Constraint handling with uncertain and noisy multi-objective evolution
E.J. Hughes · 2002
Many real world problems are constrained and have multiple objectives that must be satisfied. To compound the optimisation challenge, systems are often noisy or uncertain, leading to errors in the objective calculations. This paper develops theory to help reduce the effects of noise and uncertainty on constrained evolutionary optimisation processes. Experimental results are presented for generating Pareto surfaces with two different types of noise and also with constraints and designer preferences.