Pengembangan Ensemble Kalman Filter-Proporsional Integral Derivatif (EnKF-PID) Untuk Navigasi Dan Guidance Pada Autonomous Car

Fiddina, Qori Afiata (2023) Pengembangan Ensemble Kalman Filter-Proporsional Integral Derivatif (EnKF-PID) Untuk Navigasi Dan Guidance Pada Autonomous Car. Masters thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 6002201011_6002201011_Master_Thesis.pdf] Text
6002201011_6002201011_Master_Thesis.pdf - Accepted Version
Restricted to Repository staff only until 1 April 2025.

Download (1MB) | Request a copy

Abstract

Autonomous car disebut dengan mobil tanpa pengemudi. Autonomous car memiliki kemampuan untuk perencanaan jalur, mendeteksi kondisi lingkungan sekitarnya dan mengenali hambatan. Salah satu tugas penting yang harus dilakukan oleh autonomous car adalah melakukan perencanaan jalur sehingga dibutuhkan estimator untuk navigasi autonomous car. Ketika autonomous car berjalan atau bergerak, kendaraan akan melalui lintasan yang diinginkan. Akan tetapi pada keadaan tertentu, autonomous car bergerak jauh dari lintasan yang diinginkan. Sehingga diperlukan sebuah kontrol agar kendaraan berjalan sesuai dengan lintasan yang diinginkan. Pada penelitian ini, estimator yang digunakan adalah metode Ensemble Kalman Filter (EnKF) sedangkan untuk pengendalian gerak autonomous car menggunakan metode Proporsional Integral Derivatif (PID). Penelitian terdahulu mengenai avigasi dan kontrol dilakukan secara terpisah, akan tetapi kondisi di lapangan gerak autonomous car membutuhkan navigasi dan kontrol bekerja beriringan. Pada penelitian ini ditunjukkan bahwa navigasi dan kontrol saling berkesinambungan dengan menunjukkan presentase peningkatan metode Ensemble Kalman Filter-Proporsional Integral Derivatif (EnKF-PID). Pada penelitian ini simulasi yang dilakukan menggunakan 2 variasi yaitu simulasi tanpa referensi lintasan dan simulasi dengan referensi lintasan. Hasil yang diperoleh dari simulasi tanpa referensi lintasan metode Ensemble Kalman Filter ditunjukkan dengan perolehan nilai RMSE yang semakin kecil apabila pengambilan N-ensemble semakin besar. Peningkatan performa dari metode Ensemble Kalman Filter-Proporsional Integral Derivatif (EnKF-PID) sebesar 56,18%. Selanjutnya simulasi menggunakan referensi lintasan. Referensi lintasan yang digunakan yaitu lintasan lurus dan belok. Pada lintasan lintasan lurus performa dari Ensemble Kalman Filter- Proporsional Integral Derivatif (EnKF-PID) sebesar 94% sedangkan pada lintasan belok sebesar 79,44%. Dengan presentase peningkatan yang sangat besar sehingga Ensemble Kalman Filter-Proporsional Integral Derivatif (EnKF-PID) menambah performa dari Ensemble Kalman Filter (EnKF) dan dapat ditunjukkan bahwa navigasi dan kontrol bersifat resiprosikatif (saling mempengaruhi)
======================================================================================================================================
Autonomous car is called a car without a driver. Autonomous car has the ability to plan paths, detect environmental conditions and recognize obstacles. One of the important tasks that must be carried out by an autonomous car is to carry out path planning so that an estimator is needed for autonomous car navigation. When an autonomous car is running or moving, the vehicle will follow the desired trajectory. However, in certain circumstances, the autonomous car moves far from the desired trajectory. So we need a control so that the vehicle runs according to the desired trajectory. In this study, the estimator used was the Ensemble Kalman Filter (EnKF) method, while the Proportional Integral Derivative (PID) method was used to control the motion of the autonomous car. Previous research on navigation and control was carried out separately, but the conditions in the field of autonomous car movement require navigation and control to work hand in hand. In this study it was shown that navigation and control are mutually continuous by showing the percentage increase in the Ensemble Kalman Filter-Proportional Integral Derivative (EnKF-PID) method. In this study, the simulation was carried out using 2 variations, namely a simulation without a track reference and a simulation with a track reference. The results obtained from the simulation without the track reference of the Ensemble Kalman Filter method are shown by the acquisition of smaller RMSE values when the N-ensemble selection is greater. The performance improvement of the Ensemble Kalman Filter-Proportional Integral Derivative (EnKF-PID) method is 56.18%. Furthermore, the simulation uses the track reference. The track references used are straight and turning paths. On the straight track the performance of the Ensemble Kalman Filter- Proportional Integral Derivative (EnKF-PID) is 94% while on the turning track it is 79.44%. With a very large percentage increase, the Ensemble Kalman Filter-Proportional Integral Derivative (EnKF-PID) increases the performance of the Ensemble Kalman Filter (EnKF) and it can be shown that navigation and control are reciprocal (influence each other)

Item Type: Thesis (Masters)
Uncontrolled Keywords: Autonomous Car, Ensemble Kalman Filter (EnKF), Proporsional Integral Derivatif (PID). Autonomous Car, Ensemble Kalman Filter (EnKF), Proporsional Integral Derivatif (PID).
Subjects: Q Science > QA Mathematics > QA402.3 Kalman filtering.
T Technology > TJ Mechanical engineering and machinery > TJ223 PID controllers
T Technology > TL Motor vehicles. Aeronautics. Astronautics > TL152.8 Vehicles, Remotely piloted. Autonomous vehicles.
Divisions: Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44101-(S2) Master Thesis
Depositing User: Qori Afiata Fiddina
Date Deposited: 16 Feb 2023 06:03
Last Modified: 16 Feb 2023 06:03
URI: http://repository.its.ac.id/id/eprint/97393

Actions (login required)

View Item View Item