A hybrid particle swarm algorithm for assembly line balancing problem of type 1

To solve the assembly line balancing problem of type 1 (ALBP-I), a hybrid particle swarm algorithm (HPSA) is presented through incorporating the reduced variable neighborhood search (RVNS) into the discrete PSA (DPSA) previously developed by the author. In order to enhance the explorability and facilitate finding global optimum, the RVNS is utilized to perform a local search within the neighbors of the best particle of the swarm gBest at each generation. The HPSA is tested on a set of problems taken from the literature and compared with existing DPSA and continuous PSA (CPSA). The comparative study shows that the RVNS facilitates the convergence to global optimum and improves the solution quality. In addition, the computational comparisons indicate that HPSA is superior to existing PSAs and DPSA is better than CPSA with respect to convergence speed and explorability. The computational results illustrate the effectiveness and advantages of HPSA for ALBP-I.