Scheduling networked state estimators based on Value of Information. (December 2019)

• Record Type: Journal Article
• Authors: Molin, Adam; Esen, Hasan; Johansson, Karl Henrik
• Is Part Of: Automatica. Volume 110(2019)
• Publication Date: 2019-12
• Subjects: Networked control -- Scheduling algorithms -- Sensor fusion -- State estimation; Automatic control -- Periodicals; Automation -- Periodicals; 629.805
• DOI: 10.1016/j.automatica.2019.108578
• Languages: English
• ISSNs: 0005-1098

Abstract

Abstract: This paper proposes a novel method for information management of a networked system for the purpose of efficient state estimation. Data shared between entities is transmitted over a common communication network. Based on the notion of Value of Information (VoI), we design a prioritizing scheduler in which systems determine the benefit of sharing their real-time data to improve state estimation accuracy. The system with the highest VoI is granted a time slot to provide its data. By using a rollout strategy, feasible algorithms are developed for computing the VoI-based priorities. For decoupled subsystems, performance certificates of the VoI-based strategy are derived. An automotive case study evaluates the proposed approach.