引用本文:张珩,肖歆昕.间隙引发的变载非线性及其闭环辨识[J].控制理论与应用,2019,36(5):673~679.[点击复制]
ZHANG Heng,XIAO Xin-xin.Onload-offload cycling nonlinearity caused by backlash and its closed-loop identifications[J].Control Theory and Technology,2019,36(5):673~679.[点击复制]
间隙引发的变载非线性及其闭环辨识
Onload-offload cycling nonlinearity caused by backlash and its closed-loop identifications
摘要点击 2552  全文点击 948  投稿时间:2018-03-29  修订日期:2018-07-23
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2018.80216
  2019,36(5):673-679
中文关键词  间隙变载非线性  间隙闭环辨识  非线性效应  描述函数  自激振荡
英文关键词  backlash onload-offload cycling nonlinearity  closed-loop backlash identification  nonlinear effect  describing function  self excited oscillation
基金项目  国家重点基础研究发展计划
作者单位邮编
张珩 中国科学院力学研究所 100190
肖歆昕* 中国科学院力学研究所 100190
中文摘要
      阐明了齿轮传动位置伺服系统间隙除产生熟知的前向迟滞非线性之外, 还将因系统往复运动中加卸载相交替的反向调制致使驱动电机动力学参数同步跳变所出现的变载非线性现象. 推导建立了这类非线性在一阶谐波意义下的串联式等效频域模型及其近似计算公式, 分析了相应的幅相特性. 进一步面向服役中系统提出了直接通过闭环弱自激振荡控制所实现的间隙辨识新方法, 给出了控制器的结构及其参数整定策略. 此外, 文中还给出了含高斯白噪声测量的仿真算例, 充分验证了闭环间隙辨识方法的有效性和便捷性.
英文摘要
      It is clarified in this paper that backlash in position servo system with gear mechanisms, apart from causes the well-known forward hysteresis characteristics, leads to the onload-offload cycling nonlinearity as a result of synchronous jumps of driving dynamics parameter due to the naturally backward modulation result from repetitive changes between loading and unloading state in reciprocating motions. An equivalent frequency-domain model and its approximate formula are established for this nonlinearity within the first harmonic wave sense, and the magnitude frequency and phase frequency characteristics of the model are correspondently analyzed. Furthermore, a new backlash identification method based on closed-loop weakly self-excited oscillation is presented for systems in active service, hence a controller tuning algorithm is suggested. The effectiveness and accessibility of the proposed method are additionally verified via numerical simulations on specific example with different backlash values under white Gaussian noise measurements.