This study explores methods to improve the accuracy of satellite maneuver detection by comparing traditional statistical methods with machine learning and neural network-based methods. The two alternative methods explored are a Random Forest algorithm and a Deep Neural Network. Simulation data is generated and appropriate features are selected and computed. Data is split into training and validation. Next, alternative methods are trained and their performance is validated. Optimization techniques are then applied to achieve better results. Finally, the three methods are compared to each other. Conclusions and future work are described.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Satellite Maneuver Detection Using Machine Learning and Neural Network MethodsBehaviors


    Contributors:


    Publication date :

    2022-03-05


    Size :

    5222020 byte




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Geosynchronous Satellite Maneuver Classification via Supervised Machine Learning

    Roberts, Thomas G. | British Library Conference Proceedings | 2021


    Satellite Maneuver Detection Using Two-line Elements Data

    Kelecy, T. | British Library Conference Proceedings | 2007



    Satellite Maneuver Detection with Optical Survey Observations

    Pastor, Alejandro | British Library Conference Proceedings | 2020


    Satellite Maneuver Evaluation Tool

    C. I. Noriega | NTIS | 1990