Modeling the pavement serviceability ratio of flexible highway pavements by artificial neural networks

Terzi, Serdal

doi:10.1016/j.conbuildmat.2005.11.001

Cited by 82 publications

(35 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Many techniques using a geographic information system (GIS) have been developed to determine pavement quality [1,5,[8][9][10][11]. In particular, the real-time data of highway pavement conditions were recorded from highway service vehicles (IVECO Daily) [12].…”

Section: Introductionmentioning

confidence: 99%

Riding Quality Model for Asphalt Pavement Monitoring Using Phase Array Type L-band Synthetic Aperture Radar (PALSAR)

Suanpaga

Kamiya

2010

Remote Sensing

View full text Add to dashboard Cite

There are difficulties associated with near-real time or frequent pavement monitoring, because it is time consuming and costly. This study aimed to develop a binary logit model for the evaluation of highway riding quality, which could be used to monitor pavement conditions. The model was applied to investigate the influence of backscattering values of Phase Array type L-band Synthetic Aperture Radar (PALSAR). Training data obtained during 3-7 May 2007 was used in the development process, together with actual international roughness index (IRI) values collected along a highway in Ayutthaya province, Thailand. The analysis showed that an increase in the backscattering value in the HH or the VV polarization indicated the poor condition of the pavement surface and, of the two, the HH polarization is more suitable for developing riding quality evaluation. The model developed was applied to analyze highway number 3467, to demonstrate its capability. It was found that the assessment accuracy of the prediction of the highway level of service was 97.00%. This is a preliminary study of the proposed technique and more intensive investigation must be carried out using ALOS/PALSAR images in various seasons.

show abstract

Section: Introductionmentioning

confidence: 99%

Riding Quality Model for Asphalt Pavement Monitoring Using Phase Array Type L-band Synthetic Aperture Radar (PALSAR)

Suanpaga

Kamiya

2010

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Çalışmalarda, bağımsız değişken sayısının çok olması bazı değişkenlerin sayısal değerlerle ifade edilememesi gibi sebeplerle yapay zeka tekniklerinin sıklıkla tercih edildiği görülmektedir. Araştırmacılar tarafından, sayısal ve sözel verilerin bir arada değerlendirilmesinde oldukça kolaylıklar sağlayan bulanık mantık [10][11][12] ve YSA [13][14][15][16] yaklaşımları ile her iki yöntemin bir arada kullanıldığı ANFIS yaklaşımının [17] [3]. IRI ölçümleri ve değerlendirmeleri ASTM E 950 standardında tanımlanan profilometre cihazı ile çeyrek araç sisteminin (Quarter Car System -QCS) simüle edilmesi ile sağlanmaktadır [18].…”

Section: Ayrıcaunclassified

Bitümlü Sıcak Karışım Üstyapılarda Görülen Yüzey Bozulmaları ile Düzgünsüzlük Arasındaki İlişkilerin Modellenmesinde Bazı Yaklaşımlar

