Abstract: In cloud manufacturing platform, a multi-plants and multi-periods collaborative production problem was discussed by introducing two adjustive variables, which are the maximal number of enterprises that process an order and the minimal order splitting ratio, respectively. We established a collaborative production planning model with constrains of production time window and production capacity. Then, a self-adaptive simulated annealing genetic algorithm using bi-level code was designed. Our proposed approach was compared with a commercial optimization software CPLEX under different data sizes in numerical simulation. Lastly, the sensitivity analysis of the major parameters in the model was carried on. The experimental results provide a decision-making basis for operation decision makers of cloud manufacturing platform.

Key words: cloud manufacturing, collaborative production planning, order splitting, genetic algorithm