A Federated Derivative Cubature Kalman Filter for IMU-UWB Indoor Positioning.

Area of Science:

• Robotics
• Sensor Fusion
• Navigation Systems

Background:

• Indoor positioning systems often combine Inertial Measurement Units (IMU) and Ultra-Wideband (UWB) signals.
• These systems face challenges with nonlinear state equations and computational efficiency.
• Traditional Kalman filtering methods can lead to cumulative errors in IMU data.

Purpose of the Study:

• To propose a novel Federated Derivative Cubature Kalman Filtering (FDCKF) method.
• To enhance the computational efficiency and localization accuracy of IMU-UWB indoor positioning systems.
• To address the issue of cumulative errors in IMU data through continuous UWB calibration.

Main Methods:

• Development of the Federated Derivative Cubature Kalman Filtering (FDCKF) algorithm.
• Fusion of IMU and UWB observations using the FDCKF method.
• Continuous calibration of IMU data by UWB signals within the FDCKF framework.
• Numerical simulations comparing FDCKF against Federated Cubature Kalman Filter (FCKF) and Federated Unscented Kalman Filter (FUKF).

Main Results:

• The FDCKF method effectively fuses UWB and IMU observations.
• Continuous UWB calibration prevents cumulative errors in IMU data.
• FDCKF demonstrates superior localization performance and computational efficiency compared to FCKF and FUKF in simulations.

Conclusions:

• The proposed FDCKF algorithm significantly improves indoor positioning accuracy and efficiency.
• FDCKF offers a robust solution for IMU-UWB integrated systems by mitigating cumulative errors.
• This method provides a valuable advancement for precise indoor localization applications.