Robustness analysis of smart manufacturing information systems

Robustness plays a vital role in the operation of smart manufacturing information systems (SMISs). This article aims to establish a robust analysis theory, to explain how companies can perform robust analysis of information systems in real time during normal operations and perform robust control over unstable information systems. (1) Using of maximum variance unfolding (MVU) algorithm to fuse secondary index data into primary index data. (2) Using data-driven modeling to establish a robust evolutionary dynamics model of SMISs. (3) Using synergetic and self-organization theory to analyze the dynamic model of the robustness of SMISs. (4) Using dynamic programming and stability theory to conduct robust optimal control. (1) The primary indicator data can maintain the characteristics of the secondary indicator data by using the improved MVU method; (2) Dynamic programming and stability theory used to conduct optimal robust control can give investment strategy to improve the robustness of SMISs; (3) The combination of the synergetic theorem and machine learning can quantitatively describe the SMISs, effectively solve the complex calculation problems caused by the increase of the variable dimension. (1) This paper proposes for the first time the use of maximum variance unfolding (MVU) algorithm to fuse secondary index data into primary index data; (2) for the first time combining complex system theory with machine learning and optimization theory to analyze the robustness of information systems.