The Research Of Corrosion Prediction Of The Circulating Cooling Water System

With the benefits of the recycle of cooling water to economic development,the circulating cooling water system is used more frequently in petrochemical enterprises.The cooling water is becoming worse because the increased frequency and other factors such as production conditions.Because of this condition,the circulating cooling water system is likely to break down.The heat transfer efficiency and service life is probably reduced.Serious corrosion is likely to cause perforation and material spill which does harm to economic and water resources and even people's living environment.The current detection means is inconvenience.It can not achieve the comprehensive monitoring and judge the trend of corrosion.Therefore,corrosion prediction of the circulating cooling water system is very important.The research of corrosion prediction can guide the determination of high risk area of corrosion and the prevention measures.It is hard to find a precise maths method for the corrosion prediction of circulating cooling water system.The weighted gray correlation analysis is proposed to select the corrosion prediction parameters of the circulating cooling water system.This method provides scientific theoretical basis for the selection of the parameters of the corrosion prediction and reference for the research and application of similar projects in the field of petrochemical industry.Aiming at the problem that the lower prediction accuracy of the corrosion prediction model of the circulating cooling water system established by BP neural network,the genetic algorithm is introduced to optimize the prediction model.The main contents of this paper are as follows.(1)Learn the research status of corrosion prediction of water quality at home and abroad and prediction methods.Lay the foundation of the research on the corrosion prediction method of the circulating cooling water system.(2)Summarize the basic knowledge of the circulating cooling water system,such as the composition,the common faults and its factors.Understand the basic condition of the researched petrochemical enterprise.(3)The order of the influence factors to the corrosion rate of the circulating cooling water system are analyzed by the weighted gray correlation analysis with the two departments of the researched petrochemical enterprise and tested by another department.The weighted gray correlation analysis provides a method to select parameters for corrosion prediction and it is good to improve the accuracy of corrosion prediction.(4)The BP neural network and genetic algorithm optimizing BP neural network are used topredict the corrosion rate of the circulating cooling water system of the researched enterprise.(5)Summarize the content and the shortage of the paper,and make the next work plan.