A method for optimizing dynamic performance parameters of high-speed trains

Abstract

The invention provides a high-speed train dynamic performance parameter design optimization method, which relates to the field of parameter design optimization based on high-speed train dynamics simulation analysis, which can effectively use the overall target neural network proxy model to replace the high-speed train dynamics simulation model , combining the design analysis of high-speed trains in the multidisciplinary field with multi-objective optimization algorithms, the analysis and optimization based on the approximate model of high-speed train dynamics simulation are carried out. The specific steps of the method are as follows: establish a high-speed train multi-rigid body dynamics simulation model, determine the important input and output design space involved, select a sampling strategy to obtain a suitable design space sample set for high-speed train dynamic performance analysis; use Bayesian regularization The method improves the generalization accuracy of the overall objective neural network, and constructs an overall objective neural network proxy model with error control within a certain range by adjusting the number of nodes in the hidden layer; using the improved overall objective neural network proxy model, combined with differential evolution intelligence The multi-objective optimization algorithm is used to obtain the optimized high-speed train design parameters. It is mainly used for dynamic analysis and design optimization of high-speed trains.

Description

technical field [0001] The invention relates to the technical field of simulation analysis of high-speed train operation performance-related systems and high-speed train parameter design optimization, in particular to parameter design optimization based on high-speed train dynamics simulation analysis. Background technique [0002] From the perspective of dynamics, the high-speed train system is a nonlinear complex space coupling system, which involves track (bridge) system, high-speed train rolling stock system and pantograph-catenary system. In order to solve the design optimization problem of high-speed trains, it is necessary to conduct complex and in-depth research on structural dynamics, vehicle system dynamics, coupling dynamics, aerodynamics, and nonlinear theories. [0003] Structural dynamics is the use of related methods to study the unique problems in the structure of high-speed trains, including car body structural dynamics and track bridge structural dynamics, ...

AI Technical Summary

Problems solved by technology

[0006] At present, no relevant literature has been found to solve the multi-objective optimization problem of dynamic performance parameters of high-speed trains in a complex design space that fully considers the operating indicators. This problem has a high degree of complexity.

Images