一种新的双基阵纯方位机动目标跟踪算法

New Algorithm for Bearings-only Maneuvering Target Tracking Using Bistatic System

针对方位测量方程的非线性问题,提出了一种基于改进的自构建神经模糊网络(SON-FIN)的双基阵纯方位机动目标跟踪算法。该算法首先利用目标方位角信息,对其进行目标特征提取,然后利用小波基函数所具有的时频局部分析能力,提出了用小波基函数来代替高斯基函数这一策略。仿真结果表明,神经模糊网络可实现双基阵纯方位机动目标的实时跟踪,并且改进后的网络跟踪性能要优于高斯基网络。

In terms of the nonlinearity of the bearing measurement equation, an improved self-constructing neural fuzzy inference network （SONFIN） based algorithm is studied and applied to the case of the bearings-only maneuvering target tracking using bistatic sonar system. First the proposed algorithm extracts target features based on the bearing information, then a new strategy is proposed, i.e. the gauss membership function is replaced by wavelet basis function because of its good time- frequency analysis ability. Simulation results show that the SONFIN can track the maneuvering target instantly and additionally, and that the improved SONFIN has better tracking performance than that of SONFIN with gauss membership function.