Fourier ptychographic microscopy (FPM) is a new, developing computational imaging technology. It can realize the quantitative phase imaging of a wide field of view and high-resolution (HR) simultaneously by means of multi-angle illumination via a light emitting diode (LED) array, combined with a phase recovery algorithm and the synthetic aperture principle. However, in the FPM reconstruction process, LED position misalignment affects the quality of the reconstructed image, and the reconstruction efficiency of the existing LED position correction algorithms needs to be improved. This study aims to improve the FPM correction method based on simulated annealing (SA) and proposes a position misalignment correction method (AA-C algorithm) using an improved phase recovery strategy. The spectrum function update strategy was optimized by adding an adaptive control factor, and the reconstruction efficiency of the algorithm was improved. The experimental results show that the proposed method is effective and robust for position misalignment correction of LED arrays in FPM, and the convergence speed can be improved by 21.2% and 54.9% compared with SC-FPM and PC-FPM, respectively. These results can reduce the requirement of the FPM system for LED array accuracy and improve robustness.