Abstract: In order to solve the problems of high nonlinearity, hysteresis and low precision of outlet tobacco moisture content in the control system of SH26 drum-airflow dryer, the Relief-F algorithm was used to analyze the relationship between the outlet tobacco moisture content and the process parameters, and determine the critical control parameters; orthogonal tests were used to search for the optimization of the critical parameters, and match the best combinations of the process parameters; and the combination of fuzzy and PID control was used to predict and accurately regulate the critical parameters to achieve optimization control of the critical parameters. Test conducted using Hongtashan expanded tobacco shreds as the subject, After optimization, the amount of over-dried cut tobacco is reduced from 40.51 kg to 14.29 kg, a decrease of 64.72%; the standard deviation of the outlet tobacco moisture content is reduced from 0.186 to 0.128, a decrease of 31.18%, which improved the process moisture stability; the physical indexes of the tobacco are improved significantly, and the overall score of sensory quality is increased by 0.83 points. This research method effectively improves the process control ability of drum-airflow dryer, and improves the overall quality control level of the drying process and the quality of cigarette products.