摘要
架空管道受大气环境影响会发生腐蚀增重现象。为提高工业大气环境下管道的腐蚀增重预测精度,提出了一种基于改进海鸥优化算法的区间灰数预测模型(MSSOA-INGM(1,1,δ))。首先通过Kent映射和Levy飞行策略优化海鸥优化算法(Seagull Optimization Algorithm,SOA)的种群,避免SOA过早收敛,并改进收敛因子B以提高全局搜索能力;然后建立区间灰数NGM(1,1)模型,根据反余割函数变换、平均弱化缓冲算子和自适应初始条件优化对模型进行改进,并利用多策略下的海鸥优化算法(Multi-Strategy Seagull Optimization Algorithm,MSSOA)对改进后的INGM(1,1,δ)模型中的时间参数δ进行寻优;最后以某架空管道腐蚀数据为基础进行仿真试验。优化后新模型的预测结果与实际值高度吻合,表明该模型的预测精度得到较大提高,在架空管道腐蚀增重预测方面具有较好的鲁棒性。
A pipeline is the lifeline of oil and gas transportation,but the overhead pipeline exposed to the atmospheric environment for a long time will have a corrosion weight gain phenomenon under the influence of many factors.To improve the corrosion weight gain prediction accuracy of pipelines in an industrial atmospheric environment,an interval gray number prediction model based on an improved Seagull Optimization Algorithm(MSSOA-INGM(1,1,δ)) was proposed.Firstly,the population of the Seagull Optimization Algorithm(SOA) was optimized by the Kent mapping strategy and Levy flight strategy,and the convergence factor B was improved to improve the global search ability.Among them,the Kent mapping strategy initializes the seagull population to make the population distribution uniform,and the Levy flight strategy selects the best position of the seagull population to improve the convergence speed.Then the interval gray number NGM(1,1) model is established,and the model is improved according to the inverse cosecant function transformation,average weakening buffer operator,and adaptive initial condition optimization.The inverse cosecant function transformation can improve the smoothness of the original sequence,and the average weakening buffer operator can weaken the impact disturbance of the original sequence.The adaptive initial condition optimization overcomes the defect of fixed weight in the initial condition selection process of the INGM(1,1) model.The time parameters in the improved INGM(1,1,δ) model were optimized by using a Multi-Strategy Seagull Optimization Algorithm(MSSOA).Finally,based on the corrosion weight gain data of overhead pipelines containing mass fraction 0.12% Si in an industrial atmosphere,the optimized new model was compared and verified with the NGM(1,1) model and SOA-INGM(1,1,δ) model respectively.The results show that the prediction results of MSSOA-INGM(1,1,δ) model are highly consistent with the actual values.The gray correlation degree is 98.66%,the posterior difference ratio is 44.09%,and the average absolute percentage error is 2.56%,which are better than the other two models,indicating that the prediction accuracy of the model has been greatly improved.It has good robustness in predicting the corrosion weight gain of overhead pipelines.The research results provide a reference for pipeline corrosion prediction and decision support for pipeline inspection and maintenance under the atmospheric environment.
作者
骆正山
汪静静
骆济豪
王小完
LUO Zhengshan;WANG Jingjing;LUO Jihao;WANG Xiaowan(School of Management,Xi'an University of Architecture and Technology,Xi'an 710055,China;Ruixin Institute of Beijing Institute of Technology,Beijing 102488,China)
出处
《安全与环境学报》
CAS
CSCD
北大核心
2023年第12期4318-4324,共7页
Journal of Safety and Environment
基金
国家自然科学基金项目(41877527)
陕西省社科基金项目(2018S34)。
