Research on FJSP based on CUDA parallel cellular particle swarm optimization algorithm

This paper raised a Cellular particle swarm optimization, (CPSO) algorithm based on GPU parallel, which could reduce the computing time for processing amounts of data and solving Flexible job shop scheduling problem(FJSP) complex problems. The implementation of proposed algorithm combines the GPU parallel technology and traditional CPSO algorithm. The method accelerates the convergence rate of the particle swarm by using a large amounts of GPU threads to process each particle. In the experimental part, we use the benchmark data to test the performance of proposed algorithm. Comparing with serial CPSO, the proposed method can achieve higher computing speed ratio under the premise that ensure the optimal solution.