On-Orbit Identification And Active Vibration Suppression For Flexible Appendages Of Complex Satellites

On-orbit identification and active vibration suppression are very important technologies for satellites with high precision pointing or targeting requirements. Based on these requirements, dynamic modeling of large flexible appendages with piezoelectric sensors/actuators is studied deeply in this dissertation. Then on-orbit identification and active vibration suppression method are realized based on the model. The main contents of this dissertation are as follows:An approximate analytical dynamical model of flexible appendages with piezoelectric sensors and actuators is derived and then converted to the state-space form for the purpose of parameter identification and active vibration suppression design.Based on the characteristics of on-orbit situation, the principle and application of identification method are discussed in depth. The architecture of on-orbit identification is designed preliminarily. ERA algorithm and its improved algorithm OKID are studied in detail and the excitation method and parameter tuning for the two algorithms are discussed. Simulation results show that ERA and OKID can identify the dynamic parameters quickly with high accuracy and have certain noise immunity. Note that the identification accuracy for OKID algorithm is affected by the excitation signal, an auxiliary excitation signal generated by actuators bounded on flexible appendages is used. Simulation results show that this method can improve the identification results significantly and can reduce the computation at the same time.A scheme that combined piezoelectric actuators and PPF controller is proposed for the active vibration suppression for large flexible appendages.The principle of PPF controller is studied in detail and the effectiveness,robustness and parameter tuning is discussed. To improve the vibrationsuppressioneffect, PPF-DVFB is studied as an improved controller. Thenanalgorithm combined PPF and PD is proposed. Theoretical analysis andsimulation results show that PPF-PD algorithm is better than the other twoalgorithms and has a very good stability.