Study On The Issues Of Parameter Identification And Friction Compensation Control For Precise Electro-optical Stabilized Platforms

Electro-Optical Stabilized Platforms(EOSPs) is the key component for the missions such as target acquisition, tracking, and location on the moving vehicles, which is widely applied in different fields such as electro-optical reconnaissance, precision guidance and advance fire control systems. The ability of reconnaissance and strike of the weapons depend on the performance of EOSPs. With the development of weapon systems, the higher of the lightweight and miniaturization requirements for the EOSPs are proposed. Friction is the main factor which limits improving the control performance of the lightweight EOSPs. Aiming at the issues caused by nonlinear friction in the practice experience on developing EOSPs, the thesis carries on the research work on several aspects including the working principles, the parameter identification for EOSPs, the characteristic of the prestiction friction modeling and frictional parameters identification, and the friction compensation strategies so as to find a comparative perfect method on system modeling and friction compensation control for Electro-optical Stabilized Platforms.The following studies are included in this thesis:1. Introduce the structure and main functions of the Electro-optical Stabilized Platforms. Establish the system components’ mathematical models. And based on the dynamic equation of the EOSPs, the impacts of the mechatronic and friction parameters variation on the servo control performance of the systems are analyzed by theory and computer simulation. The results show that the low-frequency characteristic and corner frequency of the EOSPs depend on the inertia and torque coefficient, and the current closed-loop control can suppress the effects for the variation of armature resistance, inductance and back-EMF. The influence of the friction on the performance of the characteristics of the system depends directly on the system operating state. Decrease of the static friction, the difference between the static friction and Coulomb friction and the Increase of the Stribeck velocity will enhance the control capability and suppress the limit circle.2. Aiming at the shortages of frequency domain identification methods with friction, the effect principles of the friction on the velocity response are analyzed, and the reason of the phase lead phenomenon with the friction increasing in the frequency-characteristic is deduced. Based on the dynamic equation of stabilized platform, a time domain identification method is proposed, which transforms the problem of complex parametric estimation into the issues of simplified parametric model optimization though designing the current command signal. Within a limited range, this method can accurately estimate the parameters including the armature resistance, inductance, torque coefficient, inertia, damping and Coulomb friction, making use of the existing control loop and feedback signals based on the multi-level coordinate search.3. Aiming at the non-dissipative frictional behaviors of the prestiction regime in the direct-drive EOSPs, the thesis investigates the existing problem of the characteristic in the friction regime with different frictional loads. The Lu Gre friction model is proved that it can not character the friction in the prestiction regime based on the dissipative system theory. The novel prestiction friction model is proposed, which is a modified Lu Gre model including the presliding, the sliding and the prestiction regimes and can characterize friction behaviors in the prestiction regime. The comparison studies between the proposed friction model and the Lu Gre model are conducted in simulation and experiment, the results validate the prestiction friction model is effective in depicting the prestiction friction.4. To identified the parameters of the proposed friction model, a identification strategy divided into three steps such as the static, the dynamic and the prestiction frictional parameters is presented. A modified version of the Stribeck friction model is developed to satisfy modeling the static frictional characteristic in diverse friction. With known the static frictional parameters, the dynamic parameters of the Lu Gre model can be identified by using the data of the driving force and the velocity in the closed loop control based on the Multilevel Coordinate Search algorithm. Then, the remain parameters can be estimated with the velocity response in the prestiction regime. At last, the effectiveness of the identification methods are validated in several typical frictional conditions.5. To suppress the undesired effect of friction, the issue of friction compensation in servo control loop of EOSPs is studied. The compensation control method based on the prestiction friction model is developed to overcome the shortcoming of the traditional static friction model-based compensation. Aiming at the performance degraded issues of disturbance observer for the frequency characteristic of the plant effected by the nonlinear friction, a control algorithm of perturbation rejector + DOB is proposed to improve the robust performance of whole control system and the performance of disturbance suppression. The above control methods are validated in dual-rate loop structure under different friction load and effectively improve the stabilize accuracy of EOSPs.