Project Scheduling Tool
Abstract
Optimally scheduling a team of developers on a large software project is an NP- complete problem. The scheduling algorithm employed by the Evolutionary Control System (ECS) portion of the Computer-Aided Prototyping System (CAPS) does near-optimal scheduling using an algorithm that runs in Order N2 space and time. The problem addressed by this thesis is to improve the performance of the algorithm and make it more useful for scheduling software developers. The thesis accomplished three things: (1) Modified the algorithm to run in order N time and space, preserving its near-optimal behavior; (2) implemented a calendaring package that computes federal holidays for any year after 1970 and schedules tasks only on non-holiday workdays; and (3) incorporated a more realistic capability model to better match programming tasks with each developer's abilities.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1997
- Accession Number
- ADA341531
Entities
People
- John Evans
Organizations
- Naval Postgraduate School