Metaheuristics for Real-Time Optimization

The rapid progress of computational technologies and their widespread adoption have resulted in a growth of complex, real-world optimization problems in all fields, including logistics, industrial scheduling, clustering, and classification. Exact methods usually have an appealing theory but can often be computationally impractical, and greedy heuristics can provide quick but often poor solutions. In addition to offering a means of escaping local optima, metaheuristics can offer a pleasant balance between constructive procedures and population- based and local search techniques. While many new metaheuristics that are now emerging from natural, social, and human processes are broadening the definition of optimization, many traditional techniques, including Genetic Algorithms, Simulated Annealing, Tabu Search, and Ant Colony Optimization, can be used effectively for many real-life scheduling, clustering, and allocation problems. In this study, we discuss the role of metaheuristics and describe five recent works that represent an adequate combination of flexibility and robustness in solving real- world optimization problems. These include berth allocation in ports, a fisherman-like global search method, ensemble construction for multiclassifier systems, gravitational clustering, and course timetabling. We show that metaheuristics can not only provide effective solutions to problems that have not been solved previously or are too computationally expensive to solve but continue to change, adapt, and hybridize through new inspiration.