- 匡 浩1，俞阿龙2，徐新元3
2020年第6期
摘要
引用本文
摘 要：针对混凝土内部钢筋腐蚀程度判别难、精确度低等问题，提出了将改进粒子群算法（PSO)与BP神经网络结合起来，通过对钢筋锈蚀机理及其影响因素的分析，建立了以混凝土内部温度、湿度、pH值、Cl-浓度和腐蚀电位为输入，钢筋腐蚀率为输出的改进PSO-BP监测模型，并将实测输入数据与仿真结果进行了对比。结果表明，改进PSO-BP算法的收敛性与准确性均优于PSO-BP算法和BP算法。Abstract:Inviewofthedifficultyinjudgingthecorrosiondegreeofsteelbarsinconcreteandthelowaccuracy,animprovedparticleswarmoptimizationalgorithm(PSO)wascombinedwithBPneuralnetwork.Basedontheanalysisofthecorrosionmechanismofsteelbarsanditsinfluencingfactors,animprovedPSO-BPmonitoringmodelwiththeinternaltemperature,humidity,pH,chlorideionconcentrationandcorrosionpotentialoftheconcreteasinputandthecorrosionrateofsteelbarastheoutputwasestablished,andthemeasuredinputdatawerecomparedwiththesimulationresults.TheexperimentalsimulationcomparisonshowsthattheconvergenceandaccuracyofimprovedPSO-BPalgorithmarebetterthanthoseofPSO-BPalgorithmandBPalgorithm.
匡浩,俞阿龙,徐新元.改进PSO-BP算法在钢筋腐蚀监测中的应用研究[J]混凝土与水泥制品,2020(6):77-81.
KUANG H,YU A L,XU X Y.Research on the Application of Improved PSO-BP Algorithm in Reinforcement Corrosion Monitoring[J].China Concrete and Cement Products,2020(6):77-81.
