Search academic texts

Information × Registration Number 0224U002509, 0122U000817 , R & D reports Title Development the innovative methods and models of designing the industry-oriented information and communication systems for upgrading the digital industrial infrastructures popup.stage_title Head Beshlei Halyna V., Кандидат технічних наук Registration Date 17-02-2024 Organization Lviv Polytechnic National University popup.description2 Comprehensive solutions have been proposed to enhance the efficiency of industrial devices and ensure data transmission security in the context of manufacturing digitization. An intelligent monitoring system has been developed to detect anomalies in the information system traffic, employing a combination of consensus, signature, statistical, and fractal analyses. This system is capable of identifying potential external attacks, network malfunctions, and short-term channel failures, ensuring high service quality and network protection. A method for detecting abnormal behavior in complex industrial production systems has been developed, utilizing recurrent neural networks for exchanging experience among autonomous controlled vehicles. This method improves the accuracy of failure prediction and troubleshooting in industrial mechanisms, while also enabling the early detection and elimination of malfunctions, thus enhancing production efficiency and reducing maintenance costs. For the first time, a method of secure federated learning has been developed, employing encrypted neural network weights for information exchange among intelligent autonomous vehicles. This method ensures a high level of information confidentiality, preventing data reproduction without special access keys. Furthermore, the optimization, structuring, and protection of data have been improved using modifications of matrix decomposition algorithms and federated learning. This enables effective reduction of data dimensionality and protection of information from unauthorized access, leading to rapid anomaly detection and increased data processing efficiency in industrial systems.  Product Description popup.authors Hordiichuk-Bublivska Olena V. Medvetskyi Mykhailo B. Peleh Nazar V. Pyrih Yuliia V. Shkoropad Yurii I. Shubyn Bohdan P. popup.nrat_date 2024-02-17 Close