Optimization Modeling Assistant (OMA) is an expert-system computer program that assists its users in the development of mathematical and algorithmic models for optimization. As used here, "optimization" refers to a suite of powerful decision-support techniques that enable the modeling of any decision-making environment in terms of the objectives, the decisions that potentially influence the objectives, and a set of constraints that bind the decisions. Optimization analysis helps decision makers in identifying decisions that optimize their objectives. Domains in which optimization modeling are applied include real-time scheduling, logistics, planning, and financial management.

The Conceptual Architecture of OMA accommodates two distinct processes — acquisition of knowledge and optimization modeling. The architecture is modular, so that OMA can evolve and be integrated with other decision-support software tools.

Unfortunately, difficulties inherent in the construction of optimization models have prevented the realization of the full potential of optimization modeling in these and other domains. These difficulties include the following:

  • Often, optimization models for a given domain address aspects of the domain that are invisible to casual observers.
  • Optimization analysis is expensive because it demands highly specialized skills. A major hurdle is that domain experts are typically not expert in the techniques of optimization, and vice versa.
  • Because of a lack of a comprehensive decision-support software environment, optimization-based decision-support systems typically depend on building custom solutions for different domain situations. It is often necessary to redesign such a system as the environment evolves or the requirements change.

OMA (see figure) helps to overcome the difficulties by utilizing knowledge-based-systems techniques to automate much of the model-design process. Though OMA can be used by itself, it is designed to be integrated into a more comprehensive decision-support software environment. OMA assists users in developing structured optimization models from possibly unstructured domain descriptions. The intended users of OMA include both domain experts and optimization experts who lack expertise in each others' specialties.

OMA reuses domain models to automatically generate optimization models. Domain models can be process models or ontology models. In a given application, OMA guides its user(s) through a step-by-step process, according to a structured methodology, to generate relevant and solvable optimization models from structured domain models. OMA extracts, from both the user(s) and the domain models, the information (e.g., goals and parameters) necessary for generating optimization models. OMA then automatically generates and validates the optimization models for the application. The results of the validation subprocess can be used to identify information missing from the optimization models, to identify unsatisfiable constraints, and to identify circular definitions.

The main benefits of using OMA include the following:

  • Improved automated support for capturing domain-expert descriptions of systems, situations, and problems;
  • Increased productivity in the selection of optimization techniques and in the design and execution of optimization models;
  • Enhanced communication between domain experts and optimization analysts; and Reduced dependence on optimization experts to design and use optimization models for decision-support activities.

This work was done by Perakath C. Benjamin, Satheesh Ramachandran, and Madhav Erraguntla of Knowledge Based Systems, Inc., for Kennedy Space Center. For further information, access the Technical Support Package (TSP) free on-line at www.nasatech.com/tsp  under the Information Sciences category.

In accordance with Public Law 96-517, the contractor has elected to retain title to this invention. Inquiries concerning rights for its commercial use should be addressed to

Perakath Benjamin
Knowledge Based Systems, Inc.
1408 University Drive East
College Station, TX 77840

Refer to KSC-12135, volume and number of this NASA Tech Briefs issue, and the page number.