Wavelet neural network prediction of electric vehicle chaotic vibration signals#br#
NIU Zhidong 1 WU Guangqiang1, 2
1. College of Automotive Studies, Tongji University, Shanghai 201804, China;
2. Institute of Industrial Science, the University of Tokyo, Tokyo 153-8505, Japan
A vehicle experiment was carried out to study the chaotic dynamics of electric vehicles, and chaotic time series were predicted by using wavelet neural network. First, the experiment of electric vehicle on mediumBelgian road was carried out, the vertical vibration acceleration signal of the right front wheel center and battery bottom center were obtained. Second, timefrequency analysis, threedimensional phase diagrams and the Poincaré sections of the signals were obtained. The time delay was calculated by using the mutual information method, and also minimum embedding dimension was got with the Cao method, the largest Lyapunov exponent was got with Wolf method. The presence of chaotic motions in the vertical acceleration signal was found. Finally, chaotic time series of the right front wheel center vertical signal was predicted using wavelet neural network. It is found that the use of wavelet neural network to predict chaotic time series can achieve better results.
[1] 姜爱萍. 混沌时间序列的小波神经网络预测方法及其优化研究[M]. 上海：上海大学出版社, 2013.
JIANG Aiping. Wavelet neural network prediction of chaotic time series and its optimization [M]. Shanghai: Shanghai University Press, 2013.
[2] 韩敏. 混沌时间序列预测理论与方法[M]. 北京：中国水利水电出版社, 2007.
Han Min. Chaotic time series prediction theory and method [M]. Beijing: China Water & Power Press, 2007.
[3] 许传华, 任青文, 房定旺. 基于神经网络的混沌时间序列预测[J]. 水文地质工程地质, 2003, 30:30-32.
XU Chuanhua, REN Qingwen, FANG Ding-wang. Prediction of chaotic time series based on neural network [J]. Hydrogeology & Engineering Geology, 2003, 30:30-32.
[4] JOSE C. Principe, ALOK Rathie, KUO JM. Prediction of chaotic time series with neural networks and the issue of dynamic modeling [J]. International Journal of Bifurcation & Chaos, 2011, 2(04):989--996.
[5] 陈帝伊, 柳烨, 马孝义. 基于径向基函数神经网络的混沌时间序列相空间重构双参数联合估计[J]. 物理学报, 2012, 61(10):22-31.
CHEN Diyi, LIU Ye, MA Xiaoyi. Parameter Joint Estimation of Phase Space Reconstruction in Chaotic Time Series Based on Radial Basis Function Neural Networks [J]. Acta Phys. Sin, 2012, 61(10):22-31.
[6] 徐东辉, 李岳林, 解福泉. 汽油机瞬态工况油膜参数的混沌径向基神经网络预测模型研究[J]. 汽车工程, 2015:353-358.
XU Donghui, LI Yuelin, XIE Fuquan. A Research on the Prediction Model for the Fuel-film Parameter of Gasoline Engine in Transient Conditions Based on Chaotic RBF Neural Network [J]. Automotive Engineering, 2015:353-358.
[7] 黄润生. 混沌及其应用[M]. 武汉：武汉大学出版社, 2000.
HUANG Runsheng. Chaos and Its Applications [M]. Wuhan: Wuhan University Press, 2000.
[8] 吕金虎, 陆君安, 陈士华. 混沌时间序列分析及其应用[M]. 武汉：武汉大学出版社, 2002.
LV Jinhu, LU Junan, CHEN Shihua. Chaotic time series analysis and its application [M]. Wuhan: Wuhan University Press， 2002.
[9] A. Wolf, J. B. Swift, H. L. Swinney, and J. A. Vastano．Determining Lyapunov Exponents from a Time Series．Phys. D, 1985, 16.
[10] Rosenstein M T, Collins J J, Luca C J D. A practical method for calculating largest Lyapunov exponents from small data sets [J]. Physica D-nonlinear Phenomena, 1993, 65(93):117-134.
[11] 王小川. MATLAB神经网络43个案例分析[M]. 北京：北京航空航天大学出版社, 2013.
WANG Xiaochuan. MATLAB neural network 43 Cases Studies [M]. Beijing: Beihang University Press, 2013.
[12] 张菁, 樊养余, 李慧敏等. 相空间重构中延迟时间选取的新算法[J]. 计算物理, 2011, 28(3): 469-474.
ZHANG Jing, FAN Yangyu, LI Huimin, etal. An Improved Alogorithm for Choosing Delay Time in Phase Space Reconstruction [J]. Chinese Journal of Computational Physics, 2011, 28(3): 469-474.
[13] CAO Liangyue. Practical method for determining the minimum embedding dimension of a scalar time series. Physica D, 1997, 110:43-50.
