rotocraft Ground Noise Exposure Prediction System Using Neural Networks

Period of Performance: 12/30/1998 - 06/30/1999

$120K

Phase 1 SBIR

Recipient Firm

Applied Aero, LLC
48967 Ventura Drive
Fremont, CA 94539
Principal Investigator

Research Topics

Abstract

The objective of Phase I is to develop and demonstrate a neural networks system to predict, map, and interpolate rotorcraft ground noise exposure data. It is well-known that the relationships between ground noise exposures. Helicopter flight paths and operation conditions are extremely complicated and involved with many parameters, such as type of helicopters, power setting, approach path, and meteorological conditions. The formulation of such a function relationship with neural networks using experimental data seems the only viable approach. A neural network system, which is based on some existing data under theoretical guidance, will be designed and implemented, and be used to demonstrate the feasibility of the selected neural network algorithms. With the prediction system developed in Phase I, a prototype system will be designed and constructed in Phase II for a helicopter cockpit to display acoustic noise characteristics, noise level variability of different flight operation conditions, including various meteorological environments. The definition of Phase II will be formulate in Phase I in order to accomplish the implementation of prototype system under various flight test-conditions on a real-time base.BENEFITS: The rotorcraft ground noise exposure can be predicted by the newly developed system on a real-time base. In military application, the prototype system in the helicopter cockpit can guide a pilot to maneuver the vehicle to avoid potential threats in a hostile environment. In civil application, inner-city operations, police activities, and emergency operations will benefit from the technology in the approach terminal area.