Deep Sensing Approach to Single-Sensor Vehicle Weighing System on Bridges

Kawakatsu, Takaya; Aihara, Kenro; Takasu, Atsuhiro; Adachi, Jun

doi:10.1109/jsen.2018.2872839

Cited by 18 publications

(12 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The results confirmed the viability of a new strategy for axle detection. Kawakatsu et al [142] proposed a single strain sensor-based BWIM. The obtained data were automatically optimized by consulting a surveillance camera.…”

Section: Vehicle-classification-based Methodsmentioning

confidence: 99%

Vehicle-Assisted Techniques for Health Monitoring of Bridges

Shokravi

Bakhary

et al. 2020

Sensors

View full text Add to dashboard Cite

Bridges are designed to withstand different types of loads, including dead, live, environmental, and occasional loads during their service period. Moving vehicles are the main source of the applied live load on bridges. The applied load to highway bridges depends on several traffic parameters such as weight of vehicles, axle load, configuration of axles, position of vehicles on the bridge, number of vehicles, direction, and vehicle’s speed. The estimation of traffic loadings on bridges are generally notional and, consequently, can be excessively conservative. Hence, accurate prediction of the in-service performance of a bridge structure is very desirable and great savings can be achieved through the accurate assessment of the applied traffic load in existing bridges. In this paper, a review is conducted on conventional vehicle-based health monitoring methods used for bridges. Vision-based, weigh in motion (WIM), bridge weigh in motion (BWIM), drive-by and vehicle bridge interaction (VBI)-based models are the methods that are generally used in the structural health monitoring (SHM) of bridges. The performance of vehicle-assisted methods is studied and suggestions for future work in this area are addressed, including alleviating the downsides of each approach to disentangle the complexities, and adopting intelligent and autonomous vehicle-assisted methods for health monitoring of bridges.

show abstract

Section: Vehicle-classification-based Methodsmentioning

confidence: 99%

Vehicle-Assisted Techniques for Health Monitoring of Bridges

Shokravi

Bakhary

et al. 2020

Sensors

View full text Add to dashboard Cite

show abstract

“…Algohi et al (2020) derived a modified influence area method for BWIM to consider the effect of variable vehicle speed. Kawakatsu et al (2019Kawakatsu et al ( , 2020 presented a "fully neural" BWIM method that uses onedimensional deep CNNs to learn the relationship between vehicle parameters and bridge responses from monitoring data. In addition, considerable efforts have also been dedicated to the development of moving force identification (MFI) methods, which seek to find the time histories of vehicle axle forces exerted on the bridge, to improve the identification accuracy and robustness (Z.…”

Section: Introductionmentioning

confidence: 99%

“…Kawakatsu et al. (2019, 2020) presented a “fully neural” BWIM method that uses one‐dimensional deep CNNs to learn the relationship between vehicle parameters and bridge responses from monitoring data. In addition, considerable efforts have also been dedicated to the development of moving force identification (MFI) methods, which seek to find the time histories of vehicle axle forces exerted on the bridge, to improve the identification accuracy and robustness (Z. Chen, Chan et al., 2019; Z. Chen, Qin et al., 2019; Z. Chen & Chan, 2017; Feng et al., 2015; Liu et al., 2020; C.‐D.…”

Section: Introductionmentioning

confidence: 99%

Probabilistic vehicle weight estimation using physics‐constrained generative adversarial network

