Sensor Placement Optimization And Modal Identification Of Flexible Spacecraft

The low stiffness, large flexible characteristics brings on-orbit spacecraft own huge flexible, this flexible take a significant impact on the spacecraft dynamics and control. Modal identify parameters can be used for the satellite attitude control, is also useful for the stability of the attitude performance. It is necessary to place acceleration or displacement sensors on spacecraft for information required, and then to the modal identification. In the process of on-orbit test the number of measuring point will be far less than the freedom degrees of structure, so it is every significant to choose the reasonable sensor installation position and modal parameter identification method for identifying the modal parameters of structure accurately. To research the issues mentioned above, the main contents and conclusions as follows:The traditional sensor placement methods was used in the structure of flexible spacecraft, including Modal Kinetic Energy, Effective Independence Algorithm and Modal Assurance Criterion. Draw a conclusion that:the MKE place rapid, oversimplified and a higher SNR, the El ensure the vibration mode of linear independence, the MAC enhanced separately of each modal.As it is difficult to obtain the optimal value through traditional methods, a variant of the improved genetic algorithm is used to place sensors optimally on flexible spacecraft for the purpose of modal identification. The algorithm proposed by Yao was improved to combine elite selection and proportional selection. A feature called forced mutation is included to improve the convergence of the fitness value. By applying both the improved GA and Yao method to sensor placement for a satellite array, it has been shown that the improved GA required nearly half of Yao to converge to the optimal value. With the convergence effectively, some suboptimal value was generated around the optimal value. It provides more choices for the actual project.Different optimal sensor placement methods was used on a certain type of flexible satellite model, then identify the system modal parameters using Eigen-system Realization Algorithm. The study found, using satellite attitude information and point vibration sensor placement information as output data, can identify more order system modal information than that using only attitude information for the output. No matter whether there is a noise, three placement programs have the similar abilities of identification.The EI is slightly better than the other two methods, it can accurately identify the frequency of the system in the case of noise. Take the identify mode as the target mode of sensor placement, it can accurately identify the goal mode.