样条小波用于在线误跟踪检测及其特性分析

Mistracking Detection On-Line with Spline Wavelet andIts Characteristic Analysis

本文提出的小波变换误跟踪检测方法 ,以 L ipschitz指数作为波门有重叠的交叉目标数据分离的依据 ,用推导出的在线快速小波分解算法提取局部 L ipschitz指数特征 ,从而提高数据处理的实时性。理论分析与仿真结果表明 ,采用三次样条小波进行误跟踪的在线检测 ,不仅能达到精确定位时间序列的阶跃边缘 ,而且在加性噪声和乘性噪声环境下均可以做到无偏检测。

This paper presents a mistracking detection method with wavelet transformation. The data of the crossing targets are separated by means of Lipschitz exponent when tracking gates are aliasing. The realtime wavelet decomposition is derived and employed in the extraction of local Lipschitz exponent to get better property of realtime processing. It is demonstrated that the mistracking detection algorithm with spline wavelet is accurate for the step edge of a time series and is unbiased under the situation of both additive and multiplicative noise.