Logarithmic mean optimization a metaheuristic algorithm for global and case specific energy optimization
Idriss Dagal, Ambe Harrison, Wulfran Fendzi Mbasso, Fritz Nguemo Kemdoum, Mohammad Khishe, Pradeep Jangir, Aseel Smerat, Saad F. Al‐Gahtani, Z. M. S. Elbarbary, Emmanuel Fendzi-Donfack, Hassan Abouobaida
Abstract
This study introduces a novel metaheuristic optimization algorithm named Logarithmic Mean-Based Optimization (LMO), designed to enhance convergence speed and global optimality in complex energy optimization problems. LMO leverages logarithmic mean operations to achieve a superior balance between exploration and exploitation. The algorithm's performance was benchmarked against six established methods-Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), Grey Wolf Optimizer (GWO), Cuckoo Search Algorithm (CSA), and Firefly Algorithm (FA)-using the CEC 2017 suite of 23 high-dimensional functions. LMO achieved the best solution on 19 out of 23 benchmark functions, significantly outperforming all comparison algorithms. It demonstrated a mean improvement of 83% in convergence time and up to 95% better accuracy in optimal values over competitors. In a real-world application, LMO was employed to optimize a hybrid photovoltaic (PV) and wind energy system, achieving a 5000 kWh energy yield at a minimized cost of $20,000, outperforming all other algorithms in both efficiency and effectiveness. The results affirm LMO's capability for robust, scalable, and cost-effective optimization in renewable energy systems.