The Performance Analysis of the Map-Aided Fuzzy Decision Tree Based on the Pedestrian Dead Reckoning Algorithm in an...

Area of Science:

• Computer Science
• Robotics
• Geomatics Engineering

Background:

• Smartphones offer potential for mobile navigation due to widespread adoption and built-in sensors.
• Indoor pedestrian navigation faces challenges like sensor inaccuracy, unpredictable motion, and Global Navigation Satellite System (GNSS) unavailability.
• Existing Pedestrian Dead Reckoning (PDR) methods suffer from accumulated errors, limiting their real-world application.

Purpose of the Study:

• To improve the accuracy and stability of indoor pedestrian navigation systems.
• To reduce reliance on external infrastructure for positioning.
• To develop a low-complexity algorithm for real-time navigation solutions.

Main Methods:

• A fuzzy decision tree (FDT) algorithm was developed, integrating map information for enhanced navigation.
• An Indoor Mobile Mapping System (IMMS) was used for rapid map surveying.
• Bluetooth beacons were implemented for initial position determination.

Main Results:

• The proposed map-aided FDT algorithm demonstrated significant improvements in accuracy and stability compared to traditional PDR systems.
• The FDT approach showed good performance across various smartphones and users without pre-calibration or post-processing.
• The system achieved real-time navigation solution estimation.

Conclusions:

• The developed fuzzy decision tree (FDT) approach offers a robust and accurate solution for indoor pedestrian navigation.
• Map-aided FDT reduces dependency on infrastructure and overcomes limitations of standard Pedestrian Dead Reckoning (PDR).
• This method presents a stable and accurate navigation strategy suitable for diverse indoor environments and user conditions.