Output details
15 - General Engineering
University of Sheffield
Convergence Acceleration Operator for Multiobjective Optimization
Working with industrial partners, Rolls-Royce (contact: Dr I Griffin, i.griffin@rolls-royce.com) and Ford (contact: Dr R Lygoe, blygoe@ford.com), amongst others, Fleming is successfully applying multiobjective evolutionary optimisation algorithms (MOEAs) to industrial problems. Such problems are often many-objective, i.e. the number of objectives exceeds 2-3. Fleming and Purshouse (Sheffield) have pioneered research into many-objective optimisation, identifying that many-objective problems cannot be satisfactorily addressed using conventional MOEAs. This paper addresses one of two key solution quality measures, convergence, and describes an approach that significantly improves solution accuracy and speed.