Automobile accident prediction and avoidance system using multilayer perceptron neural networks
Other Title(s)
النظام الذكي للتنبؤ بالحوادث و تجنبها باستخدام الشبكات العصبية
Dissertant
Abd al-Aziz, Ibrahim Umar Abd al-Aziz
Thesis advisor
Comitee Members
al-Hamami, Ala Husayn
Naum, Riyad S.
University
Middle East University
Faculty
Faculty of Information Technology
Department
Computer Science Department
University Country
Jordan
Degree
Master
Degree Date
2013
English Abstract
Road accidents are the most known cause of death.
Many organizations and transportation manufacturers are considering transportation safety improvement as one of their top priorities.
This study contributes to the area of transportation safety by identifying the roads and intersection dangerous sections that plays a role in different types of road accidents and use these information to warn the clients in the real time about the possible danger.
The drivers mistakes that can lead to accidents are also identified based on the previous driver mistakes to help another clients and road users to avoid them in the future.
The first phase of this study was to classify the accidents types into collision, pedestrian and turn over accidents, and accidents loses into slight, serious injuries and fatalities.
Then build a model for a system to predict these accidents, its types and loses using the clients smart devices capabilities and communicating with external servers that manages the operation and the relations between clients and road users.
The system utilizing a strong prediction method called ―Artificial Neural Networks‖ with ―Back-propagation‖ learning algorithm that is considered one of the supervised learning algorithms to analyze the previous accident conditions and reasons to predict them in the future.
A database for the previous accidents conditions and reasons was built for each type of accident that contains various information about different accidents and results to learn the neural network that can predict the relation between them and the accident result to help the road user avoiding the predicted danger at real time.
The system also enables its administrator to define a danger places manually to warn system users about these dangers before they reach the danger places so they can avoid these dangers.
Main Subjects
Information Technology and Computer Science
Topics
- Islamic jurisprudence
- Artificial intelligence
- Computing
- Simulation methods
- Traffic accidents
- Automobiles
- Data processing
- Logic machines
- Neural networks(Computer science)
- Bionics
No. of Pages
143
Table of Contents
Table of contents.
Abstract.
Abstract in Arabic.
Chapter One : Introduction.
Chapter Two : Literature survey.
Chapter Three : Artificial neural network and geographical information system.
Chapter Four : Proposed model and methodology.
Chapter Five : Conclusion and future work.
References.
American Psychological Association (APA)
Abd al-Aziz, Ibrahim Umar Abd al-Aziz. (2013). Automobile accident prediction and avoidance system using multilayer perceptron neural networks. (Master's theses Theses and Dissertations Master). Middle East University, Jordan
https://search.emarefa.net/detail/BIM-698563
Modern Language Association (MLA)
Abd al-Aziz, Ibrahim Umar Abd al-Aziz. Automobile accident prediction and avoidance system using multilayer perceptron neural networks. (Master's theses Theses and Dissertations Master). Middle East University. (2013).
https://search.emarefa.net/detail/BIM-698563
American Medical Association (AMA)
Abd al-Aziz, Ibrahim Umar Abd al-Aziz. (2013). Automobile accident prediction and avoidance system using multilayer perceptron neural networks. (Master's theses Theses and Dissertations Master). Middle East University, Jordan
https://search.emarefa.net/detail/BIM-698563
Language
English
Data Type
Arab Theses
Record ID
BIM-698563
