以期待法則設計移動式機器人之人物跟隨行為與實驗評估

Abstract

本論文發展一個期待法則之人物跟隨行為在移動機器人上，本系統由人物追蹤、人物行為模型以及控制策略等組成。人物行為模型是基於人類行走模型以及社交互動而設計，該模型可以用來預測人物行走軌跡以協助機器人決定合適的行為。採用以下三種人物跟隨情境來驗證本系統效能：背後跟隨、並肩跟隨以及前方跟隨，每個情境有其特殊性以及應用。從這些實驗中可以觀察到多種人與機器人之間的互動與反應，並且從實驗數據中發現，考慮人物行走模型可提升人物跟隨的表現。

In this thesis we propose an anticipative human following behavior for the mobile robot. The system is composed by: a human tracking system, a human model and a control strategy. The design of the human model was based on relevant scientific material about human walk and social interaction. This model was used to predict the human trajectory and allow the robot to take preemptive action. The system performance was assessed through experimental evaluation of three specific human following scenarios: following from behind, following on the side and following in front. Each of these scenarios has its particularities and applications, thus providing insight about the effectiveness and usability of anticipative behavior.