Human-Knowledge-Augmented Gaussian Process Regression for State-of-Health Prediction of Lithium-Ion Batteries With Charging Curves

Abstract Lithium-ion batteries have been widely used in renewable energy storage and electric vehicles, state-of-health (SoH) prediction is critical for battery safety reliability. Following the standard SoH routine based on charging curves, a human-knowledge-augmented Gaussian process regression (HAGPR) model proposed by incorporating two promising artificial intelligence techniques, i.e., (GPR) adaptive neural fuzzy inference system (ANFIS). Human knowledge voltage profile during degradation first modeled with an ANFIS feature extraction that helps reduce need physical testing. Then, integrated GPR to enable prediction. Using as baseline, comparison study conducted demonstrate advantage of HAGPR model. It indicates can at least 12% root-mean-square error 31.8% less aging testing compared