A New Design of Swarm Intelligence Based Metaheuristics for Constrained Multi-objective Optimization

Swarm intelligence metaheuristics (SIM) are a special kind of population-based algorithms, inspired by a collective and social behavior of some species. SIM have become mostly common used for multi-objective optimization problems (MOPs). However, almost all of them are designed for unconstrained optimization problems. In this work, we intend to present a new design to handle constraints with SIM to deal with constrained multi-objective problems (CMOPs). This new design aims to investigate the potential effect of leader concept that may help the population to reach Pareto optimal solutions in constrained space. Two simple constraint handling techniques are employed to check the feasibility of candidate solutions. In addition, we use crowding distance to maintain diversity in the population. This new design is implemented with multi-objective cuckoo (MOCS) search and multi-objective particle swarm optimization (MOPSO). Promising results are obtained using well-known test functions from the specialized literature.