NR-UIO: NLOS-Robust UWB-Inertial Odometry Based on Interacting Multiple Model and NLOS Factor Estimation.

Area of Science:

• Robotics
• Sensor Fusion
• Indoor Navigation

Background:

• Global Navigation Satellite System (GNSS) is unavailable indoors.
• Existing vision- or LiDAR-based sensing struggles in feature-poor environments.
• Ultra-wideband (UWB) sensors offer an alternative for indoor robot localization.

Purpose of the Study:

• To develop a robust robot localization framework for indoor environments.
• To address limitations of UWB positioning, including orientation acquisition and multipath errors.
• To improve pose estimation performance using integrated UWB and Inertial Measurement Unit (IMU) sensors.

Main Methods:

• Proposed a framework integrating an Interacting Multiple Model (IMM) filter with UWB and IMU sensors.
• Developed a non-line-of-sight (NLOS) robust UWB ranging model to mitigate multipath errors.
• Validated the framework through experiments in real indoor environments.

Main Results:

• The proposed framework demonstrated improved localization performance compared to existing methods.
• The NLOS robust UWB ranging model effectively reduced errors in multipath environments.
• Successful integration of UWB and IMU sensors enabled accurate pose estimation.

Conclusions:

• The integrated UWB-IMU framework with a robust ranging model enhances indoor robot localization.
• This approach overcomes key limitations of UWB-only and other sensor-based methods.
• The study provides a viable solution for reliable robot navigation in GNSS-denied areas.