Hoeffding Evolutionary Algorithm and its Application in Building Equivalent Circuit Models for Lithium-Ion Battery

Abstract

There has been a growing interest in building equivalent circuit models for lithium-ion battery with evolutionary algorithms. One of the well-known algorithm is the Hoeffding bound based evolutionary algorithm (HEA). In this paper, we first introduce the definitions of the Hoeffding bound, selection operations and Hoeffding evolutionary algorithm. And then, we introduced the method for building equivalent circuit models, which composed of four fundamental functions, connection-modifying function, component-creating function, arithmetic-performing functions and automatically defined function. In this way, a Hoeffding-evolutionary-algorithm based equivalent-circuit-model for lithium-ion battery is modeled. With the model-based state estimation approaches, the obtained model can assess the state-of-charge (SOC) of cells precisely.