BUS TRAVEL TIME PREDICTION IN THE MIXED TRAFFIC BY USING STATISTICA NEURAL NETWORK

This paper presents the assessment of a number of factors affecting bus travel time and study of relationship
model between the factors and bus travel time. The model is aimed at bus travel time prediction purpose. In fact,
bus travel time is fluctuated based upon many factors of bus operation system. Fluctuation of bus travel time is
an example of complex and multidimensional phenomenon in mixed traffic. Statistica Neural Network (SNN)
tool is proposed to solve this complex phenomenon. Data collected include bus travel time, distance, average
speed, and number of bus stop. The existing bus system is an intercity regular bus which is operated in the mixed
traffic. Therefore, the bus travel time is complexity fluctuated due to the road and traffic conditions in which
buses are operated. The result show that bus travel time from current bus stop to target bus stop is good predicted
by using the variables such as distance, average speed, and number of bus stop. The bus travel time increase due
to the increase of distance and number of bus stop from current bus stop to target bus stop. Meanwhile, the
higher the average speed approaching current bus stop, the bus travel time between current bus stop and target
bus stop is lower.