Road crashes are one of the serious problem causing significant losses to society and may result in fatalities, disabilities, injuries and property damage. The World Health Organization estimates that road traffic crashes represent the third leading cause of unnatural death worldwide by the year 2020. To mitigate the risk of road crashes at urban roads, it is important to fully understand the factors affecting the road crash severity. A better understanding of the risk factors affecting crash severity can be used to reduce the level of crash severity, locate the hazardous road sites and to take suitable countermeasures. Significant efforts have been made to investigate road crash severity, but the relationship between urban road crash severity and influencing factors have not yet appropriately identified. This study deals with the models to illustrate the influence of Human Behavior Factors, Road Network Infrastructure Factors, Environmental Factors, Vehicle Factors and Traffic Factors on the road crash severity level. Real crash experience is the best source to identify the most critical factors that affects the crash severity. In this study crash, weather and traffic flow data for the seven years from year 2006 to 2012 for the city of Turin Italy, were used to develop the crash severity models based on Artificial Neural Networks (ANN). Machine learning techniques such as Artificial Neural Networks (ANN) in engineering sciences play a vital role in recent years. The ANN models are capable to predict and present the desired results including the non-linear behavior of variables through proper data sets. ANNs gather their knowledge by understanding the relationship and patterns in the dataset and train (learn) through experience, so the network does not require a priori relationships between variables and offers the opportunity to investigate the problems where phenomena are not well known. Continued...
Gli incidenti stradali sono uno dei problemi più seri che causano perdite significative alla società e possono causare decessi, invalidità, lesioni e danni alla proprietà. L'Organizzazione Mondiale della Sanità stima che gli incidenti stradali siano la terza principale causa di morte innaturale in tutto il mondo entro il 2020. Per mitigare il rischio di incidenti stradali sulle strade urbane, è importante comprendere appieno i fattori che influenzano la gravità dell'incidente automobilistico. Una migliore comprensione dei fattori di rischio che incidono sulla gravità dell'incidente può essere utilizzata per ridurre la gravità dell'incidente, identificare i siti stradali pericolosi e adottare le contromisure appropriate. Sono stati compiuti sforzi significativi per indagare sulla gravità degli incidenti stradali, ma la relazione tra la gravità della crisi della strada urbana e i fattori di influenza non è stata ancora identificata in modo appropriato. Questo studio si occupa dei modelli per illustrare l'influenza dei fattori di comportamento umano, i fattori dell'infrastruttura della rete stradale, i fattori ambientali, i fattori del veicolo e i fattori di traffico a livello di gravità degli incidenti stradali. continua ....
Artificial neural networks to investigate factors affecting crash severity over years
HAZOOR, ABRAR
2017/2018
Abstract
Road crashes are one of the serious problem causing significant losses to society and may result in fatalities, disabilities, injuries and property damage. The World Health Organization estimates that road traffic crashes represent the third leading cause of unnatural death worldwide by the year 2020. To mitigate the risk of road crashes at urban roads, it is important to fully understand the factors affecting the road crash severity. A better understanding of the risk factors affecting crash severity can be used to reduce the level of crash severity, locate the hazardous road sites and to take suitable countermeasures. Significant efforts have been made to investigate road crash severity, but the relationship between urban road crash severity and influencing factors have not yet appropriately identified. This study deals with the models to illustrate the influence of Human Behavior Factors, Road Network Infrastructure Factors, Environmental Factors, Vehicle Factors and Traffic Factors on the road crash severity level. Real crash experience is the best source to identify the most critical factors that affects the crash severity. In this study crash, weather and traffic flow data for the seven years from year 2006 to 2012 for the city of Turin Italy, were used to develop the crash severity models based on Artificial Neural Networks (ANN). Machine learning techniques such as Artificial Neural Networks (ANN) in engineering sciences play a vital role in recent years. The ANN models are capable to predict and present the desired results including the non-linear behavior of variables through proper data sets. ANNs gather their knowledge by understanding the relationship and patterns in the dataset and train (learn) through experience, so the network does not require a priori relationships between variables and offers the opportunity to investigate the problems where phenomena are not well known. Continued...| File | Dimensione | Formato | |
|---|---|---|---|
|
2018_07_Hazoor_Abrar.pdf
accessibile in internet per tutti
Descrizione: Thesis
Dimensione
2.39 MB
Formato
Adobe PDF
|
2.39 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/142142