Classifying aggressive drivers for better traffic signal control
| Autor: | Shaimaa M. Hegazy, Mohamed N. Moustafa |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
050210 logistics & transportation
Artificial neural network business.industry Computer science 05 social sciences Crash 02 engineering and technology Machine learning computer.software_genre Traffic flow Fuzzy logic Support vector machine Traffic signal Aggressive driving 0502 economics and business 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Artificial intelligence business computer Classifier (UML) |
| Zdroj: | ITSC |
| DOI: | 10.1109/itsc.2017.8317930 |
| Popis: | Traffic signal control, in a given single road intersection, primarily aims at reducing the average delay among a group of vehicles and avoiding crash accidents. Commonly, a driver's behavior is an essential cause of these accidents. There exist many solutions for automatic traffic signal control focusing on reducing average delay. However, to our knowledge, we are the first to integrate aggressive driving behavior classification with the traffic signal controller to enhance the safety and to reduce delays. Firstly, we present an artificial neural network as an aggressive driver behavior classifier trained using the benchmark Virginia Tech 100-car naturalistic driving study data. Secondly, we propose a modification to the popular Mamdani's fuzzy logic signal controller to accommodate for the integration of our aggressive behavior classifier. Multiple experiment sets were conducted to provide an indication to the effectiveness of our approach. The integration results yielded significant improvements at higher traffic flow volumes when compared against the baseline controller. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|