Research of Amelioration of Genetic Algorithms

The paper conducts a research on genetic algorithms, and due to the shortcomings of genetic algorithms, an improved algorithms based on genetic algorithms with simulated annealing algorithm mixing natural cognation exclusion and steepest descent algorithm is put forward. This new algorithm can avoid inbreeding and resolve the premature convergence problem perfectly by keeping population diversity and enlarging space searched. Finally, experimental results of practical data are given to show feasibility and validity of our method.