Optimal design of variable suspension parameters for variable-gauge trains based on the improved CRITIC method

ABSTRACT

The gauge differences of international rail lines have led to the emergence of variable-gauge trains. High-speed variable-gauge trains operating on different lines should meet the dynamic requirements. In order to achieve the above purpose and obtain better dynamic performance, a parameter optimization method for variable suspension systems is proposed. The SIMPACK dynamic model of high-speed variable-gauge trains is established. Six key parameters of the suspension system are chosen, and different suspension parameters are combined using the optimal Latin hypercube. Seven dynamic indices of the vehicle running on three different gauge tracks are solved separately. Weighted SNR is used to perform the multi-objective optimization. The improved CRITIC method is used to solve for the weight coefficients of each index, and find the optimal values of the key suspension parameters. The objective of adjusting the variable suspension parameters of variable-gauge trains is presented. Results from simulation show that the optimized suspension parameters can improve the dynamic performance of the variable-gauge vehicle, especially the derailment coefficient, the wheel load reduction rate, and the axle lateral force. The vehicle can operate with better dynamic performance on both gauge lines by using the designed variable suspension parameters.

COEDITOR-IN-CHIEF:

• Hsiau, Shu-San

ASSOCIATE EDITOR:

• Hsiau, Shu-San

KEYWORDS:

• Variable-gauge vehicle
• dynamic performance
• parameter optimization
• the improved CRITIC method

Nomenclature

$C_j$	=	Amount of information of index $j$
$C_{Pz}$	=	The primary vertical damping
$C_{Sy}$	=	The secondary lateral damping
$C_{yd}$	=	Damping of the anti-snaking damper
$f_i(x)$	=	SNR of the dynamic indices
$K_{Px}$	=	The primary longitudinal stiffness
$K_{Py}$	=	The primary lateral stiffness
$K_{Sy}$	=	The secondary lateral stiffness
$R_j$	=	Conflict indicators of index $j$
$r_{jk}$	=	Correlation coefficient of $j$ and $k$
SNR	=	The Signal Noise Ratio
$y_i$	=	Dynamic indices
${\alpha }_j$	=	Coefficient of variation of index $j$
${\omega }_j$	=	Weight of index $j$

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Key Technology R&D Program of Jilin Provincial Department of Science and Technology (grant no. 20200401108GX); Major R&D projects of Jilin Provincial Department of Science and Technology (grant no. 20210301006GX); and High-speed Intelligent Multiple Unit Major Project of Changchun (grant no. 21GD04).