Expectation Maximization Algorithm for a Battery State of Health Model with Missing Data

Abstract

With the popularity of lithium-ion batteries, battery state of health (SOH) estimation has become one of the current research hotspots. Due to network congestion, collected data usually encounter time-delay or packet loss. In this paper, an expectation maximization (EM) algorithm is proposed for the SOH model which is approximated by a polynomial model. Based on the EM method, the missing data are computed in the E step, and the parameters are updated in the M step. Compared with the least square method, the proposed algorithm has more accurate estimation accuracy. The simulation example shows the effectiveness of the proposed algorithm.