本文建立了预测碳钢在LiBr溶液中腐蚀速率的神经网络模型.该模型拟合了碱度,温度,LiBr和Na2MoO4浓度变化对碳钢全面腐蚀速率的影响,可用于准确预测不同温度下,在含有不同缓蚀剂的LiBr溶液中的碳钢腐蚀速率,其预测值和实验值完全吻合,为研究溴冷机中金属材料的腐蚀和现场监测提供了新的思路和方法.
Neutral network model has been developed for the computation of corrosion rate of carbon steel in LiBr solutions.The results indicate that the model is capable of reproducing the effects of changes in alkalinity,temperatures,LiBr and Na2MoO4 concentrations on the rates of general corrsion and good agreement between calculated and experimental corrosion rate is obtained.The model can be used to satisfactorily predict the corrosion rate of carbon steel in different concentrations of LiBr solutions containing different inhibitors at different temperatures.It also provides a novel method for corrosion monitoring of metals used in LiBr absorption chiller.
[1] KazuoTanno,ShiroSaaito,MasahikoIto,etal.Corrosioninhibitionofcarbonsteel,copperand70/30cupro＿nickelbybenzotriazoleinconcentratedlithiumbromidesolutionatelevatedtemperature[J].BoshokuGijtsu,1984,33:516～519.
[2] MasahikoItoh,AkiraMinato,MichihikoAizawa,etal.Corrosioninhibitionofcarbonsteelandcupro＿nickelbylithiummolybdateinconcentratedlithiumbromidesolutionatelevatedtemperature[J].BoshokuGijtsu,1984,33:504～508.
[3] TannoK,ItohM,TakahashiT,etal.Thecorrosionofcarbonsteelinlithiumbromidesolutionatmoderatetemperatures[J].CorrosionScience,1993,34(9):1441～1451.
[4] TannoK,M.ItonM,HekiyaH,etal.Thecorrosioninhibitionofcarbonsteelinlithiumbromidesolutionbyhydroxideandmolybdateatmoderatetemperatures[J].CorrosionScience,1993,34(9):1453～1461.
[5] HitoshiYashiro,AkijikoSai,NaoakiKumagai,etal.CorrosionbehaviorofcarbonsteelinconcentratedLiBrsolutionsupto473K[J].Zairoy＿to＿Kankoy,1999..48(6):369～374.
[6] GuoJian＿wei,LiangCheng＿hao.Mechanismoftheformationofpassivestatefilmsoncarbonsteelinahightemperatureandhighconcentrationlithiumbromidesolution[J].MaterialsScienceandTechnology(inChinese),2001,9(4):413～416..
[7] GuoJian＿wei,LiangCheng＿hao.Corrosionbehaviorofcarbonsteelinhightemperatureandhighconcentrationlithiumbromidesolution[J].J.ofDalianUniversityofTechnology(inChinese),2000,40(5):554～556.
[8] GuoJian＿wei.Researchofinhibitionmechanismonabsorptionlithiumbromiderefrigerator[D].Dalian:DalianUniversityofTechnology,2000.
[9] Chun＿genYin,Zhong＿yangLuo,MingjiangNi,etal.Predictingcoalashfusiontemperaturewithaback＿propagationneuralnetworkmodel[J].Fuel,1998,77(15):1777～1782.
[10] SmetsHMG,BogaertsWFL.SCCanalysisofaustieniticstainlessinchloride＿bearingwaterbyneuralnetworktechniques[J].Corrosion,1992,48(8):618～623.
[11] RosenEM,SilvermanDL.Corrosionpredictionfrompolarizationscansusinganartificialneuralnetworkintegratedwithanexpertsystem[J].Corrosion,1992,48(9):734～744.
[12] TrasattiSP,MazzaF.Crevicecorrosion,aneuralnetworkapproach[J].BritishCorrosionJournal,1996,31(2):105～112.
[13] HoskinsJC,HimmelblauDM.Processcontrolviaartificialneuralnetworksandreinforcementlearning[J].Comput.Chem.Eng.,1992,16(4)141～151.
[14] MirchadaniG,CaoW.Onhiddenunitsinneuralnets[J].IEEETransCircuitsSyst,1989,36(15):661～664.
