IWOA-BP neural network for cyber-security assessment in industrial IoT-enabled smart manufacturing systems

This study proposes a cyber-security assessment in industrial Internet of Things (IIoT)-enabled smart manufacturing systems, Improved Whale Optimization Algorithm-Back propagation Neural Network (IWOA-BP), based on swarm intelligence optimization and machine learning. This study aims at the poor convergence of the traditional Whale Optimization Algorithm (WOA) algorithm and proposes improved strategies from multiple aspects to construct the IWOA algorithm. In view of the problem of network performance degradation caused by random initialization of the initial parameters of the traditional BP network, this study uses the IWOA algorithm to improve it to construct the IWOA-BP algorithm. This study uses the IWOA-BP algorithm for the network data security assessment. A systematic experiment is conducted for this method in this study, and the experimental data verifies the superiority of this method. The core contributions of this study are: (1) an IWOA algorithm that integrates chaos search, reverse learning, etc. is proposed to solve the premature convergence problem of WOA; (2) an IWOA-BP model is built and the initial parameters of BP are optimized to improve the accuracy of IIoT security assessment; (3) an 8-dimensional security assessment index system is designed to provide a practical security assessment solution for intelligent manufacturing IIoT systems.