Combined with the actual situation of the production line of order-oriented enterprises, the important factor of order rejection cost due to customer priority was added to the traditional order acceptance model, and the new whale optimization algorithm (WOA) was used to solve the problem. WOA was proposed to solve real number domain problems and had the defect of easily falling into local optimum. To solve this problem, an improved whale optimization algorithm (IWOA) was proposed. The coding method based on ranking and deviation degree was used to solve the integer domain problem of the order acceptance model. Adding a dynamic learning strategy from historical individuals could avoid premature algorithms to a certain extent. In order to prevent whale individuals from deviating from the optimal direction in the process of random search and thus affecting the convergence speed, the cross-selection strategy of genetic algorithm was used to eliminate inferior individuals. The experiment compares IWOA with WOA and the improved gray wolf algorithm (HGWO), which proves the advantages of IWOA in solving the order acceptance model, the stability of result, the convergence speed of the algorithm itself and the quality of initial solution, etc.