Enhancing process state monitoring in energy storage systems: A robust design synthesis with physics-informed generalized observer

Abstract

Energy storage systems (ESSs) are crucial for managing renewable energy fluctuations. Knowing ESSs’ states is vital for thermal management. This paper presents a robust design synthesis approach, leveraging a physics-informed generalized observer (GO), for enhancing the process state monitoring in ESSs. The proposed GO takes advantage of the inherent characteristics of ESSs as interconnected subsystems with distinct physical significance. By utilizing measured inputs and outputs, the GO accurately estimates the unmeasurable and unknown process states. GO employs a physics-informed model and a tracking feedback correction mechanism for accurate state estimation. It offers high accuracy, robustness, and insensitivity to measurement noise. To validate the effectiveness of the proposed approach, two case studies are conducted on ESSs. The first involves monitoring the core temperature of lithium batteries, and the second involves an experimental study on monitoring the inner chamber temperature of heat buffer tanks. Experimental results demonstrate approximately 10% improvement compared to conventional methods. GO provides a simple design method, clear physical interpretation, and strong interpretability. Overall, GO is an effective and versatile observer for process monitoring with significant advantages and potential for various applications.