摘要
误差补偿是提高机械加工精度的有效途径,在分析测试数控直线电机进给定位精度的基础上,提出基于最小二乘支持向量机的直线电机进给定位精度误差回归建模和预测方法。最小二乘支持向量机具有完备的统计学习理论基础和学习功能,它用核函数建立预测模型,再用已知数据为学习样本训练学习机,用检验样本进行验证、预测系统未来误差。采用径向基核函数的最小二乘支持向量机对不同速度加速度下的定位误差进行了预测,并进行误差补偿。研究结果表明,采用LSSVM的方法可以较大地提高直线电机进给的定位精度。
Error compensation is an effective approach to enhance machining accuracy. Firstly, feed location accuracy of linear motor were analyzed and tested. Then least square support vector machines (LSSVM) based method of feed location accuracy error modeling and prediction was put forward. LSSVM has excellent learning and statistical functions. Predicting model was founded by kernel function, learning machine was trained with learning sample used known data, further error of the system was tested and predicted with testing sample. Position error prediction about various speed and acceleration were put forward based on least square support vector machines introduction radial group kernel function, and error compensation was processed. Research results indicate that LSSVM based method can improve feed location accuracy of linear motor significantly.
出处
《机械工程师》
2008年第4期49-51,共3页
Mechanical Engineer