Abstract: An improved adaptive genetic algorithm is presented.The disruptive selection is brought(into) selection operator.The selection probability avoids the emergency of super-individual and maintains the diversity in population.In this algorithm,crossover operator and mutation operator are redesigned in order to increase the convergence speed and avoid premature phenomenon.At the same time,through constructing the new generation within phyle,the searching rate in optimization of the algorithm is increased.(Experimental) results show that the improved adaptive genetic algorithm can sustain diversity in population(efficiently) and find the optimal individual quickly.