School Timetabling—A Case in Large Binary Integer Linear Programming
将排课问题建模为大规模0-1整数线性规划，提出基于拉格朗日松弛和次梯度优化的求解方法，并融入分支定界过程，针对包含900门课程的一年制研究生项目给出计算结果。
A timetabling problem is formulated as a large integer linear programming problem in 0-1 variables. A solution method based on Lagrangean relaxation coupled with subgradient optimization is presented. The solution method also incorporates a branch and bound procedure which takes advantage of special ordered sets of variables. We present computational results for a large timetabling problem involving 900 subjects in a one-year graduate program.