Trajectory optimization of pneumatically actuated, redundant continuum manipulators

For the planning of time-critical paths for redundant continuum manipulators, dynamic input constraints such as the maximum air mass flow on the pressure dynamics of a pneumatically actuated system have a significant impact on the transition time. This paper presents a trajectory generation framework by using optimal control strategies that consider not only the mechanical dynamics, but also the pressure dynamics of a pneumatically actuated system. For the formulation of the optimal control problem, a decoupled dynamic model of the actuators and their pressure dynamics including a maximal mass flow estimation is used. The kinematic model of the manipulator is based on the constant curvature approach. The nonlinear optimal control problem with dynamic input constraints is discretized and solved numerically. Optimization results from an example manipulator show a significant reduction of the transition time, compared to simple distance-based optimizations.