Open Access
Review
Table 1
Comparison of various AI techniques used for power quality improvement.
| Attributes | AI optimization techniques |
||||
|---|---|---|---|---|---|
| PSO | ACO | GWO | GOA | BFOA | |
| Influence | Inspired by the social behavior of birds and fish. | Inspired by ant foraging behavior. | Inspired by the social and hunting behavior of grey wolves. | Mimics grasshopper movements during food search. | Incorporates chemotaxis, swarming, reproduction, and elimination-dispersal of bacteria. |
| Complexity | Low complexity, ~50–100 iterations, fast execution (~0.5–2 s) | Moderate complexity, ~100–200 iterations, runtime (~2–5 s) due to pheromone updates | Low to moderate complexity, ~80–150 iterations, runtime (~1–3 s) | Moderate complexity, balances exploration and exploitation, ~100–200 iterations | High complexity, ~200–500 iterations, longer runtime (~5–10 s) due to multiple simulation stages |
| Suitability | Effective in handling non-linear and multiobjective optimization. | Suitable for solving combinatorial and routing problems. | Ideal for nonlinear and optimization challenges. | Suitable for dynamic and nonlinear optimization problems. | Suitable for multiobjective and complex optimization problems. |
| Convergence | Provides fast convergence with moderate computational time. | Produces high-quality solutions. | Maintains a strong balance between exploration and exploitation. | Exhibits rapid convergence under favorable conditions. | Explores solution space extensively. |
| Limitations | Prone to local optima while handling high-dimensional, intricate situations. | Requires significant computational time and resources. | Requires parameter adjustments for best performance. | Prone to instability in highly dynamic scenarios. | Requires high computational resources and converges slowly. |
| Parameter Sensitivity | Needs precise parameter adjustment. | Risk of pheromone path stagnation. | Efficiency may diminish in very high-dimensional spaces. | Requires thorough parameter tuning. | Resource-intensive for large-scale systems. |
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.