Prediction of permanent deformation in asphalt pavements using a novel symbiotic organisms search–least squares support vector regression

The prediction of asphalt performance can be very important in terms of increasing service life and performance while saving energy and money. In this study, a new hybrid artificial intelligence (AI) system, SOS–LSSVR, has been proposed to predict the permanent deformation potential of asphalt pavem...

Full description

Saved in:
Bibliographic Details
Published in:Neural computing & applications 2019-10, Vol.31 (10), p.6261-6273
Main Authors: Cheng, Min-Yuan, Prayogo, Doddy, Wu, Yu-Wei
Format: Article
Language:English
Subjects:
Artificial Intelligence
Asphalt mixes
Asphalt pavements
Computational Biology/Bioinformatics
Computational Science and Engineering
Computer Science
Data Mining and Knowledge Discovery
Decision making
Deformation
Energy conservation
Hybrid systems
Image Processing and Computer Vision
Least squares
Original Article
Prediction models
Probability and Statistics in Computer Science
Service life
Support vector machines
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The prediction of asphalt performance can be very important in terms of increasing service life and performance while saving energy and money. In this study, a new hybrid artificial intelligence (AI) system, SOS–LSSVR, has been proposed to predict the permanent deformation potential of asphalt pavement mixtures. SOS–LSSVR utilizes the symbiotic organisms search (SOS) and the least squares support vector regression (LSSVR), which are seen as a complementary system. The prediction model can be established from all input and output data pairs for LSSVR, while SOS optimizes the system’s tuning parameters. To avoid sampling bias and to partition the dataset into testing and training, a cross-validation technique was chosen. The results can be compared to those of previous studies and other predictive methods. Through the use of four error indicators, SOS–LSSVR accuracy was verified in predicting the permanent deformation behavior of an asphalt mixture. The present study demonstrates that the proposed AI system is a valuable decision-making tool for road designers. Additionally, the success of SOS–LSSVR in building an accurate prediction model suggests that the proposed self-optimized prediction framework has found an underlying pattern in the current database and thus can potentially be implemented in various disciplines.
ISSN:0941-0643
1433-3058
DOI:10.1007/s00521-018-3426-0