Abstract: To address the issues of strong sampling randomness, low efficiency in random tree expansion, and poor path smoothness in the traditional Informed RRT^* algorithm for path planning, an improved Informed RRT^* algorithm is proposed. The algorithm adopts a hybrid sampling strategy that combines Gaussian-biased sampling with uniform random sampling to reduce sampling randomness and improve the efficiency of initial path discovery. Bayesian learning is introduced to adaptively adjust the sampling weights and the Gaussian standard deviation. During node expansion, an improved artificial potential field and a greedy step-size expansion strategy are incorporated to enhance the directionality of random tree growth and increase expansion efficiency. In the path post-processing stage, a dual-scale sliding- 引用格式:冯忠居,于明威,张聪,等.冲刷场地地震波形对大直径变截面单桩时程响应的影响［J］.河北大学学报(自然科学版),2026,46(4):337-348.引用格式:张振利,谢飞,袁达凯,等.基于改进Informed RRT^*的移动机器人路径规划算法［J］.河北大学学报(自然科学版),2026,46(4):424-436.DOI:10.3969/j.issn.1000-1565.2026.04.010基于改进Informed RRT^*的移动机器人路径规划算法张振利^1,2,谢飞^1,2,袁达凯^1,2,韩树人^1,2(1.江西理工大学 电气工程与自动化学院,江西 赣州 341000;2.江西理工大学 多维智能感知与控制江西省重点实验室,江西 赣州 341000)摘 要:针对传统的Informed RRT^*算法在路径规划中存在采样随机性强、随机树扩展效率低和路径不平滑等问题,提出一种改进Informed RRT^*算法.该算法采用高斯偏置与均匀随机相结合的混合采样策略来降低采样随机性和提高初始路径搜索效率,并利用贝叶斯学习自适应调整采样权重与高斯标准差;在节点扩展阶段融合改进的人工势场与贪婪步长扩展策略,以增强随机树扩展的目的性并提高扩展速度;在路径后处理中,采用双尺度滑动窗口裁剪与欧拉螺线-圆弧-欧拉螺线(Clothoid-Arc-Clothoid, CAC)曲线平滑方法,以减少冗余节点并提高路径平滑性.通过仿真实验对比表明,改进Informed RRT^*算法获取初始路径时间更短,算法收敛效率更高,所得路径的长度更短且更平滑.关键词:Informed RRT^*; 路径规划; 高斯偏置; 人工势场; CAC曲线中图分类号:TP242.2 文献标志码:A 文章编号:1000-1565(2026)04-0424-13DOI:10.3969/j.issn.1000-1565.2026.04.010Path planning algorithm of mobile robot based on improved Informed RRT^*ZHANG Zhenli^1,2, XIE Fei^1,2, YUAN Dakai^1,2, HAN Shuren^1,2(1.School of Electrical Engineering and Automation, Jiangxi University of Science and Technology, Ganzhou 341000, China; 2.Jiangxi Provincial Key Laboratory of Multidimensional Intelligent Perception and Control, Jiangxi University of Science and Technology, Ganzhou 341000, China)Abstract: To address the issues of strong sampling randomness, low efficiency in random tree expansion, and poor path smoothness in the traditional Informed RRT^* algorithm for path planning, an improved Informed RRT^* algorithm is proposed. The algorithm adopts a hybrid sampling strategy that combines Gaussian-biased sampling with uniform random sampling to reduce sampling randomness and improve the efficiency of initial path discovery. Bayesian learning is introduced to adaptively adjust the sampling weights and the Gaussian standard deviation. During node expansion, an improved artificial potential field and a greedy step-size expansion strategy are incorporated to enhance the directionality of random tree growth and increase expansion efficiency. In the path post-processing stage, a dual-scale sliding- 收稿日期:2025-12-01;修回日期:2026-04-07 基金项目:国家自然科学基金项目(62363013);多维智能感知与控制江西省重点实验室项目(2024SSY03161) 第一作者:张振利(1976—),男,江西理工大学副教授,主要从事人工智能、检测技术与控制方向研究.E-mail:47717770@qq.com 通信作者:韩树人(1983—),女,江西理工大学讲师,主要从事网络化测控和物联网技术方向研究.E-mail:18911107@qq.com 第4期张振利等:基于改进Informed RRT^*的移动机器人路径规划算法河北大学学报(自然科学版) 第46卷window pruning method and a Clothoid-Arc-Clothoid(CAC)curve smoothing technique are employed to remove redundant nodes and improve path smoothness. Simulation results demonstrate that the improved Informed RRT^* algorithm achieves shorter initial path search time, higher convergence efficiency, and generates shorter and smoother paths.

Key words: Informed RRT^*, path planning, Gaussian offset, artificial potential field, Clothoid-Arc-Clothoid curve