Study On Control Techniques Of Inertial Stabilization On Moving Platform

Compared with ground based Electro-Optical tracking system, the design ofElectro-Optical tracking system based on moving bed is more and more difficult due toplatform moving as well as vibration. So a LOS stabilization subsystem must bedesigned to make sure the LOS stable in inertial space. In order to make sure the LOSstable to improve the tracking precision, Gyro stabilization platform is used to insulatethe effect of platform moving and vibration. The stabilizing principle of inertialstabilization platform is analyzed and this dissertation emphasizes on control methodsdesign to enhance disturbance rejection ability.Firstly, the inertial stabilized platform and the principle of inertial stabilizedplatform isolation carrier disturbances are analyzed. These works establish thetheoretical foundation for subsequent reach and implementation of control method. Inorder to modeling the inertial platform, adaptive modeling method and mechanismmodeling method are introduced respectively. The former do not need to know a priorknowledge of the object, it belongs to the black-box modeling and the later belongs toaccurate modeling. The carrier disturbance isolation equation and the transfer functionof inertial stabilized platform rejection carrier disturbance are established by analyzingthe transfer characteristics of disturbances in time domain and in frequency domain.The two-port Internal Model Control is proposed for stability control. Disturbances(including internal disturbance and external disturbance) are estimated based on InternalModel principle and feedforward them by internal model controller. The feedforwardloop is in parallel with the feedback loop and this has the equivalent effect of feedback+feedforward. In order to verify the ability of two-port Internal Model Control method,two inertial stabilization platform experiments system are set up. The experimentprinciple based on RTW is introduced and described its implementation process in detail.In order to reduce the effect of gyro noise, the Kalman filter, H∞filter and low pass filter are analyzed and compared. Simulation and experiment results both show that thetwo-port Internal Model Control effectively for carrier disturbance and the errorproduced by friction. Simulation and experiment results also show that the two-portInternal Model Control has better disturbance rejection at low and mid frequency thanPI method. It improves more than26dB disturbance rejection ratio at0.1Hz.Because of using the internal model principle to estimate the disturbance, theaccuracy of the model determines the disturbance estimation precision. So thecomposite internal model control is designed to overcome the shortcoming. A feedbackloop is added to the inner of the IMC loop and uses the feedback loop transfer functionas the equivalent control object. The inner loop decreases the model error and improvesdisturbance rejection ability. Experimental results show that, the composite internalmodel control achieves better disturbance rejection than feedback control in allsuppression bandwidth. The PI method achieve just39dB disturbance rejection ratio at3Hz but the composite internal achieve72dB, stable capacity increased by33dB.The composite adaptive inverse control is designed based on composite internalmodel control by adaptive tuning its external loop controller. The performance ofcontrol system by adopt different signal as error function to adjust adaptive filtercoefficients is analyzed in detail. When using the system output as error function, themodel must be associated with the control object to ensure the adaptive filterconvergence. The more accuracy of the model the faster convergence achieves. But theaccuracy of model does not affect the steady-state values of adaptive filter. Due to theadoption of the system output as the error function of adaptive filter, the control systemcan achieve good disturbance rejection when the adaptive filter converges to its optimalsolution even the characteristics of the controlled object and the disturbancecharacteristics changes.