Work

MEPEOA: A Multi-Strategy Enhanced Preschool Education Optimization Algorithm for Real-World Problems

Year: 2026

Type: article

Abstract: To address the limitations of the original Preschool Education Optimization Algorithm (PEOA) in population diversity preservation and late-stage convergence accuracy, this paper proposes a Multi-strategy Enhanced Preschool Education Optimization Algorithm (MEPEOA). The proposed algorithm integrates an improved population initialization strategy, a multi-strategy collaborative search mechanism, adaptive regulation, and boundary control to achieve a more effective balance between global exploration and local exploitation. The performance of MEPEOA is comprehensively evaluated on IEEE CEC2017 and CEC2022 benchmark suites and compared with several state-of-the-art metaheuristic algorithms, including EWOA, MPSO, L_SHADE, BKA, ALA, BPBO, and the original PEOA. Experimental results demonstrate that MEPEOA achieves superior optimization accuracy and stability on the majority of benchmark functions. For example, on CEC2017 with 30 dimensions, MEPEOA reduces the average fitness value of multimodal function F9 by approximately 73.6% compared with PEOA and by more than 47% compared with EWOA. In terms of stability, the standard deviation of MEPEOA on function F6 is only 4.13 × 10−3, which is several orders of magnitude lower than those of EWOA, MPSO, and BKA, indicating highly consistent convergence behavior. Furthermore, MEPEOA exhibits clear advantages in convergence speed and robustness, achieving the best Friedman mean rank across all tested benchmark suites. In addition, MEPEOA is applied to a two-dimensional grid-based path planning problem, where it consistently generates shorter and more stable collision-free paths than competing algorithms. Overall, the proposed MEPEOA demonstrates strong robustness, fast convergence, and superior stability, making it an effective and extensible solution for complex numerical optimization and practical engineering problems.

Source: Symmetry

DOI: 10.3390/sym18010154

Authors: Shuping Ni, Chaofang Zhong, Yi Zhu, Meng Wang

Institutions: Zhejiang Normal University, Jinhua University of Vocational Technology, Zhejiang University

Language: English

Topic: Metaheuristic Optimization Algorithms Research

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: gold

Estimated APC paid: $2,896

OpenAlex ID: W7124201367

FWCI:

Cites:

Cited by:

Related to: