Publicação

Low cost sonar based on echolocation: an embedded system approach

Ver documento

Detalhes bibliográficos
Resumo:One of the greatest challenges of mobile robot applications is obstacle detection and its localization. Most systems applied to address this problem can be branched into two segments: vision-based and non-vision based. In the former, camera detection and image processing are sensitive to variations in environmental conditions as well as their high cost. For this reason, this option is not interesting. In the second segment, on the other hand, ultrasonic sensors stand out for their attractive features at an affordable cost and, unlike other systems, are capable of operating in a variety of environments. The main problem using these devices is the difficulty of correctly handling the data acquired, limiting their efficiency. The present work focuses on the improvement and employment of a low cost location system based on ultrasonic sensors, combining trigonometric and signal processing techniques. The work is divided into two parts: processed version in MATLAB and application of embedded system. For the version in MATLAB, two optimal techniques — envelope detection and cross-correlation — were evaluated. For the embedded system approach, the technique with the best performance was implemented. In both parts, the validation of results obtained occurred through a set of measurements with various objects in assorted configurations.
Autores principais:Moreira, Thiago Fernandes Moya
Assunto:Ultrasonic sensor Obstacle detection Mobile robot Signal processing Embedded system Low cost
Ano:2021
País:Portugal
Tipo de documento:dissertação de mestrado
Tipo de acesso:acesso aberto
Instituição associada:Instituto Politécnico de Bragança
Idioma:inglês
Origem:Biblioteca Digital do IPB
Descrição
Resumo:One of the greatest challenges of mobile robot applications is obstacle detection and its localization. Most systems applied to address this problem can be branched into two segments: vision-based and non-vision based. In the former, camera detection and image processing are sensitive to variations in environmental conditions as well as their high cost. For this reason, this option is not interesting. In the second segment, on the other hand, ultrasonic sensors stand out for their attractive features at an affordable cost and, unlike other systems, are capable of operating in a variety of environments. The main problem using these devices is the difficulty of correctly handling the data acquired, limiting their efficiency. The present work focuses on the improvement and employment of a low cost location system based on ultrasonic sensors, combining trigonometric and signal processing techniques. The work is divided into two parts: processed version in MATLAB and application of embedded system. For the version in MATLAB, two optimal techniques — envelope detection and cross-correlation — were evaluated. For the embedded system approach, the technique with the best performance was implemented. In both parts, the validation of results obtained occurred through a set of measurements with various objects in assorted configurations.