电动缸举升伺服机构运动学误差建模与辨识

Kinematic error modeling&identification of EMA lifting servo mechanism

针对电动缸举升伺服机构在大长径比、大惯量和大质量武器站运动学模型精度不高的问题,利用所提误差分离辨识方法对电动缸举升伺服机构的运动学模型进行了高精度建模。在明确了运动学误差来源为零位、几何、间隙和趋势误差的基础上,结合闭环矢量法与滞环模型建立了包含全项误差理论运动学模型,并定性的总结了每项误差的不同特征。通过提出的基于误差解耦的静、动态误差分离流程,有效的分离出了间隙、零位、几何和趋势误差,并分别利用双传感器法、非线性最小二乘法、傅里叶级数法和小波变换法对间隙、零位、几何和趋势误差进行了有效辨识,定量的分析了每项误差的影响度。结果表明:提出的基于搜索区间的解耦化非线性最小二乘法,在有效的分离出重现误差(几何误差),并定量的明确了几何误差为最大误差来源,将辨识出来的静态误差代入运动学模型,在实验平台的上,将平台的运动学模型精度提高96.6%。该误差分离辨识方法,对电动缸举升伺服机构的高精度运动学建模具有理论参考价值。

Aiming at the problem that the kinematics model accuracy of EMA(Electromechanical Actuator)lifting servo mechanism is not high in large length-diameter ratio,large inertia,and large mass weapon station,this paper uses the error separation identification method to model the kinematics model of EMA lifting servo mechanism with high accuracy.Based on clarifying that the kinematic error sources are zero position,geometry,clearance,and trend errors,combined with the closed-loop vector method and hysteresis model,the theoretical kinematic model including all errors is established,and the different characteristics of each error are qualitatively summarized.Through the proposed static and dynamic error separation process based on error decoupling,the gap,zero position,geometric,and trend errors are effectively separated.The gap,zero position,geometric,and trend errors are effectively identified by using the double sensor method,nonlinear least square method,Fourier series method,and wavelet transform method respectively,and the influence degree of each error is quantitatively analyzed.The experimental results show that the decoupled nonlinear least-squares method based on search interval can effectively separate the reproduction error(geometric error),and it is clear that the geometric error is the largest error source.The identified static error is substituted into the kinematic model.On the experimental platform,the accuracy of the kinematic model of the platform is improved by 96.6%.The error separation and identification method have theoretical reference values for the high-precision kinematic modeling of the EMA lifting servo mechanism.