Trustworthiness in Industrial IoT Systems Based on Artificial Intelligence
Trustworthiness in Industrial IoT Systems Based on Artificial Intelligence
复制标题
基于人工智能的工业物联网系统的可信度。
DOI：
10.1109/tii.2020.2994747
复制
发表时间：
2021-02-01
影响因子：
12.3
通讯作者：
Lv, Haibin
中科院分区：
文献类型：
--
作者：
Lv, Zhihan;Han, Yang;Lv, Haibin
在新一代网络信息物理系统（CPS）支持下发展起来的智能工业环境，可以实现信息资源的高度集中。为了对CPS的可靠性进行分析和量化，提出了一种CPS可靠性自动在线评估方法。基于机器学习的知识构建了评价框架，设计了在线排名算法，实现了真实的在线分析和评价。预防措施能及时采取，系统能正常连续运行。其可靠性大大提高。基于互联网和物联网的可信性，分析了一种典型的基于时空相关检测模型的CPS控制模型，确定了综合可靠性模型分析策略。基于此，本文提出了CPS可信鲁棒智能控制策略和可信智能预测模型。通过仿真分析，得到了攻防资源的影响因素和分布式协同控制的动态过程。分布式协同控制模式下的CPS防御者可以根据CPS攻防环境进行指导和选择合适的防御资源投入。
The intelligent industrial environment developed with the support of the new generation network cyber-physical system (CPS) can realize the high concentration of information resources. In order to carry out the analysis and quantification for the reliability of CPS, an automatic online assessment method for the reliability of CPS is proposed in this article. It builds an evaluation framework based on the knowledge of machine learning, designs an online rank algorithm, and realizes the online analysis and assessment in real time. The preventive measures can be taken timely, and the system can operate normally and continuously. Its reliability has been greatly improved. Based on the credibility of the Internet and the Internet of Things, a typical CPS control model based on the spatiotemporal correlation detection model is analyzed to determine the comprehensive reliability model analysis strategy. Based on this, in this article, we propose a CPS trusted robust intelligent control strategy and a trusted intelligent prediction model. Through the simulation analysis, the influential factors of attack defense resources and the dynamic process of distributed cooperative control are obtained. CPS defenders in the distributed cooperative control mode can be guided and select the appropriate defense resource input according to the CPS attack and defense environment.