Research on oil-gas Pipeline Leakage Detection Method Based on Particle Swarm Optimization Algorithm Optimized Support Vector Machine

Abstract In order to solve the problem of low accuracy in oil-gas pipeline leak detection, a detection method based on Particle Swarm Optimization (PSO) algorithm optimized Support Vector Machine (SVM) is introduced. This uses PSO penalty factor ‘c’ and kernel function parameter ‘g’, constructs leakage model SVM. We set up an experimental platform collect negative pressure wave signals under different working conditions. After wavelet domain denoising data preprocessing, four eigenvalues Mean, Standard Deviation, Kurtosis Skewness are extracted from form eigenvector samples, which taken as input SVM, conditions normal, leakage, rise fall output. Through verification, comprehensive performance PSO-SVM better than that traditional Genetic Algorithm SVM grid search The POS-SVM can be applied pipeline.