Many-Objective Production Scheduling Optimization Method for Aluminum Alloy Creep Forming Operation

Abstract

Due to the increasingly customized product requirements of customers, the production characteristics of the aluminum alloy creep forming components production line have changed to multiple varieties and small batches. As a result, it has the problems of complex production organization, low production efficiency. As far as we know, this is the first time to study the production scheduling problem of the aluminum alloy creep forming component production line. The whole-process production scheduling model is established. Multi-population coevolutionary optimization algorithm (MPCOA) is proposed. The MPCOA method uses three subpopulations for collaborative optimization. The sharing and integration of search information is realized through the interaction between subpopulations. For the optimization of the first subpopulation, an improved VAEA (vector angle-based evolutionary algorithm) which integrates the large neighborhood search is proposed. It can help to mine global information, and avoid local optimization. For the optimization of the second subpopulation, to get uniformly distributed solutions and improve the convergence, an improved decomposition-based method is proposed, which uses an improved Tchebycheff function and adaptive update weight vector mechanism. Computational experiments are performed by using industrial datasets and engineering production data. The competitiveness and superiority of the whole-process production scheduling model and the MPCOA method are demonstrated.