The effect of cameras recording unauthorized speed violations in reducing accidents and the impact of drivers on urban highways using time series and regression

The effect of cameras recording unauthorized speed violations in reducing accidents and the impact of drivers on urban highways using time series and regression

Mohammad Aghaei Asl1 Ebrahim Salami2 Mostafa akbari3

1) Assistant Professor, Department of Engineering, Payame Noor University, Iran
2) Assistant Professor, Department of Engineering, Payame Noor University, Iran
3) MSc Student in Road Engineering - Transportation Payame Noor University Assalavi International Center

IOI Code : AADK-AAEECH
Publication : 3rd International Congress on Science, engineering & New Technologies(secongress.com/3rd)
Abstract :
    In order to predict urban traffic, the use of the method of time series and its use for intercity and suburban cameras has become common in recent years. In this study, the effect of cameras recording unauthorized speed violations in reducing accidents and the impact of drivers on urban highways was studied. The data used were related to violations recorded by Hemmat Highway Speed Control Cameras in Tehran during the period 1400-1401. The results of the descriptive analysis using SPSS software showed that in the 7-month selected timeframe in 1400 and 1401, there was a decrease in highway violations. The average number of active cameras increased in 1401 compared to the previous year. These findings show a positive effect of increasing the number of cameras on reducing speed violations on this highway. Also, comparing the equations of the violations process and the average number of cameras showed that the slope of the violations process was negative, which indicates the reduction of violations in the time frame under investigation. In the modeling section, first using the Box-Jenkins time series method, the ARIMA model for variable violations was obtained. Then, by analyzing the time series components, seasonal patterns and the decreasing trend of violations were identified. Also, regression modeling between research variables was carried out and nonlinear power model was selected as the best model. The results of the evaluation of the models showed that the time series model had less prediction error than the regression model. Therefore, the time series model performed better in predicting violations.

Keywords : camera infractions registration unauthorized speed crash reduction time series regression