Estimating Passenger Car Equivalent Factors for Heterogeneous Traffic Using Occupancy-Density Linear Regression Model

A variety of methods have been proposed in the existing literature for the estimation of passenger car equivalent (PCE) factors. These methods are based on the comparison of selected attributes of different vehicles. This research, for the first time, utilizes the basic notion of the linear relationship between road area occupancy and density for the estimation of PCE factors for different vehicle types in heterogeneous traffic. Aerial photographs obtained from an unmanned aerial vehicle (UAV) were analyzed to estimate the road area occupancy and the number of vehicles classified in seven selected groups. A linear least-squares regression model was developed between road area occupancy and classified vehicle count. The coefficients of the occupancy-density linear regression model were used to estimate PCE and motorcycle equivalent (MCE) factors. The comparison of the estimated set of PCE values with the values reported in the literature shows that PCE factors estimated using the proposed method are reasonable and produce a better occupancy-density relationship than the other studies. In comparison with the existing methods that rely on lane-based measurements, the proposed method is well suited for traffic with weak/no lane discipline, as it considers the entire road width and the dynamics of lateral movement of different types of vehicles. The proposed method does not need extensive traffic data of speeds, headways, flow rates, and so forth, and is applicable on aerial photographs obtained from other sources, such as satellites.

Automated summary

This research proposes a method that estimates the passenger car equivalent (PCE) factors for different types of vehicles in heterogeneous traffic by utilizing the linear relationship between road area occupancy and density. Aerial photographs obtained from an unmanned aerial vehicle (UAV) were used to analyze road area occupancy and the number of vehicles in different categories. A linear regression model was developed to estimate the PCE factors based on the occupancy-density relationship. The results show that the proposed method produces reasonable PCE values and outperforms existing methods.