Abstract
The reuse and maintenance of engineering modeling source code is dependent on the effective communication of the complex ideas on which the software is based. These ideas, usually referred to as the model of the software, are independent of a particular implementation. The documentation, which should clearly state the model, frequently assumes that the reader is fluent with the terminology and concepts in the modeling domain. Often, the domain knowledge is not shared by the reader, which leads to confusion, miscommunication, and undue effort in understanding the code. If the domain knowledge can be captured and stored with the model, and there is a means for automatically implementing the models, the reuse of models will enable software reuse and reduce the effort in its maintenance. This exploratory research proposes the Engineering Modeling Method for generating and documenting models. The method captures domain knowledge used in creating the models in such a way that they can often be automatically implemented. A prototype system was constructed to provide automated support for the method and was used to test the method using a lengthy example in automotive air conditioning. The method includes a set of object class definitions to represent the domain knowledge. These objects provide a cognitive framework for the domain knowledge and aid in its communication. The objects also allow the domain knowledge to be manipulated by a computer, enabling its use in testing the consistency of the model and providing information concerning its accuracy and sensitivity. Storing the domain knowledge provides for the incremental construction of a knowledge base concerning a specific domain and provides a means to maintain the knowledge within an organization despite the loss of individuals.
Lockledge, Jeffrey Carleton (1993). A method for engineering modeling using automotive air conditioning as a case study. Texas A&M University. Texas A&M University. Libraries. Available electronically from
https : / /hdl .handle .net /1969 .1 /DISSERTATIONS -1474644.