Abstract: The Integral Field Spectrograph (IFS) is an important scientific instrument on the Chinese Space Survey Telescope (CSST), and designed for high spatial resolution spectroscopic observations. During both the ground testing and future in-orbit operation, wavelength calibration for IFS is always required for precise test result and high-quality astronomical data. IFS has two observation bands, ultraviolet and visible, with the visible band containing 32 data channels, each of which has about 30 lines emitted by the Hg-Ar lamp used for calibration. Traditional manual method for calibration will be time-consuming and strenuous. To improve the efficiency of wavelength calibration, an automated wavelength calibration method must be developed. First, the 1D spectra of all IFS channels are automatically extracted from the 2D data using Connected Component Analysis (CCA). Then, Particle Swarm Optimization (PSO) and Gaussian fitting are applied to automatically identify the emission lines, meanwhile the parameters of PSO are given. Finally, the dispersion curve is fitted using a fifth-order polynomial. Implementing this method to all 32 channels in IFS visible band, the resulting calibration accuracy (root-mean-squre value of residual error) is better than 0.35 Å, meeting the accuracy requirement. This demonstrates that the automated wavelength calibration of IFS is achieved.