Identification And Control Design For Mobile Onboard Antenna Servo Systems

The Mobile onboard antenna is an important satellite communication(sat-com)terminal today,which is installed on carriers such as vehicles,ships and aircrafts,and can communicate with satellite during the movement of carrier.The servo system of antenna is the key to ensure reliable communication in the movement,which drives the antenna to quickly point to and continuously track the target satellite,and attenuates the disturbance caused by the carrier.Mobile sat-com tasks in the new era put forward strict requirements on the performance of the antenna servo system,which exerts challenges for the control design.This dissertation studies the flexible resonance model identification and multi-objective control design of the mobile onboard antenna servo system.The main contents include the following aspects:Precise kinematic and dynamic models of the mobile onboard antenna servo system are established.For the antenna servo system with the azimuth-elevation structure used in the study,the kinematic model is derived based on the forward and inverse kinematics analysis.The rigid body dynamic model is established using the Lagrangian method,and several mechanism design suggestions about dynamics decoupling are given.A flexible shaft dynamic model is established using a three-mass model structure,which explains the mechanism of the flexible resonance.By modeling the identification problem as an approximation of a noise-affected data model,a two-step identification strategy consisting of nonparametric model estimation and parametric model approximation is proposed.Compared with the optimal parameter estimation based identification methods,this strategy avoids the influence of the accuracy of the prior model,and considers the fitting of the main modes,which is suitable for the identification of flexible resonance models.In the nonparametric model estimation part,an fractional decomposition based impulse response identification method is proposed,which improves the identification accuracy of the light damping model while reducing the data length.In the part of parameter model approximation,the Hankel norm approximation problem of data model is discussed,and the approximate solution is given based on Ho-Kalman system realization algorithm.In the experiment,a practical identification scheme integrating experimental design,model calculation and model verification is presented,which is used for the flexible resonant model identification of the mobile onboard antenna.The implementation details of the scheme is introduced,system models and parameters are obtained,and the superiority is proven by comparative experiments.A bi-objective control strategy is proposed to meet the requirements on dynamic tracking performance and robustness of servo systems.Necessary and sufficient conditions about stability and tracking performance are discussed for the proposed control system,and then a robust asymptotically tracking parameterized controller is presented based on the Youla-Kucera parameterization results to solve the bi-objective control problem.Compared with traditional robust control methods,the bi-objective control considers both the center performance and the boundary performance,making the system performance more controllable at all working points.The bi-objective control strategy is applied to solve the servo control engineering problems of vehicle-mounted and airborne antennas,and corresponding design examples are exhibited.In the control design of the vehicle-mounted antenna,the decoupling properties of the bi-objective task are discussed,the existence of the solution to the bi-optimal problem is proved,and corresponding controller design scheme is given.In the design process,the constraints about the Bode integral theorem on the design are discussed,and the weight tuning work is detailed.In experimental part,many platform verification and field test results are showed,and a real satellite tracking test is conducted to verify the practical value of the design.In the control design of airborne antenna,considering the potential keyhole effect and the flexible resonance mode drift of the antenna,the control design is modeled as a robust tracking problem.Based on bi-objective control,a novel speed-position dual-loop control system structure is proposed.Combining with the H∞ loop shaping method,the corresponding design scheme is developed,and the detailed design method of shaping weight functions is discussed.Considering the ease of engineering,the bi-objective controller is decomposed into multiple modules,and the control design is carried out as a gradually upgrade process to the original controller.In the experimental part,the advantages of the design are illustrated by comparing the developed controller with some commonly used optimal and robust controllers.