Parallel processing of robotics inverse dynamics computational tasks of a two-rotational-link robot

A parallel processing scheme is described for implementing intensive computational tasks involved in the Newton-Euler (NE) equations of motion of robotics inverse dynamics problem of a two-rotational-link robot. This paper shows a formalism to efficiently solve the robotics inverse dynamics problem of the above mentioned manipulator. The aim is to use a DATG (Directed Acyclic Task Graph) to calculate efficiently precedence-constrained strongly NP-complete computational tasks in the NE equations and attain highest possible parallelism. The goat of this paper is also to determine the speed-up of parallelized computations compared to the computations involved in the uniprocessor (sequential) version of the same robot manipulator.