The principal focus of this project is the development of computational schemes and algorithms for the real- time simulation of complex multibody systems on multiprocessor computing systems. The results of this project may have a direct impact on the design, simulation and control of large multisegment, multi- joint structures as well as the design of software for the newly-emerging multi-processing hardware. The project utilizes graph theory and operations research methods to investigate inherent parallelisms in computational schemes and their suitability for execution on multiprocessing systems. Automated symbolic algebra packages are employed to generate novel theoretical formulations for describing the dynamics of complex multibody systems which incorporate rigid and flexible bodies. The work is fundamental in nature and is applicable of problems in other scientific and engineering disciplines, in particular the study of motor control is biosystems using neural net computational models.
