Interaction decision method, device and equipment of unmanned vehicle and storage medium

By employing multi-agent nested game decision-making and dynamic risk field screening, the problem of efficiency and safety of autonomous vehicles in complex interactive scenarios has been solved, achieving efficient and safe passage in complex scenarios.

CN122232671APending Publication Date: 2026-06-19DONGFENG LIUZHOU MOTOR +1
0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
DONGFENG LIUZHOU MOTOR
Filing Date
2026-05-14
Publication Date
2026-06-19

AI Technical Summary

⚠Technical Problem

Autonomous vehicles struggle to navigate efficiently and safely in complex interactive scenarios. Existing trajectory prediction methods lack a deep understanding of the behavioral intentions and interactive game relationships of traffic participants. Furthermore, rule-based methods suffer from insufficient safety redundancy in complex scenarios, making it difficult to balance traffic efficiency and driving safety.

⚗Method used

A multi-agent nested game decision-making method is adopted. By acquiring vehicle status data and environmental perception data, a high-level decision set of interaction mode is constructed. A low-level nested game model is used to predict risks and efficiency. Combined with dynamic risk field and preset risk constraints, the target driving trajectory is screened to achieve a dynamic balance between risk and efficiency in complex interaction scenarios.

🎯Benefits of technology

It improves the safety and traffic efficiency of autonomous vehicles in complex interaction scenarios. By explicitly outputting intent and incorporating quantitative calculations of risk and efficiency, and reserving safety redundancy, it solves the technical challenges of autonomous driving in complex scenarios.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

This application discloses an interactive decision-making method, device, equipment, and storage medium for autonomous vehicles, relating to the technical field of vehicle driving control. This application acquires vehicle state data and environmental perception data of surrounding targets, including vehicles, pedestrians, and roads; performs multi-agent nested game decision-making based on the vehicle state data and environmental perception data to determine a target interaction mode from different interaction modes; predicts the occupancy risk of the target interaction mode based on the vehicle state data and environmental perception data to obtain a dynamic risk field; and filters candidate trajectories of the vehicle based on the dynamic risk field and preset risk constraints to obtain the target driving trajectory. This achieves joint optimization of traffic efficiency and safety in chaotic road conditions, solving the current technical problem of autonomous driving's difficulty in efficiently and safely navigating complex interaction scenarios, and improving the safety and traffic efficiency of autonomous driving.
Need to check novelty before this filing date? Find Prior Art