Solving Assembly Line Balancing Problems with Fish School Search algorithm

Abstract

Assembly lines constitute the main production paradigm of the contemporary manufacturing industry. Thus, many optimization problems have been studied aiming to improve the efficacy of its use. In this context, the problem of balancing an assembly line plays a key role. This problem is of combinatorial nature and also NP-Hard. For this reason, many researchers on computational intelligence and industrial engineering have been conceiving algorithms for tackling many versions of assembly line balancing problems using different procedures. In this paper, the Fish School Search algorithm and a variation of it that incorporates a routine to avoid stagnation of the search process were applied in order to solve the Simple Assembly Line Balancing Problem-type 1. The results were compared with an exact solution procedure named SALOME and also with the Particle Swarm Optimization algorithm. Both proposed procedures were able to achieve good results and the stagnation avoidance routine incorporated to FSS allowed more uniform distributions of tasks among workstations in the assembly line and converged faster to optimal solutions.