融合黄金正弦与sigmoid连续化的海鸥优化算法

针对海鸥算法(SOA)在迭代寻优过程中容易陷入局部最优、收敛速度慢以及寻优精度低等缺陷,提出一种黄金正弦引导与sigmoid连续化的海鸥优化算法(GSCSOA)。在海鸥迁徙阶段,使用sigmoid函数作为非线性收敛因子引导海鸥搜寻过程,使得算法前期保持更强的全局寻优能力,后期更快收敛。在海鸥扑食阶段,引入禁忌搜索的思想,使得海鸥始终向着置信度更高的区域移动,并且在一次迭代中最优位置持续变化,从而提高寻优精度。之后使用黄金正弦机制指引种群位置更新,缩小搜索范围,提高局部寻优能力。最后,用12个测试函数和CEC2014函数集对改进算法进行测试,并与原算法以及其他算法的实验结果进行对比,结果证明改进的海鸥算法在收敛速度和精度上的表现更优。

Seagull optimization algorithm combining golden sine and sigmoid continuity

Aiming at the problems in the iterative process of seagull optimization algorithm(SOA), such as local optimization, slow convergence speed and low optimization accuracy, this paper proposed a golden sine guide and sigmoid continuous seagull optimization algorithm(GSCSOA). In the seagull migration stage, the algorithm used the sigmoid function as a nonlinear convergence factor to guide the seagull search process, so that the algorithm maintained a stronger global optimization ability in the early stage and converged faster in the later stage. In the seagull rushing stage, it introduced the idea of Taboo search, so that the seagulls always moved to the area with higher confidence, and the optimal position continued to change in one iteration, which improved the optimization accuracy. After that, it used the golden sine mechanism to guide the population position to update, which narrowed the search range, and improved the local optimization ability. Finally, this paper used 12 test functions and CEC2014 function set to test the improved algorithm, and the results prove that the improved seagull algorithm has better convergence speed and accuracy than original algorithm and other algorithms.