In the field of optimization, metaheuristic algorithms are widely used to solve complex problems in real life and engineering, such as resource allocation, drone path planning, and supply chain management. The dung beetle optimizer (DBO), as a type of metaheuristic algorithm, exhibits fast convergence and strong search capabilities. However, when dealing with large-scale complex optimization problems, it tends to fall into local optimal solutions and has poor convergence accuracy, which limits its application effectiveness.
Against this background, a research team consisting of scholars from the Key Laboratory of Agricultural Big Data of Hebei Province at Hebei Agricultural University, the School of Computer Science and Engineering at Beihang University, and the School of Electronic Information Engineering at Beihang University conducted a study titled “An Adaptive Dung Beetle Optimizer Based on an Elastic Annealing Mechanism and Its Application to Numerical Problems and Optimization of Reed–Muller Logic Circuits”.
This study proposes an adaptive dung beetle optimizer (ADBO) based on an elastic annealing mechanism to address the shortcomings of the original DBO. ADBO adopts three key improvement strategies: first, it adjusts the convergence factor in a nonlinearly decreasing manner to balance the needs of global exploration and local exploitation, thereby improving convergence speed and search quality; second, it introduces a greedy difference optimization strategy to increase population diversity, enhance global search capabilities, and avoid premature convergence; third, it uses an elastic annealing mechanism to perturb randomly selected individuals, helping the algorithm escape local optimal solutions and thus improving solution quality and algorithm stability.
Extensive experiments on the CEC 2017 and CEC 2022 benchmark function sets, as well as MCNC benchmark circuits, verify the effectiveness, superiority, and universality of ADBO. In terms of numerical problems, ADBO outperforms 12 state-of-the-art algorithms (including particle swarm optimization (PSO), grey wolf optimizer (GWO), and the original DBO) in 10-dimensional, 20-dimensional, 30-dimensional, 50-dimensional, and 100-dimensional spaces of the benchmark function sets, showing better convergence accuracy and global search capabilities. In the optimization of Reed–Muller (RM) logic circuits, when applied to the area optimization of 11 fixed-polarity RM (FPRM) circuits and 15 mixed-polarity RM (MPRM) circuits, ADBO can quickly find optimal polarities, and in most circuits, it achieves better optimization results than comparison algorithms, effectively reducing circuit area and improving circuit performance.
The paper “An Adaptive Dung Beetle Optimizer Based on an Elastic Annealing Mechanism and Its Application to Numerical Problems and Optimization of Reed–Muller Logic Circuits” is authored by Lixin MIAO, Zhenxue HE, Xiaojun ZHAO, Kui YU, Limin XIAO, Zhisheng HUO, Yijin WANG, and Xiaodan ZHANG. Full text of the paper: https://doi.org/10.1631/FITEE.2400967.