One PhD position is available in Surrey at the Surrey Space Centre of the University of Surrey (UK),
Context
This project builds upon an already established collaboration between the University of Surrey, Surrey Satellite Technology Ltd (SSTL) and the European Space Agency (ESA) where a ground-based automated planning & scheduling system for the imaging campaign of an Earth Observation constellation has been built. Multi agent systems represented a fruitful approach to model this dynamic context. The novelty of this project was to apply Ant Colony Optimisation techniques to achieve optimisation and coordination in the constellation planning. This mechanism offers high-level of adaptability and scalability. The project used mathematical models of non-linear dynamical systems to model the self-organising long-term system’s behaviours. Finally, the system developed has been applied to the ESA GENSO (Global Educational Network for Satellite Operations) network and to the SSTL Disaster Monitoring Constellation.
Project Title
Application of Dynamical Systems Theory to Multi Objective Automated Space Mission Planning based on Ant Colony approach
Project Description
The work proposed here represents the natural step forward. The innovation of this project regards three main research directions:
Multi-Objective Optimisation. The P&S problem can be treated as pareto problem, ie. separate optimisation of objectives to find a set of pareto-optimal solutions (which might have conflicting objective goals). This is becoming a key requirement for goal-oriented planning because extends the way how the objectives can be defined and it could greatly improve the operators’ evaluation of the solutions generated.
Dynamical System Theory. The model previously developed can be extended in a number of directions: the problem representation affects the range of applications of the planning system; the ACO techniques implemented affects its performance.
Hybrid architecture. This research should explore the benefits of hybrid architecture where deliberative agents are set to regulate the collective behaviours expressed by the self-organising multi agent system. This could extend the capabilities and the reliability of the overall system.
Requirements
The PhD last 3 years.
The applicant will be based in Surrey at the university. He can spend periods abroad up to 1 year, in general he should spend 6 months in ESOC but it's not mandatory.
The supervisor will be Prof Phil Palmer, head of the group of Astrodynamics at the Surrey Space Centre. Alessandro Donati and Nicola Policella will be his tutor at the European Space Operation Centre (ESOC) of Darmstadt (Germany) and Claudio Iacopino will be his tutor at the Surrey Satellite Technology Ltd (SSTL)
Applicants should hold a good honours undergraduate or Masters degree in computer science or a closely related discipline such as mathematics.
Applicants should have a background in artificial intelligence or advanced computer technologies, ideally automated planning and scheduling and an understanding/experience of the Space Operations/ Earth Observation field.
Applicants' nationality has to be ones of the member states of the European Space Agency
Funding: The applicant will receive full funding for tuition fee, international travels and a salary for 3 years.
Deadline: Until filled.
Start date: 1st October 2014- 1st January 2015
For information on how to apply and full application requirements please email c.iacopino@sstl.co.uk
Tidak ada komentar:
Posting Komentar