An improved adaptive artificial bee colony algorithm

Aiming to the main disadvantages of artificial bee colony algorithm: slow searching speed arisen from the low exploitation of global information and the high possibility to be caught into local convergence, an improved algorithm is proposed in this paper. Two different location updating strategies are separately designed for employed bees and onlooker bees (when they are changed into employed bees). At first, the location (food source) updating formula for employed bees is developed by obtaining the current global optimal information so as to accelerate the convergence speed. In addition, an adaptive strategy is introduced to increase the probability by which the employed bees are changed into scouts, thereby strengthening the randomness and the possibility to find a better solution. After that, a new location updating method is presented for onlooker bees in order to keep the randomness in a reasonable degree. The results of experiments on several test functions show that this new algorithm has a satisfactory performance such as high quality of solutions and quick convergence speed.