Back to Search Results

Fsheetx Msg Complex

POC: David M. Hetrick (865)576-7556/hetrickdm@ornl.gov http://computing.ornl.gov/cse_home/cms/cmshome.shtml Scientific Computing Within the ORNL Modeling & Simulation Group: Analysis of Complex Systems The ORNL Modeling and Simulation Group (MSG) develops sophisticated numerical solutions for a wide range of scientific, engineering, and operational applications. MSG's core competency is computational physics and engineering, and within this context we have extensive expertise in the analysis of complex systems. This problem solving scope encompasses a diverse set of topics, analysis techniques, and applications.

Human Cognitive Modeling ? Models of human performance, human/machine interaction, emulation of human intelligence, predictive workload, supervisory decision making, and trust in automation.

? Analysis methodologies such as artificial neural nets, support vector machines, nonlinear dimension reduction, discrete event simulation, and agent-based simulations.

? Applications that include nuclear-reactor operators, the U.S. Army's Joint Virtual Battlespace and Future Combat System Programs, and the Missile Defense Agency.

Processing of Chaotic Time-serial Data for Medical Diagnosis and Event-forewarning ? Physics-based analysis techniques such as phase-space analysis, higher-order spectral analysis, Bayesian parameter estimation, principal component analysis, and wavelets.

? The ability to utilize a wide variety of data sources including acoustic, mechanical accelerations, as well as electrical and electromagnetic signals.

? Medical and industrial applications that include epileptic seizure forewarning from scalp EEG data, classification of lung sounds for diagnosing pneumonia, forewarning of fatigue-crack failure in structures using transient loading and strain-rate data, predicting pump cavitation using acoustic data, and detection of high-frequency arcing events using magnetic field data.

Distributed Software Agents Application of intelligent agents to the design of network-centric process-management software using a methodology based on collaborating software agents and their organization into an intelligent decision-making collective.

Potential applications include measurement and control systems and anticipatory diagnostic and prognostic systems.

These applied scientific computing capabilities represent an extremely versatile modeling and simulation resource for a broad range of R&D, industrial, homeland defense, and military applications. We welcome the opportunity to discuss your potential applications and ways MSG can contribute to a solution.

Multi-Agent Collective Historical Data Business Rules Quality Requirements Local Process Data COLLECTIVE INTELLIGENCE THAT POSSESSES: Set of Invariant Truths Dynamic Decision Making Ability Autonomous Behavior Organizational Responsibility Decision Making Democracy Suspend Beliefs & Extract Motive Mission Agent Environmental Agent Self Reference Agent Decision Agent Physics Agent Understands & executes physics-based models of process being controlled Understands decision-making goals of collective Understands environmental interactions of process being controlled Understands its local functionality & its effect on global process Provides expert decision- making assistance to other agents Global Decisions Collaborating software agents produce a collective decision-making intelligence by utilizing consistent application-specific set of languages, ontology's, interaction protocols, definitions for complex behavior, data-security protocols Schematic illustration of the intelligent decision-making collective developed by MSG for management of complex network-centric processes

Physics
Phone: (865)576-7556
Search Home Help About InSpire