考虑多因素影响的进给系统等效惯量辨识方法  

Inertia identification method of feed drive system considering multiple factors

作　　者：湛承鹏 杨啸 赵万华[1] ZHAN Cheng-peng;YANG Xiao;ZHAO Wan-hua(School of Mechanical Engineering,Xi′an Jiaotong University,Xi′an 710054,China)

机构地区：[1]西安交通大学机械工程学院,西安710054

出　　处：《组合机床与自动化加工技术》2020年第9期99-103,共5页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：高档数控机床与基础制造装备"国家科技重大专项资助项目"(2017ZX04013001);中国博士后科学基金面上项目(2018M633498)。

摘　　要：为准确辨识进给系统等效惯量值,提出一种考虑电流环滞后及Stribeck摩擦影响的进给系统等效惯量离线辨识方法。首先建立考虑电流环滞后及Stribeck摩擦力影响的伺服进给系统离散模型。而后根据所建模型构造适应度函数,将参数辨识问题转化为参数优化问题。最后,基于差分进化算法求解该优化问题,获得进给系统等效惯量值。所提方法的有效性通过实验及仿真进行了验证。结果显示,与现有仅考虑库仑摩擦力的辨识方法相比,本文方法在不同幅值辨识指令下惯量辨识结果差异更小,惯量辨识结果更加准确。In this paper, an accurate offline identification method for the equivalent inertia of feed drive systems is proposed, in which current loop dynamics and Stribeck friction are considered. Firstly, A discrete transfer function model of feed drive system that considers the first-order dynamics of the current loop and Stribeck friction is established. Secondly, the fitness function is constructed by the use of discrete transfer function model. Thirdly, the fitness function is solved based on the differential evolution algorithm, and the inertia of feed drive system is obtained. The advantages of proposed method were verified on a test bench. The experimental results show that compared with the existing methods, the variation of the inertia value identified by the proposed method is smaller under different amplitude input commands. And Inertia identification accuracy is also greatly improved.

关 键 词：伺服进给系统 惯量辨识 电流环滞后 Stribeck摩擦力 差分进化算法 

分 类 号：TH16[机械工程—机械制造及自动化] TG506[金属学及工艺—金属切削加工及机床]