|
|
Analysis of Pilot's Behaviour Models During Flight
Jirgl, Miroslav ; Štefek, Alexandr (referee) ; Blecha, Petr (referee) ; Bradáč, Zdeněk (advisor)
This thesis deals with human – pilot behaviour modelling during a flight in terms of automatic control systems. For these purposes, the introduction to the issue of description and modelling of individual components of the whole pilot – aircraft interaction is presented. Based on that, the simulation models obtained from real measured data are designed. However, the acquisition of the real flight data is quite difficult. Therefore, the flight simulator at Brno University of Defence is used for the purposes of this work. Several experimental measurements were taken using this simulator. These were focused on measuring pilot’s reactions (responses) to visual stimulus with emphasis on obtaining judgements about their current state of training (in terms of dynamic behaviour) as well as attitude to aircraft control. In this phase, two sets of measurements with eight pilots were taken. On average, the pilots had 60 flight hours before the first set of measurements and about 80 flight hours before the second set. The obtained results are analysed using mainly the theory of automatic control approaches in order to evaluate the actual state of pilots’ abilities considering the effects of flight training.
|
|
|
Autonomous Locomotive Robot Path Planning on the Basis of Machine Learning
Krček, Petr ; Bělohoubek, Pavel (referee) ; Štefek, Alexandr (referee) ; Žalud, Luděk (referee) ; Dvořák, Jiří (advisor)
As already clear from the title, this dissertation deals with autonomous locomotive robot path planning, based on machine learning. Robot path planning task is to find a path from initial to target position without collision with obstacles so that the cost of the path is minimized. Autonomous robot is such a machine which is able to perform tasks completely independently even in environments with dynamic changes. Path planning in dynamic partially known environment is a difficult problem. Autonomous robot ability to adapt its behavior to changes in the environment can be ensured by using machine learning methods. In the field of path planning the mostly used methods of machine learning are case based reasoning, neural networks, reinforcement learning, swarm intelligence and genetic algorithms. The first part of this thesis introduces the current state of research in the field of path planning. Overview of methods is focused on basic omnidirectional robots and robots with differential constraints. In the thesis, several methods of path planning for omnidirectional robot and robot with differential constraints are proposed. These methods are mainly based on case-based reasoning and genetic algorithms. All proposed methods were implemented in simulation applications. Results of experiments carried out in these applications are part of this work. For each experiment, the results are analyzed. The experiments show that the proposed methods are able to compete with commonly used methods, because they perform better in most cases.
|
|
|
Online Identification of Trailer Parametry using Ultrasond Sensors
Vejlupek, Josef ; Štefek, Alexandr (referee) ; Vlach, Radek (referee) ; Grepl, Robert (advisor)
This thesis deals with utilizing "the common ultrasonic parking sensors" for assisting the driver with backing-up a trailer. Key issues solved in this thesis are "Online trailer parameter estimation:" determining the estimate of angle between the car and the trailer, and determining the estimate of the length of the trailer shaft (distance from trailer coupling to trailer axle). Thesis contains the model of kinematics of the car with coupled trailer and ultrasonic sensor model together with the trailer viewd as an obstacle.
|
|
|
New Hybrid Methods for Robust and Automated Parameter Estimation of Mechatronic Systems
Najman, Jan ; Štefek, Alexandr (referee) ; Opluštil, Vladimír (referee) ; Grepl, Robert (advisor)
The thesis deals with the development of a new hybrid optimization algorithm for mechatronic models. The introductory chapters are devoted to a general description of the problem of estimating unknown system parameters, based on the developed mathematical model and measured data. Furthermore, an overview and a brief description of available optimization algorithms that are suitable for solving this type of problem is given. Based on the research study, the specific objectives of the paper are then formulated. In the second part of the thesis, a newly developed set of mechatronic models created using physical modelling tools is described. Subsequently, a comparative test of the speed and success rate of the selected optimization algorithms is performed using these models. Based on the results of this test, the design of a new hybrid algorithm is proposed, which is then tested and compared with the other algorithms. Finally, several new auxiliary functions and tools are presented to detect and analyze improperly designed parameter estimation problems.
