An implementation of the Adaptive Multi-objective Evolutionary Algorithm based on Grid Subspaces (AGMOEA), designed to enhance both convergence and diversity in solving multi-objective optimization problems. This implementation follows the methodology proposed by Li and Wang in their 2021 research.
AGMOEA integrates a grid-based decomposition of the objective space into subspaces and dynamically allocates evolutionary effort using adaptive selection strategies. Key components include:
- Subspace dominance ranking for guided evolution
- Adaptive selection strategy to prioritize high-quality subspaces
- External archive management for solution elitism and diversity
- Multiple crossover operators to balance exploration and exploitation
It is especially useful for tackling complex multi-objective problems where maintaining solution diversity is challenging.
- Language: Python 3.8+
- Library: pymoo – A multi-objective optimization framework
No setup is required beyond installing pymoo
.
If you use this code or base your research on this work, please cite the following paper:
Linlin Li, Xianpeng Wang. "An adaptive multiobjective evolutionary algorithm based on grid subspaces", Memetic Computing (2021), 13:249–269. DOI: 10.1007/s12293-021-00336-7