Research on Online Identification of Lithium-ion Battery Equivalent Circuit Model Parameters

The selection of the lithium-ion battery equivalent model and the identification of battery state parameters are the focus of its research field. The battery parameter identification includes two methods, offline and online. Offline recognition is not only a time-consuming process, but also produces insufficiently accurate results. In order to achieve accurate estimation of the power battery state and accurate identification of parameters, the accuracy of the battery model needs to be improved. In this paper, a ternary lithium-ion battery is used as the research object, and a second-order RC equivalent circuit model is established. First, the HPPC cycle test experiment is used to identify the battery parameters offline, and then the forgetting factor recursive least square method is used to calculate the equivalent circuit model. Conducted online parameter identification, and built a simulation model in the MATLAB/Simulink software platform to verify the results of online identification, compared with offline identification methods, and analyzed the advantages and disadvantages of online identification. The research results show that the model parameters obtained by the genetic factor recursive least square method are more accurate than offline identification, and can reflect the internal state parameters of the battery in real time. It is a good online identification method, which lays the foundation for practical application.

I. Introduction

In today’s environment of advocating green, low-carbon, energy conservation and environmental protection, energy conservation and emission reduction have become the focus of development of any industry. In the automotive industry, electric vehicles have received widespread attention worldwide because of their low emissions, low noise, economic and practical advantages, and more and more researchers have started to study electric vehicles [1]. As the engine of electric vehicles, the research of power batteries has also made rapid progress. As the current leader of power batteries for electric vehicles, the selection of its equivalent model and the identification of battery state parameters are the focus of its research field [2]. The parameter identification of battery includes offline and online methods. Offline identification is not only a time-consuming process, but also a phenomenon of “data saturation”. In order to realize the accurate estimation and identification of the state parameters of power batteries, it is necessary to establish an online identification model of battery parameters with high accuracy, fast convergence and fast data processing, so as to realize the real-time reflection and update of the internal state parameters of batteries in operation.