Dynamic Data Driven Methods for Self-aware Aerospace Vehicles

Abstract

A self-aware aerospace vehicle can dynamically adapt the way it performs missions by gathering information about itself and its surroundings and responding intelligently. Achieving this DDDAS paradigm enables a revolutionary new generation of self-aware aerospace vehicles that can perform missions that are impossible using current design, flight, and mission planning paradigms. To make self-aware aerospace vehicles a reality, fundamentally new algorithms are needed that drive decision-making through dynamic response to uncertain data, while incorporating information from multiple modeling sources and multiple sensor fidelities.In this work, the specific challenge of a vehicle that can dynamically and autonomously sense, plan, and act is considered. The challenge is to achieve each of these tasks in real time executing online models and exploiting dynamic data streams–while also accounting for uncertainty. We employ a multifidelity approach to inference, prediction and planning an approach that incorporates information from multiple modeling sources, multiple sensor data sources, and multiple fidelities.