The paper deals with vehicular safety in urban road networks. The complexity of the vehicular flow process and accident dynamic is considerably high because of the large number of variables necessary to describe them. To obtain a good statistical significance only four branch intersections are considered for the present analysis. Accident data were collected over a period of about two years in a medium-large city in the Northern part of Italy (with about 100.000 inhabitants) for a total amount of about 400 accidents distributed over 25 intersections. Geometrical characteristics of the intersections were collected, too: in particular, the number of lanes for each branch, the area of the intersection, traffic lights the presence of stop or give way signals. Data were processed by means of using a feedforward neural network with backpropagation learning algorithm. Input variables are eleven (daytime/night-time, area of the intersection, type of accident, road bed condition, weather, type of vehicle, type of violation and flow); output is the accident index which is related to the total accident number of each intersection. Results show non-linear relationships of the accident index with flow, area of intersection, meteorological conditions and type of violations.

AN ACCIDENT ANALYSIS FOR URBAN VEHICULAR FLOW

MUSSONE, LORENZO;RINELLI, SAVINO
1996

Abstract

The paper deals with vehicular safety in urban road networks. The complexity of the vehicular flow process and accident dynamic is considerably high because of the large number of variables necessary to describe them. To obtain a good statistical significance only four branch intersections are considered for the present analysis. Accident data were collected over a period of about two years in a medium-large city in the Northern part of Italy (with about 100.000 inhabitants) for a total amount of about 400 accidents distributed over 25 intersections. Geometrical characteristics of the intersections were collected, too: in particular, the number of lanes for each branch, the area of the intersection, traffic lights the presence of stop or give way signals. Data were processed by means of using a feedforward neural network with backpropagation learning algorithm. Input variables are eleven (daytime/night-time, area of the intersection, type of accident, road bed condition, weather, type of vehicle, type of violation and flow); output is the accident index which is related to the total accident number of each intersection. Results show non-linear relationships of the accident index with flow, area of intersection, meteorological conditions and type of violations.
Urban Transport and the Environment for the 21st Century: 2nd
9781853124518
File in questo prodotto:
File Dimensione Formato  
URBAN96_Mussone_Rinelli.pdf

accesso aperto

: Post-Print (DRAFT o Author’s Accepted Manuscript-AAM)
Dimensione 71.7 kB
Formato Adobe PDF
71.7 kB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11311/538906
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? 2
social impact