A model to represent motorcycle behaviour at signalised intersections incorporating an amended first order macroscopic approach

Previous research suggests that the passenger car unit (pcu) value of motorcycles (hereafter referred to as motorbikes to avoid confusion with signal cycles) varies depending on the time during the signal cycle that the motorbike crosses the stopline. Based on data from Bangkok, (May A.D., Montgomery F.O., 1986. Working paper 222, Institute for Transport Studies, University of Leeds) reported that motorbikes crossing the stopline in the first 6 s of effective green time had a pcu value of 0 and those crossing the stopline later in the cycle had a pcu value that varied from 0.53 to 0.65, depending on the lateral positioning of the motorbike and its eventual turning movement. In order to take into account this dual motorbike pcu factor, it is essential to estimate accurately the number of QFLIERS2. A model has been derived that describes motorbike behaviour at signalised intersections. An amended first order macroscopic model was used to represent motorbike behaviour and multiple regression analysis explained inaccuracies resulting from this technique. The model was tested against video data of motorbikes collected at intersections in Indonesia, Malaysia and Thailand, where the proportion of motorbikes is high (up to 70%), and predicted the number of QFLIERS per cycle with a high degree of accuracy. The factors shown to be important in predicting the number of QFLIERS were both temporal (inputs for the amended first order macroscopic model) and spatial (average lane width, number of lanes and number of buses and trucks per lane per cycle). The importance of the storage space at the front of the queue was noted and included within the model.