一种基于博弈马尔可夫过程安全检测的无人车辆导航控制系统及方法
By employing a safety detection method based on game-theoretic Markov processes, combined with positioning perception, map generation, and integrated control modules, the relationship between line following and obstacle avoidance control in unmanned vehicle navigation was resolved. This enabled the safety assessment of intelligent agent action interactions, thereby improving the safety and reaction speed of unmanned vehicles.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHINA CONSTR THIRD ENG BUREAU GRP CO LTD
- Filing Date
- 2024-01-12
- Publication Date
- 2026-07-17
AI Technical Summary
Existing navigation algorithms struggle to handle the relationship between line following and obstacle avoidance control, and traditional safety detection methods fail to consider the impact of agent actions on safety as well as the impact of interactions between agents.
A safety detection method based on game-theoretic Markov processes is adopted, which combines localization perception, map generation, global path planning, safety assessment and integrated control modules. The safety assessment module analyzes vehicle safety in real time, and uses game-theoretic Markov processes and partially observable Markov decision processes (POMDP) to predict agent action interactions. Lateral, longitudinal and speed safety detection is designed, and a potential field model is introduced for obstacle avoidance control.
By effectively handling the relationship between tracking and obstacle avoidance control, and considering the impact of agent actions on safety and the interaction of actions between agents, the safety and reaction speed of autonomous vehicles are improved.
Smart Images
Figure CN117864168B_ABST