Ph.D. Student Position (AIO)
                   Evolutionary Computation
                            for
                    Smart Energy Networks

            http://www.cwi.nl/jobs/phd-student-0
    (see also http://www.cwi.nl/jobs/ for other positions)


Project description
-------------------
Network Capacity Planning (NCP) is important for technical and
economic efficiency and adequacy of future electricity networks.
The aim of efficient NCP is to enable the required transition
towards a more flexible and sustainable energy system combined
with higher participation on the decentralized electricity
market.

The planning of resource investments in electrical power systems
has always been a hard task, as the generation capacity should
meet demand, but not exceed it, causing unnecessarily high costs.
Network capacity planning (NCP) is also an inherently
multi-objective problem. It deals with deciding on investments at
several locations with various types of network components and
configurations, as well as different time paths for investing.
Important objectives are optimal network capacity (so that future
demand and production of electricity can be facilitated), minimal
total costs, optimal technical performance of the network (meet
technical requirements with respect to power quality and network
stability), and robustness of investments in many possible future
developments scenarios.

Key to successful decision support is the ability to study
different scenarios for future network and economics
developments, and to optimize the various objectives that are of
importance within these scenarios. Due to the high complexity of
measuring the effects of investments in the network, analytical
methods are not feasible. For this, novel and efficient
simulation models and tools for the network and its actors are
required. Thus, the objectives can be measured by a complex
simulation that is essentially a black box, with the parameters
such as the network structure, prices, and capacity levels as
input, and the objective values as output. Given such results,
the subsequent problem is how to optimize the input parameters.

In the project 2 PhD students will be active, one based at TU
Eindhoven, one based at CWI and both part-time at KEMA. The
project focuses on computational optimisation in NCP and develops
novel ICT tools (Information and Communication Technology) for
scenario simulation and decision-making in NCP. Within the
project, the PhD student based at TU Eindhoven will primarily
focus on the development of network models and simulation
platforms for realistic network and usage outcomes from parameter
and scenario choices. The PhD student based at CWI (to which this
vacancy pertains) will primarily focus on the development of
advanced and dedicated multi-objective optimization techniques in
the field of evolutionary computation in order to deal with the
high complexity of the problem at hand.

The project consists of fundamental research with an eye on
application and is carried out together with the partners.


Position
--------
Within this project it is your job to develop advanced
evolutionary algorithms (EAs) in order to automatically optimise
parameter settings for a collection of (stochastic) future
development scenarios using the simulation platform as an
evaluator. Multiple objectives that contain uncertainties are to
be taken into account (like network constraints and performance,
costs, robust return-on-investment, and economic accessibility).
Key tasks are the design of proper representations for the
optimization parameters, tackling uncertainty in the presence of
multiple objectives, the investigation of the reduction of the
number of required network simulations as input to the EAs, and
the incorporation of domain knowledge on electricity networks in
the form of constraints and local optimization combined with EAs.
As it is to be expected that available classic EA techniques are
not sufficiently scalable and efficient for NCP, the development
of more sophisticated and dedicated EA techniques for NCP are
required. Of particular interest will be
Estimation-of-Distribution Algorithms (EDAs). Characteristic of
EDAs is the use of probability distributions to model and
successfully exploit variable interdependencies, contrary to
classic EA methods.

Keywords are evolutionary algorithms, estimation-of-distribution
algorithms, black box optimization, multi-objective optimization,
uncertainty, and simulation systems.


Requirements
------------
PhD candidates are required to have a master degree in computer
science, artificial intelligence, mathematics, or in comparable
areas. Candidates that are expected to finish their M.Sc. thesis
in the near future can also apply. Candidates should have a clear
interest in fundamental research, should be creative and solid in
their research, should have (potential) interest in the
application area of electricity networks, and should be able to
cooperate with experts from different disciplines.

It's essential that you have good academic writing and
presentation skills. Candidates are expected to have an excellent
command of English.


Institute
---------
The CWI (Centrum Wiskunde & Informatica) in Amsterdam, founded in
1946, is the national research institute for mathematics and
computer science in the Netherlands. CWI performs frontier
research in mathematics and computer science and transfers new
knowledge in these fields to society in general and trade and
industry in particular.


Terms and conditions
--------------------
The terms of employment are in accordance with the Dutch
Collective Labour Agreement for Research Institutes
("CAO-onderzoeksinstellingen"). Moreover CWI offers attractive
working conditions, including flexible scheduling and help with
housing for expat employees.

The gross monthly salary, for an employee on a full time basis,
is 2,037 Euro during the first year and increases to 2,610 Euro
over a four year period. CWI also offers excellent and flexible
terms of employment, including an employee pension fund.


Information
-----------
Additional information can be obtained from:

Dr. Peter A.N. Bosman (Evolutionary Computation)
Tel. +31 (0)20 592 4238
e-mail: [log in to unmask]

Prof.dr.ir. Han La Poutré (Group Leader)
Tel. +31 (0)20 592 4082
e-mail: [log in to unmask]

Additional information about working at the CWI can be seen in
the video "A Fundamental Difference":

http://www.cwi.nl/general/movies-about-cwi

or

http://www.youtube.com/watch?v=2e-VYBdBKis


Application
-----------
Please send your application before 26 June 2011 to: [log in to unmask]

Applications should include a detailed CV, a motivation letter, a
list of your MSc courses and grades.

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
  Dr. Peter A.N. Bosman
  Scientific Staff Member
  Centrum Wiskunde & Informatica (CWI)
  (Centre for Mathematics and Computer Science)
  P.O. Box 94079
  1090 GB Amsterdam
  The Netherlands

  [log in to unmask]
  http://homepages.cwi.nl/~bosman
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-