LSSVM Based Covert Attack Method Research on Pressure Control System of Pressurizer in Nuclear Power Plant
-
摘要: 隐蔽攻击对大型核电厂稳压器压力控制系统的安全、稳定、高效运行构成了严重威胁，实现隐蔽攻击的关键是建立高精度的对象估计模型。本文提出了一种基于最小二乘支持向量机（LSSVM）的隐蔽攻击方法，通过LSSVM算法进行系统辨识，获得稳压器受攻击区域高精度的估计模型，随后利用该估计模型结合隐蔽控制器实施攻击，实现稳压器压力控制系统在无噪声、有噪声及包含非线性环节情况下的隐蔽攻击。仿真结果表明，该攻击方式对稳压器压力控制系统造成一定破坏的同时具有高度的隐蔽性。
-
关键词:
- 稳压器压力控制系统 /
- 隐蔽攻击 /
- 最小二乘支持向量机（LSSVM）算法 /
- 噪声分析
Abstract: Covert attacks pose a serious threat to the safe, stable and efficient operation of the control system of pressurizer in a large nuclear power plant. The key to realize covert attack is to establish a high-precision object estimation model. This paper proposes a covert attack method based on least square support vector machine (LSSVM). The system identification is carried out by LSSVM algorithm, and the high-precision estimation model of the attacked area of the pressurizer is obtained. Then the estimation model is combined with the hidden controller to realize the covert attack of the pressurizer pressure control system in the case of no noise, noise and with nonlinear links. The simulation results show that the attack mode not only causes some damage to the pressure control system of the pressurizer, but also has a high degree of covertness.-
Key words:
- Pressurizer pressure control system /
- Covert attack /
- LSSVM /
- Noise analysis
-
图 1 反馈型隐蔽攻击结构
n—噪声信号值；y—被控对象的输出信号值；u —被控对象的输入信号值; η—隐蔽攻击器反馈输出信号值；ηref—隐蔽控制器的设定值；Π—辨识模型；Θ—隐蔽控制器数学模型；μ—隐蔽攻击器前馈输出信号值；ycm—实际控制器接收到的反馈信号值；yref—实际控制器的设定值；uc—实际控制器发出的控制指令；C—实际控制器数学模型；D—异常检测器数学模型
Figure 1. Feedback Covert Attack Structure
图 4 不同系统模型下闭环控制系统的响应结果
Figure 4. Response Results of Closed-Loop Control System under Different System Models
图 5 有攻击和无攻击情况下被控对象输出响应结果
Figure 5. Output Response Results of Controlled Object with and without Attack
图 6 不同攻击情况下稳压器观测器捕获数据
Figure 6. Data Captured by Pressurizer Observer under Different Attacks
图 7 不同攻击情况下实际控制器反馈信号差异值
Figure 7. Feedback Signal Difference of Actual Controller under Different Attacks
图 8 不同攻击情况下稳压器观测器捕获的数据
Figure 8. Data Captured by Pressurizer Observer under Different Attacks
图 9 有噪声情况下控制器反馈信号差异值
Figure 9. Difference Value of Controller Feedback Signal in the Presence of Noise
图 10 不同攻击情况下稳压器观测器捕获的数据
Figure 10. Data Captured by Pressurizer Observer under Different Attacks
表 1 不同系统模型下ycm和uc的拟合误差对比
Table 1. Comparison of Fitting Errors between ycm and uc under Different System Models
辨识模型 对比信号 平均绝对值误差/MPa 最大误差/MPa LSSVM ycm 0.0018208 0.0107367 uc 0.1886899 1.0912147 SVR ycm 0.0527724 0.3051055 uc 6.2264503 32.1575928 
下载: 导出CSV
表 2 不同攻击情况下差异值的数学统计结果
Table 2. Mathematical Statistical Results of Difference Values under Different Attacks
模型 差异值最
大值/MPa差异值均
值/MPa差异值方
差/MPa2SVR算法模型 0.0649761 0.0108955 0.0006183 LSSVM算法模型 0.0015456 3.4836334×10−5 4.4064589×10−7
下载: 导出CSV
-
[1] SMITH R S. Covert misappropriation of networked control systems: presenting a feedback structure[J]. IEEE Control Systems Magazine, 2015, 35(1): 82-92. doi: 10.1109/MCS.2014.2364723 [2] DE SÁ A O, DA COSTA CARMO L F R, MACHADO R C S. Covert attacks in cyber-physical control systems[J]. IEEE Transactions on Industrial Informatics, 2017, 13(4): 1641-1651. doi: 10.1109/TII.2017.2676005 [3] DE SÁ A O, DA COSTA CARMO L F R, MACHADO R C S. Evaluation on passive system identification and covert misappropriation attacks in large pressurized heavy water reactors[C]//2018 Workshop on Metrology for Industry 4.0 and IoT. Brescia: IEEE, 2018: 203-208. [4] 张妍,樊登宁,黄宇,等. 大型加压重水反应堆隐蔽攻击方法研究[J]. 核动力工程,2021, 42(3): 132-139. [5] XU Z B, WU J, QUAN Z T, et al. Model and simulation study on pressurizer pressure system[C]//Proceedings of 2015 International Conference on Electrical, Automation and Mechanical Engineering (EAME 2015). Phuket: Atlantis Press, 2015: 420-423. [6] 刘向杰, 方月建. 基于移动最小二乘算法的核电站稳压器机理建模与仿真[C]//第30届中国过程控制会议(CPCC 2019). 昆明: 中国自动化学会过程控制专业委员会, 2019. [7] WANG P F, HE J J, WEI X Y, et al. Mathematical modeling of a pressurizer in a pressurized water reactor for control design[J]. Applied Mathematical Modelling, 2019, 65: 187-206. doi: 10.1016/j.apm.2018.08.006 [8] 钱虹,苑源. 核电站稳压器压力非自衡系统的预测控制[J]. 核动力工程,2019, 40(3): 87-92. [9] 唐瑶. 基于GMFAC的核电站稳压器压力优化控制[J]. 核科学与工程,2018, 38(3): 487-493. doi: 10.3969/j.issn.0258-0918.2018.03.021 -
计量
- 文章访问数: 194
- HTML全文浏览量: 136
- PDF下载量: 29
- 被引次数: 0
