论文部分内容阅读
通常的柔顺性控制采取连续的力伺服控制。我们提出用开关力监控来进行柔顺性控制、它只反映机械手各关节受力是否达到允许阈值,因而检测手段大大简化;这样的柔顺性控制过程成了离散的动作序列,只要通过分析与示教,提出状态分析与运动规划规则,并运用规划生成系统和归纳学习方法来自动编制动作序列并转为程序,就能在同一类型操作中适应各种变化。这种机械手不是简单地再现人的示救动作,它能总结示教中的规律,以适应多变的操作过程因而其示教具有智能。
The usual compliance control takes continuous force servo control. We propose to use the switch force monitoring to carry out compliance control, which only reflects the manipulator joints force to reach the allowable threshold, which greatly simplifies the detection means; this flexibility control process has become a discrete sequence of actions, as long as the analysis and teaching , Presents the rules of state analysis and motion planning, and uses the planning generation system and inductive learning method to automatically compile sequences of actions and turn them into programs to adapt to changes in the same type of operations. This manipulator is not a simple reminder of human salvage action, it can summarize the teachings of the law to adapt to changing operation so that its teaching is intelligent.