MSSSA: a multi-strategy enhanced sparrow search algorithm for global optimization

Abstract

The sparrow search algorithm (SSA) is a recent meta-heuristic optimization approach with the advantages of simplicity and flexibility. However, SSA still faces challenges of premature convergence and imbalance between exploration and exploitation, especially when tackling multimodal optimization problems. Aiming to deal with the above problems, we propose an enhanced variant of SSA called the multi-strategy enhanced sparrow search algorithm (MSSSA) in this paper. First, a chaotic map is introduced to obtain a high-quality initial population for SSA, and the opposition-based learning strategy is employed to increase the population diversity. Then, an adaptive parameter control strategy is designed to accommodate an adequate balance between exploration and exploitation. Finally, a hybrid disturbance mechanism is embedded in the individual update stage to avoid falling into local optima. To validate the effectiveness of the proposed MSSSA, a large number of experiments are implemented, including 40 complex functions from the IEEE CEC2014 and IEEE CEC2019 test suites and 10 classical functions with different dimensions. Experimental results show that the MSSSA achieves competitive performance compared with several state-of-the-art optimization algorithms. The proposed MSSSA is also successfully applied to solve two engineering optimization problems. The results demonstrate the superiority of the MSSSA in addressing practical problems.

摘要

麻雀搜索算法(SSA)是一种新的元启发式优化方法，具有简单和灵活的优点。然而，在处理多模态优化问题时，该算法仍存在早熟收敛、探索与开发不平衡等缺陷。针对上述问题，本文提出一种多策略增强的麻雀搜索算法(MSSSA)。首先，引入混沌映射以获取高质量的初始种群，并采用对立学习策略增加种群的多样性。其次，设计了一种自适应参数控制策略，以在全局探索与局部开发之间保持适当的平衡。最后，在个体更新阶段嵌入混合扰动机制，以避免算法陷入局部最优。为了验证所提方法的有效性，在IEEE CEC2014和IEEE CEC2019测试集的40个函数，以及10个不同维度的经典函数上进行了大量的实验。实验结果表明，与一些先进的算法相比，所提出的MSSSA表现出突出的优化性能。该算法还成功地应用于两个工程优化问题，证明了MSSSA在解决实际问题方面的优越性。