Corrosion Prediction Of Atmospheric And Vacuum Distillation Unit Based On Big Data Analysis

Atmospheric and vacuum distillation unit is the most important equipment in petrochemical industry.The improvement of anti-corrosion technology of atmospheric and vacuum distillation unit can ensure the stable operation of equipment and improve product quality,and achieve the effective utilization of financial resources and material resources and guarantee the safety of workers.Providing efficient and accurate anti-corrosion strategies and theoretical support based on big data technology for atmospheric and vacuum distillation unit is a major hot issue to be solved in the industry.Through the in-depth analysis and research on the massive historical data of atmospheric and vacuum unit,the corrosion circuits division,abnormal or degraded conditions diagnosis and identification,corrosion prediction model and corrosion database system are realized in the operation process of the equipment,which provides an innovative and complete corrosion detection process for atmospheric and vacuum unit.The main work of this paper is as follows:?1?Through the research and analysis of atmospheric and vacuum distillation unit and its main corrosion mechanism,the relevant data are collected and indexes of each corrosion circuit are determined.?2?Based on big data technology,the corrosion prediction system analysis method of atmospheric and vacuum distillation unit is proposed and verified by Tennessee Eastman process.Among them,the analysis methods include data preprocessing,linear correlation analysis,K-means clustering,support vector machine regression prediction model of corrosion indexes.The Tennessee Eastman process is a representative and convincing simulation industrial process.The advanced and practical methods are verified by the data of unknown fault types.A feature selection method based on support vector machine is proposed to solve the problem of fault free data in chemical process data sets.?3?The corrosion data analysis and prediction of atmospheric and vacuum distillation unit.Firstly,the corrosion big data set that can be directly used for analysis is established;then,the influencing factors of indexes are identified by correlation analysis.The influencing factors show that p H value and Fe²⁺ion concentration have negative correlation,steam quantity and temperature are the main factors that affect pipeline corrosion,and the results conform to the corrosion mechanism knowledge.K-means clustering analysis method is used to analyse the corrosion.The massive historical data without labels can be quickly classified,and the advantages and disadvantages of each working condition can be determined according to the amount of various data and the stability of the data,so as to establish an integrity operation window for the indexes.Finally,the prediction models are established for the indexes through the support vector machine regression method,and the mean square error between the predicted index quantity and the predicted value and the actual value is output,which is used to monitor the change trend of each working condition of each circuit on line,predict whether an exception will occur.?4?The database system of atmospheric and vacuum distillation unit is developed,which realizes the functions of adding,modifying,deleting and checking data,and completes the batch import?Excel format?and export of data generated in the operation process of the equipment.In this paper,through a complete idea of data analysis and corresponding data mining technology,we diagnose,identify and predict the abnormal or fault condition data during the operation of atmospheric and vacuum distillation unit,and establish an intelligent system platform based on big data technology,which provides scientific and effective theoretical basis and technical support for enterprise anti-corrosion.