University of Twente Student Theses
Predicting asphalt temperature for construction : a contribution to the analysis of the cooling rate of asphalt during construction process
Ayuquina Suarez, K.J. (2022) Predicting asphalt temperature for construction : a contribution to the analysis of the cooling rate of asphalt during construction process.
|
PDF
2MB |
| Abstract: | The Netherlands is a country that has one of the largest road networks, therefore the quality of the roads is essential. An important factor during road construction is the compaction of the Hot Mix Asphalt which needs to be done in a specific time and temperature window in order to obtain a better quality. However, the cooling rate of asphalt cannot be easily calculated since it is influenced by factors such as types of mixture, thickness of the layer, underlying layer, underlying temperature, initial temperature, wind speed, outside temperature, weather conditions, and rain. To solve this problem, a tool named ASPARiCool is being developed, which is a Dutch application that aims to predict the cooling rate using empirical measurements based on Machine Learning. However, the main drawback of a previous version is that it provides inaccurate cooling rate prediction due to the limited data obtained in field measurements. In this study, the ASPARiCool tool was studied to better understand its operation. This software works using the Multi-layer perceptron. This model is based on training the program with historical data and then making predictions. To determine if using more data improves accuracy, during the months of the study, data was obtained from projects carried out in the Twente region. The first results showed an increase in accuracy compared to when the amount of historical data was less. However, the RSME was still significant. To increase the accuracy of ASPARiCool, a sensitivity analysis of each parameter with which the MLP works was performed. In addition, a calibration of the program was then made to find the best performance. The predictions obtained using the new parameter values were better than the results using the default values by 65%. |
| Item Type: | Essay (Bachelor) |
| Faculty: | ET: Engineering Technology |
| Programme: | Civil Engineering BSc (56952) |
| Link to this item: | https://purl.utwente.nl/essays/93527 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page