Research on the problem of scheduling multi-product batches in the process industry

With the diversification and personality of customer demand the modern chemical Polyester production has shifted to multi-product small batch production. The problem of the scheduling on multi-product small batch production is a new unresolved issue which is that people have been concerned about. This paper firstly analyzes the nature of the polyester engineering process and the optimal characteristics of the problem. Based on the supply of materials and inventory constraints, the precedence relationships of the batches between the various stages are drawn. This paper applies continuous-time modeling strategy to establish mixed integer programming model of the integrated problem of batching and scheduling. The model consider the integration of batching and scheduling in decision-making and consider multi-product, conversion rate, set-up time and inventory constraints, and then give out an improved particle swarm optimization algorithm. The algorithm use two-dimensional coding on behalf of batch quantity to enhance the ability to resolve constrains, and brings out rep airing strategy, drifting strategy and scheduling strategy to decode a particle into a feasible solution of the problem. The introduction of the systolic and divergence operators can maintain particle diversity and convergence. The examples from enterprises verify that the model and the algorithm have good performance. This paper extends the batching and scheduling theory of the small batch production in the process industry, and provides methodological support and theoretical basis for the practices of the chemical polyester production.