Análise de algoritmos de alinhamento em AHRS

• Autor(a) principal: Paiva, Lucas Pimenta Silva
• Data de Publicação: 2020
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFLA
• Texto Completo: http://repositorio.ufla.br/jspui/handle/1/45454
• topic: Sistemas de navegação; Sistemas de Referência de Orientação e Rumo (AHRS); Alinhamento; Navegação autônoma; Navigation system; Alignment; Attitude and Heading Reference Systems (AHRS); Autonomous navigation; Veículos e Equipamentos de Controle

In most autonomous navigation systems, the initial orientation of the vehicle is not known. The alignment is a phase that precedes the navigation step, and is responsible for determining the orientation of the vehicle. However, low-grade inertial sensors are not recommended to perform the alignment process, since the readings of their angular rate sensors are not capable of providing accurate measurements of the Earth’s rotation rate. Therefore, some authors propose the use of magnetometers, and the observation of the Earth’s magnetic field density vector in the alignment. This project investigates the problem of stationary alignment for low-grade Attitude and Heading Reference Systems (AHRS). In this work, we present the error analysis for six alignment algorithms. These are: Three-Axis Attitude Determination (TRIAD), QUaternion ESTimator (QUEST), Euler angle-Based Algorithm(EBA), Factored Quaternion Algorithm (FQA), Algebraic Quaternion Algorithm (AQUA) e Super Fast Attitude Determination Algorithm for Consumer-Level Accelerometer and Magnetometer (SAAM). All of them propose the use of accelerometers and magnetometers to estimate the system orientation, but using different methodologies. As a contribution of this project, we highlight the study of algorithms capable of determining precise orientation values using low-grade sensors. Simulated and experimental results corroborate the proposed errors analysis, evidencing the main differences, advantages and disadvantages of the use of each of these algorithms. Thus, this work serves as a basis for future research, especially those aimed at autonomous navigation using low-grade sensors.