电子测量与仪器学报

欧阳鹏,邝先验,叶景贞.基于改进天牛群算法和三次样条插值的路径规划[J].电子测量与仪器学报,2025,39(7):259-268

基于改进天牛群算法和三次样条插值的路径规划

Path planning based on improved beetle swarm optimizationalgorithm and cubic spline interpolation

DOI：

英文关键词:good point set initialized adaptive elite mutation beetle swarm optimization algorithm cubic spline interpolation path planning

基金项目:江西省教育厅科技项目（GJJ210885）资助

作者	单位
欧阳鹏	江西理工大学电气工程与自动化学院赣州341000
邝先验	江西理工大学电气工程与自动化学院赣州341000
叶景贞	江西理工大学电气工程与自动化学院赣州341000

Author	Institution
Ouyang Peng	School of Electrical Engineering and Automation, Jiangxi University of Science and Technology, Ganzhou 341000, China
Kuang Xianyan	School of Electrical Engineering and Automation, Jiangxi University of Science and Technology, Ganzhou 341000, China
Ye Jinzhen	School of Electrical Engineering and Automation, Jiangxi University of Science and Technology, Ganzhou 341000, China

摘要点击次数: 757

全文下载次数: 592

中文摘要:

为更好地解决移动机器人路径规划问题，对天牛群算法的性能进行改进，并拓展其应用领域，提出了一种基于自适应精英变异的改进天牛群算法(AEM BSO)。首先运用佳点集对天牛种群进行初始化操作，使种群分布更加均匀，降低陷入局部最优解的风险；其次使用非线性递减惯性权重策略，以提升算法的初期探索能力和收敛速度；然后引入精英变异策略，在每次迭代时对精英个体进行变异以产生新的、可能更优的候选解，避免早熟现象和陷入局部最优解。将AEM-BSO应用于求解移动机器人全局路径，借助三次样条插值方法，对规划得出的全局路径节点进行平滑化处理，使路径更贴合实际运动需求。最后在10个测试函数上和不同环境地图上评估AEM-BSO的有效性。实验结果表明，AEM-BSO在不同测试函数中具有较好的寻优精度和稳定性能，在机器人路径规划中路径长度较原始BSO算法、粒子群算法和蝙蝠算法分别减少了0.24%、18.12%与8.41%，标准差分别减少了25.8%、96.73%、14.13%，表明了AEM-BSO算法的有效性。

英文摘要:

To address the optimization requirements in mobile robot path planning and enhance the performance limitations of the beetle swarm optimization algorithm regarding convergence precision and application scope, this paper introduces an adaptive elite mutation-based beetle swarm optimization (AEM-BSO) algorithm. The methodological innovations manifest in three principal aspects. Firstly, the implementation of good point set initialization ensures uniform population distribution, effectively mitigating the risk of local optima entrapment. Subsequently, a non-linearly decreasing inertia weight strategy enhances global exploration capabilities during initial iterations while accelerating convergence rates in later stages.Furthermore,incorporation of elite mutation mechanisms that strategically perturb high-performing individuals during iterative processes to prevent premature convergence.For practical implementation in mobile robot navigation, cubic spline interpolation optimizes waypoint connections in generated paths, ensuring kinematic feasibility and smooth trajectory formation.Comprehensive validation across 10 benchmark functions and diverse environmental maps demonstrates the algorithm’s superior optimization precision and robust stability.Experimental comparisons reveal that AEM-BSO achieves respective path length reductions of 0.24%, 18.12%, and 8.41% compared to primitive BSO, PSO and BA, accompanied by significant standard deviation decreases of 25.8%, 96.73%, and 14.13%.These quantitative improvements substantiate the proposed algorithm’s effectiveness in balancing exploration-exploitation trade-offs and enhancing solution quality for complex path planning tasks.

查看全文查看/发表评论下载PDF阅读器