High-fidelity Optimization for Next-generation Shipboard Power Systems
Abstract
Vision and Significance: Massive penetration of power electronic devices, peculiar aspects of pulsed loads, laser weapons, and elect""romagnetic rail guns, hostile and uncertain operational environments, and mission- and safety-critical nature of shipboard power sys""tems have introduced major computational challenges for high-fidelity planning, optimization, and decision making. This challenge is" recognized by the U.S. Navy as highlighted in 2015 Naval S and Tstrategy plan and 2015 Naval Power and Energy Systems Technology D"evelopment Roadmap. Future fleets will have to manage spatially-distributed resources (e.g., power, fluid, thermal). For example, th"ey require an intelligent energy distribution grid to optimally deliver power to loads and weaponry systems without access to the system-level information. The ONR challenge iscentered on control continuity for shipboard engineering and damage control systems under hostile conditions and with limited resources. Resilient energy distribution systems enable the Navy to maintain its mission read"iness despite increased power demand, sensory complexities, and introduction of directed energy weaponry. In response, we seek to cr""eate scalable and distributed optimization methods to address long standing naval problems including survivability assessment, netwo""rk reconfiguration, early-stage design evaluation, and fault detection and monitoring. The proposed research contributes to Ship Sys""tems and Engineering Research thrust and Naval Energy Resiliency and Sustainability thrust in ONR Code 33. Synergistically, it also"" pertains to the Science of Autonomy thrust in ONR Code 35, and Mathematics, Computers and Information Sciences thrust in ONR Code 3""1. Technical Contributions: We will tackle a broad class of computationally hard shipboard power system problems, and create distrib""uted control and optimization methods that are scalable, canbe implemented in real-time, and are robust to parametric, topological,"" and dynamical uncertainties. First, we offer a rigorous mathematical formulation for a variety of shipboard optimization problems b""ased on comprehensive models of hybrid power networks. Next, a family of distributed and scalable computational methods solve result"ing optimization problems. We then leverage distributed computational platforms to achieve orders-of-magnitude improvements in problems scalability. The direction to be pursued advances the area ofoptimization theory and introduces a variety of mathematical tools to the power systems literature for the study of navy-centric energy problems. Our methodology will be manifested on long-standing" navy problems including survivability assessment, early-stage design evaluation, and network reconfiguration. We leverage modern pl""atforms for distributed computation, toachieve orders-of-magnitude improvements in problems scalability, which allows for the desig"n of perfect planning and real-time strategies based on accurate physical models.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Mar 26, 2018
- Source ID
- N000141812186
Entities
People
- Ramtin Madani
Organizations
- Office of Naval Research
- United States Navy
- University of Texas at Arlington