Disassembly scheduling of complex products using parallel heuristic approaches

Abstract

A Greedy Randomized Adaptive Search Procedure (GRASP) algorithm for Disassembly Sequence Planning has recently been proposed. For this problem, feasibility-checking is rather time consuming due to the large number of constraints. The same happens to the local search used within GRASP to improve solutions. In this paper two parallel implementations of the algorithm are proposed for effectively solving larger problems. Both parallel algorithms use cooperative-search plus Path-Relinking and have been tested on a set of 48 benchmark problems. The results show that the parallel algorithms outperform the sequential GRASP algorithm and that migration-based cooperative search may be more effective in this problem than a shared global pool. Average efficiency is around 0.4.