Research Topic 1: Design of Tactile Sensors and Data Fusion:

研究方向1：触觉传感器的设计及数据融合算法：

We designed biomimetic tactile fingers to perceive static and dynamic forces and thus realized surface recognition.

我们设计了多种仿生触觉手指，以感受静态力和动态力，并实现表面识别。

Research Topic 2: Processing Algorithm for Spiking Signals.

研究方向2：神经脉冲信号的处理算法

A reference spike train-based computation framwork was proposed to tackle tactile spiking signals.

我们开发了一种基于参考脉冲序列的神经脉冲信号计算框架。

Research Topic 3: Design of Haptic Glove.

研究方向3：触觉手套的设计

A lightweight haptic glove was designed in combination with flexible circuit.

与柔性电路相结合，我们设计了一款轻量化触觉手套。

Detection of stimulus location ​

作用力位置检测

Measurement matrix of the haptic glove

触觉手套针对不同收拾的力测量矩阵

Practical appearance of the glove

触觉手套实物

Research Topic 4: Design of Dexterous Hand.

研究方向4：灵巧手的设计

Biomimetic fingers and hands are designed for dexterous manipulation.

我们开发了多种仿生手指与机械手，用于灵巧操作。

Three fingers' hand driven by PAMs

人工肌肉驱动的三指触觉灵巧手

Five fingered hand driven by PAMs

人工肌肉驱动的五指触觉灵巧手

Flexible grasp of different daily objects by the tactile hand

人工肌肉驱动的三指触觉灵巧手实现不同日常物体的抓取

Mechanocal Design of Dexterous Hand

多自由度灵巧手的机械结构

Different gesture by the motor-driven dexterous hand

多自由度灵巧手的机械结构

Different gestures by the hand

灵巧手展示不同手势

Tactile gripper

两指触觉夹爪

Research Topic 5: Structural and Parameter Optimization

研究方向5：结构与参数优化

Based on DER theory, we developed the mechanical models for the deformation of elastic rods. A inverse solvingalgorithm was proposed based on Genetic Algorithm. A path planning strategy for robotic knot tying was developedfurther. In addition, a bottom-up inverse shape finding algorithm was proposed for 2D rods.

基于DER模型，我们开发了绳状结构变形时的力学模型，基于遗传算法提出了弹性绳网络的反解算法，利用神经网络给出了协作机器人打结时的路径规划，并提出了一种自下而上的二维弹性绳形状反解方法。

Inverse Design

弹性体的逆向设计

Automatic generation of robotic trajectory

机器人轨迹的自动生成

Knot tying by collaborative robots

协作机器人完成打结

Research Topic 6: computer Vision

研究方向6：计算机视觉

We developed related contour extraction algorithm and data augmentation algorithms ect. to recognize soft objects.

我们提出了轮廓提取与数据增强等算法，以识别易变形物体。

Research Topic 7: Humanoid Robot and Manipulations

研究方向7：人形机器人及灵巧操作

Based on multimodal perception, e.g., vision, tactus and so on, dexterous manipulation of humanoid robot is enabled.

基于视觉、触觉等多模态感知信息，我们实现了人形机器人的多种操作。

Biomimetic design of a tactile finger with fingernail

弹性体带指甲触觉手指用于电子元件装配

Our designed humanoid robot with multimodal perception

我们设计的具备多模态感知功能的复合型人形机器人