基于深度学习的轴承加工机械臂自适应控制优化  

Adaptive Control Optimization of Bearing Processing Robot Arm Based on Deep Learning

作　　者：石纯一 SHI Chunyi(Huaibei Institute of Technology,Huaibei,Anhui 235000,China)

机构地区：[1]淮北理工学院,安徽淮北235000

出　　处：《自动化应用》2025年第6期22-24,共3页Automation Application

摘　　要：传统机械臂轨迹控制方法往往依赖于固定的轨迹规划,在面对复杂多变的轴承加工环境时,难以达到理想的控制效果。为此,提出基于深度学习的轴承加工机械臂自适应控制优化。采集并预处理轴承加工机械臂运动特征数据,基于深度学习构建LSTM模型,输入机械臂运动特征数据,输出运动轨迹预测结果。设计一个滑膜鲁棒控制器,根据预测轨迹实现轴承加工机械臂运动轨迹的自适应跟踪控制优化。实验结果表明,设计方法下的轴承加工机械臂运动轨迹控制误差仅为0.2 cm,证明该方法具有良好的控制性能。Traditional robotic arm trajectory control methods often rely on fixed trajectory planning,it is difficult to achieve ideal control effects in the face of complex and changing bearing processing environments.Therefore,a deep learning based adaptive control optimization for bearing machining robotic arms is proposed.Collect and preprocess motion feature data of bearing processing robotic arm,construct LSTM model based on deep learning,input motion feature data of robotic arm,and output motion trajectory prediction results.Design a sliding membrane robust controller to achieve adaptive tracking control optimization of the motion trajectory of the bearing processing robotic arm based on predicted trajectories.The experimental results show that the motion trajectory control error of the bearing processing robotic arm under the design method is only 0.2 cm,which proves that this method has good control performance.

关 键 词：深度学习 轴承加工 机械臂控制 自适应控制 

分 类 号：TP273.4[自动化与计算机技术—检测技术与自动化装置]