Analysis Of Operation Status Of Complex Electromechanical System Based On System Map

In the country’s industrial production activities,the process industry represented by the energy heavy chemical industry,power enterprises,and the pharmaceutical industry occupies an important position.The process industry production system contains a large number of equipment,and the equipment is complicated with each other,which has the characteristics of non-linearity,high coupling and huge.The distributed control system(DCS),the core control part of the process industry system,contains a large variety of sensors,thousands of loops for monitoring and control,and a huge amount of data collection.DCS stores real-time production data collected during the operation of the system,which is the DCS data set.The DCS data set consists of thousands or even tens of thousands of time series,which contains all the status information during the operation of the system.The analysis of DCS data set is essentially the analysis of massive high-dimensional nonlinear time series.Because the DCS monitoring data set has the characteristics of massiveness,nonlinearity and high coupling,it is difficult to analyze it by establishing mathematical analytical expressions.The traditional analysis method of system operation status based on DCS data set is a data-driven multivariate statistical analysis method represented by Principal Component Analysis(PCA).The core idea of this type of method is to achieve dimensionality reduction of high-dimensional data by filtering key variables.Due to the high coupling characteristic of the process industry system,there is no such thing as critical data,and a lot of useful information is often ignored in the screening process,which affects the accuracy of the analysis results.Therefore,how to analyze all the data in the DCS data set without dimensionality reduction,dig out the information about the health of the system contained therein,and grasp the health of the overall system is a huge challenge facing the process industry.To address this challenge,this paper first puts forward the concept of system state characteristic spectrum.By formulating coloring rules,using different colors to represent the degree of abnormality of the data,and replacing the changes of data with color changes,the DCS data set is converted into a flat digital color image,which is called the state characteristic spectrum of the DCS system.By observing the presentation position and distribution range of colors representing different degrees of abnormality in the characteristic spectrum of the system state,some key information that does not seem to be regular and hidden in the massive data is extracted.Understand the operational health status of each variable in each sampling period,and further grasp the overall operational health status of the system.Secondly,this paper applies digital image processing technology to quantitatively analyze the system state characteristic spectrum,and realizes the rapid tracing of system faults and the trend analysis of overall operational health.This method can issue an early warning before a major accident occurs in the system,and provide data guarantee for enterprise safety production decision-making.Finally,the full text takes the Tennessee simulation data set and the actual DCS data set of a chemical company as examples to verify the effectiveness and feasibility of the proposed method from both qualitative and quantitative aspects.