Catalogue of Artificial Intelligence Techniques
Author(s): Roy Featherstone
Robot dynamics addresses the problems of calculating the acceleration of a robot from a given set of forces (forward dynamics), and calculating the forces required to produce a given acceleration (inverse dynamics). Forward dynamics is used for simulation, inverse dynamics for control. A robot is considered to be a system of rigid bodies, or links, connected together by joints. The laws of motion for rigid bodies are used to find the equations of motion for the robot. The two main approaches are via Newton's and Euler's equations, and via Lagrange's equation. The tricky part is to calculate the answers efficiently. This is done by expressing the equations of motion in terms of recurrence relations between the motion of one link and that of one of its neighbours in such a way that the answers may be calculated recursively.
- Paul, R.P., Robot manipulators
, MIT Press, Cambridge, Mass., 1981.