Heuristic solutions for the economic manpower shift planning problem (original) (raw)

Consideration is given to the economic manpower shift planning (EMSP) problem, a capacity planning problem which seeks the workforce needed in each workday shift over a given planning horizon in order to complete a specific set of jobs at minimum cost. Since the problem has been shown to be NP-hard, we establish an optimal solution lower bound and develop heuristic algorithms for its solution. Specifically, two greedy algorithms and one meta-heuristic algorithm are developed. The meta-heuristic, which constitutes a hybrid genetic algorithm, combines the advantages of a micro-genetic algorithm (μGA) for fast solutions evolution with a variable neighbourhood search (VNS) technique for improving these solutions. Experiments over three different operating environments were performed to assess the heuristics efficiency. Comparative results from a standard integer linear programming optimiser and the lower bound proposed here show the meta-heuristic to perform very well in terms of solution quality and CPU-time requirements, particularly for large-sized problems (where it clearly outperforms both greedy algorithms). [