Solving kinematics problems by efficient interval partitioning

In this study, a generic solution methodology is presented for solving constraint satisfaction problems with smooth continuous constraint functions. We apply the proposed method in solving small scale kinematics problems with the goal of identifying all real solutions to a given problem. The developed approach assumes a collaborative methodology that integrates interval partitioning, a new interval inference method, and local search. The resulting methodology aims at reducing the search space and discarding infeasible sub-spaces effectively and reliably during the early stages of the search so that exact solutions can be identified faster by local methods.