A Novel Artificial Electric Field Algorithm for Solving Global Optimization and Real-World Engineering Problems | JoVE Visualize

Area of Science:

Computational Intelligence
Optimization Algorithms
Metaheuristic Computing

Background:

The Artificial Electric Field Algorithm (AEFA), inspired by physics, faces convergence and suboptimal solution challenges in high-dimensional problems.
Existing metaheuristics require enhancements for improved exploration, exploitation, and diversity to tackle complex optimization tasks.

Purpose of the Study:

To introduce a modified Artificial Electric Field Algorithm (mAEFA) that integrates Lévy flights, simulated annealing, and Adaptive s-best Mutation and Natural Survivor Method (NSM).
To enhance the search space, exploration potential, and robustness of AEFA, aiming for a better balance between local intensification and global diversification.
To evaluate the performance and practical compatibility of mAEFA on diverse constraint and engineering benchmark problems.

Main Methods:

Development of the modified Artificial Electric Field Algorithm (mAEFA) by incorporating Lévy flights for enhanced exploration.
Integration of simulated annealing for improved search exploitation and Adaptive s-best Mutation and Natural Survivor Method (NSM) for increased diversity.
Comprehensive quantitative and qualitative assessment using 29 CEC'17 constraint benchmarks and five engineering design problems.

Main Results:

mAEFA demonstrated superior performance against the LCA algorithm on all 29 CEC'17 test functions (100% superiority).
mAEFA outperformed SAO, GOA, CHIO, PSO, GSA, and AEFA in a significant percentage of CEC'17 test cases (96.6% down to 58.6%).
mAEFA achieved top performance in three out of five engineering design problems and secured second place in the remaining two.

Conclusions:

The proposed mAEFA effectively addresses the limitations of the original AEFA, showing improved performance and robustness.
The integration of Lévy flights, simulated annealing, and Adaptive s-best Mutation and NSM enhances the algorithm's ability to handle diverse optimization problems.
mAEFA establishes itself as a competitive and effective metaheuristic for complex optimization tasks in both theoretical and practical engineering domains.