Construction Resource Allocation and Leveling Using a Threshold Accepting-Based Hyperheuristic Algorithm

In this study we propose a threshold accepting based hyperheuristic for solving in a single run both the resource-constrained project scheduling problem or resource allocation, and the resource leveling problem. Having their roots in the field of artificial intelligence, hyperheuristics operate in the "low-level" heuristics domain rather than in the solutions domain. The hyperheuristic has been implemented within a commercial project management software package. Low-level heuristics operate on the solution domain defined by the priority values that the software uses for resource allocation. A case example from the literature and computational experiments on randomly generated projects demonstrate that the hyperheuristic achieves good performance in a timely manner, improving the results provided by the software. DOI: 10.1061/(ASCE)CO.1943-7862.0000492. (C) 2012 American Society of Civil Engineers.