Particle Swarm Optimisation Algorithm for Crack Shape Reconstruction in Magnetic Flux Leakage Nondestructive Testing

Yong Zhang,Zhongfu Ye
DOI: https://doi.org/10.1080/10589750802375962
2009-01-01
Nondestructive Testing And Evaluation
Abstract:Magnetic flux leakage testing is widely used to examine ferromagnetic materials. Considering the importance of estimating the size of surface crack in metals, which is used in nuclear power, railway and piping industries, a method based on particle swarm optimisation algorithm is proposed in this article. This approach maps the size of crack to the location of particles. After initialising the locations and velocities of the particle population, it evaluates the fitness value of the individual particle and tracks the optimal one. It then updates the positions and velocities of particles with the previous fitness values and repeats these steps until the best possible result is obtained. The fitness value is obtained by summing the absolute measurement error of the magnetic leakage field intensity of the crack. The results of simulation experiments demonstrate the effectiveness of the proposed method.
What problem does this paper attempt to address?