作者: Valentin Schmidt , Bertram Müller , Andreas Pott
DOI: 10.1007/978-94-007-7214-4_12
关键词:
摘要: This paper investigates a new approach for solving the forward kinematics of cable-driven parallel robots. combines an interval algorithm with neural networks to provide fast but accurate initial guess. The increase computation speed by factor 200 or more, while provides guaranteed convergence and definite solution any chosen degree accuracy. Iterative techniques are faster still, proposed is considered real-time feasible.