基于证据推理的动态系统安全性在线评估方法

发布时间：2018-02-09 14:35

本文关键词： 动态系统安全性在线评估证据推理 Volterra滤波器权重信息融合　出处：《自动化学报》2017年11期 　论文类型：期刊论文

【摘要】：综合考虑动态系统历史记录、当前状态以及未来退化趋势信息来对其安全性进行在线评估是极其重要的.本文提出了一种基于证据推理(Evidential reasoning,ER)的安全性在线评估方法.该方法先融合多个安全性指标获得各个时刻的安全性状态,而后融合系统"历史"、"当前"、"未来"时刻的安全性状态,评估得到系统的综合安全性水平.首先,建立了基于三阶Volterra滤波器的在线预测模型,预测指标未来信息;然后,建立了指标最优自适应权重求取模型,计算并更新指标实时权重;最后,提出了基于证据推理方法的融合框架,对"历史"、"当前"、"未来"时刻的信息进行融合,得到系统当前时刻的综合安全性评估结果.通过对某惯性平台系统的安全性评估实例验证了所提方法的有效性.
[Abstract]:Taking into account the history of dynamic systems, It is very important for present state and future degradation trend information to evaluate its security online. In this paper, a method of online security evaluation based on evidential reasoning is proposed. The metric obtains the security state at all times, Then the system "history", "present" and "future" security status are fused, and the comprehensive security level of the system is evaluated. Firstly, an online prediction model based on third-order Volterra filter is established to predict the future information of the index. The optimal adaptive weight model of the index is established to calculate and update the real-time weight of the index. Finally, a fusion framework based on evidence reasoning method is proposed to fuse the information of "history", "present" and "future". The comprehensive security evaluation results at the present time of the system are obtained. The effectiveness of the proposed method is verified by an example of the safety evaluation of an inertial platform system.
【作者单位】：火箭军工程大学控制工程系;火箭军工程大学管理工程系;
【基金】：国家自然科学基金(60736026,61370031,71601180,61773388) 飞行器海上测量与控制联合实验室开放基金(FOM2014OF14,FOM2015OF017) 中国博士后科学基金面上项目(2015M570847) 陕西省自然科学基金项目(2015JM6354) 装备预研基金(9140A19030314JB47276)资助~~
【分类号】：TP202

【相似文献】