A high-fidelity human-computer interaction driving test method based on digital twins

By combining digital twin technology with force feedback devices and simulation software, a human-machine-in-the-loop driving test closed-loop platform was constructed, which solved the problems of insufficient graphics rendering and dynamic accuracy in autonomous driving simulation testing, and realized high-fidelity and low-cost driving testing.

CN121580573BActive Publication Date: 2026-07-17BEIJING UNIV OF CIVIL ENG & ARCHITECTURE
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
BEIJING UNIV OF CIVIL ENG & ARCHITECTURE
Filing Date
2025-10-17
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

Existing autonomous driving simulation testing solutions suffer from insufficient graphics rendering capabilities or inadequate vehicle dynamics accuracy. Furthermore, high-performance real-time controllers increase economic costs and make it difficult to achieve high-fidelity human-machine interaction driving tests.

Method used

By using digital twin technology, force feedback steering wheel and pedal devices are used to capture driver operations and generate control commands. Combined with vehicle dynamics simulation software and 3D graphics engine, real-time dynamics calculation and data synchronization are achieved, avoiding the need for the 3D graphics engine to build a built-in physics engine for calculation, thus constructing a human-machine-in-the-loop driving test closed-loop platform.

Benefits of technology

It achieves high-fidelity, low-cost human-computer interaction driving testing, ensuring consistency between the virtual environment and physical simulation, and providing an immersive driving experience and high-precision test results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121580573B_ABST
    Figure CN121580573B_ABST
Patent Text Reader

Abstract

This specification discloses a high-fidelity human-machine interactive driving test method and device based on digital twins. The scheme includes: the method collects driver operations through a force feedback steering wheel and pedal devices and generates operation signals, which are received by a data processing and control software environment, converted into vehicle control commands, and sent to vehicle dynamics simulation software. The simulation software performs real-time dynamic calculations on the digital twin model according to the commands and outputs vehicle state data. This data is processed into binary data packets and synchronously sent to a 3D graphics engine via the UDP protocol. The graphics engine parses the data packets and directly drives the vehicle model in the virtual scene to update its position, attitude, and wheel state based on their content. This process avoids using the engine's built-in simple physics engine. In this scheme, all software modules run on the same IP address or networked computing device, and achieve data interaction through the UDP protocol, jointly constructing a human-machine-in-the-loop driving test closed-loop platform.
Need to check novelty before this filing date? Find Prior Art