Improved AFSA for Solving Intelligent Test Problem

Abstract

Intelligent test paper generation problem is a multi-objective parameter optimization problem under certain constraints, which has been realized by many algorithms. However, the existing intelligent test paper generation mostly adopts a single algorithm, and each algorithm has its own shortcomings. Sometimes it is inevitable to fall into the defects of the single algorithm in the process of test paper generation. Therefore, a mathematical model of intelligent test paper generation is proposed, which combines the advantages of artificial fish swarm algorithm and genetic algorithm to form a hybrid intelligent algorithm. At the beginning of intelligent test paper generation, the artificial fish swarm algorithm is used to quickly approach the test paper goal. In the process of test paper generation, when the optimal individual has no change or extremely small change in consecutive iterations, the genetic algorithm is used to jump the artificial fish individual to improve the convergence speed. The simulation results show that the hybrid intelligent algorithm can effectively optimize the effect of intelligent test paper generation by using a single algorithm alone.