Application Of Intelligent Modeling Methods In Prediction Of Buried Pipeline Corrosion Rate

The buried pipelines are always buried throughout the year,so are inevitably corroded by soil.With the increasing of the pipeline age limit,corrosion cause worse threat to the operational safety.so researching the corrosion rate prediction of buried pipeline are more and more urgent for operational safety of pipeline.The author roundly analyzed the mechanism of corrosion of buried pipeline and factors of corrosion by collecting data of field research,and the intelligent model is applied to predict the corrosion rate of buried pipeline.Due to the corrosion factors is complex,the traditional modeling method is difficult to apply.In order to improve the pipeline corrosion rate prediction precision,using principal component analysis(PCA)and radial basis function(RBF)neural network to predict the corrosion rate of buried pipeline.The input dimension of the model is reduced by using PCA.The application shows that the PCA and RBF model is effective in buried pipeline corrosion rate prediction.To improve the predictive accuracy of buried pipeline corrosion rate,the adaptive particle swarm optimization(APSO)and least squares support vector machine(LSSVM)model is built for buried pipeline corrosion rate prediction.The LSSVM has a faster learning speed,and the APSO is used to optimize the LSSVM model parameter to improve the model prediction accuracy.The simulated results show that LSSVM method which is applied to the buried pipeline corrosion rate prediction has higher prediction accuracy.For reducing the influence of the sample datum,the adaptive immune genetic algorithm weighted least squares support vector machine(AIGA-WLSSVM)is used to build the corrosion rate prediction model of buried pipeline.In order to improve the prediction accuracy of the models,the WLSSVM model parameters were optimized by using AIGA.The results show that AIGA-WLSSVM model is validity to buried pipeline corrosion rate prediction.The corrosion factors of buried pipeline are nonlinear correlation,and it is difficult to establish the mechanism of corrosion rate model.However,using intelligent modeling method can effectively solve this problem,and the intelligent modeling method is proposed in this thesis.It is important for pipeline maintenance and replacement decision.