A review of non-probabilistic machine learning-based state of health estimation techniques for Lithium-ion battery

An improved feedforward-long short-term memory modeling method for the whole-life-cycle state of charge prediction of lithium-ion batteries considering current-voltage-temperature variation

Remaining useful life and state of health prediction for lithium batteries based on empirical mode decomposition and a long and short memory neural network

Overview of batteries and battery management for electric vehicles

An optimized ensemble learning framework for lithium-ion Battery State of Health estimation in energy storage system

As a favorable energy storage component, lithium-ion (Li-ion) battery has been widely used in the battery energy storage systems (BESS) and electric vehicles (EV). Data driven methods estimate the battery state-of-health (SOH) with the features extracted from the measurement. However, excessive features may reduce the estimation accuracy and also increases the human labor in the lab. By proposing an optimization process with nondominated sorting genetic algorithm II (NSGA-II), this article is able to establish a more efficient SOH estimator with support vector regression (SVR) and the short-term features from the current pulse test. NSGA-II optimizes the entire process of establishing a SOH estimator considering both the measurement cost of the feature and the estimation accuracy. A series of nondominated solutions are obtained by solving the multiobjective optimization problem, which also provides more flexibility to establish the SOH estimator at various conditions. The degradation features in this article are the knee points at the transfer instants of the voltage in the short-term current pulse test, which is fairly convenient and easy to be obtained in real applications. The proposed method is validated on the measurement from two LiFePO4/C batteries aged with the mission profile providing the primary frequency regulation (PFR) service to the grid.

https://vbn.aau.dk/ws/files/331090316/Multi_objective_Optimization_of_Data_driven_Model_for_Lithium_ion_Battery_SOH_estimation_with_Short_term_Feature.pdf

Multiobjective Optimization of Data-Driven Model for Lithium-Ion Battery SOH Estimation With Short-Term Feature

Current pulses are convenient to be actively implemented by a Battery Management System (BMS). However, the Short-Term Features (STF) from current pulses originate from various sensors with uneven qualities, which hinders one powerful and strong learner with STF for the battery SOH estimation. This paper thus proposes an optimized weak learner formulation procedure for Lithium-ion (Li-ion) battery SOH estimation, which further enables the automatic initialization and integration of the weak learners with STF into an efficient SOH estimation framework. A Pareto Front-based Selection Strategy (PFSS) is designed to select the representative solutions from the non-dominated solutions fed by a Knee point driven Evolutionary Algorithm (KnEA), which guarantees both the diversity and accuracy of the weak learners. Afterwards, the weak learners, whose coefficients are obtained by Self-adaptive Differential Evolution (SaDE), are integrated by a weight-based structure. The proposed method utilizes the weak learners with STF to boost the overall performance of SOH estimation. The validation of the proposed method is proved by LiFePO4/C batteries under accelerated cycling ageing test including one mission profile providing Primary Frequency Regulation (PFR) service to the grid and one constant current profile.

/pdf/an-automatic-weak-learner-formulation-for-lithium-ion-1qu8ntf1pw.pdf

An Automatic Weak Learner Formulation for Lithium-ion Battery State of Health Estimation

Refined Instrumental Variable (RIV) estimation is applied to online identify the parameters of the Equivalent Circuit Model (ECM) for Lithium-ion (Li-ion) battery in this paper, which enables accurate parameters estimation with the measurement noise. Since the traditional Recursive Least Squares (RLS) estimation is extremely sensitive to the noise, the parameters in the ECM may fail to converge to their true values under the measurement noise. The RIV estimation is implemented in a bootstrap form, which alternates between the estimation in the system model and the noise model. The Box-Jenkins model of the Li-ion battery transformed from the two RC ECM is selected as the transfer function model for the RIV estimation in this paper. The errors of the two RC ECM are independently generated by the residual of high-order Auto Regressive (AR) model estimation. With the benefit of a series of auxiliary models, the data filtering technology can prefilter the measurement and increase the robustness of the parameters against the noise. Reasonable parameters are possible to be obtained regardless of the noise in the measurement by RIV. Simulation and experimental tests on a LiFePO4 battery validate the efficiency of RIV for parameter online identification compared with traditional RLS.

/pdf/online-parameter-identification-of-the-lithium-ion-battery-3i3gl06nby.pdf

Jichang Peng

Papers

Multiobjective Optimization of Data-Driven Model for Lithium-Ion Battery SOH Estimation With Short-Term Feature

An Automatic Weak Learner Formulation for Lithium-ion Battery State of Health Estimation

Online Parameter Identification of the Lithium-Ion Battery with Refined Instrumental Variable Estimation