Abstract: To address the difficulty of maintaining stable outlet moisture control using conventional PID control in the casing process of tobacco primary processing where multivariable coupling and significant time lag are present, an intelligent control method based on case-based reasoning (CBR) was proposed. This method employs a three-level control architecture that integrates the clustering of historical operating conditions, real-time data matching, and dynamic parameter optimization. Firstly, historical production data was clustered to construct a multidimensional feature case base. Secondly, an improved dynamic optimization algorithm was used to match real-time operating conditions with historical cases and determine the optimal control input for the compensation steam valve opening. Finally, a closed-loop feedback mechanism was introduced to adaptively correct the valve opening according to the deviation between the actual and set-point values of outlet moisture content, thereby enabling dynamic optimization of the control parameters. Industrial validation was conducted using cigarette products with four specifications produced by Golden Leaf Production and Manufacturing Center. The results showed that the proposed method significantly improved the control performance of outlet moisture content in the casing process. The process capability index (C_pk) increased from 1.098-1.523 to 1.574-2.248, representing an improvement of 15.4%-65.7%, while the standard deviation of moisture content decreased by 3.4%-23.8%. The proposed method provides a practical approach for improving process stability and intelligent control in tobacco primary processing.