Abstract
The aim of this work is to propose and validate a novel multi-objective optimization algorithm based on the emulation of the behaviour of the immune system. The rationale of this work is that the artificial immune system has, in its elementary structure, the main features required by other multi-objective evolutionary algorithms described in the literature, such as diversity preservation, memory, adaptivity, and elitism. The proposed approach is compared with three multi-objective evolutionary algorithms that are representative of the state of the art in multi-objective optimization. Algorithms are tested on six standard problems (both unconstrained and constrained) and comparisons are carried out using three different metrics. Results show that the proposed approach has very good performances and can become a valid alternative to standard algorithms for solving multi-objective optimization problems.
Notes
†It is here assumed that the algorithm must minimize the fitness.
‡The original notation of Equation(8) has been changed by introducing σ instead of 1/β by de Castro and Timmis Citation(2002).
§A mutation is considered successful if the fitness of the offspring is better than or equal to the fitness of the parent.