基于多元线性回归方法的海上平台用钢量分析

Analysis of Steel Weight for Offshore Platform Based on Multi-Element Linear Regression

对海上平台传统用钢量预测模型进行分析,建立一元回归模型。为了提高平台用钢量预测精度,构造多元线性回归预测模型,引入灰色关联理论,利用灰色关联分析方法确定影响中心平台用钢量的主要因素为设备干重和甲板面积,影响井口平台用钢量的主要因素为设备干重、高峰年用电负荷和井槽数。在研究海上平台用钢量变化规律的基础上,运用多元线性回归的方法建立海上平台用钢量预测方程。以渤海平台统计数据为基础,进行传统的预测模型预测和多元线性回归模型预测,给出各模型预测值与实际用钢量误差,结果表明:基于多元线性回归预测模型的预测精度可达90%,优于传统预测模型,能够较准确地反映平台用钢量。

The traditional prediction model of steel weight is analyzed,and the single-element regression model is established.In order to improve the prediction accuracy of steel weight of platform,a multi-element linear regression prediction model is constructed.Grey relation theory is introduced.The main factors influencing the steel weight of the center platform are the dry weight of the equipment and the deck area,and the main factors influencing the steel weight of wellhead platform are the dry weight of equipment,peak annual power load and the number of wells.Based on the study of the variation rule of the steel weight used in offshore platforms,a multi-element linear regression method is used to establish a prediction equation for the steel weight used in offshore platforms.Based on the statistical data of Bohai platform,the prediction by traditional prediction model and the multi-element linear regression model is performed,and the prediction value of each model and the actual steel quantity error are given.The results show that based on the multi-element linear regression model,the prediction accuracy of the multi-element linear regression model is better than that of the traditional prediction model,and the multi-element linear regression method can rather accurately reflect the steel consumption of the platform,which can reach 90%.