Abstract
The time-varying ratio error estimation (TREE) is crucial for accurately assessing the real-time status of voltage transformers. However, the majority of multi-objective algorithms (MOEAs) are unable to achieve good performance on the TREE problems due to their huge search space with complicated decision variable interactions. In this paper, we propose a problem knowledge driven bi-population cooperative framework, where the knowledge (i.e., the optimal values of most decision variables in each TREE problem are similar) is exploited to enhance the convergence of algorithm for the huge search space. The proposed framework holds a weak cooperation between two populations to achieve the assistance in tackling the original TREE problems. To this aim, when the proposed framework evolves one population by a general MOEA, it evolves the other population by a sparse MOEA to assist generation of solutions that have many similar decision variables, thus promoting the convergence towards the Pareto optimal fronts of TREE problems. In the experiments, we investigate the performance of the proposed framework on five TREE problems with their number of decision variables ranging from 3000 to 15000. Experimental results show that the proposed framework outperforms seven competitors, both in terms of final results and convergence performance.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Similar Papers
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.