论文部分内容阅读
对雷达目标高分辨距离像进行特征加权可以解决高分辨距离像各距离单元因目标姿态变化而导致的稳定性不一致问题。针对已有加权系数求解方法存在的不足,提出一种加权系数优化方法。该方法通过定义目标函数来度量不同雷达目标之间加权高分辨距离像的可分性,并采用梯度下降算法优化加权系数值,从而达到增强高分辨距离像稳定距离单元作用,减小不稳定距离单元影响的目的。基于5种飞机目标模型高分辨距离像的仿真实验表明,该方法可以有效优化加权系数,并提高雷达目标识别率。
The weighted feature of radar target high resolution range images can solve the problem of inconsistent stability caused by the change of target attitude of each distance element of high resolution range. Aiming at the shortcomings of existing methods for solving the weighting factors, a method of weighting coefficient optimization is proposed. By defining the objective function, this method measures the separability of weighted high resolution range images between different radar targets and uses the gradient descent algorithm to optimize the weighted coefficient values, so as to enhance the effect of stable distance elements of high resolution range images and reduce the unstable distance The purpose of the cell impact. Simulation experiments based on five kinds of aircraft target model high resolution range images show that this method can effectively optimize the weighting coefficients and improve the radar target recognition rate.