Proposed in the thesis is an object-oriented modeling(OOM) framework, called a JR-net modeling framework, for the design and analysis of automated manufacturing system(AMS). The proposed modeling framework, which consists of a graphical modeling tool called JR-net (job resource relation-net) and a model building procedure, is the same as current practice of AMS design and is based on the OOM paradigm of Rumbaugh(1991). In other words, a virtual prototype of AMS is constructured step-by-step based on a three-phase modeling approach: static layout modeling (object model); job flow modeling (functional model); control modeling (dynamic model). A functional JR-net model is a directed graph in which the standard resources of an AMS become nodes and job transfers among the resources become the arcs of the graph. A control JR-net model is obtained by specifying control requirements using tokens and database in the functional JR-net model.
In the middle of the thesis, JR-net modeling framework is applied to various AMSs, such as AS/RS, FMS, FAS, AMHS and the modeling characteristics of each AMS are analyzed. During the modeling of these AMSs, the usefulness of the proposed JR-net modeling framework is demonstrated in terms of ease of model building, ease of communication and modeling power.
For the guidelines of constructing a control JR-net model, which is a sophisticated one among the three JR-net models, the control problems, which occur frequently in AMSs, are structured. And the resulting control patterns of AMS are mapped into JR-net building blocks to make the construction of a control JR-net model easy. Also, a step-by-step procedure of control JR-net model construction is described.
Finally, the implementation issues of JR-net model are addressed. One approach is the conversion of JR-net model into a commercial simulation package. The other is a development of integrated AMS prototying tool.