Based On Parallel Particle Swarm Optimization Algorithm In Sorting Techniques Applications

Supercomputer (High-performance computers) is a class of the most powerful, fastest, largest storage capacity computer. It used for national high-tech fields and cutting-edge technology research, it's not only a country's economic and technological power of expression, but also to promote technology and economic development, social progress and an important tool for national security. Has become the strategic high ground of countries in the world competing.Parallel computing is synonymous with high-performance computing, because any of the high-performance computing is inseparable from parallel technology. It's very important that how to make high-performance parallel computer system fully in the national economy, scientific research and society play a role, and also attracted attention of many researchers. MPI is the most important message-passing parallel programming platform, has become the standard platform of message passing parallel programming, the mainstream tools of parallel programming. MPICH is one of the most important implementations of MPI, in the LAN connection existing computer by MPICH to compose of high-performance parallel computers group, carry on the Large-scale calculations for complex optimization problems.Particle Swarm Optimization is from the research of human to birds prey behavior. It's based on swarm's intelligence algorithm and optimization tool, and in their own algorithms exist inherent parallelism. But in dealing with large amounts of data and large-scale complex optimization problems, PSO still requires a lot of processing time, while the parallel PSO can significantly reduce problem's solving time and simplify solving the problem. Therefore the study of the parallel PSO has also been many researchers'attention.Sorting technique often used in data processing, and also frequently used operations in the internal computer system. Sorting can be simply understood as:To set a disorderly number of columns into orderly series. Sorting usually consume a large amount of computing time, while the parallel PSO added to the sorting algorithm can effectively improve the solution time. In this paper, through building MPI-based computer cluster environment, implemented based on parallel PSO quick sort, rank sort algorithm. And through on serial sorting algorithm and sorting algorithm based on parallel PSO, comparison and analysis in solution time, solution complexity, and parallel speedup. Illustrates the application of parallel PSO in the sorting algorithm are feasible and effective.