适应值的逼近决定子代个体的位置,能快速找出最优个体,但容易陷入局部解.提出了一种改进方法,确定交叉的父代个体,使进化向高适应值区域搜索.通过引入进程实现个体不等几率地产生新个体,保持种群多样性.实验证明该方法可行有效,提高了遗传算法的收敛率和收敛速度. The approximation of fitness value determines the position of the offspring individuals, and can quickly find the optimal individuals, but it is easy to fall into the local solution. An improved method is proposed to determine the crossover parent individuals and make the evolution search to a higher fitness region. Processes to achieve the individual unequal probability of generating new individuals, to maintain population diversity experiment proves that the method is feasible and effective, and improve the convergence rate and convergence rate of genetic algorithm.