融合神经网络与证据理论的发射场试验信息处理方案设计

为了提高发射场试验信息处理结果的可靠性，减小冲突信息的影响，通过对发射场试验信息特点的分析及对其中关键技术的研究，提出融合BP神经网络与DS证据理论的信息处理方案。基于神经网络的输出结果建立证据理论的识别框架，通过证据理论相关合成法则对得到的生成融合决策，提高发射场设备状态评估与决策的准确性。

The Design of Launch Site Information Processing Based on The BP Neural Network and D-S Evidence Theory

Based on the analysis on characteristics of launching sites’ testing information and the study on the key technology involved, this paper proposes an information solution incorporated with BP neural network and DS evidence theory in order to increase the reliability of launching sites’ testing information results as well as decrease the impact of conflict information. Moreover, this paper builds a frame of discernment of evidence theory according to the output of neural networks. With related combination rules of evidence theory, it also decides on obtained combination so as to improve the accuracy of assessment and decision upon launching sites’ device status.