Weld Defect Recognition Method of Pipeline Based on Improved Least Squares Twin Support Vector Machine

To improve the recognition accuracy of the weld and weld detect in the pipeline inspection, an identification method based on least squares twin support vector machine (LSTSVM) is proposed. The magnetic flux leakage (MFL) signal features are extracted according to the fluctuation and shape features, which are input into LSTSVM, LSTSVM to recognize weld, weld detect, detect and normal condition. Particle swarm optimization is used to optimize the penalty parameters and kernel parameters of LSTSVM to achieve high identification accuracy. The experimental results show that the recognition accuracy of the proposed method is 98.7%, which is higher than those of back propagation neural network, support vector machine and LSTSVM.