Evolutionary algorithm for aerospace shell product digital production line scheduling problem

Abstract

The production of aerospace shell products is directly related to the safety and reliability of aerospace products. In this work, the aerospace shell product digital production line scheduling problem (ASPDPLSP) was studied, and a solution was developed. In the production process, it is necessary to decide the processing machine and time of each operation. In order to create a scientific shell product production plan, we propose an operation scheduling algorithm (OSA). Based on the constraints of the inspection process, the OSA has two heuristic task scheduling rules. Then, in order to further optimize the product production plan, an improved genetic algorithm (IGA) is proposed. Considering the repeatability caused by random search, a method for the initial population generation with similar and diverse characteristics is proposed. Two of these generation rules retain symmetry and randomness. IGA was used to optimize the order in which the products were processed, resulting in lower costs. Simulation experiments showed that the proposed algorithm solved ASPDPLSP well and provided suggestions to produce aerospace shell products.