Wireless Sensor Networks And Data Fusion For Structural Health Monitoring Of Aircraft
Abstract
This thesis discusses an architecture and design of a sensor web to be used for structural health monitoring of an aircraft. Also presented are several prototypes of critical parts of the sensor web. The proposed sensor web will utilize sensor nodes situated throughout the structure. These nodes and one or more workstations will support agents that communicate and collaborate to monitor the health of the structure. Agents can be any internal or external autonomous entity that has direct access to affect a given system. For the purposes of this document, an agent will be defined as an autonomous software resource that has the ability to make decisions for itself based on given tasks and abilities while also collaborating with others to find a feasible answer to a given problem regarding the structural health monitoring system. Once the agents have received relevant data from nodes, they will utilize applications that perform data fusion techniques to classify events and further improve the functionality of the system for more accurate future classifications. Agents will also pass alerts up a self-configuring hierarchy of monitor agents and make them available for review by personnel. This thesis makes use of previous results from applying the Gaia methodology for analysis and design of the multiagent system.