A Novel Practical Technique to Integrate Inequality Control Objectives and Task Transitions in Priority Based Control

The task priority based control is a formalism which allows to create complex control laws with nice invariance properties, i.e. lower priority tasks do not affect the execution of higher priority ones. However, the classical task priority framework (Siciliano and Slotine) lacked the ability of enabling and disabling tasks without causing discontinuities. Furthermore, tasks corresponding to inequality control objectives could not be efficiently represented within that framework. In this paper we present a novel technique to integrate both the activation and deactivation of tasks and the inequality control objectives in the priority based control. The technique, called iCAT (inequality control objectives, activations and transitions) task priority framework, exploits novel regularization methods to activate and deactivate any row of a given task in a prioritized hierarchy without incurring in practical discontinuities, while maintaining as much as possible the invariance properties of the other active tasks. Finally, as opposed to other techniques, the proposed approach has a linear cost in the number of tasks. Simulations, experimental results and a time analysis are presented to support the proposed technique.