Knowledge Model for Selecting and Producing Reaching Movements

Abstract

This article presents a new model of reaching control. The aim of the model is to characterize the computations underlying the selection of coordinated motion patterns among the limb segments. When a spatial target is selected, stored postures are evaluated for the contributions they can make to the task, and a special weighted average (the gaussian average) is taken of the postures to find a single target posture. Movement to the target posture is achieved without explicit planning of the trajectory. Rather, the reaching motion is driven by error correction (reducing the discrepancy between the current and target posture) shaped by inertia. The model solves the degrees-of-freedom problem for reaching. It also allows joints to compensate automatically for reduced mobility of other joints and explains established effects of practice, speed–accuracy trade-off, and kinematics. The model can be extended to other tasks and motor subsystems because of the generality of its underlying concepts.

Key Words:

• compensation
• inverse kinematics
• learning
• reaching
• speed–accuracy trade-off