Many real-world applications, such as industrial manufacturing systems and water distribution networks, are complex systems, which may be hard to describe with explicit mathematical models. These are commonly labelled as black-box problems. Due to the increasing complexities of today’s fast changing environment, the demands for solving problems without any explicitly defined mathematical functions have dramatically increased. Traditional mathematical solvers, for example, the gradient descent method and the quasi-Newton methods, cannot be easily used to solve complex black-box optimization problems without gradient information. Evolutionary optimization as such provides the appropriate gradient-free alternative for solving black-box problems. However, with the rapid development of complex systems, the optimization problems become much larger. For example, the dimension of objective functions, decision variables or constraints is high, which poses challenges to existing evolutionary algorithms.
In light of the increasing demand for new innovative ways of solving real-world large complex black-box problems, this special issue aims to promote high quality research outputs in the latest advancement of evolutionary optimization, and offers a timely collection of the state-of-the-arts in the field to benefit the researchers and practitioners. To be specific, particular interest is on the interdisciplinary research of evolutionary optimization, using modern computational intelligence theories, methods and practices. Through this special issue, we believe that this would help push the boundaries of evolutionary optimization research for addressing large complex problems of the real world.