关键词: 产品服务系统, 随机调度, 工期指派, 启发式算法

Abstract: With the market competition growing ever more tense and the profit space of the product compressing continuously,the traditional manufacturing industry begins to transform from product-oriented manufacturing to service-oriented manufacturing. Instead of selling products only, the service-oriented manufacturing (SOM) firms offers the customers the integrated solutions of products and services which are called product service systems (PSSs). In order to delivery PSS orders on time and to reduce operating costs incurred by inventory and tardiness, the SOM firms need to assign reasonable due dates to the orders and make effective scheduling plans. However, the production times and service times of PSS orders may be stochastic due to customized demand, which adds more difficulty and complexity to the due date assignment in sequencing PSS orders. Therefore, how to jointly schedule PSS orders and assign due dates in a stochastic environment is a significant problem for SOM firms.
In this paper, the stochastic scheduling problem of PSS orders with due date assignment is addressed for a SOM firm consisting of one single manufacturing plant and multiple regional service centers. It is assumed that the production times and service times are independent and normally distributed with the mean and variance provided. The objective is to minimize the total expected earliness, tardiness and due date assignment costs. To solve this problem, the mean value theorem of integrals is applied to obtain an approximate objective function of the original one and the optimality conditions are analyzed. A sequencing rule called the weighted shortest average production time is presented, based on which three heuristics are proposed. In the numerical experiments, the sensitivity analysis is carried out and the effectiveness and robustness of the proposed heuristics is illustrated.
The results indicate that the unit earliness cost deviation has little influence on the total expected costs, hence there is no need for the dispatchers to ensure the accuracy of the unit inventory cost. Whereas the unit due date assignment cost has a significant impact on the results of the scheduling decision, which suggests that the dispatchers should pay great attention to the accuracy of this cost. Our research sheds light on reducing the operating costs and increasing the service levels of the SOM firms, since it can assist the dispatchers of SOM firms in scheduling PSS orders and setting due dates effectively.

Key words: product service system, stochastic scheduling, due date assignment, heuristic algorithms