Criteria for Recognition of Wheel Skid of Trains Based on Time Series Analysis

Abstract:

Current criteria to judge wheel skid of trains such as velocity difference often cannot recognize wheel skid timely and have no uniform critical value for different trains or railway lines. Aiming at the disadvantages, new criteria based on time series analysis are proposed. With appropriate method of order determination and parameter estimation, AR time series model is established for the data series of velocity difference. Then, Greens function and characteristic equation are constructed with the parameters of the model to determine wheel skid by the convergence state of Greens function or the value of characteristic equations roots. Simulation result shows that the two criteria based on time series model can recognize wheel skid earlier than velocity difference. Moreover, the roots of characteristic equation can also be used as a criterion with a uniform critical value under different application conditions.

• Periodical: Applied Mechanics and Materials (Volume 440)
• Pages: 237-242
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.440.237
• Online since: October 2013
• Authors: Jun Bin Peng*, Xiao Yi Hu, Yong Jun Liu
• Keywords: Criteria, Recognition, Time Series Analysis, Train, Wheel Skid

References

[1] Junbin Peng: Traction and Brake of EMUs, China Railway Publishing House, Beijing, (2009).

[2] Aiqin Jia, Jianjun Chen: Intelligent Control of the Anti-skid System in a Brake System, Mechanical Science and Technology, Vol. 29, No. 7 (2010), pp.956-961.

[3] Yan Wang: Applied Time Series Analysis, China Renmin University Press, Beijing, (2005).

[4] Yanzhi Dong: Analysis and Modeling of Time Series Based on Matlab, Computer Engineering, Vol. 29, No. 12 (2003), pp.170-172.