| 陈继府*,伍鼎灿*,谢习华**.基于PSO-RBF神经网络的智能钻机机械臂误差研究[J].高技术通讯(中文),2025,35(10):1133~1144 |
| 基于PSO-RBF神经网络的智能钻机机械臂误差研究 |
| Research on the error of the manipulator of an intelligent drilling rig based on PSO-RBF neural network |
| |
| DOI：10. 3772 / j. issn. 1002-0470. 2025. 10. 010 |
| 中文关键词: 智能钻机； 机械臂； 误差补偿 |
| 英文关键词: intelligent drilling rig, manipulator, error compensation |
| 基金项目: |
| 作者 | 单位 | | 陈继府* | (*易普力股份有限公司长沙 410000)
(**中南大学高性能复杂制造国家重点实验室长沙 410000) | | 伍鼎灿* | | | 谢习华** | |
|
| 摘要点击次数: 18 |
| 全文下载次数: 20 |
| 中文摘要: |
| 随着基础设施建设和矿山开采的智能化需求不断增长，智能钻机已成为提高隧道施工效率和精度的关键装备。本文围绕智能钻机机械臂的运动学误差补偿技术，针对其高精度定位和自动化施工需求，提出了一套系统的解决方案，并从理论分析到实验验证进行研究。探讨机械臂误差来源及其对末端定位精度的影响，利用微分变换理论推导运动学参数误差模型。在误差补偿方法上，提出基于粒子群算法(particle swarm optimization，PSO)优化的径向基函数(radial basis function，RBF)神经网络模型。相比传统的参数补偿方法，PSO-RBF神经网络能够更高效地逼近非线性误差模型，并能提升补偿精度和鲁棒性。实验结果表明，该方法可使机械臂末端定位误差平均降低93.66%，验证了其应用于复杂机械臂补偿误差的潜力。以双臂智能钻机为实验平台，对比参数误差补偿法和PSO-RBF神经网络补偿法，进一步证明了本文智能算法的优越性和适用性。 |
| 英文摘要: |
| With the continuous growth of the intelligent requirements for infrastructure construction and mining, intelligent drilling rigs have become key equipment for improving the efficiency and accuracy of tunnel construction. Focusing on the kinematic error compensation technology of the manipulator of an intelligent drilling rig, aiming at its requirements for high-precision positioning and automated construction, this paper proposes a systematic solution and conducts in-depth research from theoretical analysis to experimental verification. It explores the sources of manipulator errors and their influence on the end-positioning accuracy, and derives the kinematic parameter error model using the differential transformation theory. In terms of error compensation methods, a radial basis function (RBF) neural network model optimized by the particle swarm optimization (PSO) algorithm is proposed. Compared with traditional parameter compensation methods, PSO-RBF neural network can approximate the non-linear error model more efficiently and significantly improve the compensation accuracy and robustness. Experimental results show that this method can reduce the average end-positioning error of the manipulator by 93.66%, verifying its potential for compensating errors in complex manipulators. Using a dual-arm intelligent drilling rig as the test platform, by comparing the parameter error compensation method and the PSO-RBF neural network compensation method, the superiority and applicability of the proposed intelligent algorithm are further proven. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
