- 作者: 黃國勝; 邱正毅
- 作者服務機構: 國立中正大學電機工程學系
- 中文摘要: 人類的肢體運動具有卓越的順服控制能力,因此機器人的運動控制機制的研發,若可從研究人類的肌肉運動而獲得啟發,必有益於增進機器人在適應控制方面的能力。本文探討模擬生理上肌肉控制的方法而發展出機器人順服控制機制策略之可行性,並將此控制機制實際應用於PUMA 560機器手臂控制上。在非拘束運動實驗中,該機制實具有良好的運動位置控制能力。在受拘束運動的實驗結果中,更說明了此類肌肉順服控制模組在無力量感測器之情況下,也具有很好的順服能力。同時該模組也具有兩種運動控制模式,除一般自主性控制外,在非自主性控制模式下,機器人可行力量追隨之操作,雖然在操控模式轉換時,控制參數值需作更替,但由實驗結果顯現,此並不會造成系統的不穩定性。因此藉由機器人在各種運動狀況下的實驗結果得知,本文所提出之類神經肌肉模組對於機器人在順服控制上具有優越的能力。
- 英文摘要: Motivated by a desire to improve the capability of robotic adaptation, a neuromuscular-like modelwith nonlinear damping is developed for robotic control. To implement a robot with neuromuscular control,patterns of motor commands for coordinating multiple joints are studies. In this paper the compliantcapability of the model is demonstrated by controlling movements of a PUMA 560 robot to explore possiblerobotic applications. To demonstrate the free motion, different movements involving joint planning andCartesian planning were tested on a PUMA 560 robot. The experimental results for constrained movementsalso show that neuromuscular control has very good compliant capability to adapt to different constrainedconditions with no force sensor involved. Having the ability to control both constrained and unconstrainedmovements, neuromuscular control is a good mechanism for robotic compliance control.
- 中文關鍵字: compliance control; neuromuscular control; voluntary/involuntary movement; constrained movement
- 英文關鍵字: --