Multi-objective optimisation design of inspection vehicle for indirect bridge measurements

Indirect bridge measurements have attracted growing attention in recent years, with its performance strongly influenced by vehicle parameters. To facilitate efficient and accurate parameter selection for bridges with varying characteristics, an enhanced multi-objective Particle Swarm Optimisation (PSO) algorithm is proposed. Chaotic sequence and Lévy flight are introduced to improve global search performance, while cluster computing is employed to boost computational efficiency. Depending on the chosen objective functions, multiple bridge mode frequencies can be identified. Importantly, complete prior knowledge of bridge parameters is not required; optimisation can be performed using estimated values, making the method well-suited for ageing bridges with incomplete records. A laboratory experiment is conducted to validate the proposed approach and to investigate the effects of vehicle speed and wheelbase. Results show that the algorithm achieves fast convergence and robust global optimisation. It is observed that excessively low vehicle speeds reduce energy transfer in the vehicle–bridge interaction system, and a non-dimensional speed parameter of at least 0.005 is recommended. The influence of wheelbase on frequency identification is found to be negligible.