Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Deep reinforcement learning for collaborative aerial transportation
Collaborative object transportation using micro aerial vehicles (MAVs) is a promising drone technique. It is challenging from a control perspective, since multiple MAVs are mechanically coupled, imposing hard kinematic constraints. Traditional model-based methods often require linearization of the nonlinear problem which restrains the performance such as transporting speed and the payload. The goal of his project aims at exploring the possibility of using the deep reinforcement learning approach to obtain a centralized control policy for collaborative aerial transportation, which is more efficient than the state-of-the-art methods. The policy will be trained in a simulation environment and then transferred to real-life experiments. Applications should have strong experience in C++, Python. Applicants with reinforcement learning and flight control background are favored.
Collaborative object transportation using micro aerial vehicles (MAVs) is a promising drone technique. It is challenging from a control perspective, since multiple MAVs are mechanically coupled, imposing hard kinematic constraints. Traditional model-based methods often require linearization of the nonlinear problem which restrains the performance such as transporting speed and the payload. The goal of his project aims at exploring the possibility of using the deep reinforcement learning approach to obtain a centralized control policy for collaborative aerial transportation, which is more efficient than the state-of-the-art methods. The policy will be trained in a simulation environment and then transferred to real-life experiments. Applications should have strong experience in C++, Python. Applicants with reinforcement learning and flight control background are favored.
Collaborative object transportation using micro aerial vehicles (MAVs) is a promising drone technique. It is challenging from a control perspective, since multiple MAVs are mechanically coupled, imposing hard kinematic constraints. Traditional model-based methods often require linearization of the nonlinear problem which restrains the performance such as transporting speed and the payload. The goal of his project aims at exploring the possibility of using the deep reinforcement learning approach to obtain a centralized control policy for collaborative aerial transportation, which is more efficient than the state-of-the-art methods. The policy will be trained in a simulation environment and then transferred to real-life experiments. Applications should have strong experience in C++, Python. Applicants with reinforcement learning and flight control background are favored.
The goal of this project is to use deep reinforcement learning on collaborative aerial transportation. The method needs to be validated in real flight tests.
The goal of this project is to use deep reinforcement learning on collaborative aerial transportation. The method needs to be validated in real flight tests.
Sihao Sun (sun at ifi.uzh.ch), Yunlong Song (song at ifi.uzh.ch)
Sihao Sun (sun at ifi.uzh.ch), Yunlong Song (song at ifi.uzh.ch)