A Plug-and-Play (PnP) Tool based on Online Machine Learning for Real-Time Monitoring and Control of Mechanical Systems

The goal of the project is to develop and demonstrate a plug-and-play (PnP) platform based on online neural network (NN) learning and modeling for real-time monitoring, prognostics, and control of mechanical systems. In Phase I, key technology elements of real-time HUMS data analysis including “around-theclock” NN learning, feature selection, model predictive control, and embedded computing platform were developed in a modular environment. By way of extensive US Army relevant case studies, critical evidence of our technology was established to enable in-process performance prediction and control reconfiguration for mission completion under degraded system performance. In the proposed Phase II effort, the platform will be optimized in performance and functionality. The NN engine will be refined in efficiency and reliability. Fault detection and identification algorithms will be developed for rapid responses. Control modification to mitigate anomaly will be improved to accelerate online synthesis. Software-hardware integration will be realized on the autonomous embedded AI computing platform with hybrid CPU/CPU architecture for enhanced efficiency and deployability. The technology will be integrated with Armydesignated workflow of HUMS data analysis to enable real-time situational responses. Both software and AI-enabled hardware will be extensively validated and demonstrated via selected case studies of Army interest.