Joint Online Estimation of Multiple Operating States of Lithium-Ion Batteries Considering Safety Constraints Based on Physics-Informed Neural Networks

Lithium-ion batteries have various operating states, including state of power (SOP), state of health (SOH), state of charge (SOC), and remaining useful life (RUL). Exact estimation of these parameters, while maintaining battery operational safety, is crucial for enhancing battery efficiency and developing scheduling strategies. In view of the problems in current research, such as the inability to quantify battery safety indicators, the single-parameter estimation, and the need for improved estimation accuracy, this paper first designs constraints by quantifying the state of safety (SOS). Then, an online joint estimation framework for multiple operating states of the battery is constructed based on the physics-informed neural network(PINN). Finally, verification is carried out using an actual battery operation dataset. Results indicate that the method proposed in this study can effectively guarantee battery operational safety and enhance the precision of operational state estimation.