The Material of the curved steel plate that is a part of hull is made through the forming process. In order to save the inventory and human resource related cost of forming shop, a plan that reflects the characteristics of the forming process should be established. However, scientific planning is not possible because decisions made by worker experience are predominant in the forming process. In addition, the prediction of mid to long term demand and the creation of an optimization plan for the production site based on it do not take into account the variability at the site, resulting in an error between the production activities and the plan. In order to improve this problem, this study constructed a model of cutting, cold, and hot molding processes using DES( discrete event simulation ) to simulate production activities performed in actual processes by reflecting the characteristics of mechanical processes such as failure and repair. In addition, a model hardening (ME, model enhancement) process was used to improve the equation by reflecting the result values of simulation repeatedly in planning using the water purification plan optimization method, which was used to reflect the relationship between the operator for each period calculated using the least squares method. This allowed the establishment of an optimal plan with a constant error rate considering the variability of the molding process.