Abstract:  Aiming at the shortcomings of poor local development ability， slow convergence rate and low solution accuracy in solving the function optimization problems of standard SCA， a sine cosine algorithm based on the honey gathering mechanism（SCAHGM） was proposed. To begin with， the positions of parameters r2， r3 and r4 were changed， so as to adopt the same parameters r1， r2， r3 and r4 for each individuals， meanwhile the parameters r1 and r3 were adjusted respectively by power decreasing function adaptively and in oneform dynamically， which reduced randomness and improved search efficiency of the algorithm. Then， the greedy selection strategy， the operators of employed bees and reconnaissance bees were utilized， which sped up the convergence rate， improved the optimization accuracy of the algorithm， increased diversity of the population and protected the algorithm from falling into local optimum. Besides， the sine cosine operator or employed bees’ operator was executed alternately with a certain probability in the iterative processes of the algorithm， which better balanced global exploration and local development capability of the algorithm. At last， simulation results with 23 standard benchmark functions illustrate that SCAHGM performs better optimization performance than that of basic SCA， improved algorithms for SCA and other meta-heuristic algorithms， and the feasibility and applicability were verified by optimizing two mechanical design examples. 

Key words: swarm intelligence, sine cosine algorithm（SCA）, artificial bee colony algorithm, honey gathering mechanism, greedy selection, mechanical design optimization

摘要： 针对标准正弦余弦算法（SCA）在求解函数优化问题时存在局部开发能力差、收敛速度慢和求解精度低等问题，提出了一种基于采蜜机制的正弦余弦算法（SCAHGM）。首先，更改参数r2、r3、r4的位置，使每个个体采用相同的参数r1、r2、r3和r4，按幂递减函数自适应调整参数r1，并动态调整参数r3，减少随机性，提高算法搜索效率；其次，利用贪婪选择策略、采蜜蜂算子、侦察蜂算子，加快算法收敛速度，提高算法优化精度，增加种群多样性，防止算法陷入局部最优；然后，在算法迭代过程中，以一定概率交替执行正余弦算子或采蜜蜂算子，更好地平衡算法的全局探索与局部开发能力。最后，选取23个标准测试函数进行仿真实验，结果表明SCAHGM算法较标准SCA、改进SCA和其他元启发式算法具有更佳的寻优性能，并通过优化2个机械设计实例，验证了SCAHGM算法的可行性和适用性。

关键词: 群体智能, 正弦余弦算法, 人工蜂群算法, 蜂群采蜜机制, 贪婪选择, 机械设计优化