Evaluation Method of Pipeline Corrosion Level Based on Multilayer Convolutional Sparse Coding

Abstract

The present invention provides a pipeline corrosion level assessment method based on multi-layer convolution sparse coding, which includes the following steps: S1: Install the first acceleration sensor and the second acceleration sensor on the outer surface of the pipeline, and use a hammer to tap the concerned area , to obtain the shock response signal; S2: Build a grade evaluation model based on multi-layer convolutional sparse coding; S3: Collect the actual hammer test data of the pipeline to be tested, preprocess and use the time domain signal as a test set, and input it to step S2 for training Among the good models; S4: According to the evaluation model based on multi-layer convolution sparse coding, the evaluation results of pipeline corrosion level are given. The present invention adopts multi-layer convolutional sparse coding as a model, uses the oscillation attenuation function to extract effective features in multiple frequency bands, can effectively extract weak damage features in the vibration signal under pipeline corrosion, and completes automatic corrosion level evaluation, avoiding traditional methods. The dependence of threshold selection is suitable for online pipeline corrosion monitoring.

Description

technical field [0001] The invention relates to the technical field of non-destructive testing, in particular to a pipeline corrosion grade evaluation method based on multi-layer convolution sparse coding. Background technique [0002] Corrosion damage of pipelines is one of the main problems faced by petroleum, construction and other industries. During the service process, the pipeline will inevitably be affected by external forces and environmental changes, causing aging problems such as corrosion and wear, and then causing accidents. Therefore, it is of great significance to carry out research on nondestructive testing and evaluation methods of pipelines to ensure the safe and reliable service of pipelines. At this stage, the non-destructive testing technologies for pipeline corrosion mainly include ultrasonic testing, magnetic flux leakage testing, and radiographic testing. However, the service conditions of the pipeline are complex, and the use conditions of the above...

AI Technical Summary

Problems solved by technology

However, the working mechanism of the network is difficult to clarify, lacks interpretability, and a model with perfect performance often requires a large number of parameters and a huge amount of calculation

Images