The next Inside Diamond open day will feature stalls and activities, a short introduction to Diamond and a tour of the machine. We expect the visit will last around two and a half hours. Booking for open days opens 6-8 weeks in advance of the event. Click here for more details.
The overall performance of the Diamond electron storage ring is critically dependant on a large number of machine variables or parameters. This performance can be characterised in a number of different ways, such as by measuring electron beam lifetime, injection efficiency, injection transients, and so on, giving a series of objectives to be optimised by varying the machine parameters. Quite frequently, improving one objective comes at the cost of harming another. Similarly, given the large number of variables involved in optimising the ring performance, the true, global optimum solution may be difficult to identify using simple parameter scans.
In order to address this problem, a flexible optimisation tool has been developed that is capable of optimising several objectives at once (Multi-Objective optimisation), using an arbitrary number of variables to achieve this goal (typically the strength of various magnets). The application can be used with any online optimisation problem, being not tied to any particular set of optimisation variables. The tool is also able to cope with measurement noise on the parameters to be optimised. At present, the tool has been developed and tested with several optimisation algorithms: a Multi Objective Genetic Algorithm (MOGA), a Multi Objective Particle Swarm Optimization (MOPSO) and a Multi Objective Simulated Annealing (MOSA).
So far emphasis has been put on the Online Optimization aspects. Depending on the inclination and skills of the candidate, the present project is open to a couple of possible developments:
1) the realization of a new Python based optimization algorithm, to be added to the present list (e.g. Artificial Bee Colony in a Multi Objective fashion), and a thorough study of the different performances of all the algorithms,
2) the connection of the present optimization suite with a particle tracking code used to simulate the machine (Virtual Machine – VM).
Both tasks should be complemented with a series of tests on the machine (Storage Ring or Booster or Transfer Lines).
Suitable Subjects: Physics, Mathematics, Computing, Engineering
Project duration: 12 weeks
TO APPLY PLEASE FOLLOW THE LINK BELOW
Diamond Light Source is the UK's national synchrotron science facility, located at the Harwell Science and Innovation Campus in Oxfordshire.
Copyright © 2018 Diamond Light Source
Diamond Light Source Ltd
Harwell Science & Innovation Campus