Application of Genetic Algorithms in Academic Course Scheduling Systems

Abstract

Resource allocation optimization problems are prevalent in dynamic and constraint-rich real-world scenarios, posing significant challenges due to their complexity and conflicting objectives. This study proposes a novel multi-objective genetic algorithm (MOGA) designed to effectively balance trade-offs in resource allocation problems. The algorithm employs Pareto-based selection to manage conflicts among objectives, adaptive crossover and mutation operators to dynamically adjust search strategies, and a dynamic constraint-handling mechanism to enhance feasibility and solution diversity. Experiments using real-world data demonstrate the algorithm's superior performance in terms of convergence speed, solution quality, and Pareto front diversity compared to traditional approaches such as single-objective genetic algorithms and heuristic-based simulated annealing. The results further validate the algorithm's general applicability to a wide range of multi-objective optimization problems, including resource allocation and scheduling. This study provides a robust and scalable optimization framework, offering valuable insights for addressing complex multi-objective problems across various domains.