A particle swarm optimization for assembly line design with station paralleling and equipment selection

A 0–1 integer linear programming model is presented for the assembly line design problem (ALDP) with station paralleling and equipment selection. The presented model overcomes the incompleteness of a well-known model for ALDP by adding new constraints. Due to the NP-hard essence of ALDP, a particle swarm optimization (PSO) algorithm is proposed. In the PSO, a permutation encoding scheme along with zero in-degree topological sorting is adopted to make the particle correspond to a feasible task sequence (FTS). Then, a graph approach is utilized to identify the assembly line with lowest fixed cost associated with given FTS. To update the particle, a position based crossover operator for permutation encoding is adopted. Case study illustrates the validity of the 0–1 integer linear programming model, and also shows the effectiveness and efficiency of the PSO algorithm.