2D defect reconstruction of pipeline based on PSO combined with BP neural network

Aiming at the defect magnetic flux leakage signals of complex characteristics, and the difficulty of magnetic flux leakage signals described defect geometrical characteristics. A new approach based on particle swarm optimization and back-propagation neural network algorithm was proposed to reconstruct pipeline 2D defect. Combined particle swarm optimization with back-propagation neural network, this method utilized easy to realize, fast convergence speed and high accuracy merit of particle swarm algorithm to optimize the structure of neural network, and solve the problem in BP neural network which is sensitive with the initial weights, easy to fall into the local least value. The proposed algorithm apply to 2D defect reconstruction of pipeline The experiment results show that the validity to improving the reconstruction accuracy based on PSO-BP Neural Network method, with a highly practical value.