Search for: All records

This work presents an integrated architecture for a prognostic digital twin for smart manufacturing subsystems. The specific case of cutting tool wear (flank wear) in a CNC machine is considered, using benchmark data sets provided by the Prognostics and Health Management (PHM) Society. This paper emphasizes the role of robust uncertainty quantification, especially in the presence of data-driven black- and gray-box dynamic models. A surrogate dynamic model is constructed to track the evolution of flank wear using a reduced set of features extracted from multi-modal sensor time series data. The digital twin's uncertainty quantification engine integrates with this dynamic model along with a machine emulator that is tasked with generating future operating scenarios for the machine. The surrogate dynamic model and emulator are combined in a closed-loop architecture with an adaptive Monte Carlo uncertainty forecasting framework that allows prediction of quantities of interest critical to prognostics within user-prescribed bounds. Numerical results using the PHM dataset are shown illustrating how the adaptive uncertainty forecasting tools deliver a trustworthy forecast by maintaining predictive error within the prescribed tolerance. 

This paper considers path planning with resource constraints and dynamic obstacles for an unmanned aerial vehicle (UAV), modeled as a Dubins agent. Incorporating these complex constraints at the guidance stage expands the scope of operations of UAVs in challenging environments containing path-dependent integral constraints and time-varying obstacles. Path-dependent integral constraints, also known as resource constraints, can occur when the UAV is subject to a hazardous environment that exposes it to cumulative damage over its traversed path. The noise penalty function was selected as the resource constraint for this study, which was modeled as a path integral that exerts a path-dependent load on the UAV, stipulated to not exceed an upper bound. Weather phenomena such as storms, turbulence and ice are modeled as dynamic obstacles. In this paper, ice data from the Aviation Weather Service is employed to create training data sets for learning the dynamics of ice phenomena. Dynamic mode decomposition (DMD) is used to learn and forecast the evolution of ice conditions at flight level. This approach is presented as a computationally scalable means of propagating obstacle dynamics. The reduced order DMD representation of time-varying ice obstacles is integrated with a recently developed backtracking hybridA∗ graph search algorithm. The backtracking mechanism allows us to determine a feasible path in a computationally scalable manner in the presence of resource constraints. Illustrative numerical results are presented to demonstrate the effectiveness of the proposed path-planning method.