(Georgia Institute of Technology, 2002-06)
Roth, Bryce Alexander; Graham, Matthew; Mavris, Dimitri N.; Macsotai, Noel I.
The objective of this paper is to describe a method for selecting optimal engine technology solution sets while simultaneously accounting for the presence of technology risk. This method uses a genetic algorithm in conjunction with Technology Identification, Evaluation, and Selection methods to find optimal combinations of technologies. The unique feature of this method is that the technology evaluation itself is probabilistic in nature. This allows the performance impact and associated risk of each technology to be quantified in terms of a distribution on key engine technology metrics. The resulting method can best be characterized as a concurrent genetic algorithm/Monte Carlo analysis that yields a performance- and risk-optimal technology solution set. This solution set is inherently a robust solution because the method will naturally strive to find those technologies representing the best compromise between performance improvement and technology risk. Finally, a practical demonstration of the method and accompanying results is given for a typical commercial aircraft engine technology selection problem.
(Georgia Institute of Technology, 2002)
Roth, Bryce Alexander; Graham, Matthew; Mavris, Dimitri N.; Macsotai, Noel I.
Successful selection of propulsion system technologies for development and incorporation into new engine designs requires careful balance among many competing design objectives (i.e. performance, cost, risk, etc.). One seldom has sufficient development resources available to fully explore all promising concepts and must therefore choose a few technologies that show the greatest promise to meet program objectives. This paper describes a method of selecting optimal combinations of engine technologies. This method employs a technology impact forecasting environment in conjunction with genetic algorithms to find Pareto-optimal technology solution sets. These results are illustrated using Technology State Transition Diagrams to show how technologies move into and out of the Pareto-optimal sets. An edge search procedure is introduced as a means to efficiently characterize the objective space, the results of which are presented in the form of ternary plots. These plots show how technologies benefit multiple (oftenconflicting) objectives and help find robust or compromise technology combinations. Finally, these methods are applied to select engine technology combinations for a commercial engine system of current interest.