Abstract: Gasoline accounts for 60%-70% of refineries' revenue. Therefore, optimizing the scheduling problem of the gasoline blending process attracts significant attention from both the industry and the academe. The scheduling problem is a mixed-integer nonlinear programming (MINLP) problem that essentially includes comprehensive constraints. we propose a novel modeling and optimization method for the scheduling problem of a continuous gasoline blending process. First, we propose a nonlinear programming (NLP) model for output verification to validate the feasibility of the initial production plan. Second, we establish a recipe optimization NLP model to obtain the recipes in blending periods, with the property fluctuations taken into consideration. Finally, we formulate an MINLP model for scheduling to obtain the optimal scheduling plan for operation. In addition, we present a batch moving optimization strategy to solve the three-level model. Case analysis results show that the proposed method is suitable for the continuous blending process and can provide reasonable production plans for refineries.