Two approaches for solving the buffer allocation problem in unreliable production lines

A parallel tabu search for solving the primal buffer allocation problem in serial production systems

“…The algorithm structure is identical to the BS proposed by Demir et al [14], the only difference being the initial solution generation mechanism that is based on the proposed ISG method. Actually, the aforementioned authors used BS as an alternative technique to validate a novel tabu search (TS).…”

“…In most literature about metaheuristics applied to BAP, the designer arbitrarily fixes an initial total buffer capacity sensibly lower than (J À 1) Â (n À1), regardless of any rationale or quantitative method (See for example [14,18,57]). As a result, the search ability of any metaheuristics may be strongly supported by such a forced choice and the global performance may be significantly improved.…”

“…Due to the structure of the BS algorithm, no limit may be fixed in terms of either generation or evaluations. Therefore, it has been implemented according to the same setting used in the literature (See [14]). In order to emphasize the computational effectiveness of PTS, the maximum number of evaluations has been deliberately reduced with respect to the other competitors.…”

“…A parallel tabu search (PTS) algorithm has been developed to address the primal buffer allocation problem in reliable production lines. Differently from the previous tabu search procedures, as for example those proposed by Demir et al [14] and Lutz et al [31], PTS exploits the parallel configuration inherited from the parallel computing field to enhance the search strategy over the available space of solutions. In addition, an adaptive neighborhood generation mechanism that tunes the entity of buffer size variations according to the number of iterations allows a quick convergence of the parallel branches towards feasible and promising near-optimal solutions.…”

“…Also, it is worthy to point out that they made use of a decomposition method to address the aforementioned issue related to unreliable production lines. Two research contributions emphasizing the effectiveness of tabu search in solving the BAP are ascribable to [13,14]. First, they presented an adaptive tabu search (ATS) for designing unreliable non-homogenous production lines with the objective of throughput rate maximization.…”