Study on the Detection of Welding Defects in Grounding Leads Based on Ultrasonic Guided Waves and Improved the Subtraction-Average-Based Optimizer Algorithm

As an important part of the grounding system, the detection of grounding leads is of great significance to the safety of operation and maintenance. In view of the problem that the faults of grounding leads cannot be efficiently recognized when they occur during actual operation, this paper proposes a method for recognizing defects in welded seams of grounding leads based on the ultrasonic guided wave detection method and the improved Subtraction-Average-Based Optimizer (ISABO) algorithm. Firstly, the ultrasonic guided wave detection of weld damage in soil-free/soil-containing grounding leads lap structure is analyzed, and numerical simulation and signal acquisition are performed; secondly, Subtraction-Average-Based Optimizer (SABO) is improved by circle mapping with inverse learning and Cauchy variation strategy, and the optimization of variational mode decomposition (VMD) parameter finding based on the improved SABO algorithm is proposed and compared with the support vector machine (SVM) combination of a grounding leads defect detection method; Finally, the application of the improved SABO algorithm is verified by the detection signal data set, the improved SABO algorithm in the parameter optimization is better than the traditional method, through the detection of multiple defects in the absence of soil-free/soil-containing detection comparison, the defect identification accuracy in the two conditions has reached 93.3% and 86.67% respectively, indicating that compared with the SABO algorithm, the method is more effective in detecting defect types and has good practical value.