Vetter Philipp, Flash Tamar, Wolpert Daniel M
Sobell Department of Motor Neuroscience, Institute of Neurology, University College London, United Kingdom.
Curr Biol. 2002 Mar 19;12(6):488-91. doi: 10.1016/s0960-9822(02)00715-7.
There are infinitely many different combinations of arm postures which will place the hand at the same point in space. Given this abundance, how is one configuration chosen over another? Two main hypotheses have been proposed to solve this problem. Postural models suggest that the posture adopted is purely determined by the desired hand position (known as Donders' law). Transport models suggest that the adopted posture depends on where the hand has moved from. A specific transport model, the minimum work model, has been proposed in which the adopted posture is the one that minimizes the amount of work required to move the hand to the new location. The postural model predicts that the posture will be independent of where the hand has moved from, whereas the transport models predict that the posture will depend on the previous posture. We have devised a simple redundant task-touching a target bar using a hand-held virtual stick-to examine these models. The results show that neither model alone can account for the data. We propose a control planning strategy in which there is a combined cost function that has both a postural term as well as a transport term.
存在无数种不同的手臂姿势组合，这些组合都会使手处于空间中的同一点。鉴于有如此多的组合，那么如何在其中选择一种配置而非另一种呢？为解决这个问题，人们提出了两种主要假设。姿势模型表明所采用的姿势完全由期望的手部位置决定（这被称为东德斯定律）。运动模型表明所采用的姿势取决于手从何处移动而来。一种特定的运动模型，即最小功模型，被提出来了，其中所采用的姿势是使将手移动到新位置所需的功量最小化的姿势。姿势模型预测姿势将与手从何处移动而来无关，而运动模型预测姿势将取决于先前的姿势。我们设计了一个简单的冗余任务——使用手持虚拟棒触碰目标杆——来检验这些模型。结果表明，单独任何一个模型都无法解释这些数据。我们提出一种控制规划策略，其中存在一个组合成本函数，它既有姿势项又有运动项。
