Abstract: Regarding the issues of long response time and low planning efficiency in the path planning algorithms for smart vehicles facing diverse working scenarios, a multi-element collaborative optimization strategy was proposed. Firstly, the vigilance mechanism of SSA was integrated with the population optimization characteristics of PSO, optimizing the inertia weight factor and learning factor in the PSO algorithm. Secondly, a “triangular wiring” search rule was introduced to perform bidirectional optimization(RRT-Connect) on the RRT algorithm. Subsequently, a complex environmental road simulation model was established using MATLAB software, and simulation tests were conducted on the proposed optimization solutions. The results demonstrate that, compared to single optimization approaches, the collaborative optimization algorithm exhibits significant advantages in terms of path length and planning time. Finally, real-vehicle tests are conducted on the application scenarios of the two collaborative optimization solutions, showing that in local path planning, the SSA-PSO algorithm has a shorter response time and higher planning efficiency, while in global path planning, the “triangular wiring” RRT-Connect algorithm exhibits greater advantages.

Key words: path planning, sparrow search algorithm（SSA）, particle swarm optimization（PSO） algorithm, triangular wiring, rapidly-exploring random tree（RRT） algorithm

摘要： 针对智能汽车在面对多样化工作场景时其路径规划算法存在响应时间长、规划效率低的问题，提出了多元协同优化策略。首先，融合麻雀搜索算法(SSA)的警惕机制与粒子群算法(PSO)的种群寻优特性，对PSO算法中的惯性权重因子和学习因子进行优化；其次，提出“三角布线”搜索规则，对快速扩展随机树算法（RRT）进行双向优化（RRT-Connect）；然后，基于MATLAB软件建立了复杂环境道路仿真模型，对上述优化方案进行了仿真验证。结果表明，相较于单一的优化方案，协同优化算法在路径长度与规划时间上均具有显著的优势。对两种协同优化方案的应用场景进行了实车试验，结果显示：在局部路径规划中，SSA-PSO算法响应时间更短，规划效率更高，而在全局路径规划中，“三角布线”RRT-Connect算法更具优势。

关键词: 路径规划, 麻雀搜索算法, 粒子群算法, 三角布线, 快速扩展随机树算法