面向眼内手术的主操作器设计与零力控制方法研究

以视网膜血管注药为代表的眼内手术需要医生在眼内狭小空间中实现长时间、微米级精准操作。受限于医生手部的生理性颤抖、力感知能力,目前这一术式难以应用于临床实践。主从式手术机器人为解决此问题提供了可行途径,该方法中医生通过主操作器控制从机器人完成手术,可规避前述不足对手术效果的影响。但常用的主操作器面向的是通用环境,自由度类型、控制方法均不符合眼内手术的应用场景,导致医生学习成本过高。针对此问题,提出一种面向眼内手术的主操作器,进行了构型和结构设计、样机制作和装配,并完成了对主操作器的动力学参数辨识和基于转矩控制的零力控制算法设计与验证。

Development of a Haptic Interface with Proximity Sensors and Vibration Motors

As described in this paper, we propose a new haptic interface for a service robot. For safety with service robots working in a space where people live, some notification before collision with an obstacle is desirable. To achieve such a function, we developed a master-slave manipulator system in which the slave manipulator surface is covered with many proximity sensors. Additionally, we developed a haptic device that feeds back proximity sense information to the operator using small vibration motors. We attached the haptic device to the arm of the operator and vibrated the vibration motor corresponding to the sensor. Thereby, the operator was able to ascertain the position of an object near the manipulator, and to make the robot maneuver to avoid it before collision. To confirm the system usefulness, we equipped subjects with the developed proximity sense presentation device and performed a detection-position-specific experiment and an obstacle avoidance experiment in a narrow space. As results of the detection position specific experiment on five subjects, four subjects reported the detection position correctly. The remaining one person failed because of his particular arm shape. Operation experiments conducted in a narrow space showed that all subjects＇ work was successful when given feedback of proximity sense information. Nobody was successful without proximity sense information. Results of these two experiments demonstrate that this proposed system is useful for obstacle avoidance of a master-slave manipulator system.

Manipulating active layer morphology of molecular donor/polymer acceptor based organic solar cells through ternary blends

The development of molecular donor/polymer acceptor blend(MD/PA)-type organic solar cells(OSCs) lags far behind other type OSCs. It is due to the large-size phase separation morphology of MD/PAblend, which results from the high crystallinity of molecular donors. In this article, to suppress the crystallinity of molecular donors, we use ternary blends to develop OSCs based on one polymer acceptor(P-BNBP-f BT) and two molecular donors(DR3 TBDTT and BTR) with similar chemical structures.The ternary OSC exhibits a power conversion efficiency(PCE) of 4.85%, which is higher than those of the binary OSCs(PCE=3.60% or 3.86%). To our best knowledge, it is the first report of ternary MD/PA-type OSCs and this PCE is among the highest for MD/PA-type OSCs reported so far. Compared with the binary blends, the ternary blend exhibits decreased crystalline size and improved face-on orientation of the donors. As a result, the ternary blend exhibits improved and balanced charge mobilities, suppressed charge recombination and increased donor/acceptor interfacial areas, which leads to the higher shortcircuit current density. These results suggest that using ternary blend is an effective strategy to manipulate active layer morphology and enhance photovoltaic performance of MD/PA-type OSCs.

A new approach for teleoperation rendezvous and docking with time delay

Teleoperation rendezvous and docking can be used as a backup for autonomous rendezvous and docking (RVD) for an unmanned spacecraft or for guiding the chaser docking with an uncooperative target.The inherent teleoperation time delay is a rigorous problem,especially when the chaser is teleoperated on the ground.To eliminate the effect of time delay,a new approach for teleoperation RVD is studied.The characteristics of teleoperation RVD are analyzed by comparisons with the teleoperation robot and with manually controlled RVD;the relative motion of the chaser is predicted based on the C-W equation;and the processed measure information with time delay through the Kalman filter is utilized to correct the current prediction.Experimental results verify that the approach produces an 18% enhanced success rate of teleoperation RVD compared with direct visual feedback,and consumes less time and fuel.The developed approach also solves the time delay problem effectively.Teleoperation RVD using this method can be applied as a useful backup for autonomous RVD.