A non-conventional approach for the computation of inverse time-variant functions

The main objective of this research is to validate the proposed methodology and its non-conventional implementation as an alternative for a faster and efficient computation of inverse time variant functions that can be implemented in real-time.; The main contributions of this Thesis are: (1) A novel methodology for the fast and robust computation of general inverse functions that vary in the time domain. (2) A non-conventional approach based on ANN/RBFN to improve the performance of a conventional numerical solution framework. (3) An efficient non-conventional approach (ANN and RBFN) for the motion planning and singularities avoidance of redundant manipulators that provides faster computation than conventional (numerical) techniques, in particular for (redundant) manipulators with a large number of degrees of freedom. (4) Fast and efficient conventional and non-conventional (ANN) approaches for posture/balance optimization of articulated figures in computer animation, in particular for cases containing a large number of degrees of freedom. (5) A novel geometrical approach implemented conventionally and non-conventionally, based on the null space projection operator for the balance/posture control of articulated figures. (Abstract shortened by UMI.)...