A novel multi-objective particle swarm optimization with multiple search strategies 

 Abstract

 Recently, multi-objective particle swarm optimization (MOPSO) has shown the effectiveness 
 in solving multi-objective optimization problems (MOPs). However, most MOPSO algorithms only 
 adopt a single search strategy to update the velocity of each particle, which may cause some 
 difficulties when tackling complex MOPs. This paper proposes a novel MOPSO algorithm using 
 multiple search strategies (MMOPSO), where decomposition approach is exploited for transforming 
 MOPs into a set of aggregation problems and then each particle is assigned accordingly to optimize 
 each aggregation problem. Two search strategies are designed to update the velocity of each 
 particle, which is respectively beneficial for the acceleration of convergence speed and the keeping 
 of population diversity. After that, all the non-dominated solutions visited by the particles are 
 preserved in an external archive, where evolutionary search strategy is further performed to exchange 
 useful information among them. These multiple search strategies enable MMOPSO to handle various kinds 
 of MOPs very well. When compared with some MOPSO algorithms and two state-of-the-art evolutionary 
 algorithms, simulation results show that MMOPSO performs better on most of test problems.