A Fine-grained Parallel Artificial Bee Colony Algorithm Based on GPU-Acceleration

Artificial Bee Colony(ABC) algorithm still needs plenty of computing time when it processes large-scale data and high complexity problems.In order to solve those problems in practice,we proposed a parallel ABC algorithm based on GPU acceleration with OpenCL parallel programming model.The process of parallel ABC is each employee bee mapping to OpenCL work group and onlooker using local roulette wheel selection,which makes parallel ABC speed up.The simulation results show that parallel ABC algorithm achieves an effective optimization and speeds up its running.