Cai

Liu

2021

Computer aided Civil Eng

View full text Add to dashboard Cite

Traffic information plays an important role in the design and management of civil transportation infrastructure. Bridge weigh-in-motion (BWIM) provides an effective tool for traffic information gathering by estimating vehicle parameters including its weight through bridge responses. Most existing BWIM algorithms rarely consider the epistemic uncertainty of vehicle weight in terms of the probabilistic distribution of estimated axle weights (AWs) of the vehicle. This paper proposes a novel methodology for probabilistic vehicle weight estimation using a physics-constrained generative adversarial network (GAN). Generative models are introduced to describe the probabilistic distributions of estimated AWs and bridge responses. Physics constraints on the generative models are formulated and enforced by minimizing a physics-based loss function. The generative models are then learned by training a physics-constrained GAN using the observed bridge responses. Numerical study and field testing are conducted to demonstrate the proposed method using representative highway bridges and vehicles.The results show that the proposed method can successfully capture the uncertainty in the vehicle weight estimation and provide the probabilistic distributions of the estimated AWs for different vehicle types and loading conditions considered, which can enhance the application of BWIM for relevant tasks such as traffic data collection and truck overloading enforcement. Based on the results obtained from the numerical study and field testing, the maximum coefficient of variation obtained for the AWs and gross vehicle weight of the presented cases are 0.55 and 0.11, respectively.

show abstract

“…Once one sensor of the whole strain system fails, the weights of moving vehicles may hardly be identified. The issues existing in connection and strict synchronization among multiple sensors still imply a source of instability for BWIM systems [ 17 ].…”

Section: Introductionmentioning

confidence: 99%

“…Yu et al [ 21 ] used the wavelet transform for the global bridge bending strain signal to detect the vehicle axles, but the accuracy of the transformation is affected by noises and dynamic effects. Takaya et al [ 22 ] developed a mono-sensor method which utilizes a neural network to identify the transverse position, speed, and axle count of vehicles based on the bridge strain response. Although this method cannot identify the traffic loads, it makes full use of the convenience in optically measuring the speed or axle information of vehicles to create the associated dataset, providing inspiration for a stable, compact BWIM system with mono-sensor or single measuring section.…”

Section: Introductionmentioning

confidence: 99%

BwimNet: A Novel Method for Identifying Moving Vehicles Utilizing a Modified Encoder-Decoder Architecture

Deng

2020

Sensors

View full text Add to dashboard Cite

Traffic loading monitoring plays an important role in bridge structural health monitoring, which is helpful in overloading detection, transportation management, and safety evaluation of transportation infrastructures. Bridge weigh-in-motion (BWIM) is a method that treats traffic loading monitoring as an inverse problem, which identifies the traffic loads of the target bridge by analyzing its dynamic strain responses. To achieve accurate prediction of vehicle loads, the configuration of axles and vehicle velocity must be obtained in advance, which is conventionally acquired via additional axle-detecting sensors. However, problems arise from additional sensors such as fragile stability or expensive maintenance costs, which might plague the implementation of BWIM systems in practice. Although data-driven methods such as neural networks can estimate traffic loadings using only strain sensors, the weight data of vehicles crossing the bridge is difficult to obtain. In order to overcome these limitations, a modified encoder-decoder architecture grafted with signal-reconstruction layer is proposed in this paper to identify the properties of moving vehicles (i.e., velocity, wheelbase, and axle weight) using merely the bridge dynamic response. Encoder-decoder is an unsupervised method extracting higher features from original data. The numerical bridge model based on vehicle-bridge coupling vibration theory is established to illustrate the applicability of this new encoder-decoder method. The identification results demonstrate that the proposed approach can predict traffic loadings without using additional sensors and without requiring vehicle weight labels. Parametric studies also show that this new approach achieves better stability and reliability in identifying the properties of moving vehicles, even under the circumstances of large data pollution.

show abstract

Deep Sensing Approach to Single-Sensor Vehicle Weighing System on Bridges

Cited by 18 publications

References 21 publications

Vehicle-Assisted Techniques for Health Monitoring of Bridges

Vehicle-Assisted Techniques for Health Monitoring of Bridges

Probabilistic vehicle weight estimation using physics‐constrained generative adversarial network

BwimNet: A Novel Method for Identifying Moving Vehicles Utilizing a Modified Encoder-Decoder Architecture

Contact Info

Product

Resources

About