Analisis Implementasi Filter Kalman dan Filter Low Pass sebagai State Observer pada Prototipe Kontrol Momen Giroskop

Alfito, Muhammad Fauzan (2023) Analisis Implementasi Filter Kalman dan Filter Low Pass sebagai State Observer pada Prototipe Kontrol Momen Giroskop. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 02111940000013-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2025. Download (3MB) | Request a copy

Abstract

Teknologi self-balancing menjadi solusi yang relevan dalam mempertahankan keseimbangan kendaraan roda dua. Pada penelitian terdahulu algoritma kontrol PID diterapkan pada prototipe kendaraan roda dua menggunakan CMG dengan IMU berbasis MEMS sebagai sensor orientasi. Pada penelitian tersebut sensor IMU mengalami vibrasi pada putaran motor yang tinggi sehingga menyebabkan drift dan ketidakakuratan pada pembacaan sensor. Pada penelitian ini filter Kalman, filter low pass, dan filter Kalman-low pass digunakan untuk meredam sinyal frekuensi tinggi dan meminimalkan drift pada sensor sehingga pembacaan sensor lebih akurat. Metode RMSE digunakan untuk mengevaluasi parameter desain filter yang paling optimal. Hasil penelitian menunjukkan bahwa performa filter Kalman dan filter low pass tidak dapat diukur dengan metode RMSE dengan referensi nol karena penahan tidak dapat menjaga objek pengukuran tetap pada posisi nol dalam uji statis dan sensitivitas sensor yang sangat kecil yaitu sebesar 0,06 derajat. Performa filter low pass diamati dari delay filter dan grafik magnitude response hasil transformasi Fourier. Pada penelitian ini metode filter Kalman-low pass dengan nilai R= 2×106 dan cut-off frequency 15 Hz memberikan hasil pembacaan dengan delay 0,1 detik dan drift yang minim namun dapat menghasilkan overestimate pada simpangan yang besar.
=================================================================================================================================
To maintained the balance of a two-wheeled vehicles, a self-balancing principles is used as a relevant solution. In the previous study, a PID control algorithm was applied to a prototype of a two-wheeled vehicle using CMG with a MEMS-based IMU as the orientation sensor. The result shows that the IMU sensor experienced vibrations during high motor rotations, it is resulting a sensor drift and inaccuracies of the readings. Adressing the previous issu, this study implemented Kalman, low pass, and Kalman-low pass filters to suppress high-frequency noise signals and minimize sensor drift. The RMS Error method is implemented to evaluate the most optimal filter design parameters. The results shows that the performance of the Kalman and low pass filters cannot be measured using the RMSE method with a zero reference, due to the inability of the restraint to maintain the measurement object in a static position, and the sensor's very small sensitivity, which is about 0.06 degrees. The performance of the low pass filter is observed through filter delay and the magnitude response graph obtained from the Fourier transformation. In this study, it is concluded that the Kalman-low pass filter method with an R value of 2×106 and a cut-off frequency of 15 Hz provides readings with a delay of 0.1 seconds and minimal drift, but it may lead to an overestimation in large swing conditions.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	digital signal processing, filter Kalman, filter low pass, IMU, kendaraan self-balancing, Kalman filter, low pass filter, self-balancing vehicle
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5102.9 Signal processing.
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Mechanical Engineering > 21201-(S1) Undergraduate Thesis
Depositing User:	Alfito Muhammad Fauzan
Date Deposited:	10 Oct 2023 04:11
Last Modified:	10 Oct 2023 04:12
URI:	http://repository.its.ac.id/id/eprint/102812

Actions (login required)

View Item