http://yetl.yabesh.ir/yetl1/handle/yetl/4237039 2026-07-16T23:29:45Z 2026-07-16T23:29:45Z Yuchen Xie Danfeng Du Ziqi Wang Yang Liu Mengju Bi http://yetl.yabesh.ir/yetl1/handle/yetl/4309944 2026-02-16T21:56:00Z 2025-01-01T00:00:00Z

YOLO-SFT: Road Damage Detection Algorithm Based on Feature Diffusion Yuchen Xie; Danfeng Du; Ziqi Wang; Yang Liu; Mengju Bi Pavement damage detection is an important research area in road maintenance and traffic safety, but traditional detection methods have shortcomings such as low accuracy and poor real-time performance. A pavement damage detection algorithm, You Only Look Once-spatial feature transformation (YOLO-SFT), based on feature diffusion is proposed in this paper to improve the detection accuracy and efficiency. First, a new module, StarNet-context anchor attention (Star-CAA), is designed to replace the C2f of the backbone part, which enhances the ability of feature extraction and gradient flow and optimizes the detection performance and generalization ability of the model. After that, a new pyramid network, focusing diffusion cross stage (FDCS), is independently developed to optimize the neck part. Through the unique feature-focusing diffusion mechanism, features with rich contextual information are diffused to various detection scales. Finally, the detection head part is redesigned to propose a new efficient detection head, task align dynamic (TAD), which obtains joint features by learning task interaction features from multiple convolutional layers. It strengthens the accuracy and real-time performance of pavement damage detection. Experimental results show that the F1 score of YOLO-SFT is improved by 4.0% and the mean average precision (mAP) is enhanced by 5.6%. In addition, the computational parameters of YOLO-SFT are reduced by 16.9%, the model size is reduced by 14.9%, and the running speed reaches 64.5 frames per second (FPS). The proposed algorithm has good application prospects and provides an effective solution for pavement distress detection in intelligent transportation systems.

2025-01-01T00:00:00Z Faizan Ahmad Lali Rahul Raj Singh Syed Waqar Haider http://yetl.yabesh.ir/yetl1/handle/yetl/4309942 2026-02-16T21:55:57Z 2025-01-01T00:00:00Z

Mechanistic-Empirical Design for Rubblized Pavements in Michigan Faizan Ahmad Lali; Rahul Raj Singh; Syed Waqar Haider Pavement mechanistic-empirical design (PMED) is a modern approach to designing new and rehabilitated pavements. The Michigan Department of Transportation (MDOT) follows the current guideline methodology for rehabilitation designs, utilizing hot-mix asphalt (HMA) overlays on rubblized plain cement concrete (PCC) pavements as new flexible pavement. PMED offers an alternative of HMA overlay on fractured jointed plain cement concrete (JPCP) for rubblized pavements. This paper investigates the optimal design approach and HMA input level for rubblized pavements in Michigan, comparing performance predictions using global and locally calibrated models across three input levels. Results indicated negligible differences between new and overlay designs at global performance predictions for input Levels 1 and 3. Local calibration at Level 1 produced better outcomes, but Level 3 results were also acceptable, barring the thermal cracking model. The study analyzed 11 pavement sections, revealing that HMA thicknesses were, on average, 1.02 cm thinner than current guideline designs. A new flexible pavement design with Level 1 data input is recommended for rubblized pavements in Michigan. This study aims to improve HMA overlay design on rubblized pavements, promoting efficient, cost-effective pavement rehabilitation. Although several studies are available on implementing the PMED for new pavements, research on rehabilitated pavements is limited. Moreover, different design options and input levels in the PMED and data unavailability make the design of rehabilitated pavements more challenging. By comparing PMED design options (overlay and new), calibration methods, and input levels (Levels 1, 2, and 3), the study seeks to enhance prediction accuracy for key pavement distresses. These findings will aid practitioners in selecting the optimal design and input level for effective rehabilitation, especially when data availability is challenging. The study’s findings validate the MDOT current practice of treating HMA overlays over rubblized concrete as new flexible pavements.

2025-01-01T00:00:00Z Wenbo Liu Xu Yang Hainian Wang Ling Ding Wei Xu http://yetl.yabesh.ir/yetl1/handle/yetl/4309941 2026-02-16T21:55:55Z 2025-01-01T00:00:00Z

Forward Simulation of GPR Detection for Irregular Concealed Distresses of Road Pavement Based on the FDTD Method Wenbo Liu; Xu Yang; Hainian Wang; Ling Ding; Wei Xu Detecting concealed pavement distresses by ground penetrating radar (GPR) has become popular. However, inner distresses develop irregularly, leading to the uncertainty of the feature in B-scan images. Meanwhile, the heterogeneous pavement increases the difficulty of feature determination. To determine the GPR characteristics and their variation rules of irregular features in heterogeneous pavement structures, random multiphase pavement and irregular distress models were established for forward simulation. The parameters such as antenna frequency, distress size, and filling media were set in different values to observe feature variation. Regional average gray-scale value was chosen to characterize the relation between the distress size and reflected wave energy. After forward modeling, the typical B-scan image features of inclined cracks, irregular cracks, irregular debonding areas, and loose areas were identified. The influence of antenna frequency, filling medium, and size of disease on the variation of disease characteristics was also determined. In this study, the reflected wave energy intensity was characterized by the regional average gray-scale value, and the results showed that the size of cracks and debonding areas were positively correlated with the reflected wave energy intensity. The R2 of the vertical air crack, vertical water-bearing crack, and irregular water-bearing crack is 0.977, 0.971, and 0.926, respectively. The R2 of the air debonding area is 0.930.

2025-01-01T00:00:00Z Peng Chen Karim Chatti Bora Cetin http://yetl.yabesh.ir/yetl1/handle/yetl/4309940 2026-02-16T21:55:54Z 2025-01-01T00:00:00Z

Accuracy of Pavement ME Methodology in Predicting Strain Responses under Multiple Axle Loadings in Flexible Pavements Peng Chen; Karim Chatti; Bora Cetin The mechanistic-empirical pavement design method (Pavement ME) uses linear-elastic analysis to calculate the strain responses induced by single and multiple axle loadings, and the method relies on the concept of equivalent or predominant loading frequency to determine the representative elastic modulus of the asphalt concrete (AC) layer. An investigation of the characteristics of multiple axle loading frequencies and the accuracy of predicting critical strains under both single and multiple axle loadings using the same frequency (e.g., Pavement ME) is presented. This paper compares the Pavement ME frequency with several other loading frequencies derived from single, tandem, and tridem axle loading scenarios in 12 cases with combinations of different AC layer thicknesses, moduli, and temperatures, as well as different vehicle speeds. The validity of these frequencies was determined by comparing the strain responses simulated by linear-elastic analysis using each of the frequencies with the results obtained from full dynamic viscoelastic analysis. The study showed that a lower loading frequency needs to be used for multiple axle configurations especially for AC layers with low to moderate stiffness if one wants to achieve a high accuracy of vertical stain prediction; the accuracy of maximum horizontal strains predicted by all loading frequencies is acceptable and is independent of axle configurations or frequency methods in general.

2025-01-01T00:00:00Z