查詢結果分析
來源資料
頁籤選單縮合
題 名 | 基於自組織映射類神經網路之機器腿的逆向運動分析=Inverse Kinematic Analysis of Robot Legs Based on the Self-Organizing Mapping Neural Networks |
---|---|
作 者 | 張志鋒; 廖世謨; 王芯維; | 書刊名 | 高雄應用科技大學學報 |
卷 期 | 39 2010.05[民99.05] |
頁 次 | 頁21-42 |
分類號 | 448.94 |
關鍵詞 | 機器腿; 自組織映射類神經網路; 逆向運動分析; Robot leg; Self-organizing mapping neural networks; Inverse kinematic analysis; |
語 文 | 中文(Chinese) |
中文摘要 | 機器人之腿部機構中若具有閉合迴路運動鏈,則在指定踝關節相對於髖關節之位置向量後可能有許多個可行之構形或運動分支與其對應。對於此種機構之逆向運動分析問題,本文採用自組織映射(SOM)類神經網路以快速決定一組比較適當的構形,並取得該構形所對應之驅動接頭角位移以便進行軌跡控制。首先,以具有五桿運動鏈之機器腿為例,介紹SOM類神經網路之訓練方法和該網路在逆向運動分析上之應用,然後以運動模擬和兩足機器人之步行實驗以驗證此種逆向運動分析方法之正確性,結果顯示這種方法尤其適用於求解具有閉合迴路運動鏈之腿部機構的逆向運動分析問題。 |
英文摘要 | A robot leg mechanism, with closed-loop kinematic chain in it, might have several feasible configurations or kinematic branches corresponding to a specified position vector measuring from hip to ankle. In order that the most appropriate configuration and the driving joint angles of the leg mechanism can be quickly determined for trajectory control, such an inverse kinematic problem is solved by using the self-organizing mapping (SOM) neural network. A biped robot with five-bar chain in its legs is used to demonstrate the training of the SOM neural network and its application to the inverse kinematic problem. Kinematic simulations and walking experiments are then made to verify the success of the approach. The results show that this approach is especially applicable to solve the inverse kinematic problem for those mechanisms with closed-loop chains in them. |
本系統中英文摘要資訊取自各篇刊載內容。