STUDY BUS ARRIVAL TIME PREDICTION FOR DEVELOPMENT PASSENGER INFORMATION DISPLAY SYSTEM

• Main Author: Imam Boni Mores & Muhammad Fauzan, N
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/41457
• Institution: Institut Teknologi Bandung

Indonesia is a developing country with enormous potential to become a developed country and one of the characteristics of a developed country is proper public transportation. However, public transportation is currently still lacking in public interest, one example is the bus. There are several factors that make buses less desirable. The most common complaint is that passengers do not know the position of the nearest bus and how long the bus will arrive at the bus stop. In this study an interface system will be made for bus passengers at the bus stop that can determine the position of the nearest bus and how long the bus will arrive at the bus stop. This interface system is known as the Passenger Information Display System (PIDS). The bus will be equipped with a GPS sensor to estimate position and also be used as input for predicting arrival time. After that, modeling will be carried out to get a model of the prediction of bus arrival time by using several methods approaches, namely Radial Basis Function Neural Network and Adaptive Neuro Fuzzy Inference System with training using Hybrid Learning and Particle Swarm Optimization. To improve the predicted arrival time of the bus, the Kalman filter algorithm is added to estimate the arrival time at any time. Passanger Information Display System (PIDS) can provide information about the current position and accurate estimation of bus arrival time, so that it can provide benefits to prospective bus passengers at a bus stop to streamline the waiting time at the bus stop.