Abstract: The crossover of the conventional Genetic Algorithm（GA） has two problems.On the one hand，it doesn’t make the best performance of both searching new solution space and conserving the good schemata of individuals.On the other hand，it decreases the diversity of population on some level.In this paper，an intelligent crossover is proposed to improve the conventional GA.The new refined GA is applied in the solution of numerical processing problem of mono-pulse velocity measurement system.Numerical results indicate that the new refined GA is pretty good at both precision and calculating speed，and point out that the calculating result is better when the number of signal replicating is 64.

Key words: numerical processing, velocity measurement, mono-pulse, genetic algorithm, target function

摘要： 在传统的遗传算法中，交叉算子探索新的解空间与保存个体中的优良模式的性能不能同时达到最优，且在一定程度上破坏群体多样性。提出智能交叉算子改进了传统遗传算法，用于单个脉冲测速系统的数值处理。结果表明算法具有较高的精度，且计算速度较快；并指出复制信号个数为64时计算效果较优。

关键词: 数值处理, 测速, 单个脉冲, 遗传算法, 目标函数