Distributed Multiobjective Quantum-Inspired Evolutionary Algorithm (DMQEA)

Most of the multiobjective evolutionary algorithm inherently has heavy computational burden, so it takes a long processing time. For this reason, many researches for reducing computational time have been carried out, in particular by using distributed computing such as multi-thread coding, GPU coding, etc. In this paper, multi-thread coding is used to reduce computational time and applied to multiobjective quantum-inspired evolutionary algorithm (MQEA). In MQEA, nondominated sorting and crowding distance assignment which take a long time are carried out in each subpopulation. By multi-thread coding, the processes in each subpopulation can be performed simultaneously. To demonstrate the effectiveness of the proposed distributed MQEA (DMQEA), comparisons with single-thread and multi-thread are carried out for seven DTLZ functions.