This paper focuses on the design of a model-driven control strategy for deep drawing processes with the forming of an inner door panel as application scenario. This control strategy was designed with two main functions, namely the part quality monitoring based on the sheet metal material properties and the draw-in of the part flange and the process control through actuators. To implement these functions, two supervised learning tasks were carried out in order to first identify the part quality after each forming operation and then, in case of defective parts, to provide absolute target values to adjust actuators in order to meet the quality requirements of the next part. The models trained for both tasks achieved a maximum accuracy of 96,96% for the part quality identification, which is a classification task and an R2-Score of 0.92 for the estimation of target values for the actuator control, which is a regression task.