Evolutionary Constrained Optimization with Dynamic Changes and Uncertainty in the Objective Function

Abstract

Many real-life optimization problems involve dynamic changes with uncertain parameters and data, which make the decision-making process challenging. Although there are some studies on solving dynamic or uncertain problems, there is limited work on solving problems with both dynamic and uncertain characteristics. Therefore, this paper proposes an evolutionary framework for solving constrained optimization problems where the objective function's coefficients are uncertain and changing over time. In the algorithm, a mechanism is proposed for detecting a change and predicting the magnitude of uncertainty, which helps to generate better initial solutions for the evolutionary search process that improves its performance after a dynamic change. It is evaluated on 13 benchmark problems, with the reported results demonstrating its efficiency in terms of the quality of its solutions.