LOCATION DETERMINATION SYSTEM DESIGN FOR TRUCK TRAILER BASED ON SENSOR FUSION USING UNSCENTED KALMAN FILTER ALGORITHM

• Main Author: Simanjuntak, Josavin
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/43350
• Institution: Institut Teknologi Bandung

As the number and distribution of the population increased, the activity of fulfilling logistical needs has also increased. One of the modes of logistics transportation that plays an important role in the availability of logistics is sea transportation. Logistics transportation must be made as efficient as possible, especially at the port. One of the solutions is to develop autonomous truck trailer. To develop an autonomous vehicle, an accurate Location Determination System (LDS) is needed so that the autonomous vehicle can run well. In this study, truck trailer prototype is used as object of the LDS designed here. The sensors used are encoder, potentiometer, and Marvelmind beacon which function as Indoor Positioning System (IPS) and compass. Then, the speed, orientation and position data from the three sensors are processed by using the sensor fusion technique with Unscented Kalman Filter (UKF) algorithm to produce a better estimation of the position and orientation of truck trailer. The estimation results are then displayed in a GUI application in real time. Based on the results achieved in this study, the UKF algorithm can produce estimates of the position and orientation of container trucks with an average error of 2.9 cm for position, 0.01 radians for head orientation, and 0.12 radians for container orientation.