Behavioral Decision Based on Fuzy Inference and Finite State Machine
王亮WANG Liang;苏东旭SU Dong-xu;杨兴龙YANG Xing-long;
马文峰MA Wen-feng;王子军WANG Zi-jun
(一汽奔腾轿车有限公司,长春130103 )
(FAW Pentium Car Co., Ltd.,Changchun 130103, China )
摘要:本文提出了一种应用于单向双车道的规则型行为决策算法,通过将模糊推理与有限状态机相结合,提高了算法的可行性和决策结果的准确性,基于有限状态机的决策方法具备场景遍历广度优势,但对某一特定场景缺乏遍历的深度,其状态转移条件也较为简单。使用层次状态机与模糊推理相结合的决策算法,其不仅能提升场景遍历深度,还能根据速度和距离进行自适应调节,进一步提升换道决策的准确性与合理性。先在Matlab自动驾驶工具箱中建立道路场景,然后提出并构建基于模糊推理和有限状态机的行为决策模型,采用B样条曲线进行轨迹规划、模型预测控制进行路径跟踪,最后通过Simulink仿真验证了决策系统的可行性。
Abstract:'This paper proposes a rule-based behavional decision algorithm for one-way two lane traffic. By combining furmy reasoningwih Finite-state machine, the feaibility of the algorithm and the accuracy of the decision resuls are improved.'The decision method basedon Finite-state machine has the advantage of scene traversal breadlth, but it lacks the depth of traversal for a specific scene, and is statelransition condlitions are relatively simple. The decision algorithm combining hierarchical state machines and fuzy reasoning can not onlyimprove the depth of scene traversal, but also adapltively adjust according to speed and distance, further improving the accuracy andrationality of lane changing decisions. First, the road scene is established in the Matlab autopilot toobox, then a behavioral decision modelbased on furay reasoning and Finite-stale machine is proposed and constructed. B-spline curve is used for trajectory planning, and modelpredictive control is used for path tracking.Finally, the feasibility of the decision system is verified through Simulink simulation.
关键词:模糊推理;层次状态机;换道决策
Key words: fuzzy reasoning; hierarchical state machine; lane change decision
|
