SENSOR FUSION FOR CONTAINER POSITION MONITORING USING PARTICLE FILTER ALGORITHM
<p align="justify"> <br /> <br /> Position monitoring and navigation systems become the important part of Automated Guided Vehicle (AGV). AGV is used to transport container which moves automatically from loading/unloading area in port container terminal to the temporary s...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/30509 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | <p align="justify"> <br />
<br />
Position monitoring and navigation systems become the important part of Automated Guided Vehicle (AGV). AGV is used to transport container which moves automatically from loading/unloading area in port container terminal to the temporary storage area. Mobile robot navigation system can be built using relative position sensor or absolute position sensor. Relative position sensor is not suitable for long-term navigation because the sensor has drifting bias and accumulated error. Absolute position sensor has good accuracy for long-term navigation but low update rate. Sensor fusion is a technique to fuse both types of positioning sensor to provide high accuracy estimation and high update rates. This can be achieved by fusing inertial sensor (accelerometer, gyroscope, and magnetometer) with GPS navigation sensor. However, both of sensors type might give wrong measurement data due to environmental changes. Therefore a sensor which is resilience to environmental changes such as RFID is introduced. <br />
<br />
Particle filter is an algorithm used to implement sensor fusion. The particle filter algorithm provides as an estimator for correcting data from the inertial sensor using the GPS navigation sensor. The particle filter works by weighting the data generated by each position sensor. This weight of each particle is based on the probability of true value sensor readings. The particle with the biggest weight is chosen to determine the estimated value. From this research, it is known that particle filter is able to produce good estimation of object position. Fusion of inertial sensors, GPS navigation sensors and RFID is able to provide position data accurately with faster update rate than the result of individual sensors. The position monitoring system in this study was built wirelessly and real-time. It aims to be able to observe the position of the object any time. Wireless communication is used to allow the system to be monitored remotely. <br />
<p align="justify"> |
|---|