Abstract: The market demand for hot-rolled seamless steel tube presents multi-variety and small-batch characteristics, and delivery addresses and delivery time of orders are usually different. In this paper, the problem of hot-rolled order scheduling of seamless steel tube is extracted and defined from actual production. Hot-rolled order scheduling is the process of combining production orders into rolling batches and scheduling rolling batches sequence. In order to meet customer needs and ensure production efficiency and continuity, delivery factors, such as delivery address and delivery time, are introduced to the order scheduling. Moreover, considering delivery requirements and process characteristics, a model for this order scheduling problem is built to optimize the delivery period, ensure concentrated production and distribution, and minimize the total setup time of machines. A two-stage algorithm is provided further. In the first stage, to minimize differences among delivery time and addresses of orders, a hierarchical clustering algorithm is put forward on cohesion policy to cluster orders with the same process constraints and similar delivery requirements, thereby forming an initial rolling plan. In the second stage, a hybrid variable neighborhood search algorithm is presented to improve the initial plan in terms of setup times and earlyness/tardiness. Results of experiments with actual order data show the effectiveness of the model and algorithm. This research has a useful reference for the study of order scheduling problem of other steel products or products involving complex process and delivery constraints.

Key words: order scheduling, hot-rolled, seamless steel tube, delivery, hierarchical clustering