Optimal discovery of ambient acoustic noise field
Caitlin C. Bogdan, Andrew S. Wixom, James Gregory McDaniel, Sean B. Andersson · 2017
We describe a problem in which a multi-agent acoustic sensing swarm seeks to characterize the ambient acoustic noise field as an input for selecting subsequent sensing locations. We develop a method for solving this problem by applying the Pontryagin Maximum Principle to determine an optimal control law and demonstrate this application on a single agent system. An algorithm is developed for solving the resulting mixed boundary value differential equations in real time by an eigenvalue decomposition method. This technique is demonstrated on a representative ambient acoustic noise field via simulation.