Abstract: Aiming at the defects of low performance and poor accuracy of basic sine cosine algorithm (SCA), a hybrid sine cosine algorithm (HSCA) was designed and applied to the robot path planning problem (RPP). Firstly, HSCA combined the initial solution construction method based on opposition-based learning method. Meanwhile, meme grouping and the evolutionary mechanism of TLBO (teaching-learning-based optimization) were integrated by HSCA to strengthen the information exchange of subsequent solutions, which strived to enhance search performance. Aiming at RPP problem, the spline interpolation method was introduced by HSCA in the path curve planning process to ensure the solution accuracy and reduce the optimization dimension of the current problem. Finally, function optimization and path planning tests were carried out. The experiment results show that HSCA has better performance than the comparison algorithm.

Key words: traffic and transportation engineering; robot path planning; sine cosine algorithm; opposition-based learning; teaching-learning-based optimization algorithm

摘要： 针对正余弦算法（sine cosine algorithm，SCA）性能低、精度差等缺陷，设计了混合正余弦算法（hybrid sine cosine algorithm，HSCA），并将HSCA运用于机器人路径规划（robot path planning，RPP）问题。HSCA融合了基于反向学习方法的初始解构造方法。同时，HSCA通过融入模因分组和TLBO（teaching-learning-based optimization）的进化机制来强化后续解的信息交流，力求增强搜索性能。针对RPP问题，HSCA在路径曲线规划过程中引入了Spline插值方法，旨在确保求解精度的同时降低当前问题的优化维度。最后，开展了函数寻优和路径规划测试，实验结果表明, HSCA比对比算法具有更好的性能。

关键词: 交通运输工程；机器人路径规划；正余弦算法；反向学习；教-学优化算法