关键词: 多智能体, 遗传算法, 染缸排产

Abstract: To solve the problem that unreasonable scheduling causes many high-energy consumption and pollution emissions in the dyeing and printing industry, a dynamic production scheduling strategy of dip-dye systems based on genetic algorithm and Multi-Agent and optimization methods was proposed. This method used the algorithm of hierarchical scheduling with dye vat MES, ERP and the PCS real-time data. Among them, the static layer supported GA under the constraints of multi-product batch processing, non-equivalence of dye vats, pre-orders, delivery orders, as well as the switching cost; the dynamic layer supported multi-agent coordination of dynamic optimization algorithm at running status. Through the algorithm with multi-constraints conditions of production and many factors of dynamic changes, the dynamic optimal design with dye vat homework tasks could be obtained. The simulation results compared with SGA and artificial scheduling showed that the strategy based on data-driven dynamic optimization under the premise of the static task of the workshop scheduling could improve production efficiency while achieving the objective of energy-saving & pollution emission reduction. The results in the enterprise also showed that the method was effective and feasible.

Key words: multi-Agent, genetic algorithm, vat scheduling