Search

This thesis develops a prototypic integer programming model to aid in solving the Naval Postgraduate School academic course scheduling problem. The simplified model schedules faculty members to teach their assigned courses in specific rooms at specific times and schedules groups of students to the courses they have requested. The model assures, as best possible, that room capacity is no exceeded, students and faculty have time for lunch and faculty requesting 'back-to-back' courses are accommodated. To make the problem manageable, we concentrate on just one building, Glasgow Hall, and three departments, Operations Research, Mathematics and National Security Affairs. Even doing this, the model generated in GAMS (Generalized Algebraic Modeling System) has about 287,778 variables and 148,161 constraints and is too large to solve. Consequently, a simplified model, restricted to the Operations Research Department, is solved. This problem encompasses 19 faculty members, 26 courses, 83 sections and 11 classrooms. The model has less than 32,000 variables and 17, 000 constraints and is solved using GAMS and the X-System on an Amdahl 8995-700A in 3488.4 seconds., http://archive.org/details/aprototypicmodel1094539763, Major, Chinese Army, Approved for public release; distribution is unlimited.

A heuristic approach is presented to solve the Naval Postgraduate School's quarterly scheduling problem for academic courses and final examinations. The current scheduling system is studied and an automated system for data collection is developed and implemented. An automated system for the scheduling of final examinations is designed and implemented. Results using real data from one quarter produced feasible solutions to the final examination scheduling problem of 1700 students, 267 courses and 850 sections. The academic course scheduling heuristic is discussed including an integer linear programming approach to the timetabling and distribution problem of students among segments of the same course. An example with an optimal solution to the single course scheduling problem is presented., http://archive.org/details/thenavalpostgrad1094521459, Captain, German Army, Approved for public release; distribution is unlimited.