基于GWO-LSSVM算法的海底管道腐蚀预测模型研究

Research on corrosion prediction model of submarine pipeline based on GWO-LSSVM algorithm

目的针对海底管道腐蚀影响因素存在信息叠加与相互耦合、作用机理复杂、腐蚀速率预测难度大的问题,提出一种灰狼优化(GWO)算法优化最小二乘支持向量机(LSSVM)的腐蚀速率预测新模型。方法该模型利用灰狼优化算法对最小二乘支持向量机的核参数与惩罚因子进行迭代寻优,减少参数选择的盲目性,提升预测精度,应用该模型对海水挂片腐蚀实验的50组样本进行学习与预测,并与传统最小二乘支持向量机、粒子群优化最小支持向量机进行了预测精度的比较。结果灰狼优化最小二乘支持向量机的平均绝对误差、均方误差、均方根误差均最小,其决定系数更接近于1,说明该模型的预测结果与真实值最接近,算法效率高。结论构建的模型可以用于当前油气工程大数据驱动的腐蚀预测中,其结果可以为海底管道的腐蚀与防护提供决策技术支持。

Objective Aiming at the problems of information superposition and mutual coupling of submarine pipeline corrosion factors,complex action mechanisms,and difficult corrosion rate prediction,this article proposes a corrosion rate prediction new model of gray wolf optimization(GWO)algorithm optimized least square support vector machine(LSSVM).Methods The model uses the gray wolf optimization algorithm to iteratively optimize the kernel parameters and penalty factors of the least squares support vector machine to reduce the blindness of parameter selection and improve the prediction accuracy.The model is applied to 50 sets of samples of seawater coupon corrosion experiment.The learning and prediction are carried out,and the prediction accuracy is compared with traditional least square support vector machine and particle swarm optimization minimum support vector machine.Results The average absolute error,mean square error,and root mean square error of the gray wolf optimized least squares support vector machine are all smallest,and the coefficient of determination is closer to 1,which indicate that the prediction result of the model is closest to the real value,and the algorithm efficiency is high.Conclusions The model constructed in this article can be used in the current corrosion prediction driven by big data in oil and gas engineering,and the results can provide a decision-making technical support for the corrosion and protection of submarine pipelines.