Zhejiang University of Technology
NSFC-Zhejiang Joint Fund for the Integration of Industrialization and Informatization (No. U1709213)；Key Research and Development Program of Zhejiang Province (No. 2020C01109)
Cut order planning (COP) is the first stage of the garment manufacturing process and plays an important role in the production management and cost control. COP for large scale and irregular multicolor garment orders remains a key issue, and it is an NP (Non-deterministic Polynomial)-hard nonlinear optimization problem. To deal with this issue, a hybrid optimization algorithm based on NGSAII was proposed, which is the first time that applied the multi-objective evolutionary algorithm (MOEA) to solve COP problem. First, an MOEA model for multicolor COP was established to minimize the production excess and the number of cutting table. Second, the ridge regression decoupling method was utilized to decouple the size combination scheme and the spreading layer scheme to improve the accuracy of solutions. Meanwhile, the real-number encoding strategy was used to encode COP solutions to promote the solving efficiency. Finally, application cases and comparison experiments of several algorithms were carried out. The results show that the devised algorithm has obvious advantages in accuracy and efficiency over heuristic algorithms and optimization software. As a result, the hybrid optimization algorithm can effectively optimize the production management of the cutting department, thereby reducing the cost of fabric and setup, and has significant application and reference value.