|
|
Advanced Navigation in Heterogeneous Multi-robot Systems in Outdoor Environment
Jílek, Tomáš ; Duchoň,, František (referee) ; Mazal, Jan (referee) ; Žalud, Luděk (advisor)
The doctoral thesis discusses current options for the navigation of unmanned ground vehicles with a focus on achieving high absolute compliance of the required motion trajectory and the obtained one. The current possibilities of key self-localization methods, such as global satellite navigation systems, inertial navigation systems, and odometry, are analyzed. The description of the navigation method, which allows achieving a centimeter-level accuracy of the required trajectory tracking with the above mentioned self-localization methods, forms the core of the thesis. The new navigation method was designed with regard to its very simple parameterization, respecting the limitations of the used robot drive configuration. Thus, after an appropriate parametrization of the navigation method, it can be applied to any drive configuration. The concept of the navigation method allows integrating and using more self-localization systems and external navigation methods simultaneously. This increases the overall robustness of the whole process of the mobile robot navigation. The thesis also deals with the solution of cooperative convoying heterogeneous mobile robots. The proposed algorithms were validated under real outdoor conditions in three different experiments.
|
|
|
Processing of localization data and its accuracy
Návara, Marek ; Chromý, Adam (referee) ; Jílek, Tomáš (advisor)
This work is focused on correction of localization data, relative to them location of the sensor on robot. Also solves the conversion between coordinate systems WGS-84, ETRS-89 and local coordinate system. For localization data are designed error characteristics and thier transformation.These points have been implemented in C# as individual classes for position sensors and sensors orientation. Classes were firstly developer for single-axis sensors and they are base for multi-axis sensors. The implementation has been verified in the experiment with the robot envMap. Experiment results show that the correction is implemented properly and can be further used. Error characteristics listed in the work characterize the measurement uncertainty of location data and can be used in further processing.
|
|
|
Processing of localization data and its accuracy
Návara, Marek ; Baránek, Radek (referee) ; Jílek, Tomáš (advisor)
This work is focused on correction of localization data, relative to them location of the sensor on robot. Also solves the conversion between coordinate systems WGS-84, ETRS-89 and local coordinate system. For localization data are designed error characteristics and their transformation.These points have been implemented in C#. Classes were firstly developer for single-axis sensors and they are base for multi-axis sensors. The implementation has been tested on simulated data and verified in the experiment with the robot envMap. Experiment results show that the correction is implemented properly and can be further used. Error characteristics listed in the work characterize the measurement uncertainty of location data and can be used in further processing.
|
|
|
Processing of localization data and its accuracy
Návara, Marek ; Baránek, Radek (referee) ; Jílek, Tomáš (advisor)
This work is focused on correction of localization data, relative to them location of the sensor on robot. Also solves the conversion between coordinate systems WGS-84, ETRS-89 and local coordinate system. For localization data are designed error characteristics and their transformation.These points have been implemented in C#. Classes were firstly developer for single-axis sensors and they are base for multi-axis sensors. The implementation has been tested on simulated data and verified in the experiment with the robot envMap. Experiment results show that the correction is implemented properly and can be further used. Error characteristics listed in the work characterize the measurement uncertainty of location data and can be used in further processing.
|
|
|
Processing of localization data and its accuracy
Návara, Marek ; Chromý, Adam (referee) ; Jílek, Tomáš (advisor)
This work is focused on correction of localization data, relative to them location of the sensor on robot. Also solves the conversion between coordinate systems WGS-84, ETRS-89 and local coordinate system. For localization data are designed error characteristics and thier transformation.These points have been implemented in C# as individual classes for position sensors and sensors orientation. Classes were firstly developer for single-axis sensors and they are base for multi-axis sensors. The implementation has been verified in the experiment with the robot envMap. Experiment results show that the correction is implemented properly and can be further used. Error characteristics listed in the work characterize the measurement uncertainty of location data and can be used in further processing.
|
|
|
Advanced Navigation in Heterogeneous Multi-robot Systems in Outdoor Environment
Jílek, Tomáš ; Duchoň,, František (referee) ; Mazal, Jan (referee) ; Žalud, Luděk (advisor)
The doctoral thesis discusses current options for the navigation of unmanned ground vehicles with a focus on achieving high absolute compliance of the required motion trajectory and the obtained one. The current possibilities of key self-localization methods, such as global satellite navigation systems, inertial navigation systems, and odometry, are analyzed. The description of the navigation method, which allows achieving a centimeter-level accuracy of the required trajectory tracking with the above mentioned self-localization methods, forms the core of the thesis. The new navigation method was designed with regard to its very simple parameterization, respecting the limitations of the used robot drive configuration. Thus, after an appropriate parametrization of the navigation method, it can be applied to any drive configuration. The concept of the navigation method allows integrating and using more self-localization systems and external navigation methods simultaneously. This increases the overall robustness of the whole process of the mobile robot navigation. The thesis also deals with the solution of cooperative convoying heterogeneous mobile robots. The proposed algorithms were validated under real outdoor conditions in three different experiments.
|
guest ::
