Improved Artificial Electric Field Algorithm Using Nelder-Mead Simplex Method for Optimization Problems

Artificial electric field (AEF) algorithm was developed to be an alternative method for tackling with real-world engineering problems as a physics inspired meta-heuristic algorithm. Due to its stochastic nature, AEF suffers from poor exploitation, which needs to be improved. Therefore, this study attempts to obtain an effective structure by using AEF together with Nelder-Mead (NM) simplex search method. The proposed method has been named as artificial electric field with Nelder-Mead algorithm (AEF-NM). The proposed algorithm performs global search via AEF whereas NM is used to achieve better local search ability. In this way, a better performing algorithm has been achieved. Four well-known benchmark functions (Rosenbrock, Ackley, Sphere, Schwefel) were adopted to test the proposed algorithm. Comparative statistical analyses were carried out using the state-of-the-art algorithms such as sinecosine, atom search optimization, salp swarm and original AEF algorithms to observe the capability of the proposed AEF-NM. The proposed algorithm was shown to be have better performance than other compared algorithms.