2.5 KVA Inverter with Selectable Output Voltage for Ground Mobile Vehicle Use.

The Phase II objective is to develop a Multi-agent Architecture for Robust Adaptive Decision-aiding (MARAD), specifically to support enhanced adaptive user interfaces and tactical event predictions via sequential learning. Feasibility of MARAD wasdemonstrated under the Phase I effort, in which we developed a sequential learning component that supports the capturing of user interactions with the visualization interface to identify user habits and to create preference profiles for adaptive displays.The learning component was also used to analyze battlefield event sequences to generate rules to predict significant tactical events. We embedded the learning component within our in-house BattleField Viewer (BFV), which provided visualization andinteraction services for MARAD, in a complex battlefield simulation environment.Under Phase II, we propose to fully develop MARAD by: 1) expanding its capabilities across multiple scenarios; 2) expanding the functionality of its learning component to generate enhanced user profiles and event rules; and 3) introducing into MARAD a datafusion component and a situation assessment component to facilitate enhanced learning of the battlefield context. We will develop and implement MARAD based on our in-house BattleField Viewer (BFV) and CHI Systems' C3Core. We will investigate compliance ofMARAD with the Joint Technical Architecture (JTA).