Online TCP trajectory planning for redundant continuum manipulators using quadratic programming

摘要

Continuum manipulators enable new fields of application by their inherent flexibility and continuous deformations. However, they cannot compete with conventional robots yet. This is partially caused by a lack of practical and powerful trajectory generation algorithms. Therefore, an online TCP trajectory generator for redundant continuum manipulators has been developed, that computes desired actuator states for given task variables that can be used for actuator controllers. Using the fast gradient optimization algorithm and quadratic programming, actuator velocities are optimized considering actuator constraints with respect to a cost function including the differential kinematics. Experiments prove the online capabilities of the proposed trajectory generation approach. Furthermore, the presented method can be used for controlling continuum robots by incorporating inputs from human-machine-interaction devices.