面向网络遥操作的多传感器集成手爪的设计

以网络遥操作为应用背景，设计了一种分层的多传感器集成手爪。设计中首先提出了包含了数据采集层、数据处理层、网络交互层等三个层次的手爪体系结构。然后提出了基于支持向量机的融合算法，并结合CAN总线等技术对手爪系统进行了逐层设计，使其具有模块化、智能化等特点。还搭建了多传感器集成手爪实验系统，完成了基于多传感器集成手爪的网络遥操作作业。实验结果表明：该手爪具有较强的局部自主能力，并有效提高了网络遥操作系统的作业能力。

Design of Multisensory Gripper for Internet-based Teleoperation

A layered multisensory integration gripper is designed for internet-based teleoperation. During the process of design, the gripper architecture is developed, which includes data acquisition layer, data process layer and network interface layer. A support-vector-machine-based data fusion algorithm is proposed. Combined with other technologies such as CAN bus, all layers of gripper system are finished, endowing the gripper with characteristics including modularization and high intelligence. A gripper test bed is also built on which the internet-based teleoperation jobs is based are achieved. The experimental results show that the gripper system has a good ability of local-autonomy and can improve the operation capability of internet-based teleoperation system effectively.