人工皮膚之發展與NTU-Hand IV之整合

Abstract

本篇論文之目的，是發展一套適用於實驗室已發展之NTU-Hand IV的人工皮膚，使NTU-Hand IV有觸覺的感知能力。此外並以模糊理論為基礎的抓握控制演算法，配合人工皮膚，提升NTU-Hand IV的抓握能力。 人工皮膚是由兩種感測器組成。正向力感測器是由撓性電極板和壓力導電橡膠 CSA（Pressure Conductive Rubber, CSA）結合而成，以感測正向力的大小及分佈；滑動感測器的材料是PVDF (PolyVinyliDene Fluoride)，用以感測抓握物體的滑動情形。透過兩種感測器，使人工皮膚具有感測三維觸覺的感測能力。最後亦發展適用於人工皮膚之硬體、韌體及使用者圖形介面，並與獨立的控制系統作整合。 在抓握控制演算法方面，力量/位置混合控制為抓握控制的架構，並引進了以模糊控制為基礎的力量控制以及多感測器融合理論。在抓握時，將人工皮膚所感測到的訊號作為模糊控制的輸入，根據資料庫推論出適當的馬達扭力大小，並考慮力量封閉和摩擦圓錐，使NTU-Hand IV能達到穩定抓握，並擁有近似人類抓握的能力。

The purpose of this thesis is to develop an artificial skin system that is suitable for NTU-Hand IV, which was developed in our laboratory. The artificial skin is composed of normal force sensor and slippage sensor. To enhance the grasping ability, a fuzzy-based force control algorithm integrated with artificial skin system is applied. The artificial skin is combined with two kinds of sensors. One is a normal force sensor. A flexible printed circuit board is combined with a pressure conductive rubber (CSA) to detect the distribution and magnitude of normal force. The other one is a slippage sensor, whose sensing material is PVDF (PolyVinyliDene Fluoride) piezoelectric polymer strip. Slippage sensor is used to detect the slip of grasped object. Therefore, the artificial skin has 3D tactile sensing ability via these two kinds of sensors. Moreover, the hardware-firmware system and the graphical user interface are developed and integrated with a stand-alone control system. The hybrid force/position control is the architecture of the grasping control. A Fuzzy-Based force control methodology and multi-sensor fusion are applied and integrated with the artificial skin. According to the signals sensed from the artificial skin, an inference engine will infer an adequate motor torque from data base. Friction cone and force closure are also considered while grasping. NTU-Hand IV with the artificial skin will then be simulated to grasp well as that of mankind performance.