一种基于支持度的抗干扰融合识别方法

A Fusion Recognition Method of Anti-Interference Based on Support

针对传感器在受到干扰的情况下测量数据严重失真的问题,提出一种抗干扰的融合识别方法。该方法的核心思想在于给每个传感器分配相应的权值,权值分配的依据是各传感器的支持度。其他传感器对该传感器的支持度越小,则认为该传感器被干扰的程度越大,被分配到的权值越小。将加权融合后的测量数据输入RBF神经网络获取目标的基本概率赋值,利用D-S证据理论决策规则进行识别,仿真结果表明该方法能有效抑制干扰。

According to the problem that measurement data of the sensors are distorted seriously in the condition of interference, a kind of anti-interference fusion recognition method is proposed. The core idea of this method is distributing corresponding weight to each sensor, and the distribution of weight value is in terms of support of each sensor. When the sensor gets less support from others which means the sensor is jammed more and would get less weight. The basic probability assignment（BPA） of target will be achieved when the measurement data after weighted fusion is as the input of RBF neu- ral network, finally the target will be recognized with the D-S evidence theory of decision rules. The simulation results show that the interference could be restrained by the proposed method.