Abstract

This paper proposes a novel hybrid optimization algorithm, termed Arithmetic Smell Agent Optimization (ASAO), which integrates the global search strengths of Smell Agent Optimization (SAO) with the local refinement capabilities of the Arithmetic Optimization Algorithm (AOA). The hybridization aims to overcome common limitations such as premature convergence and poor local optima entrapment. ASAO dynamically tunes AOA parameters using SAO mechanisms, ensuring a balanced exploration–exploitation process. The algorithm was evaluated using a suite of 14 benchmark functions of varying complexity. Results show that ASAO consistently outperformed or matched both SAO and AOA in terms of convergence speed, solution accuracy, and stability. Specifically, ASAO achieved improved fitness values, reduced computational iterations, and smoother convergence profiles across multiple test cases. These findings confirm ASAO’s robustness, efficiency, and reliability in solving complex optimization problems, highlighting its potential for real-world engineering applications requiring dynamic, high-dimensional, and nonlinear solution strategies.

Key words: Arithmetic Smell Agent Optimization, Arithmetic Optimization Algorithm, Objective Function, Optimization, Smell Agent Optimization.