| Corrosion and scaling cause serious economic losses and waste of water resources incirculating cooling water system, so it is particularly important to treatment with circulatingcooling water. We have to monitor of corrosion and scaling in real time so as to treatment withcirculating cooling water. In recent years, with the development of information technologyand automation in Petroleum Chemical Corporation, a lot of data can be got from the system.We can establish water quality model of circulating cooling water system with the research ofwater quality characteristics based on the data. It has important theoretical significance andapplication value, as well as the focus of this paper.In this paper, the data from professional information management system of industrialwater is collected and pre-processed. Analysis and propose of six kinds of factors which affectthe corrosion and scaling. Train samples and test samples are provided by PetroleumChemical Corporation The model structure of BP neural network was determined. Throughthe contrast of convergence speed and system stability, LM algorithm was determined as thetraining algorithm of BP neural network after a lot of test and compare simulation resultswhich using the standard BP algorithm, LM algorithm and so on.In this paper, the water quality model of circulating cooling water system was establishedand then the model was validated and tested. The results show that the model can predict thewater quality of the corrosion and scaling properties with high accuracy. The method provideda way to judge the corrosion and scaling. The judge of the tendency of corrosion and scalingof circulating cooling water with water quality index could be more accurate and reasonable.Finally, the software of intelligent auxiliary analysis for the circulating cooling water waspreliminary designed with Visual C++ which could determine the water quality characteristics,analysis the water data and auxiliary information query. |
PDF Full Text Request