Edinburgh Napier University

  The aim of this study is to estimate the Passenger Car Unit
(PCU) values of motorcycles in two traffic environments: at
the beginning of a green period and in a saturation flow.
Motorcycles are able to undertake both lane-based and
non-lane-based movements in mixed traffic. Hence their
PCU values cannot be adequately measured simply by
their headways. In addition, their PCUs could vary
according to the traffic context of urban networks, which
makes their PCU estimation even more complex. This
study employed the flow rate method to estimate
motorcycle PCU values, with the help of a recently
developed agent-based simulation model, which is capable
of representing the characteristic movement patterns of
motorcyclists. An experiment was designed to conduct a
systematic analysis, in which four conditioning variables
were considered: the advanced stop line, the number of
lanes, the width of lanes and the proportion of motorcycles.
The simulation results showed that at the beginning of
green periods, the key factor affecting the PCU values was
the number of motorcycles able to filter to the head of the
queue. However, in saturation flow, the key factor was the
opportunity for motorcycles to move alongside another
vehicle in the same lane. An ex-post analysis was
conducted using a regression model to fit the simulation
results. This model indicated that the PCU values of
motorcycles at the beginning of green periods are
averagely 0.237 lower than those in multiple-lane
saturation flows. In addition, the PCU values decrease
0.143 with every 1.0 m increase of the lane width. PCU
values for motorcycles in different traffic conditions are
suggested using this model.