Multireservoir optimisation in discrete and continuous domains

In this paper, the honey-bee mating optimisation (HBMO) algorithm, which is based on the mating procedure of honey-bees in nature, is presented and tested with three benchmark multireservoir operation problems in both discrete and continuous domains. To test the applicability of the algorithm, results are compared with those from different analytical and evolutionary algorithms (linear programming, dynamic programming, differential dynamic programming, discrete differential dynamic programming and genetic algorithm). The first example is a multireservoir operation optimisation problem in a discrete domain with discrete decision and state variables. It is shown that the performance of the model compares well with results of the well-developed genetic algorithm. The second example is a four-reservoir problem in a continuous domain that has recently been approached with different evolutionary algorithms. The third example is a ten-reservoir problem in series and parallel. The best solution obtained is quite comparable with the linear programming solution, and slightly better than the best result reported by other investigators using genetic algorithms. In all three cases, convergence of the solutions in different runs to near-global optima and its rapid convergence rate compared to genetic algorithm demonstrates the applicability and efficiency of the proposed algorithm in solving water-resource optimisation problems in both discrete and continuous domains.