DM-GWO: An Extended Grey Wolf Optimizer for Dynamic Multi-Objective Problems

In this paper, a novel dynamic multi-objective gray wolf optimizer (DM-GWO) is developed to handle the dynamic multi-objective optimization problems (DMOPs). In particular, a group of sentinel wolves is employed to detect environmental changes via fitness re-evaluation, and the idea of diversity enhancement is adopted to deal with the dynamism so as to track the time-varying Pareto front rapidly. Moreover, a non-dominate sorting based threshold-resplacement mechanism is designed to alleviate the randomness in introducing new solutions. Benchmark evaluation results on FDAI test function have demonstrated the effectiveness of proposed DM-GWO, in addition, sensitivity analysis of DM-GWO is carried out, which presents parameterization suggestions in different situations. To sum up, proposed DM-GWO is a reliable and competent dynamic multi-objective optimization algorithm, which has provided valuable experiences for extension works of other heuristic algorithms to handle DMOPs.