Parametric identification and dynamic characterisation of flexible manipulator system

Abstract

This paper presents an investigation of an intelligence modeling technique for dynamic characterization of a single-link flexible manipulator system. The flexible manipulator system was first modeled using finite difference (FD) method. A bang-bang torque was applied as an input and the dynamic response of the system was investigated. Performance of the algorithm is compared with the manipulator theoretical natural frequencies obtained analytically. Next, a parametric identification of the system is developed using the conventional Least Square (LS) algorithm and the intelligent Genetic Algorithm (GA). Comparative assessment is presented for validation of the model in characterizing the manipulator system in frequency domain. The developed genetic-modeling approach will be used for control design and development in future work.