Non-Pavement Bridge Weigh-In-Motion:Theoretical And Experimental Study

Since a large number of heavy vehicles have emerged and the phenomenon of overloading has become increasingly severe, the bridge safety incidents caused by traffic loads have happened frequently in recent years. The weigh-in-motion problem of high-speed vehicle loads that has been attracting much attention is of far-reaching significance in managing overloaded vehicles, administrating and maintaining the infrastructures such as highway and bridge, etc, as well as assessing the bridge health and safety conditions. This paper mainly focuses on a new weigh-in-motion method:non-pavement bridge weigh-in-metion (B-WIM) and takes the dynamic response of bridges generated by passing vehicles as the object and studies the fundamental theory of B-WIM from the point of influence line measurement, axle information detection and vehicle weighing-in-motion. By employing the methods of theory, numerical simulation and test, the weigh-in-motion problem of vehicle loads is investigated. Furthermore, the main work and some conclusions are drawn as follows:1. Based on the idea of modeling separately and coupling partly, the vehicle-bridge coupled system is modeled and the vehicle-bridge coupled vibration function is solved iteratively by using the Matlab program. According to the iterative results, the dynamic strain responses of the bridge are calculated when the vehicles passing which shows that this method can improve the efficiency of calculation and simulate each axle's behavior of arriving and leaving the bridge effectively.2. The theory and optimized algorithm of the non-pavement bridge weigh-in-motion are proposed. The non-pavement B-WIM algorithm is derived from the determination of bridge influence line, detecting the information of axles and vehicle weigh-in-motion. The gradient method for multi-parameter partially optimized is introduced to investigate the initial B-WIM results. By compared with the results of the vehicle-bridge coupled vibration example, the results in this paper are verified. The investigation illustrates that less system parameters are needed and has a better feasibility by taking the bridge dynamic strain response as the object of research and using the influence line fitting method to research the weigh-in-motion of vehicle loads. The B-WIM system takes the existed bridge as a steelyard, increases the length of the sensitive elements, records a long time range of the vehicle information test, obtains more data and has enough precision and obvious advantages over the pavement weigh-in-motion. Meanwhile, the partially optimized research can reduce the deviation resulted from the accidental factors, especially reduce the influence caused by the deviation of velocity identification, improve the precision of the results and meet the requirements of engineering applications.3. The wavelet transform method for axle information identification is put forward. Taking into account the time-frequency localization characteristic of the wavelet transform, the wavelet is used for the axle information identification process of the B-WIM. By extracting wavelet coefficient curve under specified scale from fuzzy signal, the peak points corresponded to the axles are highlighted, especially to axles with short distance. The numerical and test results show that by using the time-frequency localization characteristic of the wavelet transform, extracting wavelet coefficients under specified scales to identify axle information, can reduce the influence of interference signals, identify axle information of double shaft, triple shaft and relatively small wheelbase, and remedy the deficiency of traditional partial peak-picking method about the continuous judgment at the flat peak. Expand the application range of B-WIM.4. Considering the uncertainty of vehicle exciting frequency, the automatic identification algorithm for axle information is proposed. Solving two crucial problems in B-WIM, determining the initial value of exciting frequency and eliminating the interference of bridge vibration is the objective. After automatically searching the suitable excitation initial frequency, the precise excitation frequency and corresponding scale coefficient are obtained through the iterative calculation. Taking the two wavelet coefficient curves of special scale for multiplication when the cross-correlation function reaches its maximum, which can filter the interference of colored noise and automatically identify the axle information. The whole procedure only requires signals of two measuring points on different sections rather than artificially judgments.5. The applicability of B-WIM upon different kinds of bridges and the influence of factors such as the size of bridges, the position of measuring points and measuring noise were analyzed. The results demonstrate that:bridges with orthotropic plate deck, frame bridges, T-beam bridges and slab bridges can be used for B-WIM researches. It can be used in most types of bridges in highway and railway in China and is considered to have broad application prospects.6. Precision assessment standard, procedures, and precision level statistical calculation method of B-WIM system test were developed. The actual engineering application of B-WIM system was evaluated by testing its precision level, and it can provide important reference for applications.