Conflict to study safety at four leg-signalised intersections

Abstract

In the absence of reliable accident records, the traffic conflict technique is a suitable alternative method for use to conduct a study of traffic safety. The present research aimed to conduct such a study at four leg-signalised intersections using the traffic conflict technique. Surveys which included traffic conflict, volume data, spot speed and some geometric characteristics were carried out at four sites in Baghdad city. Rate quality control and cluster analysis were used to identify the hazardous approach and the abnormal conflict types, respectively. The developed models showed that percentages ranging from 70·1 to 88·4% of variation in hourly traffic conflict could be determined by linear relationships with hourly approach traffic volume. Although this finding agrees with the literature, an exponential relationship introduced high percentages of determination ranging from 75·9 to 89·7%. This might be due to an extra increase in conflicts at high traffic volume. The identification of hazardous approaches showed that the lower rank in geometric characteristics among all approaches was the most hazardous one. This agrees with the findings from the developed model. In order to enhance traffic safety, the study proposes suitable countermeasures such as creating a left-turn lane and providing a special phase for left-turning traffic.