Supervisory Control System for Ship Damage Control: Volume 1 - Design Overview
Abstract
This report presents a state-of-the-art concept of automated situation awareness for ship damage control. The solution encompasses model-based crisis recognition, model-based predictive validation, automated casualty response, and a supervisor interface console. The report details a solution approach to the Supervisory Control decision-making task for damage control management of fire, smoke, flooding, pipe rupture, and stability aboard ships. The solution is relevant to ships built in the future, as it assumes intelligent sensors and actuators that are more advanced than those that exist on any current ship. The solution encompasses all three stages of Supervisory Control: model-based crisis recognition, model-based predictive validation, and automated casualty response. This report also describes an approach to the creation and evaluation of a Supervisor-Operator Override Console because of its integral relationship to Situation Assessment. The described Supervisory Control solution is designed to function iii vitro, by use of the existing Illinois Damage Control Scenario Generator armed with simulated smart sensors and actuators. And it is designed to function in vivo, in live crisis exercises aboard a floating research laboratory called the ex-USS Shadwell. The three major research areas of the described Supervisory Control solution are knowledge-based expert systems, machine learning in noisy domains, and assessment of supervisory human-computer interfaces.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 27, 2001
- Accession Number
- ADA392798
Entities
People
- David C. Wilkins
- Frederic W. Williams
- Janet A. Sniezek
- Patricia A. Tetem
Organizations
- United States Naval Research Laboratory