基于极值分布模型的管道剩余壁厚特征参量预测方法研究

Research on the Method of Obtaining Residual Wall Thickness of Pipeline Based on Gumbel Extreme Value TypeⅠDistribution

当前对管道剩余壁厚的监测大多采用单点测量方法,因此测量得到的剩余壁厚存在较大的偶然性,无法科学客观地评估管道的腐蚀状况。为此,笔者采用多探头检测技术进行测量,将得到的结果结合Gumbel极值Ⅰ型分布模型对被测管段的最小剩余壁厚进行预测,利用现场实际采集的数据推导典型管道部件剩余壁厚的统计参数,用于管道剩余壁厚的预测。通过将预测的最小剩余壁厚值与实际测量的最小剩余壁厚值进行对比,验证了该方案的可行性。

Considering the fact that monitoring pipeline’s residual thickness adopts a single-point measurement method currently and this makes the residual thickness obtained has obvious contingency and the corrosion condition of pipelines can’t be evaluated scientifically and objectively, in this paper, a multi-point detection technology was proposed and applied and combined with Gumbel extreme value type Ⅰ distribution to predict the minimum wall thickness of the measured pipe section, and then it has the data collected on site employed to derive prediction parameters and the predicted minimum remaining wall thickness value compared with the actual measured minimum residual wall thickness values to verify feasibility of the algorithm.