Table 2 Comparison of optimization methods, objectives, algorithm parameters, and characteristics.
From: An axiomatic system engineering design method based on NSGA-II algorithm applied to complex systems
Method | Optimization objective | Algorithm parameters | Characteristics |
|---|---|---|---|
Traditional reduction algorithm | None | – | Performs strongly connected component decomposition followed by topological sorting. Deterministic approach without stochasticity, yielding a unique fixed ordering. |
GA | Single objective: minimize UV | Population size = 100; 50 generations; crossover rate = 0.7; mutation rate = 0.3 (shuffle); selection: tournament | Outputs a single best solution based on evolutionary search. |
NSGA-II | Multi-objective: minimize UV, maximize SD | Population size = 400; 350 generations; crossover rate = 0.7; mutation rate = 0.3; non-dominated sorting with crowding distance | Generates a Pareto front of trade-off solutions; three representative solutions are selected for analysis. |
Evaluation metrics | UV (Upper-triangular dependencies); SD (Structural modularity score) | – | Used for visual comparison of DSM reordering and quantitative performance evaluation. |
