Vision-based Global Path Planning and Trajectory Generation for Robotic Applications in Hazardous Environments
Ziaei, Zahra (2016)
Tampere University of Technology
2016
Rakennetun ympäristön tiedekunta - Faculty of Built Environment
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:ISBN:978-952-15-3730-1
https://urn.fi/URN:ISBN:978-952-15-3730-1
Tiivistelmä
The aim of this study is to find an efficient global path planning algorithm and trajectory generation method using a vision system. Path planning is part of the more generic navigation function of mobile robots that consists of establishing an obstacle-free path, starting from the initial pose to the target pose in the robot workspace.
In this thesis, special emphasis is placed on robotic applications in industrial and scientific infrastructure environments that are hazardous and inaccessible to humans, such as nuclear power plants and ITER1 and CERN2 LHC3 tunnel. Nuclear radiation can cause deadly damage to the human body, but we have to depend on nuclear energy to meet our great demands for energy resources. Therefore, the research and development of automatic transfer robots and manipulations under nuclear environment are regarded as a key technology by many countries in the world. Robotic applications in radiation environments minimize the danger of radiation exposure to humans. However, the robots themselves are also vulnerable to radiation. Mobility and maneuverability in such environments are essential to task success. Therefore, an efficient obstacle-free path and trajectory generation method are necessary for finding a safe path with maximum bounded velocities in radiation environments. High degree of freedom manipulators and maneuverable mobile robots with steerable wheels, such as non-holonomic omni-directional mobile robots make them suitable for inspection and maintenance tasks where the camera is the only source of visual feedback.
In this thesis, a novel vision-based path planning method is presented by utilizing the artificial potential field, the visual servoing concepts and the CAD-based recognition method to deal with the problem of path and trajectory planning. Unlike the majority of conventional trajectory planning methods that consider a robot as only one point, the entire shape of a mobile robot is considered by taking into account all of the robot’s desired points to avoid obstacles. The vision-based algorithm generates synchronized trajectories for all of the wheels on omni-directional mobile robot. It provides the robot’s kinematic variables to plan maximum allowable velocities so that at least one of the actuators is always working at maximum velocity. The advantage of generated synchronized trajectories is to avoid slippage and misalignment in translation and rotation movement. The proposed method is further developed to propose a new vision-based path coordination method for multiple mobile robots with independently steerable wheels to avoid mutual collisions as well as stationary obstacles. The results of this research have been published to propose a new solution for path and trajectory generation in hazardous and inaccessible to human environments where the one camera is the only source of visual feedback.
In this thesis, special emphasis is placed on robotic applications in industrial and scientific infrastructure environments that are hazardous and inaccessible to humans, such as nuclear power plants and ITER1 and CERN2 LHC3 tunnel. Nuclear radiation can cause deadly damage to the human body, but we have to depend on nuclear energy to meet our great demands for energy resources. Therefore, the research and development of automatic transfer robots and manipulations under nuclear environment are regarded as a key technology by many countries in the world. Robotic applications in radiation environments minimize the danger of radiation exposure to humans. However, the robots themselves are also vulnerable to radiation. Mobility and maneuverability in such environments are essential to task success. Therefore, an efficient obstacle-free path and trajectory generation method are necessary for finding a safe path with maximum bounded velocities in radiation environments. High degree of freedom manipulators and maneuverable mobile robots with steerable wheels, such as non-holonomic omni-directional mobile robots make them suitable for inspection and maintenance tasks where the camera is the only source of visual feedback.
In this thesis, a novel vision-based path planning method is presented by utilizing the artificial potential field, the visual servoing concepts and the CAD-based recognition method to deal with the problem of path and trajectory planning. Unlike the majority of conventional trajectory planning methods that consider a robot as only one point, the entire shape of a mobile robot is considered by taking into account all of the robot’s desired points to avoid obstacles. The vision-based algorithm generates synchronized trajectories for all of the wheels on omni-directional mobile robot. It provides the robot’s kinematic variables to plan maximum allowable velocities so that at least one of the actuators is always working at maximum velocity. The advantage of generated synchronized trajectories is to avoid slippage and misalignment in translation and rotation movement. The proposed method is further developed to propose a new vision-based path coordination method for multiple mobile robots with independently steerable wheels to avoid mutual collisions as well as stationary obstacles. The results of this research have been published to propose a new solution for path and trajectory generation in hazardous and inaccessible to human environments where the one camera is the only source of visual feedback.
Kokoelmat
- Väitöskirjat [4862]