A New AR Model for Detecting the Outliers of Power Battery Currents

Ruzhen Dou, Fang Liu, Chao Ma, Yitong Gou


Outliers detection for electric vehicle power battery current data, of which accuracy is related to the estimated SOC of the electric vehicle, cannot use traditional algorithms because of the character of the data. In this paper we propose a new AR model, which we called Double-Crossed AR (DCAR) model, to satisfy this demands. New AR model reserves the advantage of real-timeness and rapidity of traditional AR method and overcomes the disadvantage of dependence on mean and variance of AR model which is unsuitable to outlier detection of electric vehicle power battery current data. Experimental simulation verifies the practicability and performance of this new algorithm at the end of the paper.


Outliers detection, AR model, electric vehicle power battery current data

Full Text:



  • There are currently no refbacks.

Copyright © 2017 by Academic Research Publishers Co., Ltd. All rights reserved.