| Bridge weigh-in-motion(BWIM) technology is a method of using instrumented bridge as a large scale to identify axle weights of moving vehicles across the bridge through continuously collected bridge response. Accurate vehicle configuration(axle numbers, axle spacing and vehicle speed,) determinations are crucial for the level of accuracy of the identified axle weights and gross vehicle weight. The latest BWIM systems, called nothing on the road(NOR) or free-of-axle detector(FAD) BWIM technology, collect vehicle configuration by attaching additional transducers mounted underneath the bridge slab to induce signals of the passing vehicles so as to detect them. However field tests demonstrate that the collected signals from these FAD transducers sometimes are difficult to identify vehicle axles directly. This paper presents a wavelet-based analysis of stain signals and shows the efficacy of using wavelets in pattern recognition of these signals. Based on field tests on an approach span of the Fifth Huaihua Bridge located in Huaihua, Hunan province, China, a 42-ft long prestressed concrete T-girder bridge, located on Highway I-459 in Hoover, Alabama, and simply-supported slab bridge, located in Guilin, Guangxi province, the application of vehicular axles identification was investigated on the latest BWIM system with different bridge types under different circumstances, and wavelet technic was used to help BWIM system to improve the axle identification. Emphasis of this thesis was put on the following aspects:(1) To apply wavelet technic on signals of BWIM system for improving vehicular axle identification based on the field-test data, and then provide technical support for axle weight and gross weight identification in BWIM system;(2) To prove that the applications of all of the continue wavelet transform(CWT), stationary wavelet transform(SWT) and discrete wavelet transform(DWT) to these signals could also be used to improve the accuracy level of identification of vehicle axle. Comparison with different types of wavelet transform(WT), the paper states that different types of WT possess their own advantages and specific limitations in the scopes of application;(3) To propose an approach by jointly controlling the minimum Shannon entropy and maximum correlation to choose the best wavelet scale and the base wavelet respectively, in order to achieve the most suitable identification result with CWT;(4) To apply wavelet transform to the FAD signals from the field tests on I-459 and demonstrate that the wavelet analysis shows efficient potentiality in identifying axle information in multiple presence, avoiding the miss of some or all trucks on the bridge in the case of multiple presence and thus improving the efficiency and accuracy in the successful application of BWIM system. Furthermore, the signals of sensors installed on the soffit of diaphragms were also performed by wavelet analysis in the paper and the results demonstrated that these signals from diaphragm sensors could be used to identify the axle information as well with wavelet transform;... |
PDF Full Text Request