Abstract: A lightweight bridge modal parameter identification method based on vehicle-bridge interaction and displacement reconstruction was proposed for small- and medium-span bridges with sparse sensor deployment.Measured bridge accelerations were first double-integrated,and support displacement constraints were introduced to suppress integration drift and reconstruct bridge displacement responses.Variational nonlinear chirp mode decomposition was then used to extract the quasi-static displacement component.A quasi-static displacement observation matrix defined by sensor locations and time samples was constructed,and singular value decomposition was used to identify the first three bridge mode shapes.Natural frequencies were further obtained by combining the fundamental frequency estimated from a single-point spectrum with a singular-value constraint relationship.An ABAQUS finite element model of a three-span continuous girder bridge in the Tong’an section of the Xiamen BRT system was established,and parametric analyses were conducted for vehicle speed,vehicle mass,road roughness grade,and sensor number.Under the reference case,the first three natural frequency relative errors were 0.71%,3.43%,and 4.28%,and the corresponding modal assurance criterion (MAC) values were 0.9998,0.9974,and 0.9890,using only three accelerometers.Under different parameter conditions,all mode shape MAC values exceeded 0.988,and the frequency errors remained within an acceptable range.The method provides a feasible low-cost approach for lightweight modal identification of small- and medium-span bridges.