Integration of GPS precise point positioning and MEMS-based INS using unscented particle filter

Area of Science:

• Geomatics Engineering
• Navigation Systems
• Signal Processing

Background:

• Global Positioning System (GPS) and Inertial Navigation System (INS) integration faces challenges due to nonlinear models and potential divergence with Extended Kalman Filter (EKF), especially during GPS outages with Micro-Electro-Mechanical Systems (MEMS) inertial sensors.
• Traditional Particle Filtering (PF) offers a nonlinear estimation solution but suffers from high computational burden, limiting its practical application.

Purpose of the Study:

• To introduce and evaluate the Unscented Particle Filter (UPF) as an improved estimation technique for integrating GPS Precise Point Positioning (PPP) with MEMS-based INS.
• To compare the performance of UPF against EKF, Unscented Kalman Filter (UKF), and PF in a tightly coupled GPS-INS integration scenario.

Main Methods:

• Utilized a tightly coupled mechanization in the raw GPS and INS measurement domain.
• Employed un-differenced ionosphere-free linear combinations of pseudorange and carrier-phase measurements for PPP.
• Implemented and compared EKF, UKF, PF, and the proposed UPF using a real-world test scenario.

Main Results:

• The UPF requires fewer samples for accurate solutions compared to traditional PF, leading to reduced processing time.
• UPF demonstrated up to a 15% enhancement in positioning accuracy during GPS outages compared to EKF.
• All filters (EKF, UKF, PF, UPF) yielded comparable results when GPS measurement updates were consistently available.

Conclusions:

• The UPF is a computationally efficient and accurate method for GPS-INS integration, particularly beneficial during signal outages.
• UPF offers a significant improvement in navigation accuracy and processing efficiency over traditional filters for MEMS-based INS integration.