Profile and contact force estimation of cable-driven continuum robots in presence of obstacles

Abstract

Accurate prediction of shape and contact forces significantly improves the performance of a continuum robot during its operation in obstacle-laden environments. This paper presents an optimization-based mathematical framework to predict the bending profile of a cable-driven continuum robot in presence of obstacles. The kinematics model is derived from the concept of strain energy minimization and can easily incorporate obstacles as inequality constraints in the optimization-based approach. The location of point of contact can be identified by observing the Lagrange multipliers of the inequality constraints. Using the kinematics model and the principle of virtual work, a method to estimate the reaction forces at contact is proposed. The model shows high accuracy, with RMS error of 1.35 mm in prediction of the pose for experiments conducted on a 180 mm long robot prototype. Validation experiments are also conducted on the prototype by imposing contact at different locations on the robot. In all cases, the average error in predicting the contact force is found to be less than 1.0 g for applied loads ranging from 50 to 350 g.