Hard Prioritization Control Allocation: Dealing With the Priority Inversion Phenomenon

This paper proposes the novel hard prioritization control allocation algorithm for systems with task priorities based on a recursive feasible region intersection algorithm that guarantees strict task priority. Hard prioritization addresses the key limitation of the priority inversion phenomenon that arises in a well-known class of task prioritization methods, referred to as soft prioritization algorithms. The worst-case computational cost is shown to be linear with respect to the number of tasks for a generic case. Furthermore, soft and hard prioritization approaches are evaluated through a case study of velocity closed-loop control of an underwater vehicle, demonstrating that correcting the priority inversion phenomenon could improve the transient performance of the primary task.