Cyber-Physical System Based on Large Data Analysis and Research on Deep Fusion of Complex Systems

Abstract

In recent years, the Cyber-Physical System(CPS) has become an important direction of research in the academic and scientific & technological world home and abroad, which is thought as the third wave in the world’s information technology after the computer and the internet. CPS is a multi-dimensional, heterogeneous and deeply integrated open system involving the knowledge of several subjects, including computer, communication and control, etc., as there are obvious differences among the research theories and methods of different subjects and fields, huge challenges have been brought to the applied research on the CPS. A new modeling method named DCDM is put forward: faced with the true environment of CPS, some original discrete data is obtained from tests through sensors and networks, etc., and started from data, a continuous model is established for the system via certain mathematical methods (GMDH), to shorten the error between the discrete data and the continuous model, the method of feedback control is adopted to continuously adjust the difference between both of them, until it decreases to a certain range. DCDM has fundamentally changed the traditional modeling patterns, putting forward an idea of establishing system models starting from discrete data, thus realizing the deep integration of physical events and the computing system through feedback control.