Application of IMU/GPS Integrated Navigation System Based on Adaptive Unscented Kalman Filter Algorithm in 3D...

Area of Science:

• Robotics and Autonomous Systems
• Geomatics Engineering
• Sensor Fusion

Background:

• Reliable positioning is critical for emergency rescue personnel and operations.
• Global Navigation Satellite Systems (GNSSs) like GPS suffer signal instability in challenging environments such as dense forests.
• Integrating multiple sensors, including Inertial Measurement Units (IMUs) and GPS, is essential for robust and accurate positioning.

Purpose of the Study:

• To develop and evaluate an advanced integrated navigation system for precise 3D positioning of rescue personnel in complex environments.
• To enhance the robustness and accuracy of positioning systems, especially during Global Navigation Satellite System (GNSS) signal interruptions.
• To compare the performance of the proposed Adaptive Unscented Kalman Filter (AUKF) against traditional filtering methods in realistic rescue scenarios.

Main Methods:

• Implementation of an Adaptive Unscented Kalman Filter (AUKF) algorithm with adaptive measurement noise variance matrix.
• Integration of data from Global Positioning System (GPS), Inertial Measurement Units (IMUs), and barometric altimeters for sensor fusion.
• Performance evaluation through field tests in 2D and 3D forest road scenarios, including rugged terrain and simulated GPS signal outages.

Main Results:

• The AUKF-based integrated navigation system demonstrated superior positioning accuracy compared to Extended Kalman Filter (EKF), Unscented Kalman Filter (UKF), and Adaptive Extended Kalman Filter (AEKF).
• Significant error reductions were observed: 18.32% in the north, 8.51% in the up, and 3.85% in the east directions compared to EKF on rugged forest roads.
• The system showed excellent adaptability and maintained continuous, accurate positioning during GPS signal interruptions.

Conclusions:

• The AUKF-based integrated navigation system offers enhanced robustness and accuracy for rescue personnel positioning in challenging forest environments.
• Adaptive filtering techniques are crucial for overcoming GNSS signal limitations and ensuring reliable navigation during emergency operations.
• The proposed system provides a viable solution for improving the safety and efficiency of search and rescue missions.