Parameter identification of moving vehicles considering road roughness of bridge

Using the concept of dynamic moving influence matrix in VDM, a moving vehicle identification method in vehicle-bridge coupled system is derived and elaborated, and for which the road roughness is considered. The vehicle is simulated by a two degree of freedoms mass-spring vehicle model. In this method, vehicle parameters are taken as the variables and then optimized by minimizing the square distance between the measured responses of the bridge and its estimated responses with high identification accuracy. The method is robust to noise and moreover requires fewer sensors than that needed in moving loads identification methods. During the optimization, the repetitive construction of the variant system matrix can be avoided by the utilization of the dynamic moving influence matrix and thus the optimization efficiency is high. A numerical example of a three-span beam with rough surface is used to verify the proposed method, in which the multiple vehicle parameters and moving forces can be identified successfully by one sensor with the pollution of 5% Gaussian noise. In addition, the identifications of vehicle parameters on respective ideal smooth road surface and rough surface are discussed. Via the sensitivity analysis of the vehicle parameters, it further states the influence of the road roughness and vehicle parameters on the structural response. At last it gives the selection suggestion of the optimization variables in different identification cases.