A neural network based postattack damage assessment system

Abstract

Key elements of an automated damage assessment (ADA) will include ground-based sensors to survey and measure postattack damages, communication networks to link sensors, a survival recovery center (SRC), a runway repair team (or robots) for rapid response, and advanced signal processors to perform the 'search and optimization' processes for the 'best' airbase recovery plan. To meet the USAF ADA requirements, ITT Avionics has proposed the development of a hybrid signal processor. The system will consist of algorithmic processors and neural networks. To improve DA performance, key DA functions are implemented by neural networks. Due to the intrinsic nature of distributed processing power, the neural network not only provides the high throughput required for DA but it also achieves fault tolerance and graceful degradation, which are extremely important for the Rapid Runway Repair program.<>