| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 20.08 MB | Adobe PDF |
Authors
Abstract(s)
Os Sistemas Globais de Navegação por Satélite (Global Navigation Satellite System, GNSS) providenciaram um aumento significativo do rigor e disponibilidade de dados de Posicionamento, Navegação e Tempo (PNT), no entanto, esta vantagem expôs também vulnerabilidades nas infraestruturas críticas, pela dependência destes sistemas. A emergência dos veículos autónomos potencia ainda mais a criticidade desta dependência, em particular em cenários militares onde é fundamental dispor de dados de navegação e posição robustos e resilientes.
Este estudo apresenta uma via para aumentar a resiliência dos sistemas PNT, através da implementação de fusão sensorial utilizando um Filtro de Kalman Estendido (Extended Kalman Filter, EKF). Explora-se, também, a importância dos dados PNT em cenários de Navigation Warfare (NAVWAR). É apresentada uma visão abrangente dos sensores relevantes. Os princípios dos Filtros de Kalman são abordados com detalhe.
O trabalho desenvolvido na dissertação envolve a integração de um conjunto de sensores num veículo terrestre não tripulado (Unmanned Ground Vehicle, UGV) incorporando um recetor GNSS e um Sistema de Posicionamento Indoor (IPS) ba seado em beacons para complementar os sensores já existentes, Unidade de Medida Inercial (Inertail Measurement Unit, IMU) e encoders para as rodas. São desen volvidos modelos cinemáticos e dinâmicos, juntamente com modelos dos sensores para IMU, odometria, GNSS e IPS. O estudo centra-se na implementação de EKFs para fusão sensorial. Adquiriram-se múltiplos conjuntos de dados em ambiente interior e exterior, resultantes de diversos ensaios reais com o UGV. Subsequentemente, realizou-se uma análise qualitativa e crítica dos resultados obtidos.
Este trabalho, alinhado com um grupo de trabalho da NATO, contribui para o avanço de soluções PNT robustas e resilientes para sistemas autónomos que operam em ambientes de disrupção de sinal GNSS. As conclusões e os resultados do desempenho da fusão sensorial são apresentados nos capítulos finais da dissertação.
Global Navigation Satellite Systems (GNSS) have significantly improved the accuracy and availability of Positioning, Navigation, and Timing (PNT) data. However, this advantage has also exposed vulnerabilities in critical infrastructures due to their dependence on these systems. The emergence of autonomous vehicles further intensifies the criticality of this dependence, particularly in military scenarios where robust and resilient navigation and positioning data are essential. This study presents an approach to enhance the resilience of PNT sys tems through the implementation of sensor fusion using an Extended Kalman Filter (EKF). The importance of PNT data in Navigation Warfare (NAVWAR) scenarios is also explored. A comprehensive overview of relevant sensors is presented. The principles of Kalman Filters are addressed in detail. This research focuses on the integration of a sensor suite on an Unmanned Ground Vehicle (UGV), incorporating a GNSS receiver and an Indoor Positioning System (IPS) to complement existing Inertial Measurement Unit (IMU) and encoder systems for the wheels. Kinematic and dynamic models are developed, along with sensor models for IMU, odometry, GNSS, and IPS. The core of the study focuses on the implementation of EKF algorithms for sensor fusion. Multiple datasets were acquired from various trials conducted with the UGV. Subsequently, a qualitative and critical analysis of the obtained results was conducted. This work, aligned with a NATO working group, contributes to the advan cement of robust and resilient PNT solutions for autonomous systems operating in GNSS-denied environments. Conclusions and results of the sensor fusion perfor mance are presented in the final chapters of the thesis.
Global Navigation Satellite Systems (GNSS) have significantly improved the accuracy and availability of Positioning, Navigation, and Timing (PNT) data. However, this advantage has also exposed vulnerabilities in critical infrastructures due to their dependence on these systems. The emergence of autonomous vehicles further intensifies the criticality of this dependence, particularly in military scenarios where robust and resilient navigation and positioning data are essential. This study presents an approach to enhance the resilience of PNT sys tems through the implementation of sensor fusion using an Extended Kalman Filter (EKF). The importance of PNT data in Navigation Warfare (NAVWAR) scenarios is also explored. A comprehensive overview of relevant sensors is presented. The principles of Kalman Filters are addressed in detail. This research focuses on the integration of a sensor suite on an Unmanned Ground Vehicle (UGV), incorporating a GNSS receiver and an Indoor Positioning System (IPS) to complement existing Inertial Measurement Unit (IMU) and encoder systems for the wheels. Kinematic and dynamic models are developed, along with sensor models for IMU, odometry, GNSS, and IPS. The core of the study focuses on the implementation of EKF algorithms for sensor fusion. Multiple datasets were acquired from various trials conducted with the UGV. Subsequently, a qualitative and critical analysis of the obtained results was conducted. This work, aligned with a NATO working group, contributes to the advan cement of robust and resilient PNT solutions for autonomous systems operating in GNSS-denied environments. Conclusions and results of the sensor fusion perfor mance are presented in the final chapters of the thesis.
Description
Keywords
Filtro de Kalman Fusão Sensorial Navegação Inercial Odome tria Resiliência PNT Sistema Global de Navegação por Satélite