摘要
为了提高服务机器人的系统服务能力和智能化程度,提出一种面向机械手等器件的感认知增强模型.通过融合来自不同传感通道的感知能力和认知能力,增强机械手的感知和认知.基于此模型,以视觉、触觉为例,在视觉系统上用尺度不变特征变换算法实现目标物体的识别,并对目标物体进行定位;在触觉系统上对触觉传感器采集的数据用主成分分析算法降维,并用支持向量机算法获得分类模型;在认知系统上,机械手在抓取物体时根据物体信息自适应地规划路径并决策物体的抓取模式.通过在多类物体混杂环境中的抓取和分拣实验,验证了该感认知增强机械手系统的可用性.
A perception enhanced model for devices like robotic arm was proposed in order to improve the service capacity and intelligence of service robot.The model enhanced the perception and cognition of robotic arm by fusing perceptive and cognitive ability from different sensing passages.Taking visual and tactile sense for example,scale-invariant feature transform algorithm was applied on visual system to recognize and locate target object,principal component analysis algorithm was applied on tactile system to reduce dimensions of data collected by tactile sensors,and support vector machine algorithm was applied to obtain a classified model.In cognition system,the robotic arm was able to plan the trajectory adaptively and choose grab modes according to the object＇s information when grabbing objects.Grasping and sorting experiments was performed in a mixed circumstance with many kinds of objects,of which the results has verified the availability of the perception enhanced robotic arm system.
出处
《浙江大学学报（工学版）》
EI
CAS
CSCD
北大核心
2016年第6期1155-1159,共5页
Journal of Zhejiang University：Engineering Science
基金
中央高校基本科研业务费专项资金资助项目(2015FZA5017)
浙江省自然科学基金资助项目(LY13F020034)
关键词
感认知增强
机械手
视觉
触觉
认知
enhanced perception
robotic arm
visual sense
tactile sense
cognition
