Project 56: Reinforcement Learning Based on the Model-free Control of Multicopters
Contact Information:
Prof. Wang Hesheng
Email: wanghesheng@sjtu.edu.cn
Project Description and Objectives:
Multicopters are being applied in an increasingly wider range of tasks, such as photography, object transportation and even military, as their prices drop and their capability increases. One of the key questions that arises along with such applications is how to enhance the versatility of multicopters, so that it can perform different tasks with different payloads autonomously, without the need for manual adjusting of the multicopter control scheme. The program aims to exploit the scalability and transferability of reinforcement learning methods to extract model-free control schemes for multicopters, and validate the system in both simulation and on real-world multicopters.
Eligibility Requirements:
C, C++, Python
Main Tasks:
Designing reinforcement learning based multicopter control scheme.
Method validation in both simulation and on real-world multicopters.
Website: