| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 5.25 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
Os Unmanned Surface Vehicles (USV) são veículos autónomos que operam na superfície da água sem guarnição. O avanço futuro dos USVs depende do desenvolvimento de uma autonomia total sendo que, para este fim é indispensável a criação de algoritmos, capazes de atender aos critérios essenciais para uma utilização capaz de evitar obstáculos, eficiente, eficaz e segura no ambiente marítimo.
A presente dissertação propõe um algoritmo baseado no método que recorre ao uso do Velocity Obstacles (VO), isto é, cria cones, no espaço de velocidade, que representam as velocidades proibidas para o USV. A região de velocidades proibidas foi criada com recurso a uma figura com seis pontos sendo que, esta pode variar através da alteração de parâmetros como a distância de segurança e t_max, para criar trajetórias mais suaves e seguras.
O algoritmo desenvolvido cumpre com as regras do RIEAM através da introdução de restrições no espaço de velocidades. Este considera ainda as limitações dinâmicas do navio, como por exemplo, a velocidade máxima, a taxa de guinada e a aceleração máxima. Em situações de emergência, em que outros navios não cumprem com o Regulamento Internacional para Evitar Abalroamentos no Mar
(RIEAM) o algoritmo é capaz de encontrar soluções para evitar os obstáculos. Os resultados destacaram a capacidade do algoritmo em tomar decisões instantâneas e ajustar as trajetórias, enquanto mantém uma distância de segurança, mesmo em situações de emergência. No entanto, apesar da satisfação geral com os resultados, o algoritmo não é perfeito, sendo que por vezes apresenta manobras menos eficientes do que aquelas que seriam de esperar de uma intervenção humana.
Unmanned Surface Vehicles (USVs) are autonomous waterborne vehicles that operate without human crew on the water's surface. The future progression of USVs depends on the development of total autonomy. To achieve this objective, it is imperative to create algorithms capable of avoiding obstacles and satisfying the critical criteria for their efficient, effective, and safe deployment within the maritime domain. This dissertation presents an algorithm based on the Velocity Obstacles (VO) method. This technique generate cone-shaped obstacles in the velocity space, that represents the USV's prohibited velocities. The region of forbidden velocities is established using a figure featuring six points. This configuration is subject to variation through the adjustment of parameters, such as safety distance and t_max, with the aim of creating smoother and more secure trajectories. The algorithm in question complies with the COLREGS rules by introducing restrictions in the velocity space. Moreover, it takes into account the dynamic constraints of the vessel, including maximum velocity, yaw rate, and maximum acceleration. In case of emergencies, wherein other vessels fail to adhere to RIEAM regulations, the algorithm exhibits the capacity to maneuver tactically in order to avoid obstacles. The results highlighted the algorithm's remarkable ability to make instantaneous decisions and adapt trajectories while maintaining a safe distance, even in emergency situations. However, despite the general satisfaction with the results, it is important to note that the algorithm is not perfect and sometimes maneuvres less efficiently than would be expected from human intervention.
Unmanned Surface Vehicles (USVs) are autonomous waterborne vehicles that operate without human crew on the water's surface. The future progression of USVs depends on the development of total autonomy. To achieve this objective, it is imperative to create algorithms capable of avoiding obstacles and satisfying the critical criteria for their efficient, effective, and safe deployment within the maritime domain. This dissertation presents an algorithm based on the Velocity Obstacles (VO) method. This technique generate cone-shaped obstacles in the velocity space, that represents the USV's prohibited velocities. The region of forbidden velocities is established using a figure featuring six points. This configuration is subject to variation through the adjustment of parameters, such as safety distance and t_max, with the aim of creating smoother and more secure trajectories. The algorithm in question complies with the COLREGS rules by introducing restrictions in the velocity space. Moreover, it takes into account the dynamic constraints of the vessel, including maximum velocity, yaw rate, and maximum acceleration. In case of emergencies, wherein other vessels fail to adhere to RIEAM regulations, the algorithm exhibits the capacity to maneuver tactically in order to avoid obstacles. The results highlighted the algorithm's remarkable ability to make instantaneous decisions and adapt trajectories while maintaining a safe distance, even in emergency situations. However, despite the general satisfaction with the results, it is important to note that the algorithm is not perfect and sometimes maneuvres less efficiently than would be expected from human intervention.
Description
Keywords
USV Regras RIEAM Velocity obstacles Evitar Obstáculos