Texture decomposition with particle swarm optimization method

The newly developed optimization algorithm-particle swarm optimization (PSO) algorithm is introduced into the crystallographic texture decomposition. With the linear correlation factor as the evaluation parameter, both the PSO algorithm and the Nelder–Mead Simplex (NMS) algorithm are evaluated in this paper. The evaluation result reveals that the PSO algorithm is more effective when it comes to the complicated multi-component textures, i.e., instead of falling into the local minimum in the NMS algorithm, the PSO algorithm goes to the global minimum. So high quality of texture decomposition is obtained with the PSO algorithm.