The automated three-dimensional warehouse is an important unit in the intelligent production line, which can realize the accurate query of real-time inventory information, and its usage efficiency is directly related to the production efficiency of the intelligent production line. To address the problem of arbitrary and unintelligent allocation of storage location in the intelligent production line, a digital twin-based storage location assignment (SLA) optimization method is proposed. Firstly, a multi-dimensional modeling method of data-driven digital twin fusion is used to build a multi-dimensional model of the digital twin of the automated warehouse. Secondly, the communication mechanism and information interaction principle of the digital twin system of the automated warehouse are studied and analyzed. Then, an improved ant colony optimization (IACO) is proposed to optimize the model, and the obtained level information is integrated into the digital twin system of the automated warehouse and mapped to physical entities. Finally, it is proved through experiments that the method can meet the requirements of intelligent selection of cargo level for process-based material entry and exit in stereo warehouses, which is important for improving the intelligence of aerospace equipment production mode and increasing the productivity of intelligent production lines.