An on-line quality evaluation method of anchor chain flash butt welding based on DTW and MDS

Abstract

The invention discloses an on-line evaluation method of anchor chain flash butt welding quality based on DTW and MDS. The method comprises: creating a 2L*2L database E1 of size 2L*2L; 2, judging whether the system detects a new signal to be tested or not, if yes, then executing step 3; If no, stopping the welding and ending the system process; 3, collecting signals to be tested; 4, carrying out data preprocessing on that signal to be measured; 5, respectively calculating that dissimilar distance between the signal to be measured and the selected 2L historical data; 6, establishing a dissimilardistance matrix E2 of size 2L*1; 7, combining that matrices E2 and E1 to form a dissimilar distance matrix F with the size of (2L +1) * (2L +1); 8, carrying out dimensionality reduction treatment onthat dissimilar distance matrix F to a matrix G of (2L +1) * P; 9, carrying out clustering analysis on 2L +1 P-dimensional data in that matrix G; 10, judging whether that signal to be tested is clustered into a normal signal class or not, if not, processing the fault signal; If yes, determining the signal to be measured as a normal signal, and processing the normal signal; 11, returning to that step 1 to update the database E1; The invention can quickly and effectively identify the faulty welding product, improve the production efficiency and ensure the welding quality.

Description

technical field [0001] The invention belongs to the field of welding, in particular to an online evaluation method for anchor chain flash welding quality based on DTW and MDS. Background technique [0002] With the development of world shipping, people put forward higher requirements for shipping safety. As an important link to ensure the safety of shipping, the quality of the anchor chain has attracted more and more attention. Due to the advantages of high thermal efficiency, good welding quality, and stable quality of welded joints, the flash welding method is the mainstream of the high-quality anchor chain market at this stage. [0003] The flash welding process is a process that is affected by multiple parameters and the electrical environment is relatively harsh. The highly nonlinear welding process and the randomness of the continuous flash process make it very difficult to establish an accurate model for the flash welding process, so the establishment of a flash weld...

AI Technical Summary

Problems solved by technology

The disadvantage is that the welding quality evaluation model of this method is built by artificial neural network, and a large number of experiments under different environments and conditions are required to obtain sufficient sample data.

Moreover, the artificial neural network method requires a large number of online calculations, which takes a long time and requires high computer performance, so most of them are not suitable for industrial sites.

In addition, the welding quality evaluation model of this method needs to be established by conducting a large-scale welding test on each welding object before actual production. This is a static modeling method that does not take into account the observed data and the observed data at past moments ( sample data), without considering the dynamic characteristics of the system

Images