An Ensemble of PSO and Artificial Electric Field Algorithm for Computationally Expensive Optimization Problems

Population-based optimization algorithms are extensively studied to address various types of optimization problems. Among them, the artificial electric field algorithm (AEFA) has gained popularity. However, AEFA still has limitations such as slow convergence speed, insufficient memory utilization, and limited search capabilities. To overcome these challenges, we propose a hybrid algorithm called particle swarm algorithm (PSO) and artificial electric field (PSAEF) that introduces a novel definition of Coulomb’s constant. The new definition of Coulomb’s constant in PSAEF enhances its exploration rate, enabling it to avoid local optima. Additionally, the algorithm incorporates knowledge of the global best solution to improve its exploitation phase and convergence rate. We evaluate the performance of PSAEF on a set of bound-constrained IEEE CEC benchmarks and compare the results with 11 state-of-the-art optimization algorithms. Extensive analyses and statistical testing, including the Wilcoxon signed-rank test, are conducted to validate the results. The experimental findings demonstrate that PSAEF outperforms other state-of-the-art algorithms in terms of accuracy and statistical significance. It achieves superior performance on 92 and 80 \(\%\) (average percentile) of the problems, respectively. The results indicate that the proposed hybrid algorithm exhibits enhanced search capabilities and faster convergence rates compared to other algorithms.