An Improved Wild Horse Algorithm to Solve Facility Layout in Horizontal Cylinder Space

FLP (Facility layout problem) refers to placing a certain number of devices in a designated space under certain conditions. This is a NP-hard problem. At present, the vast majority of related research focuses on two-dimensional planes and regular cube space. This paper studies the FLP in the three-dimensional space of a horizontal cylinder. Combined with engineering practice, the overlap constraint, gravity constraint, proximity constraint and symmetry constraint are mathematically expressed. The combined explosion phenomenon and engineering complexity of FLP make it difficult for deterministic algorithms to arrive at a reasonable solution in a relatively short period of time. Unfortunately, there are also few heuristic algorithms that can effectively solve the problem of facility layout. This paper improves the latest wild horse algorithm and introduces a directed jump strategy, which greatly reduces the layout time. Meanwhile, a dominant strategy is introduced to deal with constrained multi-objective optimization scenarios. Experiments show that these strategies are feasible.