Data-driven information modeling method for complex electromechanical system in process industry

The invention discloses a data-driven information modeling method for a complex electromechanical system in the process industry. The data-driven information modeling method for the complex electromechanical system in the process industry is employed. The overall modeling process does not need prior knowledge of physical topology. In the modeling process, an information flow relationship is taken as a reference for analyses of coupling relations among different functional components of the system. By taking the anti-noise property and information loss into full consideration, a symbolization process of the system for monitoring time sequences is optimized. By binary coding and decimal decoding of symbolized phase points, the integration between the symbolization process with automated analyses and calculations is achieved. Therefore, the difficulty in building an information model of the complex electromechanical system in the process industry based on monitored data under the noise condition is resolved. 本发明