Real-time robot dynamic simulation on a vector/parallel supercomputer

The authors present the results of research performed in computational robot dynamics to achieve real-time simulation of a manipulator on a general-purpose vector/parallel computer. Once the complex dynamics equations required in the simulation have been effectively vectorized, a coarse-grain parallel block predictor-corrector method for performing the motion integration is realized on multiple CPUs to exploit a form of temporal parallelism. Results on a CRAY Y-MP8/864 show that effective use of vectorization and parallelization yields an order-of-magnitude speedup, resulting in a computational rate 50 times faster than real-time for reasonable end-effector position errors on the order of a micron. This translates to real-time performance on a less powerful parallel computing system.<
>