Prediksi Data Waktu Common GNSS Generic Time Transfer Standard (CGGTTS) dengan Metode Kalman Filter untuk Meningkatkan Akurasi Pembacaan

Muflihah, Zulfa (2025) Prediksi Data Waktu Common GNSS Generic Time Transfer Standard (CGGTTS) dengan Metode Kalman Filter untuk Meningkatkan Akurasi Pembacaan. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5009221140-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2027. Download (8MB) | Request a copy

Abstract

Penggunaan Kalman Filter pada data waktu CGGTTS untuk mengurangi noise dan meningkatkan akurasi prediksi dalam aplikasi transfer waktu berbasis GPS/GNSS. Dalam proses ini, Kalman Filter digunakan sebagai metode filter data dan multi-step forward prediction. Selain itu juga, pada penelitian ini mengevaluasi performa Kalman Filter pada prediksi 1 step ke depan hingga 10 step ke depan. Evaluasi dilakukan menggunakan metrik performansi seperti RMSE, MSE, dan MAE untuk menilai akurasi dan stabilitas hasil prediksi. Kemudian, hasil metrik performansi prediksi Kalman Filter dibandingkan dengan prediksi Moving Average. Hasil penelitian menunjukkan bahwa Kalman Filter dalam konteks filtering data mampu menghasilkan data yang lebih halus, konsisten, dan akurat, dibuktikan dengan hasil Kalman Filter dengan raw REFGPS, nilai RMSE sebesar 73.58, MSE sebesar 5414.25, dan MAE sebesar 59.77 sehingga sangat bermanfaat untuk aplikasi sinkronisasi jam atom lebih presisi. Kalman Filter secara konsisten menunjukkan performa terbaik dibandingkan metode Moving Average dalam semua metrik evaluasi, yaitu RMSE (Root Mean Squared Error), MSE (Mean Squared Error), dan MAE (Mean Absolute Error). Kalman Filter mampu menangani data dinamis dan mengurangi kesalahan secara lebih efektif dibandingkan Moving Average. Hasil multi-step forward prediction menunjukkan semakin banyak jumlah time step, semakin menurun tingkat akurasinya. Kalman Filter lebih efektif dalam menangani noise dan fluktuasi, menjadikannya metode prediksi yang lebih akurat untuk aplikasi dengan kebutuhan presisi tinggi. Kalman Filter mampu memperbaiki kualitas data dan prediksi waktu. Penelitian ini mendukung penerapan Kalman Filter dalam aplikasi yang memerlukan prediksi waktu dengan presisi tinggi, seperti telekomunikasi, navigasi, dan penentuan waktu global.
====================================================================================================================================
Application of Kalman Filter to CGGTTS time series data for noise reduction and improved prediction accuracy in GPS/GNSS-based time transfer applications. The Kalman Filter is employed for both data filtering and multi-step forward prediction. This research evaluates the Kalman Filter's performance for predictions from 1 to 10 steps ahead. Performance is assessed using metrics such as RMSE, MSE, and MAE to evaluate the accuracy and stability of the predictions. The prediction performance of the Kalman Filter is then compared with that of a Moving Average. The results demonstrate that the Kalman Filter, in the context of data filtering, is capable of producing smoother, more consistent, and accurate data. This is evidenced by the comparison of Kalman Filter results with raw REFGPS data, yielding an RMSE of 73.58, an MSE of 5414.25, and an MAE of 59.77, making it highly beneficial for more precise atomic clock synchronization applications. The Kalman Filter consistently demonstrates the best performance compared to the Moving Average method across all evaluation metrics, namely RMSE (Root Mean Squared Error), MSE (Mean Squared Error), and MAE (Mean Absolute Error). The Kalman Filter effectively handles dynamic data and reduces errors more efficiently than the Moving Average. The multi-step forward prediction results show decreasing accuracy with an increasing number of time steps. The Kalman Filter is more effective in handling noise and fluctuations, making it a more accurate prediction method for applications requiring high precision. The Kalman Filter improves both data quality and time prediction. This research supports the application of Kalman Filters in applications requiring high-precision time prediction, such as telecommunications, navigation, and global timing.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Prediksi Data Waktu, CGGTTS, Kalman Filter, Moving Average
Subjects:	T Technology > T Technology (General) > T174 Technological forecasting
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Physics Engineering > 30201-(S1) Undergraduate Thesis
Depositing User:	Zulfa Muflihah
Date Deposited:	31 Jan 2025 07:31
Last Modified:	31 Jan 2025 07:31
URI:	http://repository.its.ac.id/id/eprint/117460

Actions (login required)

View Item