University courses timetabling in an educational department in the maximum level of availability for students (regarding the time conflict between the courses) not only improves the efficiency of the educational plan and overall satisfactory, but also increases the possibility of graduating of students on predefined educational period. In most university course timetabling problems a curriculum is presented to students. So students can apply for courses that belong to their curriculum. Thus, flexibility in the selection of courses by students is not considered. However, accessibility to pre-registration figures and applicants of each course is possible by mechanized registration systems in most universities. This study is based on field studies of some of our prestigious universities and a case study of the faculty of Industrial Engineering, Isfahan University of Technology. Based on Pre-registration reports, which determines the number of students applying for each course, tow Binary Integer Programming model are presented in this paper for Faculty-Course-Time assignment problem which consider both the faculty preferences and course availability for students. Since Faculty-Course-Time assignment problems belong NP-Hard complexity class, two metaheuristics based on ant colony system and simulated annealing are presented. The efficiency of the presented algorithms is shown by solving some test problems, obtained from a case study and comparing results with the optimal solution in small scale. In large scale the relative efficiency of the algorithms is also shown via solving some random test problems. In small scale, deviation percentage from optimal solution for the ant colony system is less than simulated annealing algorithms with similar processing time. In large scale, the processing time of the ant colony system in compare with simulated annealing increases dramatically while their fitness has no considerable difference
