Abstract:

Based on the fully-informed particle swarm optimization ( FIPS) and the dynamic tabu search ( TS) ,a hybrid Pareto algorithm is proposed to solve the complex multi-objective flexible job-shop scheduling problem( FJSP) ,which takes advantage of the global search capability of FIPS to determine the search direction and then performs a local search with TS to effectively improve the search performance. In this algorithm,first,a strengthbased fitness function is adopted to evaluate the quality of particles,which makes the non-dominated solutions uniformly distribute along the Pareto front. Then,several neighbourhoods based on public key blocks are employed to keep the diversity of the swarm,which avoids the trapping in the local optimum and effectively accelerates the convergence of the algorithm. Moreover,a self-adaptive perturbation based on mutation is introduced in the algorithm to enhance the diversity of solutions. The results of comparative experiments in different scales indicate that the proposed algorithm is of good search performance and is effective in solving the multi-objective FJSPs in both large and small scales.

Key words: flexible shop scheduling, fully-informed particle swarm optimization, tabu search, multi-objective optimization