سلسله مراتبی برای کنترل مدل ایستگاه های کاری هوشمند FMS

Hierarchical planning, scheduling and control in flexible manufacturing systems (FMS) provide a systematic way to effectively allocate resources along different time horizons. This paper describes the design and development of an intelligent hierarchical control model based on a proposed tool management method. The control model consists of four levels: the process plan selection, the master scheduling, the job sequencing and the control level. The model is developed to optimize the machine utilization and balance tool magazine capacity of a flexible machining workstation (FMW) in a tool-sharing environment. Problems are identified and modeled in the level of process plan selection, master scheduling, and job sequencing. A genetic-based algorithm was developed to solve the problem domains throughout the hierarchical planning and scheduling model. Fuzzy logic technique could also be incorporated into the master production schedule (MPS) level to allow for a more realistic result in the presence of uncertainty and impreciseness in order to fit the realistic nature of actual industrial environments.

In order to gain market competitiveness, manufacturers must be able to cope with rapid change in customer demands, reduce manufacturing costs while remaining the quality of products, and shorten the product cycle time. The design and implementation of flexible manufacturing system (FMS) has been one of the popular approaches to address these aspects. Other advantages which can be of benefit from FMS have also been discussed in [1]. The inherent flexibility of FMS, to a large extent, is brought about by computer numerical control (CNC) machines containing tool magazines with multiple tool slots. These machines are capable of performing various functions, for example, a CNC milling machine can also have boring and drilling functions so long as the appropriate tool is inserted in the tool magazine. Therefore, the number of different tools and the tool magazine size determine how flexible such a system can be. On the other hand, the more flexibly the machine can perform, the more expensive and difficult it is to maintain the machine. As a result, the FMS can produce parts that are in large variety and small batches. However, with the increase in part variety, the number of tool types and hence tool costs will also increase. According to Cumings [2] and Ayres [3], tool cost contributes to about 25–30% of the total fixed cost and variable cost in a FMS. One of the most common strategies to reduce this cost is tool sharing among machines in FMS. Gaalman et al. [4] conducted a feasibility study on a FMS in a tool-sharing environment and showed that tool sharing contributes significantly to savings in the overall cost of an FMS. To adopt a tool-sharing strategy, the benefit will be the reduction in tool inventory and improvement in tool utilization. On the other hand, because of the high versatility of the CNC machines and the complexity of tool-sharing strategy, careful planning, scheduling, and control must be in place to make the implementation of such a system justifiable.

In this paper, a scalable and flexible architecture of the hierarchical intelligent workstation control model has been presented. The key elements include four functioning levels, the process plan selection level, the master scheduling level, the job sequencing level and the control level as well as the genetic-based planning and scheduling algorithm. The function of each level within the model has been described and the genetic-based planning and scheduling solution is evaluated. The GA-based mechanisms had been tested and results have been reported by the authors in [9] and [11]. It is found that the proposed GA method does not require any unrealistic assumptions on the objective functions such as linearity, convexity and differentiability. In addition, the large scale and complicated cross-level problems with multiple objectives can be solved in a relatively short time. The intelligent hierarchical planning, scheduling and control model provided a systematic way to effectively allocate resources along different time horizons. This intelligent workstation controller has been found to be an effective and efficient algorithm which provides optimization for the upper level as well as the lower level equipment controller to carry out their functions.