摘 要
随着制造业向智能化、高效化发展，自动化装配成为提升生产效率和产品质量的关键环节，而机器人路径规划作为自动化装配的核心问题备受关注。本研究旨在解决复杂环境下机器人在自动化装配中的路径规划难题，以提高装配精度与效率。基于此目的，提出一种融合深度强化学习与改进人工势场法的混合路径规划算法。该方法利用深度强化学习强大的环境适应能力学习最优策略，通过改进人工势场法快速避开障碍物并减少局部极小值点的影响。实验结果表明，所提算法能够有效应对复杂多变的装配环境，在保证高精度的同时显著缩短了路径规划时间，平均路径长度较传统方法减少约20%，规划成功率提高了35%。此外，该算法具有良好的泛化能力，可应用于不同类型的自动化装配任务中。本研究创新性地将两种不同原理的方法有机结合，为解决自动化装配中机器人路径规划问题提供了新思路，对推动智能制造技术的发展具有重要意义。
关键词：机器人路径规划 深度强化学习 改进人工势场法
Abstract
As manufacturing evolves towards intelligence and efficiency, automated assembly has become a critical factor in enhancing production efficiency and product quality, with robot path planning emerging as a core challenge. This study addresses the complexities of robot path planning in automated assembly within intricate environments to improve assembly accuracy and efficiency. To this end, a hybrid path planning algorithm that integrates deep reinforcement learning with an improved artificial potential field method is proposed. The approach leverages the robust environmental adaptability of deep reinforcement learning to learn optimal strategies while utilizing the enhanced artificial potential field method to swiftly avoid obstacles and mitigate the impact of local minima. Experimental results demonstrate that the proposed algorithm effectively handles complex and dynamic assembly environments, significantly reducing path planning time while maintaining high precision. Compared to traditional methods, the average path length is reduced by approximately 20%, and the planning success rate increases by 35%. Furthermore, the algorithm exhibits excellent generalization capabilities, making it applicable to various types of automated assembly tasks. Innovatively combining two distinct methodologies, this research offers a novel approach to solving robot path planning problems in automated assembly, contributing significantly to the advancement of smart manufacturing technologies.
Keyword:Robot Path Planning Deep Reinforcement Learning Improved Artificial Potential Field Method
目 录
1绪论 1
1.1自动化装配的背景与意义 1
1.2国内外研究现状综述 1
1.3本文研究方法概述 2
2路径规划基础理论 2
2.1路径规划基本概念 2
2.2常用路径规划算法 3
2.3约束条件与优化目标 3
3关键技术分析 4
3.1环境建模方法 4
3.2障碍物规避策略 5
3.3实时路径调整机制 5
4应用案例研究 6
4.1典型装配任务分析 6
4.2路径规划方案设计 6
4.3实验结果与讨论 7
结论 7
参考文献 9
致谢 10
