Autonomous Landing for Unmanned air vehicle (UAV) is always very challenging due to external forces acting on the systems. These forces are associated with disturbances which make the UAV difficult to control and to land safely autonomously. The purpose of this research project is to use different methods of reinforcement learning algorithm and control systems based vision. The PhD project will also look at developing techniques based on Reinforcement learning, mathematical modelling, Control theory and Machine learning, and compare with Q-learning, hindsight replay; Neural Network autonomous landing combine with learning environment is a potential avenue to the research.
Reinforcement learning is a powerful, overall approach to discovering optimal policies for complex decision-making problems. Recently, with approximations of functions such as neural networks, Reinforcement learning has significantly expanded the range of applications, from playing computer games to learning movements by simulated human. It is successfully used in areas where massive amounts of simulated data can be generated, such as robotics and games. Reinforced learning as a promising path to General AI.
Other application of reinforcement learning in this studies is on robot manipulation with an interest in real-world robot learning problems, which deal of noisy data, and the practical aspects of real-world data gathering rather than simulated tasks. There is a particular interest to focus on methods which combine learning-based methods (such as imitation learning and reinforcement learning) with classical methods (such as optimal control and vision-based state estimation). The PhD projects is at the intersection of learning-based methods (such as imitation learning and reinforcement learning) and classical methods (such as optimal control and vision-based state estimation).
Examples of current areas of research in are: (i) data-efficient model-based reinforcement learning for real-world robot manipulation, (ii) transferring control policies trained in simulation over to the real world, (iii) learning of new tasks from a small number of human demonstrations, (iv) visual state estimation for object grasping and interaction, and (v) visual scene understanding for planning multi-stage tasks.
The PhD research will develop new algorithms to achieve substantial improvements over other methodologies, but also provide novel aspect and perspectives on how to mix different branches of reinforcement learning effectively to gain better efficient response. The control algorithm would need to be demonstrated either application of on Unmanned air vehicle (UAV) or an actual 6-DoF robot arms, for example to open doors training data of real-world experience, making it one of the first demonstrations of deep Reinforcement Learning approaches on real robots.
The successful applicant should have been awarded, or expect to achieve, a Masters degree in a relevant subject with a 60% or higher weighted average, and/or a First or Upper Second Class Honours degree (or an equivalent qualification from an overseas institution) in Control Engineering; Robotics and Computer Science. Preferred skill requirements include knowledge/experience of Data Science, Control Systems; Mathematical Modelling, Artificial Intelligence and Machine Learning.
Submitting an application
As part of the application, you will need to supply:
- Your current CV
- Copies of your academic qualifications for your Bachelor degree, and Masters degree; including certificates and transcripts, and must be translated in to English
- A research proposal statement*
- Two academic references
- Proof of your English Language proficiency
Details of how to submit your application can be found here.
The application must be accompanied by a “research proposal” statement. An original proposal is not required as the initial scope of the project has been defined, candidates should take this opportunity to detail how their knowledge and experience will benefit the project and should also be accompanied by a brief review of relevant research literature.
Include the supervisor name, project title, and project reference in your application.