Multi-objective Optimisation with Uncertainty

P. Jones, A. Tiwari, R. Roy, and J. Corbett (UK)


Fuzzy Modelling, Evolutionary Engineering, Engineering, Optimisation


: This paper looks at the implications on multi objective optimisation when the objectives are subject to uncertainty. In this case the High Efficiency Deep Grinding (HEDG) process is modelled using a Fuzzy Inference Cognitive Map (FICM) with uncertainty present. Uncertainty forms an extra objective with the user having to decide which is preferable; near optimal yet highly uncertain results or certain but inefficient results. Wheel loading forms a process constraint that is modelled as an objective within the FICM. Two Algorithms are applied to the problem, NSGA-II and SPEA. The uncertainty associated with each objective is aggregated using the FICM to give an overall uncertainty value for each solution.

Important Links:

Go Back