Yaw/Heading optimization by drift elimination on MEMS gyroscope

The main goal of the paper is to achieve a highly accurate measurement of yaw/heading without the support of the Global Positioning System (GPS) and magnetometer by using a practical model based on the principle "No Motion No Integration" (NMNI). The proposed technique removes the drift significantly to optimize the Micro-Electro-Mechanical System (MEMS) gyroscope for the yaw/heading estimation. A "Renovating Model" is added to the NMNI algorithm as a real-time detector for sensor motion state. The 'NMNI' can work effectively with an independent gyroscope or collaborate with other MEMS sensors via fusion algorithms such as Madgwick, Mahony, and Kalman to overcome the limitations of the Global Positioning System (GPS) in the indoor environment. Moreover, the two other critical factors: slope and rotation speed, were examined on sensor behavior to thoroughly verify each filter's pros and cons. The experiments were carried out using a low-cost platform equipped with MEMS as gyroscope, accelerometer, and magnetometer. A Pan Tilt Unit-C46 (PTU-C46) with high accurate positioning was used as a reference angle for both static and dynamic experiments. The results show the considerable advancement of yaw estimation by implementing the NMNI model into the gyroscope thanks to the effective drift removal. Moreover, the fusions between NMNI filter with Mahony and Madgwick accomplish high yaw measurement performance when the sensor on the high slope without magnetometer.(c) 2021 Elsevier B.V. All rights reserved.