Publicação
Prevision, control and optimization of a hexapod robot posture in inclined surfaces
| Resumo: | In a world marked by natural and man-made disasters, the imperative of deploying mobile autonomous robots to replace human involvement in hazardous environments is evident. With this in mind, this dissertation focuses on developing posture control techniques that allow the robot to safely navigate these environments. Among legged robots, hexapod robots distinguish themselves as exceptional performers. Their capabilities extend to climbing, functioning with damaged limbs, and exhibiting remarkable static balance and gait movement. To comprehend the significance of hexapod robots in contrast to other legged counterparts, an extensive analysis is conducted, studying the structural attributes of hexapods, such as body composition, leg and joint arrangements, actuator and sensor configurations, thereby exposing the advantages and disadvantages intrinsic to this type of robots. The groundwork for this research is firmly established as it delves into the realm of posture adjustment. In pursuit of enhanced adaptability for the hexapod robot across various terrains, five leg path algorithms were compared, namely: triangular function, parabola function, 3rd-degree spline function and 3rd and 4th-degree Bézier curves. Compared along four different environments the preferred choice for this purpose is the 3rd-degree Bézier curve algorithm. This exploration, focusing on posture adjustment, provides a foundation for the understanding of the ATHENA hexapod model, encompassing its kinematic principles and gait generation strategies. The application of Q-Learning aided with integration of proprioceptive and exteroceptive sensors and simulation frameworks form a robust foundation for the posture adjustment problem. Through simulations with diverse control parameters in different slope environments, optimal control parameters for each slope were identified. These findings were then applied to simulate the robot navigating terrain with various slopes. A simulation lacking height control parameters resulted in failure, while the controlled simulations successfully adapted to variable slopes. |
|---|---|
| Autores principais: | Silva, Diogo Alexandre Pires da |
| Assunto: | “ATHENA” Height control Hexapod robot Posture adjustment Q-learning Ajuste de postura Controlo de altura Robô hexápode |
| Ano: | 2023 |
| País: | Portugal |
| Tipo de documento: | dissertação de mestrado |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Universidade do Minho |
| Idioma: | inglês |
| Origem: | RepositóriUM - Universidade do Minho |
| Resumo: | In a world marked by natural and man-made disasters, the imperative of deploying mobile autonomous robots to replace human involvement in hazardous environments is evident. With this in mind, this dissertation focuses on developing posture control techniques that allow the robot to safely navigate these environments. Among legged robots, hexapod robots distinguish themselves as exceptional performers. Their capabilities extend to climbing, functioning with damaged limbs, and exhibiting remarkable static balance and gait movement. To comprehend the significance of hexapod robots in contrast to other legged counterparts, an extensive analysis is conducted, studying the structural attributes of hexapods, such as body composition, leg and joint arrangements, actuator and sensor configurations, thereby exposing the advantages and disadvantages intrinsic to this type of robots. The groundwork for this research is firmly established as it delves into the realm of posture adjustment. In pursuit of enhanced adaptability for the hexapod robot across various terrains, five leg path algorithms were compared, namely: triangular function, parabola function, 3rd-degree spline function and 3rd and 4th-degree Bézier curves. Compared along four different environments the preferred choice for this purpose is the 3rd-degree Bézier curve algorithm. This exploration, focusing on posture adjustment, provides a foundation for the understanding of the ATHENA hexapod model, encompassing its kinematic principles and gait generation strategies. The application of Q-Learning aided with integration of proprioceptive and exteroceptive sensors and simulation frameworks form a robust foundation for the posture adjustment problem. Through simulations with diverse control parameters in different slope environments, optimal control parameters for each slope were identified. These findings were then applied to simulate the robot navigating terrain with various slopes. A simulation lacking height control parameters resulted in failure, while the controlled simulations successfully adapted to variable slopes. |
|---|