| Summary: | A novel electrical circuit analogy is proposed to model electrochemical systems under realistic automotive operation conditions. The model is developed for a lithium ion battery and is based on a pseudo 2D electrochemical model. It calculates the evolution of species concentration distribution and diffusion for a given current load, as a result of electrochemical reactions. The application example is an automotive system, in which results are obtained at a rate faster than real-time and well within the accuracy requirements of a typical Battery Management System (BMS). This is the first Equivalent Circuit Network (ECN)-type model that tracks directly the evolution of species inside the cell, and includes implementation of complex electrochemical phenomena usually omitted such as double layer capacitance and overpotentials due to mass transport limitations in the electrode host material. Three networks, one for each of the three species present, electrons, Li ions, and intercalated Li atoms, are connected through a triple species element, which governs the conversion between species at the resulting activation, diffusion and passivating layer overpotentials. The link between the three networks is an important feature of this model, coupling information of each of the three joining circuit networks at this node. The model is fully thermally coupled and can account for capacity fade via a decrease in the amount of active species and for power fade via an increase in a resistive solid electrolyte inter-phase layer at both electrodes. The model’s capability to simulate cell behaviour under dynamic events from 0% to 100% State-of-Charge (SoC) conditions is demonstrated with various test procedures. Examples for model output are given for various standard battery testing load cycles as well as realistic automotive drive cycle loads. Although cast in the framework familiar to application engineers, the model is essentially an electrochemical battery model: all variables have a direct physical interpretation and there is direct access to all states of the cell via the model variables (species concentrations, potentials) for the later purpose of control systems design. Further model extensions are easily implemented atop the current framework. The presented methodology is further applicable for electrochemical system performance evaluation and prediction of battery performance in any kind of portable battery-powered electronic device and application with low computational power availability and online solution requirements.
|
|---|
