A Novel Combined Data Mining Algorithm for State-of-Charge Estimation in Electric Vehicle

As lead-acid battery in electric vehicle (EV) is nonlinear system, it is difficult to establish the robust, accuracy relationship between the load voltage and the current under different temperatures and state of charge (SOC). A novel combined data mining algorithms is put forward to investigate state of charge (SOC) estimating model. The fuzzy C-Means (FCM) clustering algorithm is used to divide test data into several clusters. As a high performance learning machine based on statistical learning theory, the samples in every cluster are trained by the support vector machine (SVM) and the FCM-SVM algorithm is got. Tests are performed in different discharge current and chamber temperatures by battery evaluation and testing system. Compared with the Multiple Regression, BP Neural network, Momentum BP Neural network, SVM with polynomial, SVM with S-shaped kernel, SVM with RBF kernel, the FCMSVM with RBF kernel model can simulate the battery dynamics superior generalization accuracy and robust. The average relative error of FCMSVM with RBF kernel model is 2.55%.

[1]  Xuexia Yang.  Expansion Research on K-means Clustering Algorithm Based on Rough Set , 2012 .

[2]  Gregory L. Plett,et al.  Extended Kalman filtering for battery management systems of LiPB-based HEV battery packs: Part 3. State and parameter estimation , 2004 .

[3]  Gregory L. Plett,et al.  Sigma-point Kalman filtering for battery management systems of LiPB-based HEV battery packs Part 2: Simultaneous state and parameter estimation , 2006 .

[4]  Terry Hansen,et al.  Support vector based battery state of charge estimator , 2005 .

[5]  Chang Yan,et al.  A FCM Algorithm Based on Weighted Intuitionistic Fuzzy Set , 2012 .

[6]  Sheng Fu Shasha Cai Xiaomin Zhu.  Design of Intelligently Online Detection and Analysis System for Automobile Transmission , 2012 .

[7]  J. Suykens.  Nonlinear modelling and support vector machines , 2001, IMTC 2001. Proceedings of the 18th IEEE Instrumentation and Measurement Technology Conference. Rediscovering Measurement in the Age of Informatics (Cat. No.01CH 37188).

[8]  S. Rodrigues,et al.  A review of state-of-charge indication of batteries by means of a.c. impedance measurements , 2000 .

[9]  Chen Ning.  Fuzzy K-Prototypes Algorithm for Clustering Mixed Numeric and Categorical Valued Data , 2001 .

[10]  Andreas Jossen,et al.  Methods for state-of-charge determination and their applications , 2001 .

[11]  Hongwen He,et al.  Evaluation of Lithium-Ion Battery Equivalent Circuit Models for State of Charge Estimation by an Experimental Approach , 2011 .

[12]  K. T. Chau,et al.  A new battery available capacity indicator for electric vehicles using neural network , 2002 .

[13]  Yong Chen,et al.  RBF Kernel Based Support Vector Machine with Universal Approximation and Its Application , 2004, ISNN.

[14]  Souradip Malkhandi,et al.  Fuzzy logic-based learning system and estimation of state-of-charge of lead-acid battery , 2006, Eng. Appl. Artif. Intell..

[15]  Chen Quanshi,et al.  Support vector machine based battery model for electric vehicles , 2006 .

[16]  Shunyi Zhang,et al.  A Novel Support Vector Machine Awareness Model Based on Genetic Algorithm Optimization for Peer-to-Peer Network Traffic , 2012 .