Model of projection pursuit drought evaluation based on improved artificial fish swarm algorithm of Sanjiang Plain

摘要

The model of projection pursuit drought evaluation based on artificial fish swarm algorithm was proposed to optimize the function and seek the optimum projection for solving the problems such as discrete evaluation grade and low resolved results obtained by using other methods. In order to avoid local optimization and improve global search ability and convergence speed of the algorithm, the adaptive artificial fish step and crowded degree factor were applied to optimize artificial fish swarm algorithm. The departure ratio of rain-fall, Z index and homogenization of rain-temperature were selected to establish evaluation model of Sanjiang Plain based on above methods. The results showed that model effectively avoided the incompatibility of single index. It was feasible to evaluate actual drought situation, and provide a new way for drought evaluation.