An Improved Artificial Bee Colony Algorithm for Solving Extremal Optimization of Function Problem

Some defects of artificial bee colony algorithm such as low efficiency, slow convergence rate, may lead to a fall into local optimum. In order to deal with these problems, some thoughts in genetic algorithm are introduced to improve bee colony algorithm in this paper. Specifically, a factor, which used to memorize the current global optimal position, is added to the follower bee operator to improve the global convergence speed and accuracy of the bee colony algorithm. Additionally, inertia factor and search factor are also adopted to change the proportion between the factors that affect the global convergence speed and local convergence speed, in order to accelerate the speed of bee colony algorithm applied in function extremum optimization. The experimental results show that the improved algorithm leads to fast convergence, high efficiency and robust performance.