Method for Extracting Stationary Features During the Welding Process of Polyethylene Gas Pipelines
The stationarity of the welding process of polyethylene gas pipelines is an important factor affecting their welding quality.Taking the pressure monitoring data of the welding process as the research object,dimensionality reduction is achieved through geometric feature segmentation of the monitoring sequence.Curve fitting is performed on each stage,and a steady-state index is constructed based on the de-rivative information of the fitting function to quantitatively characterize the stationarity of the welding process parameters.At the same time,by mining historical steady-state data to determine the threshold,the stability evaluation and steady-state judgment of welding pressure at each stage were achieved.The stationary characteristics of the process was extracted to provide a process quality basis for welding quality e-valuation.Using an improved data segmentation method based on orthogonal distance feature points,the welding pressure monitoring time series is divided into different welding stages.A stationary feature extraction method based on segmented superposition fitting is proposed,and a steady-state index is constructed using first-order and second-order derivatives to quantitatively represent the stationarity of each stage.The example analysis result using the hot melt welding pressure data has verified that the method proposed in this paper can quantitatively characterize the stationarity of welding process pressure data.
polyethylene gas pipelinewelding processstationary feature extractioncurve fittingsteady-state index
万方数据
维普
