基于神经网络与D-S证据理论的目标识别

针对目标识别中基本可信度分配需要专家知识在实际中难以实现的问题,提出一种基于神经网络和D-S证据理论相结合的多传感器数据融合的方法.该方法利用D-S理论来表示和处理不精确的、模糊的信息,发挥神经网络的自学习、自适应和容错能力,提高了系统识别率.最后通过实验,利用神经网络来处理证据理论中的基本可信度分配问题,对几种空中目标进行身份估计数据融合,经计算机仿真证实了该方法的有效性.

Target Identification Based on Neural Network and D-S Evidence Theory

In this paper,a method of multisensor data fusion based on neuron network and reasoning(Dempster-Shafer evidence reasoning) is presented.The method can use D-S's evidence to deal with the inaccuracy and fuzzy information.And also it can give full play to self-study of neural net,self-adapting and fault tolerant ability.In this way it has robustness to uncertain information and can improve the system identification rate.Then the D-S evidence is used to fuse the results derived from the neural network at different time.The result of computer simulation shows that the method is efficient and correct.