A change severity degree-based dynamic multi-objective optimization algorithm with adaptive response strategy

Najwa Kouka, Rahma Fourati, Raja Fdhila, Amir Hussain, Adel M. Alimi

Abstract: Highlights•An adaptive response strategy is proposed to handle environmental changes.•LSTM prediction is used to generate an initial population of high-quality solutions.•For small change degree, solutions are updated using elite learning with knee point.•Multi-swarm with shared information is used to promote solutions' diversity.•Experiments on IEEE CEC 2018 DMOP and OL-DOPs benchmarks validated our DMOPSO-ARS.