|
|
Vehicle Control via Reinforcement Learning
Maslowski, Petr ; Uhlíř, Václav (referee) ; Šůstek, Martin (advisor)
The goal of this thesis is a creation of an autonomous agent that can control a vehicle. The agent utilizes reinforcement learning that uses neural networks. The agent interprets images from the front vehicle camera and selects appropriate actions to control the vehicle. I designed and created reward functions and then experimented with hyperparameters setup. Trained agent simulate driving on the road. The result of this thesis shows a possible approach to control an autonomous vehicle agent using machine learning method in CARLA simulator.
|
|
|
Application of Reinforcement Learning in Autonomous Driving
Vosol, David ; Zbořil, František (referee) ; Janoušek, Vladimír (advisor)
This thesis is focused on the topic of reinforcement learning applied to a task of autonomous vehicle driving. First, the necessary fundamental theory is presented, including the state-of-the-art actor-critic methods. From them the Proximal policy optimization algorithm is chosen for the application to the mentioned task. For the same purpose, the racing simulator TORCS is used. Our goal is to learn a reinforcement learning agent in a simulated environment with the focus on a future real-world application to an RC scaled model car. To achieve this, we simulate the conditions of remote learning and control in the cloud. For that, simulation of network packet loss, noisy sensory and actuator data is done. We also experiment with the least number of vehicle's sensors required for the agent to successfully learn the task. Experiments regarding the vehicle's camera output are also carried out. Different system architectures are proposed, among others also with the aim to minimize hardware requirements. Finally, we explore the generalization properties of a learned agent in an unknown environment.
|
|
|
Autonomous control of the vehicle through image processing
Fronc, Leoš ; Píštěk, Václav (referee) ; Kučera, Pavel (advisor)
This diploma thesis deals with the topic of autonomous vehicles and especially lane detection. The paper describes and compares two main approaches to the lane detection - using traditional methods of computer vision and convolutional neural networks. The aim of the work was to create a system that would be able to recognize road lanes in a real time. The proposed system consisted of a Jetson Nano computer, a ZED stereo camera and a programmed algorithm. In total, two algorithms have been developed that use completely different approaches. Finally, the whole system was tested in terms of functionality and lane recognition.
|
|
|
Application of Reinforcement Learning in Autonomous Driving
Vosol, David ; Zbořil, František (referee) ; Janoušek, Vladimír (advisor)
This thesis is focused on the topic of reinforcement learning applied to a task of autonomous vehicle driving. First, the necessary fundamental theory is presented, including the state-of-the-art actor-critic methods. From them the Proximal policy optimization algorithm is chosen for the application to the mentioned task. For the same purpose, the racing simulator TORCS is used. Our goal is to learn a reinforcement learning agent in a simulated environment with the focus on a future real-world application to an RC scaled model car. To achieve this, we simulate the conditions of remote learning and control in the cloud. For that, simulation of network packet loss, noisy sensory and actuator data is done. We also experiment with the least number of vehicle's sensors required for the agent to successfully learn the task. Experiments regarding the vehicle's camera output are also carried out. Different system architectures are proposed, among others also with the aim to minimize hardware requirements. Finally, we explore the generalization properties of a learned agent in an unknown environment.
|
|
|
Autonomous control of the vehicle through image processing
Fronc, Leoš ; Píštěk, Václav (referee) ; Kučera, Pavel (advisor)
This diploma thesis deals with the topic of autonomous vehicles and especially lane detection. The paper describes and compares two main approaches to the lane detection - using traditional methods of computer vision and convolutional neural networks. The aim of the work was to create a system that would be able to recognize road lanes in a real time. The proposed system consisted of a Jetson Nano computer, a ZED stereo camera and a programmed algorithm. In total, two algorithms have been developed that use completely different approaches. Finally, the whole system was tested in terms of functionality and lane recognition.
|
|
|
Vehicle Control via Reinforcement Learning
Maslowski, Petr ; Uhlíř, Václav (referee) ; Šůstek, Martin (advisor)
The goal of this thesis is a creation of an autonomous agent that can control a vehicle. The agent utilizes reinforcement learning that uses neural networks. The agent interprets images from the front vehicle camera and selects appropriate actions to control the vehicle. I designed and created reward functions and then experimented with hyperparameters setup. Trained agent simulate driving on the road. The result of this thesis shows a possible approach to control an autonomous vehicle agent using machine learning method in CARLA simulator.
|
guest ::
