A Multistage R & D Project Selection Model with Multiple Objectives.
Abstract
This dissertation presents a multiple objective, zero-one mathematical programming model for the stochastic, multistage R&D project selection and resource allocation problem. The model is intended for those research and development activities which have a certain degree of autonomy in the selection of their research portfolio and which concentrate efforts in applied research or development projects. Each feature of the proposed model is designed to incorporate realistic aspects of the R&D decision making process. First, the model includes an integer goal programming formulation which enables the formal incorporation of multiple, noncommensurate objectives along with integer treatment of possible projects. Second, decision tree planning techniques are used to represent the sequential aspects of allocating scarce resources over the planning horizon, to reflect various project and goal relationships, and to incorporate probabilistic future events and outcomes. Third, simulation techniques are used to convert the probabilistic parameters into deterministic inputs for the mathematical formulation. Finally, the model includes a heuristic based solution algorithm involving system constraint partitioning procedures, and iterative one-pass variable selection and solution variable exchange routines to enable the efficient solution of realistic size problems.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1979
- Accession Number
- ADA108303
Entities
People
- Thomas Craig Harrington
Organizations
- Air Force Institute of Technology