基于RS-MSWOA-LSSVM的油气管道失效压力预测

Prediction of Oil and Gas Pipeline Failure Pressure Based on RS-MSWOA-LSSVM

油气管道受多种因素的影响会发生管壁减薄、管道破裂等现象,为提高管道失效压力的预测精度,提出一种基于RS-MSWOA-LSSVM的油气管道失效压力预测模型。首先,采用粗糙集(RS)属性约简提取关键特征,以优化预测模型的输入变量;然后,采用混合策略下的鲸鱼优化算法(MS-WOA)对惩罚因子C和核函数参数σ^(2)进行寻找,并将优化后的参数代入最小二乘法支持向量机(LSSVM)进行预测,得到最优解;最后,引入均方误差、均方根误差、平均绝对误差和决定系数(R^(2))4个评价指标,对LSSVM模型、WOA-LSSVM模型和RS-MSWOA-LSSVM模型的预测精度进行了对比评价。结果表明:RS-MSWOA-LSSVM模型与另外两种模型相比,其预测结果的R^(2)提升至0.9968,均方误差降至0.0639 MPa,均方根误差降至0.2528 MPa,平均绝对误差降至0.2223 MPa,说明该模型的预测结果与实际结果的拟合度更高,且预测精度优于其他两种模型。该研究结果可为油气管道失效压力的预测与管道的安全防护提供技术支撑和决策依据。

In order to improve the prediction accuracy of pipeline failure pressure,this paper proposes oil and gas pipeline failure pressure prediction model based on RS-MSWOA-LSSVM.Rough Set(RS)attribute reduction is used to extract key features,optimize model input variables,optimize penalty factor C and kernel function parameterσ^(2)by using Whale Optimization Algorithm under Mixed Strategy(MS-WOA),and substitute them into Least Squares Support Vector Machine(LSSVM)to predict and obtain the optimal solution.The mean square error,root mean square error,mean absolute error and determination coefficient(R^(2))are introduced to evaluate and compare the prediction accuracy of LSSVM、WOA-LSSVM and RS-MSWOA-LSSVM.The results show that compared with the other two models,the R^(2)increases to 0.9968,the mean square error of the RS-MSWOA-LSSVM model drops to 0.0639 MPa,the root mean square error drops to 0.2528 MPa,and the mean absolute error drops to 0.2223 MPa.Then the prediction results of the model have a higher fit with the actual results,and the prediction accuracy is better than that of the other two models,which provides an important reference basis for pipeline safe transportation and risk prevention.The research results can provide strong technical support and decision-making basis for the prediction of pipeline failure pressure and effective safety protection of pipelines.