|
|
|
Modelling, Optimization and Control Design for Strongly Nonlinear Systems with Discrete Sensors
Bastl, Michal ; Štefek, Alexandr (referee) ; Opluštil, Vladimír (referee) ; Grepl, Robert (advisor)
The main motivation for this effort was international cooperation on modeling and control design for the stabilization system of the naval satcom antenna. The dissertation focuses on the design of models based on ODE and DAE from the perspective of mechatronic design. It mainly discusses the possibilities for simulation with a fixed step solvers, what is a necessary condition for real-time simulation and other use of modern approaches such as RCP and HIL. Furthermore, the work deals with modeling of friction, which is due to the strongly non-linear properties problematic for numerical models. The last output is the simulation comparison and verification of the possibility of implementing the incremental encoder into the Kalman filter algorithm.
|
|
|
Online Identification of Trailer Parametry using Ultrasond Sensors
Vejlupek, Josef ; Štefek, Alexandr (referee) ; Vlach, Radek (referee) ; Grepl, Robert (advisor)
This thesis deals with utilizing "the common ultrasonic parking sensors" for assisting the driver with backing-up a trailer. Key issues solved in this thesis are "Online trailer parameter estimation:" determining the estimate of angle between the car and the trailer, and determining the estimate of the length of the trailer shaft (distance from trailer coupling to trailer axle). Thesis contains the model of kinematics of the car with coupled trailer and ultrasonic sensor model together with the trailer viewd as an obstacle.
|
|
|
Online Identification of Trailer Parametry using Ultrasond Sensors
Vejlupek, Josef ; Štefek, Alexandr (referee) ; Vlach, Radek (referee) ; Grepl, Robert (advisor)
This thesis deals with utilizing "the common ultrasonic parking sensors" for assisting the driver with backing-up a trailer. Key issues solved in this thesis are "Online trailer parameter estimation:" determining the estimate of angle between the car and the trailer, and determining the estimate of the length of the trailer shaft (distance from trailer coupling to trailer axle). Thesis contains the model of kinematics of the car with coupled trailer and ultrasonic sensor model together with the trailer viewd as an obstacle.
|
|
|
Analysis of Pilot's Behaviour Models During Flight
Jirgl, Miroslav ; Štefek, Alexandr (referee) ; Blecha, Petr (referee) ; Bradáč, Zdeněk (advisor)
This thesis deals with human – pilot behaviour modelling during a flight in terms of automatic control systems. For these purposes, the introduction to the issue of description and modelling of individual components of the whole pilot – aircraft interaction is presented. Based on that, the simulation models obtained from real measured data are designed. However, the acquisition of the real flight data is quite difficult. Therefore, the flight simulator at Brno University of Defence is used for the purposes of this work. Several experimental measurements were taken using this simulator. These were focused on measuring pilot’s reactions (responses) to visual stimulus with emphasis on obtaining judgements about their current state of training (in terms of dynamic behaviour) as well as attitude to aircraft control. In this phase, two sets of measurements with eight pilots were taken. On average, the pilots had 60 flight hours before the first set of measurements and about 80 flight hours before the second set. The obtained results are analysed using mainly the theory of automatic control approaches in order to evaluate the actual state of pilots’ abilities considering the effects of flight training.
|
|
|
Autonomous Locomotive Robot Path Planning on the Basis of Machine Learning
Krček, Petr ; Bělohoubek, Pavel (referee) ; Štefek, Alexandr (referee) ; Žalud, Luděk (referee) ; Dvořák, Jiří (advisor)
As already clear from the title, this dissertation deals with autonomous locomotive robot path planning, based on machine learning. Robot path planning task is to find a path from initial to target position without collision with obstacles so that the cost of the path is minimized. Autonomous robot is such a machine which is able to perform tasks completely independently even in environments with dynamic changes. Path planning in dynamic partially known environment is a difficult problem. Autonomous robot ability to adapt its behavior to changes in the environment can be ensured by using machine learning methods. In the field of path planning the mostly used methods of machine learning are case based reasoning, neural networks, reinforcement learning, swarm intelligence and genetic algorithms. The first part of this thesis introduces the current state of research in the field of path planning. Overview of methods is focused on basic omnidirectional robots and robots with differential constraints. In the thesis, several methods of path planning for omnidirectional robot and robot with differential constraints are proposed. These methods are mainly based on case-based reasoning and genetic algorithms. All proposed methods were implemented in simulation applications. Results of experiments carried out in these applications are part of this work. For each experiment, the results are analyzed. The experiments show that the proposed methods are able to compete with commonly used methods, because they perform better in most cases.
|
guest ::
RSS feed.