Highly constrained university course scheduling using modified hybrid particle swarm optimization

At universities, course scheduling problem is an NP-hard dilemma concerned with instructor assignments and class scheduling under various constraints and restricted resources. A number of novel meta-heuristic algorithms based on the principles of particle swarm optimization, genetic algorithm, simulated annealing, tabu search, etc. were proposed and demonstrated to solve University Course Scheduling Problem (UCSP). Among several Particle Swarm Optimization (PSO) based methods, Hybrid PSO (HPSO) is the most recent one and is shown to achieve optimal solution. In those exiting algorithms including HPSO, some important features regarding co-class concept, format of classes, lab scheduling, etc. are either ignored or adopted partially. In this study the modified HPSO (MHPSO) is investigated for highly constrained environment introducing some important features with HPSO. MHPSO takes into account some significant hard and soft constraints that make it unique to solve complex UCSP with higher coverage. The proposed MHPSO has been tested on a real-world highly constrained environment and found to produce feasible solution.