ADAPTIVE DECENTRALIZED RESOURCE OPTIMIZATION
Abstract
In this project, we will develop algorithms in support of automated decentralized decisionmaking system that adapts to unknown changes in the surrounding environment. Problems of optimal decision-making among multiple agents can be formalized as adequate optimization problems. The major scientific challenge is that these optimization problems should not be solved in a centralized fashion, but rather, the agents themselves need to cooperate on a decentralized computation of the optimal decision and, furthermore, the agents have to repeatedly adapt and revise their decision as new information arrives. Our proposed work has objective to advance the state-of-the-art of distributed computations by tackling head-on the full complexity of the optimization. In particular, we will handle timevarying objectives, local constraints which involve both individual agents as well as their neighbors and, crucially, global constraints on the entire system.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 04, 2017
- Source ID
- N000141712126
Entities
People
- Angelia Nedich
Organizations
- Arizona State University
- Office of Naval Research
- United States Navy