微型机与应用
微型機與應用
미형궤여응용
Microcomputer & its Applications
2015年
19期
80-82,85
,共4页
无人车%贝叶斯网络%障碍物%分类
無人車%貝葉斯網絡%障礙物%分類
무인차%패협사망락%장애물%분류
unmanned vehicle%Bayesian network%obstacle%classification
针对无人车获得的障碍物信息的不确定性和不完整性以及贝叶斯分类器对不完整信息比较敏感等不足,选择了贝叶斯网络分类方案。该方案在构建贝叶斯网络模型时,将贝叶斯推理运用其中,提高了贝叶斯网络的识别率。该分类系统依托某研究所无人车项目,通过激光雷达和 CCD 传感器获取障碍物实时信息,为网络模型提供数据信息。
針對無人車穫得的障礙物信息的不確定性和不完整性以及貝葉斯分類器對不完整信息比較敏感等不足,選擇瞭貝葉斯網絡分類方案。該方案在構建貝葉斯網絡模型時,將貝葉斯推理運用其中,提高瞭貝葉斯網絡的識彆率。該分類繫統依託某研究所無人車項目,通過激光雷達和 CCD 傳感器穫取障礙物實時信息,為網絡模型提供數據信息。
침대무인차획득적장애물신식적불학정성화불완정성이급패협사분류기대불완정신식비교민감등불족,선택료패협사망락분류방안。해방안재구건패협사망락모형시,장패협사추리운용기중,제고료패협사망락적식별솔。해분류계통의탁모연구소무인차항목,통과격광뢰체화 CCD 전감기획취장애물실시신식,위망락모형제공수거신식。
Because of the obstacle information′s uncertainty, incompleteness and the short of Bayesian classifier that it is sensitive for incomplete information, the paper selects a Bayesian network classification scheme. When building the Bayesian network model, this program using Bayesian inference which can improve the adaptability of the Bayesian network. The classification system relies on a research institute unmanned vehicle project. Real-time information is obtained via laser radar and CCD sensor that provides data to the network model.