Compensation-based Characteristic Modeling and Tracking Control for Electromechanical Servo Systems With Backlash and Torque Disturbance.

For two-inertia servo mechanisms, the high order, backlash nonlinearity and external disturbance make the precise modeling and control difficult to implement. This article provides an easily implemented modeling and control strategy to deal with this problem. A characteristic modeling framework of disturbed nonlinear systems is proposed. To restrain the modeling error, the backlash nonlinearity and torque disturbance are observed by constructing a finite-time extended state observer (FESO) based on homogeneity properties and then the compensation action can be taken. Based on the compensated system, the discrete-time characteristic model is established using the sampled input-output data, which degrades the modeling complicacy. To estimate the model parameters, an adaptation law with projection algorithm is proposed using the tracking error and the estimation error as the excitation signal. A discrete-time second-order fast terminal sliding-mode control (DSFTSC) is proposed based on the characteristic model to stabilize the whole system, where an improved reaching law is designed to enhance the rapidity and weaken the chattering and the utilization of the fast terminal switching surface also speeds up the regression rate and decreases the tracking error. Finally, the effectiveness of the characteristic modeling, the adaptive law and the control scheme is validated by simulations in Matlab and experiments in a practical test rig, respectively. [ABSTRACT FROM AUTHOR]