| Oil and gas is an important energy source that drives China’s economic development.Oil and gas pipelines in a complex environment are susceptible to leakage by various corrosion factors,causing serious economic losses and irreversible environmental pollution.In view of the existing qualitative evaluation method of the residual strength of defective pipelines is too empirical and quantitative evaluation methods rely too much on specific parameters,and the calculation steps are cumbersome and other shortcomings.The paper applies the integrated algorithm to the calculation of residual strength of corrosion-defective pipelines and constructs a model,which is of great significance to realize the transformation of residual strength of corrosion-defective pipelines from qualitative to quantitative research and to solve the problem of leakage of pipelines containing corrosion defects.Based on the combing and analysis of domestic and foreign literature,the integrated algorithm RS-PSO-ELM model and RS-GA-BP model are constructed for the problem of predicting the residual strength of external corrosion defect pipelines by using the theories of RS theory,PSO algorithm and ELM algorithm.Focused on exploring the problem of residual strength prediction of single-point external corrosion defect pipeline and complex external corrosion defect pipeline,introducing RS attribute reduction theory to eliminate redundant information,determining the key factors affecting the residual strength,selecting PSO and GA to optimize the operational performance of ELM and BP neural network respectively,and bringing the normalized core index into the calculation.The results show that the pipe diameter,pipe wall thickness,length,depth,width of corrosion defects,the pipe’s own tensile strength and ultimate strength are all factors influencing the residual strength of defective pipes,the defect length has the greatest influence on the residual strength of single point external corrosion defective pipes,and the pipe diameter-thickness ratio and defect depth have the greatest influence on the residual strength of complex external corrosion defective pipes;the integrated algorithm PSO-ELM has an overall The performance of the integrated algorithm PSO-ELM is stable and has higher prediction accuracy compared with ELM algorithm,ASME B31 G,DNV RP-F101,SHELL92 and PCORRC;integrated algorithm GA-BP has simple operation steps and higher prediction accuracy than BP neural network and PSO-BP model and finite element method.The feasibility and accuracy of the integrated algorithm RS-PSOELM model and RS-GA-BP model are verified,which provides a new method for predicting the remaining strength of external corrosion defective pipes.The two sets of integrated algorithms constructed in this paper to predict the remaining strength of external corrosion defective pipelines have certain accuracy and feasibility compared to the evaluation criteria and finite element methods.Due to insufficient experimental data and limited time for pipeline bursting,exploring the applicability of more integrated algorithms in the pipeline residual strength prediction problem is a direction for further research in the future. |
