as described in claim 1, characterized in that, The knowledge base is pre-loaded with vehicle structure knowledge data, software operation specification data, and training step text data; the retrieval enhancement generation technology generates the response text by combining the retrieved knowledge base content with a large language model.
5. The interaction method for virtual reality software based on AI communication as described in claim 1, characterized in that, In step S40, the operation data includes the user's operation behavior, operation sequence, and compliance data during the training; the personalized evaluation report includes operation highlights, non-standard steps, error cause analysis, and improvement suggestions.
6. An interactive system for virtual reality software based on AI communication methods, characterized in that, The system for implementing the method according to any one of claims 1 to 5, the system comprising: VR hardware modules are used to provide an immersive virtual reality environment and collect user voice input; The virtual assistant module is deployed in the VR scene and serves as a visual medium for voice interaction with users; The AI processing module is used to recognize and understand user speech, and to perform intent parsing, knowledge question and answer generation, and operation data analysis through the intent parsing model; The business execution module is used to control the VR scene engine to execute the corresponding business operations based on the formatting instructions generated by the AI processing module. The data storage module is used to store knowledge base data, user operation records, and evaluation report data.
7. The system according to claim 6, characterized in that, The AI processing module includes: A speech recognition unit is used to convert user speech into text; Large model inference unit, used to parse user intent and generate operation instructions or question-and-answer text; A speech synthesis unit is used to convert reply text into speech; The retrieval enhancement generation unit is used to generate question-and-answer text by combining knowledge base content.
8. The system according to claim 6, characterized in that, The business execution module includes an AI agent unit, which is used to receive formatted business instructions, verify the current VR scene status, and control the VR scene engine to automatically perform vehicle disassembly and assembly, fault diagnosis, or tool combination operations.
9. The system according to claim 6, characterized in that, The system is integrated into a VR automotive training system, and also includes: The VR hardware layer is used to collect user voice, present 3D virtual scenes, and output voice feedback through virtual assistants. The AI interaction layer is used to parse voice commands and answer knowledge questions; The business function layer is used to connect with the automotive training business logic; The data service layer is used to store vehicle structure data, operating procedures, training operation records, and evaluation reports.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method described in any one of claims 1 to 5.