基于多传感器数据融合的档案存放安全性预警系统研究

Research on archives storage security early warning system based on multi-sensor data fusion

针对目前电力档案存放安全性预警工作存在的信息反馈滞后、预警干预机制不完善、风险管控能力较弱等问题,开发了一款基于多传感器数据融合的档案存放安全性预警系统。通过构建面向多维参数的传感器集群网络,引入改进过的多传感器数据融合算法并融入预警分类器建立面向电力档案的动态预警模型,在MATLAB2016b环境下进行模型效能仿真验证,较好解决了多维应用背景下的电力档案存放安全性预警过程中人力耗费与实际效能失衡、信息反馈滞后等问题,具有动态预警精准、泛化预警能力强、风险管控变化趋势预估效率高等优势。以国家电网某省电力公司为效能评价载体,利用VS2012平台开发了验证环境并对模型进行了实证分析,分析结果表明,所提模型可以实现全方位的电力档案存放安全性动态预警,在预警实时性、模型拟合度、信息过载处理效率等方面具有明显优势。

Aiming at the problems of lagging information feedback,imperfect early warning intervention mechanism and weak risk management and control ability in the current security early warning work of power archives storage,a security early warning system for archives storage based on multi-sensor data fusion is developed.By constructing sensor cluster network for multi-dimensional parameters,introducing improved multi-sensor data fusion algorithm and integrating early warning classifier,the dynamic early warning model for power archives is established.The effectiveness of the model is verified by simulation under the environment of MATLAB 2016b,which can better solve the imbalance between manpower consumption and actual efficiency in the process of early warning for power archives storage security under the background of multi-dimensional application.It has the advantages of accurate dynamic early warning,strong generalization early warning ability and high efficiency of risk management and control change trend prediction.Taking aprovincial power company of the State Grid as the carrier of efficiency evaluation,a verification environment is developed using VS2012 platform and the model is analyzed empirically.The analysis results show that the model proposed in this paper can realize all-round dynamic early warning of power file storage security,and has obvious advantages in real-time warning,model fitting degree and information overload processing efficiency.