CAJ | 학술논문

指挥自动化网络安全评价能够全面了解网络运行状况，及时采取防范措施，对确保指挥自动化系统发挥战斗力起至关重要的作用。从管理、环境、硬件、软件及数据等方面出发进行安全性分析，对各项指标进行专家评分。根据评分结果利用模糊层次和神经网络建立网络安全评价体系，通过神经网络存储记忆功能，快速对新数据作出响应，得出当前指挥自动化网络综合安全指数。分别对梯度下降BP算法和LMBP算法进行仿真分析，得出LMBP算法在收敛速度和较小误差方面效果更佳。
지휘자동화망락안전평개능구전면료해망락운행상황，급시채취방범조시，대학보지휘자동화계통발휘전두력기지관중요적작용。종관리、배경、경건、연건급수거등방면출발진행안전성분석，대각항지표진행전가평분。근거평분결과이용모호층차화신경망락건립망락안전평개체계，통과신경망락존저기억공능，쾌속대신수거작출향응，득출당전지휘자동화망락종합안전지수。분별대제도하강BP산법화LMBP산법진행방진분석，득출LMBP산법재수렴속도화교소오차방면효과경가。
Security assessment of command automation network which can comprehensively master the running status of the network and take immediate precautions playing an extremely important role on developing the combat effectiveness of the command automation system. The management,environment, hardware,software,data security etc and evaluation of each indicator are analyzed by expert. Fuzzy Analytical Hierarchy Process and neural network are used to build the network security evaluation system according to the score of relevant influencing factors,Comprehensive Index of Command automation network is got through the neural network’s responding to new data immediately with associational memory function,in which the traditional BP and LMBP algorithm are respectively used. Finally through contrast and analysis of simulation results,it proves that the LMBP algorithm achieves a better effect on the rate of convergence and the accuracy.