The study of efficient metaheuristics to solve optimization problems has been the subject of research by the scientific community. To find good results in terms of solutions of quality and computational time is important to define which parameter setting should be used. This process is often hard due to relation between the parameters, the best values for the parameters depend on the problem, the instances and the search time. Thus, the main idea of this project is improve the hybrid Clustering Search method (CS), using the concept of random keys by the Biased Random-Key Genetic Algorithms (BRKGA) and studying new automated ways to detect promising regions and change the parameter values during the search. We intend to design a method with easiness of implementation and robustness in terms of quality of solution and search time. To analyze the performance of the proposed method we will solve two optimization problems: berth allocation and cranes scheduling problem and point-feature cartographic label placement problem. In the evaluation we should be used instances available in public libraries and, if possible, some real-life instances. Methods from statistical analysis will be use to conduct the performance assessment of the designed method.
News published in Agência FAPESP Newsletter about the scholarship: