### 摘要

The Two-Line-Elements (TLE) of satellites contains the orbital parameters used to calculate its position at a determined time based on a referential Earth Centered Inertial (ECI) system These parameters are given publicly by the North American Aerospace Defense Command (NORAD) and used nowadays as a conveyed reference for satellite tracking By using the SGP4 algorithm and the TLE parameters the position and velocity of the satellite can be predicted with an acceptable level of accuracy However the accuracy of the TLE parameters is subject to degradation when the time of orbit prediction deviates from reference epoch Thus the TLE parameters need to be updated periodically in order to maintain the quality in orbit predication For Low-Earth Orbiting (LEO) satellites specifically targeted for Cubesats whichnowadays are used for research the update of these parameters and identification of the satellites is critical for the operation in the ground station Because the nature of its orbits the average available communication time with them is limited to some minutes in a day thus certain degree of accuracy is needed for its tracking The research focuses on the determination of a better approximation of the orbital parameters of LEO satellites based on two inputs: one is the TLE parameters that might be outdated and the other is a recent tracking data of the LEO satellite In this research an optimization process that improves the parameters for tracking purposes is investigated This process involves the analysis of the transmission affected by the Doppler Effect in two models that are included in a genetic algorithm design One model is based on the frequency variation using the previous orbital parameters as starting approximation and the other is the actual frequency variation received from a specific location In the first approach the SGP4 model is used in order to make the calculation of the position and velocity at different periods of time in a given coordinate system In the second approach the center frequency over time of the transmission is obtained directly from measurements Due to noise attenuation and local interference an algorithm is used to track the center frequency Using the difference of the prediction and the real value measured at the same time a better approximation of the position and velocity can be made by using the genetic algorithm approach The best gen that can reduce the error of the Doppler Effect measurements are used to generate a new TLE which includes better approximation of the following days獎項日期 | 2014 八月 8 |
---|---|

原文 | English |

監督員 | Jyh-Chin Juang (Supervisor) |