Battery Health Prognosis for Electric Vehicles Using Sample Entropy and Sparse Bayesian Predictive Modeling

Abstract

 Abstract-Battery health monitoring and management is of extreme importance for the performance and cost of electric vehicles. This paper is concerned with machine learning enabled battery State-of-Health (SOH) indication and prognosis. The sample entropy of short voltage sequence is used as an effective signature of capacity loss. Advanced sparse Bayesian predictive modeling (SBPM) methodology is employed to capture the underlying correspondence between the capacity loss and sample entropy. The SBPM-based SOH monitor is compared with a polynomial model developed in our prior work. The proposed approach allows for an analytical integration of temperature effects such that an explicitly temperature-perspective SOH estimator is established, whose performance and complexity is contrasted to the support vector machine (SVM) scheme. The forecast of remaining useful life (RUL) is also performed via a combination of SBPM and bootstrap sampling concepts. Large amounts of experimental data from multiple lithium-ion battery cells at three different temperatures are deployed for model construction, verification, and comparison. Such a multi-cell setting is more useful and valuable than only considering a single cell (a common scenario). This is the first known application of combined sample entropy and SBPM to battery health prognosis.

A model‐based and data‐driven joint method for state‐of‐health estimation of lithium‐ion battery in electric vehicles

“…Two lithiumion batteries, numbered A and B, are tested. Main specifications of the investigated lithium-ion batteries are shown in Table 1, and more detailed equipment parameters are referred to the Pan et al 26 According to previous studies, 17,27 the experimental procedure is shown in Figure 1B. The whole experimental period includes three characterization tests (10°C, 25°C, and 40°C) and one aging experiment (45°C).…”

“…In recent years, such methods have been widespread concerned due to their model-free characteristics and high flexibility. The HIs can be obtained through various methods, such as battery capacity, ohmic resistance, 23 model parameters, 10,13 charging/discharging parameters, 17,18,24 and incremental capacity/differential voltage analysis (ICA/DVA) 21,25 . To reduce the data size for aging feature extraction, Hu et al 17 tried to approximate the relationship between the sample entropy of discharge voltage and capacity through the sparse Bayesian predictive modelling (SBPM).…”

“…To date, the main obstacles to the data-driven approach are how to extract effective health indicators (HIs) to quantify battery capacity degeneration and how to find a simple and suitable online estimation algorithm using the microprocessor. The HIs can be obtained through various methods, such as battery capacity, ohmic resistance, 23 model parameters, 10,13 charging/discharging parameters, 17,18,24 and incremental capacity/differential voltage analysis (ICA/DVA) 21,25 . The effective HIs can significantly reduce the computing effort and improve estimation accuracy.…”

An On-Board Remaining Useful Life Estimation Algorithm for Lithium-Ion Batteries of Electric Vehicles

“…In addition to filter algorithms, there exists an abundance of algorithms for RUL estimation based on statistical theory and pattern recognition, such as Bayesian theory [25][26][27], neural networks (NN) and various transformers [16,28], support vector machine (SVM) algorithms [19,25,29,30], etc. The NN and SVM algorithms both have remarkable advantages in dealing with nonlinear modeling problems.…”

Fault-Tolerant Control of ANPC Three-Level Inverter Based on Order-Reduction Optimal Control Strategy under Multi-Device Open-Circuit Fault

“…A SBPM-based SOH monitor was proposed in ref. 23 , which provides a practical method for the realization of fault-tolerant control. An in-situ voltage fault diagnosis method based on the modified Shannon entropy was proposed in ref.…”