融合流体扰动与近端策略优化算法的路径规划研究OA
Research on Path Planning Integrating Fluid Disturbance and Proximal Policy Optimization Algorithms
路径规划算法是机器人技术以及自动化系统等领域的核心问题之一.针对三维动态空间中易出现移动目标、动态威胁等,构建智能体路径规划的仿真环境及相关模型,确定仿真流程.运用近端策略优化对改进的流体扰动算法的初始路径进行修正,实现最优决策.并进行仿真验证,结果表明在三维动态空间中,该算法在路径规划的收敛效率、路径平滑度和避障成功率等方面均优于对比算法.对于三维动态空间中易发生出现可移动障碍物等情况,此算法优化改进的流体扰动算法,具有较好的鲁棒性和良好的适应性.
Path planning is one of the key problems in robotics and automation systems.Aiming at the frequent occurrence of moving targets and dynamic threats in the three-dimensional(3D)dynamic space,the simulation environment and related models of agent path planning are constructed,and the simulation process is determined.This study applies Proximal Policy Optimization(PPO)to adjust the initial path generated by the improved Fluid Disturbance algorithm,thereby achieving optimal decision-making.Simulation results demonstrate that,in 3D dynamic space,this algorithm outperforms baseline algorithms in terms of path planning convergence efficiency,path smoothness,and obstacle avoidance success rate.For scenarios involving movable obstacles and other dynamic conditions in 3D dynamic space,this algorithm optimizes the improved Fluid Disturbance algorithm and exhibits strong robustness and good adaptability.
谷志新;贾珂怡;徐凯宏
东北林业大学计算机与控制工程学院,哈尔滨 150040东北林业大学计算机与控制工程学院,哈尔滨 150040东北林业大学计算机与控制工程学院,哈尔滨 150040
信息技术与安全科学
流体扰动算法近端策略优化算法路径规划三维动态空间
fluid disturbance algorithmPPOpath planning3D dynamic space
《火力与指挥控制》 2026 (3)
66-73,8
国家重点研究与发展计划基金资助项目(2022YFD1301104)
评论