Browsing by Author "PEREIRA, SAMUEL PEDRO DA ROCHA"
Now showing 1 - 1 of 1
Results Per Page
Sort Options
- Geração de horários para instituições de ensino superiorPublication . PEREIRA, SAMUEL PEDRO DA ROCHA; Bettencourt, Nuno Miguel GomesUniversity Course Scheduling Problem (UCSP) is classified as an NP-Hard problem, which is a combinatorial optimization problem that requires effective methods to manage the high level of complexity in constraints, optimization of time and resources, and assure its practical applicability to higher educational institutions. In an attempt to solve the UCSP, this document analyzes a wide array of methodologies, including Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Local Search (LS), among others. Strong emphasis is made on the aptitude and limitations of each approach, further to adaptability to specific scheduling scenarios. This work addresses the key research questions so that relevant insights, valuable for enhanced understanding, could be extracted to drive further improvement in automated scheduling systems. Based on the findings, a solution was designed and developed with a modular architecture, implementing GA and PSO, along with a dedicated correction module, which uses LS to explore nearby solutions and reduce constraint violations. Generated schedules were evaluated through fitness functions, identifying conflicts in resource assignments and penalizing them accordingly, which allowed comparisons with other schedules, including manually created ones, based on implemented criteria. This solution was then experimented with a real-world scenario, using data from the Bachelor’s Degree in Informatics Engineering at ISEP. The results compare various configurations and strategies, for each algorithm, demonstrating the effectiveness of the proposed solution in generating feasible and optimized schedules, highlighting its potential for practical deployment and future research in automated university course scheduling.