ESO: An enhanced snake optimizer for real-world engineering problems

Abstract

Meta-heuristic algorithms are an essential way to solve realistic optimization problems. Developing effective, accurate, and stable meta-heuristic algorithms has become the goal of optimization research. Snake Optimizer (SO) is a novel algorithm with good optimization results. However, due to the limitations of natural laws, the parameters are more fixed values in the exploration and exploitation phase, so the SO algorithm quickly falls into local optimization and slowly converges. This paper proposes an Enhanced Snake Optimizer (ESO) by introducing a novel opposition-based learning strategy and new dynamic update mechanisms (parameter dynamic update strategy, sine–cosine composite perturbation factors, Tent-chaos & Cauchy mutation) to achieve better performance. The effectiveness of ESO has been tested on 23 classic benchmark functions and the CEC 2019 function set. 13 functions of 23 classic benchmark functions have variants that belong to multiple dimensions (Dim = 30, 100, 500, 1000, and 2000). In addition, ESO is also used to solve four real-world engineering design problems. Experimental results and two statistical tests show that the proposed ESO performs better than the other 13 state-of-the-art algorithms, including SO. The MATLAB code of ESO is available from: https://ww2.mathworks.cn/matlabcentral/fileexchange/120173-eso-an-enhanced-snake-optimizer.