Estimation of battery state-of-charge using ν-support vector regression algorithm

Abstract

Accurately estimating the SOC of a battery during the electric vehicle drive cycle is a vital issue that currently remains unresolved. A support vector regression algorithm (SVR), which has good nonlinear approximation ability, a quick convergence rate and global optimal solution, is proposed to estimate the battery SOC. First, the training data and the test data required in the estimation operation are collected using the ADVISOR software, followed by normalization of the data above. Then, cross validation and grid search methodologies are used to determine the parameters in the v-SVR model. Finally, simulation experiments have been carried out in the LIBSVM simulator. The simulation results show that, compared to the BP neural network algorithm, the v-Support Vector Regression algorithm performs better in estimating the battery SOC.

Robustness analysis of State-of-Charge estimation methods for two types of Li-ion batteries

“…Many SOC estimation models have been proposed, such as artificial neural networks based models [5e7], fuzzy logic models [8,9] and support vector regression (SVR) based models [10,11]. The robustness of these models strongly relies on the quantity and quality of the training data set.…”

Estimation of State of Charge of a Lithium-Ion Battery Pack for Electric Vehicles Using an Adaptive Luenberger Observer

“…Consequently many computational intelligence-based and optimization-based approaches such as artificial neural networks based models [6][7][8][9][10][11][12][13] and fuzzy logic models [14][15][16][17][18] have been chosen to implement the processes of input selection, training and validation, to establish an adequately accurate SoC estimation model. Support vector regression (SVR) based methods were also applied to realize the SoC estimation of batteries, such as the standard  -SVR model [19], the least squares SVR model [20], the  -SVR model [21] and the fuzzy clustering based SVR model [22]. This type of method can often produce a good estimate of SoC, due to the powerful ability to approximate nonlinear function surfaces.…”

A Novel Sliding Mode Observer for State of Charge Estimation of EV Lithium Batteries

“…3. Neural network [3]- [6]: a large number of samples with comprehensive data are required for model training and the sample data as well as the training methods affect the accuracy. 4.…”