Acid rain is harmful to the environment. The cause of acid rain is mainly from air pollutants, such as SO2, NOx, etc. Due to the fact that most SO2 is emitted from power plants, the control of SO2 emissions from power plants can reduce acid rain pollution. Pressure swing adsorption (PSA) is an efficient technology for the concentration and recovery of SO2 from flue gas. This simulation study employs a three-bed PSA process designed to concentrate and recover SO2 from flue gas. The adsorbent utilized in this study is NO-treated XAD-16. There is less than 0.5% NOx in flue gas, so we used feed that is assumed to consist of 17% CO2, 0.5% SO2, and 82.5% N-2. Statistical design is utilized to find the optimum operating condition for the study. Numerical experiments using the method of orthogonal table are set up to study the effects of five factors, with three levels for each factor. The five factors are: the feed pressure, three step times (including T1 time, T2 time, and T3 time), and the P/F ratio, which is defined as the ratio of the amount of the purge gas stream to that of the feed gas stream. According to the results of the statistical analysis, following optimal operating conditions are obtained: the feed pressure is 1.8 atm; the P/F ratio is 0.03; and the three step times, T1, T2, and T3, are 32 s, 35 s, and 50 s, respectively. The concentration of SO2 in the product stream is 15.84% at optimum operating conditions, and the SO2 recovery is 99.79%.