Robust UAV Position and Attitude Estimation using Multiple GNSS Receivers for Laser-based 3D Mapping

Introduction

The utilization of small-sized Unmanned Aerial Vehicles (UAVs) for applications like remote sensing and aerial surveying has garnered substantial interest due to their cost-effectiveness and flexibility in operations. One of the critical aspects of these applications, particularly in direct 3D mapping, is obtaining accurate position and attitude data for the UAV. These are foundational for translating sensor data to a common coordinate frame, enabling the construction of precise maps.

Methodology

A novel technique for estimating the positions and attitudes of small-sized UAVs has been introduced, focusing on the use of multiple Global Navigation Satellite System (GNSS) antennas and receivers. This approach uniquely leverages the redundancy of multiple GNSS inputs to improve the performance of real-time kinematic (RTK) positioning, traditionally reliant on single-frequency GNSS receivers. Two primary strategies are employed: hybrid GNSS fix solutions and a consistency examination of GNSS signal strength.

This novel system involves outfitting a UAV with six GNSS antennas/receivers, resulting in an extended array and creating multiple possibilities for baseline vector estimations. An estimation approach, known as the "Q-Method," is utilized for calculating the optimal attitude from these vectors without necessitating an Inertial Measurement Unit (IMU), which generally adds to the cost and complexity of the system.

Experimental Results

Through static and flight tests, the proposed technique demonstrated its capacity to significantly refine the accuracy and reliability of UAV position and attitude estimations in multipath environments—areas with signal reflection and diffraction challenges. In static tests, the fix rate of RTK-GNSS solutions was notably enhanced, and multipath signals were effectively detected and mitigated.

In flight tests, the UAV's position and attitude estimates were successfully used to produce a 3D map with an accuracy of 5 cm. This impressive result was achieved using only GNSS receivers and a laser scanner, without the assistance of IMUs or other advanced sensors, which are often expensive and exceed the payload capacity of small UAVs.

Conclusion

The research demonstrates how incorporating multiple, low-cost GNSS receivers can yield precise and reliable UAV positioning and attitude estimation, even in challenging environments. This advancement could be pivotal in scenarios where direct 3D mapping is necessary but without the need for extensive groundwork or access to high-precision instruments, such as during the immediate aftermath of disasters. The system holds the potential for extensive use in various industries, bridging the gap between affordability and mapping accuracy.