International Session (Emerging Technology in Civil Engineering)

[CS2-12] Vehicle Identification on Passing Lane Road using Bridge Acceleration Responses

〇MURTUZA PETLADWALA¹, SHOHEI KINOSHITA¹, SHIGERU KASAI¹ (1.NEC CORPORATION JAPAN)

Keywords:Vehicle identification, Bridge health monitoring, Traffic status, Non-parametric Bayesian

The identification of traffic status over a bridge is important to select the free vibration region, which is used in vibration modal analysis. In conventional systems, vehicle identification is based on threshold-based approaches, which are specific to driving lane & are sensitive to identify presence of vehicle on passing lane. To solve such problem, an unsupervised approach is proposed, which includes clustering of time-repetition features, extracted from the impulsive responses of vehicle passages. Our experimental results shows 94.6% accuracy in identifying vehicles on passing lane, in a traffic status, where vehicles are present on each lane of the bridge.