Cross-Platform Computational Techniques for Analysis Code Integration and Optimization
Olds, John R.
Steadman, Kimberly B.
MetadataShow full item record
Following NASA's lead in Intelligent Synthesis Environments, advanced vehicle design communities are beginning to explore automated distributed computing frameworks for integrating disciplinary analysis tools. These design frameworks allow collaborative design teams to take advantage of distributed expertise and existing legacy codes, while retaining some of the automation and optimization capabilities of monolithic synthesis tools and simple subroutines. A key capability in making these frameworks a reality will be the ability to integrate and access contributing analysis codes running on different computing platforms and in various remote locations. This paper reports a cross-platform technique for integrating Microsoft Excel spreadsheets into UNIX-based computing frameworks. Specifically, a combination of UNIX shell scripts, telnet connections via the Internet, and Applescript is used to remotely execute an Excel spreadsheet hosted on a Macintosh computer and return the results to an executive program running on a UNIX workstation. Sample scripts and integration procedures are outlined. Examples are given in which the technique is used to remotely drive a launch vehicle costing spreadsheet under the control of grid search and genetic algorithm optimization techniques hosted on a UNIX workstation. Advantages and disadvantages of the present technique are discussed.