1. Simplified Least Squares - Support Vector Machines for Lead-Acid Batteries SOC Estimation
- Author
-
Li Chuan Zheng, Hong Tao Chen, Yi Hui Zheng, Jian Ming Zhao, Li Xue Li, and Xin Wang
- Subjects
Support vector machine ,Battery (electricity) ,Polynomial ,Engineering ,Nonlinear system ,business.industry ,Multivariable calculus ,Kernel (statistics) ,Control engineering ,General Medicine ,Lead–acid battery ,business ,Least squares - Abstract
Lead-acid batteries are widely used in all walks of life, the State-of-Charge is the most important part of the battery management system. On account of the strong coupling, multivariable, nonlinear characteristics of the batteries, this paper adopts the LS-SVM method to predict remaining battery capacity. The nonlinear POLYnomial (POLY) kernel is employed to design the LS-SVM. Aiming at predigesting the hardware requirement of the LS-SVM application, this paper uses the deduced induction of the input vector to simplify the POLY kernel, greatly reducing the memory capacity of the practical application. At last, simulation on MATLAB was done to verify the validity of the proposed model, the simulation results show that the LS-SVM based on the simplified PLOY kernel is applicable to battery SOC estimation, and the estimation error can be controlled within 5%.
- Published
- 2014