Model updating of the vehicle-track coupled system based on in-situ dynamic measurements

Abstract

With the rapid development of subway transportation, it has brought great convenience to us, but the induced vibration has seriously influenced our daily life. The simulation models are the major approach to investigate this problem. However, due to the discreteness and randomness of the physical parameters of various materials and inappropriate simplifications of the finite element (FE) model, the simulated results cannot reflect the actual dynamic responses accurately. To solve this issue, we carried out systematic investigations as follows: (1) conducting in-situ dynamic measurements of the track structure in an operating subway line; (2) establishing a simulation model of the vehicle-track coupled system; (3) proposing the multi-objective functions according to the simulated and measured results; (4) designing and implementing the sensitivity analysis based on the Latin Hypercube Sampling (LHS); (5) devising and performing the model updating process based on the multi-island genetic algorithm (MIGA). The results reveal that the updated model can simulate the dynamic responses accurately, and the simulation results have an excellent agreement with the in-situ measurements. Overall, the developed method in this work can obviously improve the accuracy of the dynamic simulation of the track structure and provide a way for the model updating and vibration analysis of the vehicle-track coupled system.