A parallel sensitive area selection-based particle swarm optimization algorithm for fast solving CNOP

Abstract

Recently, more and more researchers apply intelligent algorithms to solve conditional nonlinear optimal perturbation (CNOP) which is proposed to study the predictability of numerical weather and climate prediction. The difficulty of solving CNOP using intelligent algorithm is the high dimensionality of complex numerical models. Therefore, previous researches either are just tested in ideal models or have low time efficiency in complex numerical models which limited the application of CNOP. This paper proposes a sensitive area selection- based particle swarm optimization algorithm (SASPSO) for fast solving CNOP. Meanwhile, we adopt the self-adaptive dynamic control swarm size strategy to SASPSO method and parallel SASPSO with MPI. To demonstrate the validity, we take Zebiak-Cane (ZC) numerical model as a case. Experimental results show that the proposed method can obtain a better CNOP more efficiently than SAEP [1] and PCAGA [2] which are two latest researches on intelligent algorithms for solving CNOP.