Institutional Repository
Technical University of Crete
Technical University of Crete
EN
|
EL
| URI: | http://purl.tuc.gr/dl/dias/6D75665F-5671-4A9F-A199-3B341B815CBA | ||
| Year | 2021 | ||
| Type of Item | Diploma Work | ||
| License |
|
||
| Bibliographic Citation | Alexios Vlachos, "Use of adaptive algorithm for traffic control in large scale urban networks", Diploma Work, School of Electrical and Computer Engineering, Technical University of Crete, Chania, Greece, 2021 https://doi.org/10.26233/heallink.tuc.91178 | ||
| Appears in Collections |
In recent years, a new algorithm was developed and studied in order to achieve the automatic configuration and configuration of control systems. One of the applications investigated was traffic control systems in urban networks. The Adaptive Fine-Tuning (AFT) algorithm has been found to significantly improve the performance of fixed-time traffic control systems, but also helps to offset the effects of continuous changes affecting a traffic control system by external or internal factors (drivers’ habit changes, socio-economic change, seasonality). A large urban network also implies an increased number of parameters to be optimized, which has been observed to reduce or even nullify the effectiveness of AFT, as it would take much longer to achieve convergence of parameters to a local minimum. This diploma thesis attempts to investigate the effectiveness of AFT on large-scale networks and to find ways to separate the network into smaller, independent sub-networks. The aim is to find a way for the partial optimization of the subnets to simultaneously lead to the improvement of the overall network. The application of this technology requires the collection of data from the network to which it will be applied. The collection of data for the operation of AFT can be achieved either through the installation of sensors (e.g., monitors or cameras), or through signals received from geographical positioning systems (GPS) of mobile phones, which provide the necessary data of the connected vehicles. With the second option we achieve a significant reduction in costs and increased accuracy compared to conventional data collection methods.
Copyright © DIAS 2013
