Research on hybrid cloud particle swarm optimization for multi-objective flexible job shop scheduling problem

Abstract

Flexible job shop scheduling is an NP-hard problem. In this paper, we design a novel hybrid cloud particle swarm optimization (HCPSO) algorithm with genetic algorithm (GA) that is adopted to provide optimal solutions according to the pareto optimality principle in solving multi-objective FJSS problem. It is aimed at minimizing completion time of jobs, total workload and maximum workload. The novelty of the new proposed approach is that the whole particles are divided into three different populations respectively with different weights according to the fitness value. The weight has stable tendency and randomness properties based on the cloud model, which not only improves the convergence speed, but also maintains the diversity of the population. The simulation results show that the HCPSO algorithm has the advantages of small optimization, fast convergence, high efficiency and good population diversity, which verifies the effectiveness and the feasibility of HCPSO algorithm. The results of the instance verify that HCPSO algorithm is suitable for multi-objective optimization problems.