基于GABP神经网络的工业控制系统风险评估与对策分析

发布时间：2018-07-07 16:07

本文选题：工业控制系统 + 风险评估　；参考：《广西师范学院》2017年硕士论文

【摘要】：目前,工业互联网的发展进一步加速了我国“工业化”与“信息化”的融合,工业生产中已大范围的开始自动化生产。各种相连接的工业自动化设备组成庞大的工业自动化生产网络。但是,由于在使用中只注重工业应用的可用性和实时性要求,设备在设计安装过程中忽略了自动化设备的远程控制的潜在危险。随着互联网的发展,各种软件协议、操作平台以及管理的漏洞被挖掘出来。这是近年来工业控制系统安全事件频发的主要原因。特别是在能源、交通和关乎国家安全的关键部门的小漏洞而导致的事故,不仅会造成财产损失和人员伤亡,而且可能会威胁国家安全,所以针对工业控制系统的安全风险的评估显得十分重要。然而目前国内外针对工业控制系统的风险研究尚未有统一的评估因素以及量化评估方法。所以本文使用文献分析法,通过分析国内外已经颁布的关于ICS安全标准、规范以及相关研究文献,提出了风险评估中的相关风险因素。并通过神经网络迭代得出各评估因素的权重,再依据权重进行模糊综合评价计算风险评估的量化值和评估等级。本文针对上面论述中ICS安全风险评估现状做了下面工作:1、本文根据我国国内的工业控制系统的现状,以及我国分级保护措施,并借鉴IEC62442和NIST SP-800,提出了基于损失的ICS分级保护的定级标准。2、基于我国颁布的GB/T30976.1的安全风险评估规范,提出使用GABP神经网络的方法,将GB/T30976.1中安全风险评估因素作为风险评估的基本因素。根据GABP神经网络的训练结果得出风险评估因素权重。再基于得出的风险评估要素的权重,使用模糊综合评估方法,得出ICS的安全风险的风险量化值。3、结合上述评估方法和流程,使用PHP+Apache+MySQL的web技术进行系统设计。将上述方法在系统中实现。形成一个可扩展、可维护的风险评估模块。将该模块作为广西工业控制系统安全监测平台的风险评估方法。使数据收集和风险评估作为一个完整的系统,为区内ICS风险评估提供评估工具。
[Abstract]:At present, the development of industrial Internet has further accelerated the integration of "industrialization" and "information" in China. The industrial production has begun to automate the production of automation in a large range. All kinds of connected industrial automation equipment make up a huge industrial automation production network. However, only the availability and real time of industrial applications are paid attention to. With the development of the Internet, various software protocols, operating platforms, and management vulnerabilities are excavated with the development of the Internet. This is a major cause of the frequent security events of industrial control systems in recent years, especially in energy, transportation, and national security. Accidents caused by small loopholes in the key departments will not only cause property loss and casualties, but also threaten national security. Therefore, it is very important to assess the safety risk of industrial control systems. However, there is no unified assessment factor and quantity for the risk study of industrial control systems at home and abroad. Based on the analysis of ICS safety standards, specifications and related literature, this paper puts forward the risk factors in the risk assessment by using the literature analysis method, and obtains the weight of the evaluation factors through the neural network iteration, and then evaluates the risk assessment by fuzzy comprehensive evaluation according to the weight. In this paper, the following work is done on the status of ICS security risk assessment: 1. According to the status of domestic industrial control system in China, and the classification protection measures in our country, and using IEC62442 and NIST SP-800 as reference, the standard.2 based on the loss based ICS classification protection is proposed, which is based on the promulgation of our country. GB/T30976.1's safety risk assessment standard, the method of using GABP neural network, the security risk assessment factor in GB/T30976.1 as the basic factor of risk assessment. According to the training results of GABP neural network, the weight of risk assessment factor is obtained. Then based on the weight of the risk assessment factor, the fuzzy comprehensive evaluation method is used. The risk quantification value.3 of ICS security risk is obtained, and the system design is carried out using the web technology of PHP+Apache+MySQL. The above method is implemented in the system. A scalable and maintainable risk assessment module is formed. The module is used as a risk assessment party for the safety monitoring platform of the Guangxi industrial control system. Data collection and risk assessment as a complete system to provide an assessment tool for ICS risk assessment in the area.
【学位授予单位】：广西师范学院
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP183;TP273

【参考文献】