برنامه ریزی عملیات متوالی: روش سطح کلان هیبریدی بهینه

Increased global competition and unpredictable market changes are challenges facing manufacturing enterprises. Changes of part design and engineering specifications trigger frequent and costly changes in process plans, setups, and machinery. The paradigm shift in manufacturing systems and their increased changeability also require corresponding responsiveness in support functions; process planning is a key logical enabler that should be further developed to cope with changes encountered at the system level and to support new manufacturing paradigms and continuously evolving products. Retrieval-based planning, predicated on rigid predefined boundaries of part families, does not satisfactorily support this changeable manufacturing environment. On the other hand, pure generative planning is not yet a reality. Therefore, a sequential hybrid approach at the macro level is proposed where, initially, the part family’s master plan is retrieved, followed by application of modeling tools and solution algorithms to arrive at the plans of the new parts, whose features could exceed its respective original family boundaries. Two distinct generative methods, namely reconfigurable process planning and process replanning, are presented and compared. A genuine reconfiguration of process plans to optimize the scope, extent, and cost of reconfiguration is achieved using a 0–1 integer programming model. Also, because the problem is combinatorial in nature, a random-based evolutionary simulated annealing algorithm has been tailored for replanning. The developed methods are, conceptually and computationally, analyzed and validated using an industrial case study.

Mass customization and agile manufacturing are paradigms that have emerged recently to address the new challenges of the 21st century of highly customized and varying products. Products are continuously evolving beyond the boundaries of their original part families. At the design level, product variety and new product introductions for a dedicated manufacturing system are not usually considered. A flexible manufacturing system (FMS) overcomes this challenge by having all of the needed functionality built in a priori; however, this results in high initial capital investment as well as relatively lower utilization. To stay competitive, increasingly responsive manufacturing systems and their enablers are beginning to emerge [1]. Reconfigurability is an engineering technology that makes it possible to react quickly and efficiently to market changes [2]. A reconfigurable manufacturing system (RMS) is achieved by incorporating basic process modules that can be rearranged or replaced quickly and reliably to adjust the production capacity and functionality in response to new market conditions and process technology. Modularity, integrability, customization, convertibility, and diagnosability are its distinct characteristics [3]. When these characteristics are embedded in the system design, a high degree of reconfigurability is achieved [4]. This type of manufacturing system allows flexibility not only in producing a variety of parts, but also in changing the system itself. These systems will be open-ended and will run less risk of becoming obsolete because they will enable rapid changes of system components and rapid addition of application-specific software modules [5]. Reconfigurability aims at achieving more competitiveness by exploiting new technology and supporting business paradigms [1]. RMS is gradually becoming a reality and is being deployed by many mid to large-volume manufacturers [6]. Reconfiguration could be achieved at the system or machine levels, and it may be classified as soft (logical) or hard (physical) in nature [1]. Process planning is an important soft-type enabler for such changeable systems. It is an essential function for the smooth operation of any manufacturing system running under the variable conditions described earlier. This paper describes a hybrid planning methodology, where retrieval of master or existing plans initially takes place followed by generative processing by means of algorithmic and optimization methods. A review is presented of process planning methods, especially those that would support changeable manufacturing systems at large as well as continuously evolvable part families. The proposed sequential process planning methodology is described, and the knowledge retrieval and manipulation portion of it is detailed, followed by description of the generative mathematical programming portion. The overall developed methodology, the two developed models, and their algorithms are examined and discussed. An industrial test case of a family of single-cylinder engine front covers was employed for verification and also to illustrate the application of the presented methodology.

This paper addresses a new problem that arises due to the increased changes in products and systems and the need to manage these changes cost effectively and with least disruption of production activities and their associated high cost. It proposes novel solutions for the need to frequently plan and replan manufacturing processes. A new sequential semi-generative methodology to solve the classical problem of process planning and sequencing is developed. Under this proposed hybrid sequential notion, the process replanning scheme was compared with the recently developed innovative planning scheme through reconfiguration. The two methods are important “Changeability Enablers” for both traditional and new paradigms of manufacturing systems. The process planning approach presented in this paper is practical, easy to implement and apply. They can be readily integrated with upstream and downstream applications, through standard CAD/CAM data inputs and outputs, for both preprocessing of data prior to sequencing (including technological data and clustering of operations if needed) and further detailing of individual processes through microplanning to determine feed rate, depth of cut, etc. in a metal removal application. One of the main benefits of these new approaches is to reduce the time and cost required to generate a process plan. The overall proposed methodology is advantageous than existing methods such as nonlinear process planning or preplanning scenarios, where alternate process plans are developed and provided ahead of time in anticipation of future changes. In addition to the obvious cost and computational burden that could be avoided by the developed approach, future changes in products and technology cannot be fully predicted; hence, the usefulness of preplanned alternatives is diminished. Furthermore, preplanned processes would likely become obsolete as manufacturing resources and technologies are changed. The presented process planning methods can improve the efficiency of process planning activities and can help “manage changes” on the shop floor by introducing an important changeability enabler in the field of process planning. The planner would have the option of choosing to completely change the process plans using highly refined globally optimal replanning or to employ localized optimal reconfiguration, depending on anticipated production volume, product variability and market stability.