• A memory-integrated artificial bee algorithm for heuristic optimisation

      Bayraktar, Tugrul (University of Bedfordshire, 2014-02)
      According to studies about bee swarms, they use special techniques for foraging and they are always able to find notified food sources with exact coordinates. In order to succeed in food source exploration, the information about food sources is transferred between employed bees and onlooker bees via waggle dance. In this study, bee colony behaviours are imitated for further search in one of the common real world problems. Traditional solution techniques from literature may not obtain sufficient results; therefore other techniques have become essential for food source exploration. In this study, artificial bee colony (ABC) algorithm is used as a base to fulfil this purpose. When employed and onlooker bees are searching for better food sources, they just memorize the current sources and if they find better one, they erase the all information about the previous best food source. In this case, worker bees may visit same food source repeatedly and this circumstance causes a hill climbing in search. The purpose of this study is exploring how to embed a memory system in ABC algorithm to avoid mentioned repetition. In order to fulfil this intention, a structure of Tabu Search method -Tabu List- is applied to develop a memory system. In this study, we expect that a memory system embedded ABC algorithm provides a further search in feasible area to obtain global optimum or obtain better results in comparison with classic ABC algorithm. Results show that, memory idea needs to be improved to fulfil the purpose of this study. On the other hand, proposed memory idea can be integrated other algorithms or problem types to observe difference.