A Computerized Formal Means To Reason About Components In Simulation Models And Environments Part Ⅱ: Module Development Theory

摘要

The field of modeling and simulation is undergoing much change, with artificial intelligence being one clearly defined cause. Newer forms of numerical and symbolic modeling and combinations requiring both of these are creating new needs and difficulties along with great opportunities. In response to this, we have posited a central role for a formal logic-based reasoning system housed in a "general purpose simulation environment" (GP-SE) to support interactive reasoning about models in both numerical and symbolic forms, and also for environment components. This formalism is covered in part Ⅰ of the paper, where we argue that an associated body of knowledge, an ancillary theory, backed up by software, is needed to realize the formalism in its intended GP-SE context. In this part of the paper, we present the ancillary theory's embodiment and capabilities, e.g., an abstract machine model; a formally defined core of instructions and definitions; interactive development with I/O port orientation; "presentation processors"; incremental system change, portability and extensibility; and roles of and relationships to PROLOG, C and C++ systems. The impact of the logic-based theory on these matters and the implications they have on it, are mentioned at key points, e.g., in terms of acquiring and creating specifications, procedural and non-procedural; testing and adapting of specifications; and automating development of implementations from specifications and vice versa.