A Feasible Sequence Oriented Discrete Particle Swarm Optimization for Operation Sequencing in CAPP

Abstract

Operation sequencing is one of crucial tasks for process planning in any CAPP system. In this study, a novel discrete particle swarm optimization (DPSO) approach is proposed to solve the operation sequencing problems in CAPP. To find the process plan with lowest machining cost efficiently, the DPSO only searches the feasible operation sequences (FOSs) satisfying precedence constraints among operations. In the DPSO, a FOS is directly represented by a permutation via a particle and the fragment crossover based updating mechanism is developed to evolve the particles. Furthermore, the fragment mutation for altering FOS and the uniform mutation for changing machine, cutting tool and tool access direction for each operation are incorporated into the DPSO to improve exploration ability. A case study involving two prismatic parts are used to verify the performance and efficiency of the DPSO. The comparison between the DPSO and two existing PSOs as well as an existing genetic algorithm shows promising higher performance of the DPSO with respect to solution quality for operation sequencing.