Random-Field Estimation For Dynamics Of Robots

Abstract

Report discusses use of random-field mathematical models as alternatives to deterministic models of classical mechanics to describe dynamics of robot arms. These alternative models used to establish relationship between methods of estimation theory and robot dynamics. Approach yields new class of algorithms performing computations typical of estimation theory to solve such fundamental problems in robotics as forward and inverse dynamics and inverse kinematics.