An Improved Genetic Algorithm and Its Performance Study

As an effective global search method, genetic algorithm has obvious advantages. But it usually has problems of premature convergence and local optimum in practical application. According to this, a new algorithm with improved selection, crossover and mutation is proposed. Through the simulation experiments, the improved algorithm shows its faster convergence and better stability. It is valid which can not only avoid local optimum but also enhance the accuracy and probability of convergence.