Kırbaş¹,

Karaşahin²,

Demir³

et al. 2018

SDÜ Fen Bil Enst Der

View full text Add to dashboard Cite

Anahtar KelimelerÜstyapı performansı, Yüzey bozulmaları, IRI Özet: Esnek üstyapılarda görülen yüzey bozulmaları ile düzgünsüzlük arasında ilişkilerin araştırıldığı çalışmalar incelendiğinde sınırlı sayıda yüzey bozulma türünün dikkate alındığı görülmektedir. Literatüre katkı sağlamak amacıyla bu çalışmada, 13 adet yüzey bozulma türü ve bozulma şiddetleri ile birlikte toplam 32 adet üstyapı bozulması ile IRI arasındaki ilişkilerin matematiksel modelleme analizi yapılmıştır. Modelleme çalışmalarında doğrusal regresyon, değişkenli uyarlamalı regresyon eğrileri (MARS) ve yapay sinir ağları (YSA) yaklaşımları kullanılmıştır. Oluşturulan modellerin tahmin yetenekleri ortalama mutlak hata (OMH), kök ortalama karesel hata (KOKH), ortalama mutlak göreceli hata (OMGH) ve regresyon katsayısı (R 2 ) istatistiksel karşılaştırma yöntemleri kullanılarak değerlendirilmiştir. Tahmin yeteneği en yüksek olan modelin YSA yaklaşımı kullanılarak oluşturulan model olduğu tespit edilmiştir. Ayrıca, YSA yaklaşımında girdi değişkenlerinin çıktı değişkeni üzerindeki etkileri bağlantı ağırlıklarına göre değerlendirilmiştir. Bu değerlendirmeye göre, üstyapı bozulmaları oluşma nedenlerine (mekanizmalarına) göre incelendiğinde, IRI üzerinde % 43.8 yük kaynaklı, % 39 diğer sebepler kaynaklı ve % 17.2 iklim kaynaklı bozulmaların etkili olduğu sonucuna ulaşılmıştır. Some Approaches to the Modeling of Relationships between Surface Distresses and Roughness in Hot-Mixed Asphalts Keywords Pavement performance, Surface distresses, IRIAbstract: When studies investigating the relationship between pavement roughness and surface distresses seen in flexible pavements are examined, a limited number of surface distress types are considered to be taken into account. In order to contribute to the literature, in this study, mathematical modeling analysis of relations between IRI and a total of 32 pavement distresses with 13 types of surface distress types and severities were carried out. Linear regression, multivariate adaptive regression splines (MARS) and artificial neural networks (ANN) approaches are used in modeling studies. Estimating capabilities of the analyzed mathematical models were evaluated using statistical comparison methods such as mean absolute error (MAE), root mean squared error (RMSE), mean absolute relative error (MARE) and regression coefficient (R 2 ). The model constructed using the ANN approach was ascertained to be the model with the highest prediction accuracy. In addition, in the ANN approach, the effects on the output variable of the input variables are evaluated according to the connection weights. According to this evaluation, when pavement distresses are examined according to the cause of distress, it is concluded that 43.8 % of IRI is caused by load, 39 % is caused by other causes and 17.2 % is caused by climate.

show abstract

“…These parts are The ANN modeling consists of two steps: the first step is to train the network; the second step is to test the network with data, which were not used in training step. Neural networks have been trained to perform complex functions in various fields of application including pattern recognition, identification, classification, speech, vision, and control systems [23]. Artificial neural networks model developed in this research has four neurons (variables) in the input layer and one neuron in the output layer as illustrated in Fig.…”

Section: Fuzzy Logicmentioning

confidence: 99%

Prediction of compressive strength of heavyweight concrete by ANN and FL models

Başyiğit

Akkurt

Kılınçarslan

et al. 2009

Neural Comput & Applic

View full text Add to dashboard Cite

The compressive strength of heavyweight concrete which is produced using baryte aggregates has been predicted by artificial neural network (ANN) and fuzzy logic (FL) models. For these models 45 experimental results were used and trained. Cement rate, water rate, periods (7-28-90 days) and baryte (BaSO 4 ) rate (%) were used as inputs and compressive strength (MPa) was used as output while developing both ANN and FL models. In the models, training and testing results have shown that ANN and FL systems have strong potential for predicting compressive strength of concretes containing baryte (BaSO 4 ).

show abstract

Modeling the pavement serviceability ratio of flexible highway pavements by artificial neural networks

Cited by 82 publications

References 9 publications

Riding Quality Model for Asphalt Pavement Monitoring Using Phase Array Type L-band Synthetic Aperture Radar (PALSAR)

Riding Quality Model for Asphalt Pavement Monitoring Using Phase Array Type L-band Synthetic Aperture Radar (PALSAR)

Bitümlü Sıcak Karışım Üstyapılarda Görülen Yüzey Bozulmaları ile Düzgünsüzlük Arasındaki İlişkilerin Modellenmesinde Bazı Yaklaşımlar

Prediction of compressive strength of heavyweight concrete by ANN and FL models

Contact Info

Product

Resources

About