Abstract: Crow search algorithm is developed from imitating foraging behaviors of crows. In order to reduce the blindness of updating strategy, a crow search algorithm based on sine and cosine algorithm is proposed, which means designing sine and cosine algorithm as a local optimization operator to optimize the crow agent. The improved algorithm makes each crow agent fully absorbing the position difference information between itself and the best agent, effectively guiding the crows to approach the optimal value along the optimal direction, and improves the convergence and accuracy of the optimization. A series of standard functions are tested and the experimental results show that SCA-CSA has a better performance.

Key words: crow search algorithm, sine cosine algorithm, multimodal function

摘要： 乌鸦搜索算法模拟乌鸦觅食行为对个体位置进行更新与搜索，为降低基本乌鸦搜索位置更新策略本身存在的盲目性，将正弦余弦作为局部优化算子嵌入到基本算法中，提出了正弦余弦指引的乌鸦搜索算法。该算法通过正弦余弦操作使每一个乌鸦个体都可以充分吸收自身与最优个体的位置差信息，有效指引乌鸦个体沿最优值方向趋近最优值，改善算法的收敛效果和寻优精度。并对一系列测试函数进行寻优实验，实验结果表明该改进算法性能良好。

关键词: 乌鸦搜索算法, 正弦余弦算法, 多模态函数