基于多重分形特征的防空雷达目标分类方法

Target classification by surveillance radar based on multifractal features

在介绍常规雷达飞机目标回波数学模型的基础上,利用多重分形测度分析手段,分析了常规雷达飞机目标回波的多重分形特性及其多重分形特征的提取方法,并从模式分类的角度,提出了利用多重分形特征对喷气式飞机、螺旋桨飞机和直升机等三类飞机目标进行分类的方法。实验表明,上述三类飞机的目标回波数据具有显著不同的多重分形特征,所定义的多重分形特征参数可以作为飞机目标分类和识别的有效特征,仿真实验验证了所提方法的有效性。

On basis of introducing the mathematical model of aircraft returns in the conventional radar, by means of the multi- fractal nleasure analysis, this paper analyzed the multifractal characteristic of the aircraft returns as well as the extraction meth- od of their muhifractal signatures, and proposed the classification method for three types of aircraft containing jets, propeller aircrafts and helicopters from the angle of pattern recognition. The experimental analysis shows, the conventional radar returns from three types of aircraft targets, containing jets, propeller aircrafts and helicopters, have significantly different multifractal characteristic curves, and the defined multifractal characteristic parameters can be used as effective features for aircraft target classification and recognition. The simulation validated the validity of the proposed method.