首页 | 本学科首页   官方微博 | 高级检索  
     

具有个性化的虚拟自主汽车视觉行为模型研究
引用本文:娄燕,卢永明,何汉武,郑德涛.具有个性化的虚拟自主汽车视觉行为模型研究[J].计算机工程与应用,2006,42(17):184-187.
作者姓名:娄燕  卢永明  何汉武  郑德涛
作者单位:1. 广东工业大学机电工程学院,广州,510090;南华大学机械工程学院,湖南,衡阳,421001
2. 广东工业大学机电工程学院,广州,510090
基金项目:广东省科技攻关项目;湖南省教育厅科研项目
摘    要:构造了一个虚拟自主汽车(VirtualAutonomousVehicle,VAV)视觉行为模型:信息获取、障碍物避免、目的地追踪和行为融合四部分组成。采用模糊神经网络控制器来模拟该模型中的障碍物避免行为,此控制器共分为五层,可以避免传统的势场法产生局部最小问题,或用模糊控制器造成的主观性强的问题,使VAV轨迹更加平滑。通过神经网络学习,修正模糊隶属函数参数值,可模拟不同类型的驾驶员。

关 键 词:VAV  视觉行为模型  模糊神经网络  个性化
文章编号:1002-8331-(2006)17-0184-04
收稿时间:2005-12
修稿时间:2005-12

Study on the Visionaction Model of Virtual Autonomous Vehicle with Individuation
Lou Yan,Lu Yongming,He Hanwu,Zheng Detao.Study on the Visionaction Model of Virtual Autonomous Vehicle with Individuation[J].Computer Engineering and Applications,2006,42(17):184-187.
Authors:Lou Yan  Lu Yongming  He Hanwu  Zheng Detao
Affiliation:1School of Electro-Mechanical Engineering,Guangdong University of Technology,Guangzhou 510090; 2School of Mechanical Engineering,Nanhua University,Hengyang,Hunan 421001
Abstract:A vision-action model of virtual autonomous vehicle is presented.It consists of information receiving,objection avoidance,goal seaking and behavior fusion.The behavior of obstacle avoidance is simulated by fuzzy neural net controller which consists of five layers.It avoids the local smallest problem of the potential method and the subjective problem of fuzzy controller so that the VAV goes more smoothly.The parameter of fuzzy member function can be corrected with the neural network as result that the model can simulate the driver with individuation.
Keywords:VAV
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号