基於使用者意圖之被動式行動輔助機器人操控與導引系統研發

Abstract

隨著高齡化社會的來臨，行動輔助機器人應用在老年照護的相關研究快速發展。但在真實環境及安全的考量下，如何讓被動式行動輔助機器人能基於使用者意圖進行高效率操控與導引輔助相當重要。在導引輔助的研究上，基於使用者意圖應用在輔具的相關研究非常少，尤其是在主要依靠使用者施力來行動的被動式行動輔助機器人上。但使用者的行動意圖相當程度的影響到此類型輔具系統的運作，更突顯此方面研究的重要性。在本研究中，我們提出基於使用者意圖之被動式行動輔助機器人操控與導引系統。在導引方面，我們開發了力感測握把，用來偵測使用者作用在輔具的施力，乃透過所提出的學習方法來推得使用者握力與作用在輔具的推力/力矩之間的關係，並結合本實驗室之前所發展的導引策略[27]進行導引輔助。在操控方面，我們開發了髖關節旋轉角度偵測器，用來偵測使用者的行走朝向角，並結合力感測握把，根據兩者量測到的使用者輸入與狀態，建立使用者意圖偵測系統，接著將所偵測到的行走意圖輸入到新開發的被動式輔助控制策略，即時輸出適當的輔助煞車力，以達到基於使用者意圖的高效率行走輔助。最後，我們將所提出的系統實現在本實驗室所開發的被動式行動輔助機器人i-Go上，並邀請包括銀髮族在內的使用者進行實驗以驗證系統效能。

As the number of elderly population is rapidly growing in our society, robot walking helpers has come up as a research focus. To be practical and safe for use in daily life for the elders, one issue of interest is whether they can be effectively manipulated following user intention and provide guidance for the users when demanded. As quite a number of guidance schemes have been proposed, the study on effective manipulation and guidance based on user intention is comparatively few. These concerns become even more imperative for a passive robot walking helper, as it relies mainly on user’s applied forces to move. In this dissertation, we propose an effective manipulation and motion guidance system for a passive robot walking helper based on user intention. For motion guidance, we develop a pair of force sensing grips to identify user-applied force imposed on the passive robot walking helper via a learning scheme for the mapping between the measured applied force and driving force/torque imposed on the helper. This learning scheme, along with a guidance strategy extended from our previous work [27], provides motion guidance for the elders during walking. For effective manipulation, we propose a system to recognize motion intention and turning angle of the user from the developed hip rotation detector and the force sensing grips aforementioned, and then utilize a newly developed assistive control strategy to generate proper braking torque for effective manipulation. The feasibility of the proposed system is demonstrated via a series of experiments on motion assistance using i-Go, a passive robot walking helper developed in our laboratory. And target users, such as the elders, were invited for the experiments.