Bridge influence line identification using convex programming from Bridge Weigh-in-Motion strain signals

The influence line for a point on a structure shows how shear force or bending moment, or typical strain components, at that point changes as a unit load moves along the structure. The bridge influence line can be used in Bridge Weigh-in-Motion (BWIM) systems and Structural Health Monitoring (SHM) systems. Influence line identification from measurement data is a possible alternative to Finite Element Method-based solutions since measurement-based approaches do not need model validation, and expertise in numerical modelling. A novel convex programming-based influence line identification method is introduced in this paper. Its comparison with previous methods is also executed showing that this novel method results in 0.46% lower Over Relative Errors, while the Amplitude Relative Errors increase by 0.06% in different scenarios compared to the best state-of-the-art methods. The decrement in Overall Relative Error is even higher (0.82%) under high noise level conditions while the Amplitude Relative Error increment (0.09%) is kept low. Robustness tests are also conducted by introducing varying levels of noise to axle loads, vehicle location, and strain signals. Finally, the proposed method is also evaluated experimentally.