返回
A MODEL MANAGEMENT FRAMEWORK FOR HETEROGENEOUS ALGEBRAIC MODELS - OBJECT-ORIENTED DATABASE-MANAGEMENT SYSTEMS-APPROACH
DOI:10.1016/0305-0483(95)00007-B.png)
摘要
En 中文
Recently, a growing number of corporations are adopting diverse algebraic modeling languages (AMLs) to augment their productivity of developing mathematical models. As different AMLs penetrate various sections of an organization, a new challenge is faced: effective management of heterogeneous models that have been designed and developed with different modeling conventions and manipulation operations, but intended to solve similar business problems. In this paper, we propose a model management framework that can accommodate such heterogeneous algebraic models, and in so doing, can provide universal model manipulation commands. Three leading contemporary AMLs, namely, AMPL, GAMS, and SML, are employed to render a representative paradigm. The benefits of this framework are manifold, the more significant among which are the following: (1) providing a conceptual basis for uniformly incorporating dissimilar modeling conventions of different AMLs, the framework adapts the generic model concepts and provides a canonical representation for each model; (2) as the framework embodies the generic model concepts, it consolidates database, modelbase, and associated solvers in a single formalism using an Object-oriented Database Management System (ODBMS); (3) taking advantage of the flexibility of a database query language, the framework can provide uniform model management commands to heterogeneous models represented with different AMLs, and can effectively facilitate integration of the model management system components. Also presented in this paper is a prototype model management system that has been designed following the ideas prescribed by the framework, and developed on a commercial ODBMS, ObjectStore, and using an object-oriented programming language, C++.
Keyword:
MODEL MANAGEMENT
MATHEMATICAL PROGRAMMING
MODELING
OBJECT-ORIENTED DATABASE MANAGEMENT SYSTEMS
AI总结
对已上传原文的论文进行重点信息的提取,主要内容包括:简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
O
IF:
7.2
论文数:
3.7K
被引数:
1.4W
机构
暂无机构信息

