IRI Prediction using Machine Learning Models

Authors: ,

Abstract: Road infrastructure is the backbone of the economy of any country. The recent increase in the length of roads has never been matched in history. The increase in length comes with huge demand for the maintenance of pavements in an orderly fashion. The pavement management system is used for planning maintenance based on pavement performance evaluation. The international roughness index (IRI) is considered a standard parameter for the functional evaluation of flexible pavements. In the present study, IRI is predicted through machine learning models using the LTPP database. The main objective of the study is to find the optimal machine learning which can be used for IRI prediction. Three machine learning models, (i) linear regression, (ii) optimised trees, and (iii) optimised Gaussian process regression (GPR), has been used for predicting IRI. Different models have been compared based on various statistical parameters. The optimised GPR model performed best per the R-Squared value (0.89).

WSEAS Transactions on Computer Research, ISSN / E-ISSN: 1991-8755 / 2415-1521, Volume 11, 2023, Art. #10

PDF
• HTML
DOI
XML
Cite
×
Citation Tools
WSEAS AMA
Ankit Sharma, Praveen Aggarwal, "IRI Prediction using Machine Learning Models," WSEAS Transactions on Computer Research, vol. 11, pp. 111-116, 2023, DOI:10.37394/232018.2023.11.10
Ankit Sharma, Praveen Aggarwal. IRI Prediction using Machine Learning Models. WSEAS Transactions on Computer Research. 2023;11:111-116. 10.37394/232018.2023.11.10
Copy to Clipboard
Citation copied to Clipboard

Certification

---