A novel intelligent predictor for low-rate global positioning system (GPS) system

• Main Authors: Hasan, Ahmed Mudheher, Samsudin, Khairulmizam, Ramli, Abdul Rahman
• Format: Article
• Language: English
• Published: Academic Journals 2011
• Online Access: http://psasir.upm.edu.my/id/eprint/23487/1/23487.pdf; http://psasir.upm.edu.my/id/eprint/23487/; https://academicjournals.org/journal/SRE/article-abstract/46BBC2C21789

Global positioning system (GPS) is the most common instrument utilized for navigational purpose. Unfortunately these satellite signals may get lost due to signal blockage. On the other hand, inertial navigation systems (INSs) can address this problem and overcome the non-availability of GPS signals for a short period of time due to the inherent sensors errors. In such case, INSs can benefit from aiding such as GPS. The difference in sampling rate between the GPS and INS must be overcome to realize the integration of the two systems. In general, Kalman filter (KF) is used to predict GPS data in order to integrate signals from high data rate systems, like INSs, with GPS that have low data rate. However, KF is usually criticized for working under predefined linear dynamic error models. In this paper, adaptive neuro fuzzy inference system (ANFIS) trained using genetic algorithm (GA) was adopted to predict the mislaid reading data for GPS to be synchronized with those of INS data. Hence, the gap between the two systems reading data is solved to provide synchronization between the INS and GPS systems. So, it is possible to compare the reading data of both systems. Three strategies have been proposed and the results shows superior performance in predicting missed GPS data with lowest mean square error.