Abstract: It can be challenging to secure the cyber-physical systems (CPS) that power the Internet of Things (IoT), as security measures established for general information and technology operations (IT/OT) may not be effective in a CPS environment. In order to find attacks and techniques created specifically for CPS and more specifically for the control system, this article proposes a two-level common search. This initial phase involves designing a decision tree classifier and new deep learning model for attack detection in a variety of ICS environments. A deep neural network is created for the attack in the second layer. Real data from water pipelines and treatment facilities was used to test the projected model. The outcomes demonstrate that the suggested model performs better than other competing techniques facing comparable challenges.

#