Award Data

LongShortWay Inc. and Northeastern University propose a family of feature reduction and ensemble classifier methods based on Principal Component and Dynamic Logic feature clustering algorithms. New methods combine feature clustering with non-linear feature reduction via manifold learning, and bagging, boosting, and stacking ensemble algorithms.

The U.S. Air Force is interested in developing hypersonic vehicles including reusable transport aircraft, cruise missiles, and unmanned systems. Hypersonic flight regimes result in multifaceted and very difficult design challenges that can be encapsulated into an aerothermoelastic problem, which is a complex interaction of structural, thermal, and aerodynamic mechanisms. When a flexible structural ...

Automatic speech recognition (ASR) technology is in wide use today and has been successfully integrated into a number of applications. Although there are also many potential applications for automatic speaker identification (SID), SID accuracy rates are generally not sufficient for widespread use in practical applications. For high accuracy ASR and SID, the microphone must be located near the talk ...

Toyon Research Corp. and the Penn State Univ. propose research and development of innovative algorithms for classifying objects and activities observed in high-dimensional data extracted from multi-sensor motion imagery. The proposed algorithms include novel feature clustering techniques to enable effective characterization of intra-class and inter-class appearance variations in datasets containin ...

Experimental data from a number of DoD supported research programs have indicated that electrical systems capable of producing highly repetitive bursts of high voltage, fast risetime pulses are a critical enabling technology for applications tied to nonequilibrium physical processes, such as combustion, boundary layer flow, and the generation of high power EM waves. Recognizing the important role ...

ABSTRACT: The U. S. Air Force needs turbulent combustion models that can be used to simulate combustion in actual propulsion systems at both design and off-design conditions, not models that are only useful for highly idealized problems. With this motivation in mind, Combustion Science & Engineering, Inc. and the Computational Combustion Lab at Georgia Tech plan to enhance current capabilities to ...

Structural Health Monitoring (SHM) can be reliably used to perform on-line health monitoring of any type of structures with minimal human involvement. Current SHM systems can perform the functions required but are heavy, bulky and difficult to integrate with the structure to provide on-board real-time structural integrity assessment. Embedding of currently available sensors and electronics used by ...

This work will focus on the development of monolithic SWIR focal plane array technologies using CMOS or CMOS compatible fabrication technology. This will be realized on a Silicon substrate and incorporate APD+TIA arrays and be scalable to Megapixel arrays and coherent receiver operation for capture of the full optical wavefront vector information.

In this program, HyPerComp and University of Michigan have teamedtogether to develop a high-order grid generator for Euler and viscousmeshes. The grid generator is based on HyPerComps successful generalpurpose CAD2Mesh software and is being integrated with HyPerCompsHDphysics and U. Michigans XFlow DG high-order solvers. High-order gridgeneration methods are being implemented to accurately capture ...

Enig Associates, Inc., a small business providing advanced modeling and simulation capabilities to the DoD and DoE, is proposing an innovative and novel electrical approach, using explosive-driven flux compression generators to Joule heat light metal load in sub millisecond time scale from a solid metal state to a first ionization plasma state going through multi-phase transitions to generate an a ...