The future battlefield will include multiple dissimilar manned and unmanned aerial, ground, sea, and space vehicles working in concert with each other to support fires, logistics, maneuvers, communication, and coordination-based missions. Mission effectiveness and efficiency are often at odds, and due to the distributed and dissimilar energy flows inherent in Multi-Domain Operations (MDO) there is a need to understand, identify, and characterize the energy flows. The ability to analyze the energy flows and effectively maintain adequate energy reserves could provide strategic capabilities to the warfighters, permitting energy informed operations to maximize mission effectiveness and efficiency, while mitigating vulnerabilities. This research focuses on developing energy and exergy characterization through development of Artificial Intelligence (AI), Machine Learning (ML), and Artificial Neural Networks (ANNs) for assessing and analyzing performance of a platform. These types of tools were developed for a Mobile Ground Vehicle (MGV) for two primary reasons: (1) reconcile past state information not directly observable from the available sensor measurements, and (2) to characterize the aggregate energy flow behavior, such that Model Predictive Control (MPC) could be applied to customize the performance of the vehicle. Preliminary results indicate that provided a reference drive cycle and an estimation of the tractive force requirements, either a set of ANNs or Long Short-Term Memory (LSTM) recurrent neural networks can be used to reconcile past sensor measurements and extended to estimate energy flow characterizations of the drivetrain and other ancillary components within the vehicular platform that could affect efficiency and mission effectiveness.
