CUDA-based Hierarchical Multi-Block Particle Swarm Optimization Algorithm

In order to improve the traditional Particle Swarm Optimization (PSO) algorithm's speed and optimization ability, this paper proposes a new algorithm based on CUDA (Compute Unified Device Architecture) technology which employs the two level PSO, the bottom level PSO and the top level PSO. And in the bottom level, the particles are divided into N groups, each of which will run the PSO and send the best particle to the top level individually to achieve better convergency. And the algorithm applys the CUDA threads to run the above PSO at different levels parallel to accelerate the algorithm speed. The simulation results show that the performance of the algorithm the paper provided is better than that of the traditional PSO.