Project Type MQP Submission date 2010-04-27 Authors Ramsey D Abouzahra, ECE Cengiz Karakoyunlu, ECE Yuan Shi, ECE Billy Zhong, ECE URN E-project-042710-144538 Title Hybrid Indoor Geolocation for Robotic Applications Advisor Pahlavan, Kaveh, EE Availability unrestricted
The purpose of our major qualifying project is to develop a localization method for precise indoor geolocation. There are two phases in our project. The first involves devising an algorithm to combine the readings from various sensors. The second part is developing the user interface. By combining the readings from these sensors using the Extended Kalman Filter, 90th percentile of error is below approximately 2.7 meters. In conclusion, the location estimation of the inertial sensor is accurate in the short term, but is subject to drift. Wi-Fi localization is accurate in the long term, but is more subject to instantaneous variation. However, by sensor fusion of the inertial system and Wi-Fi localization, the advantages of each can be exploited to achieve optimal results.
Browse by Author | Browse by Department | Search all available E-projects
Questions? Email email@example.com