APPLICATION OF KALMAN FILTER AS A METHOD TO ESTIMATE TILT FROM INERTIAL MEASUREMENT UNIT ON SELF-BALANCING SCOOTER

• Main Author: (NIM : 13205159); Pembimbing : Dr. Ir. Hilwadi Hindersah, M.Sc., ZULKARNAEN
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/14581
• Institution: Institut Teknologi Bandung

The advancement in MEMS technology made possible for the spreading implementation of ineartial measurement unit which include gyroscope and accelerometer with low price and small size. One of its implementation is Segway PT. However, MEMS sensors suffer low performance compared to its macro version so that it needs method to estimate the true value. In this final year project, Kalman filter is applied as a method to estimate tilt on IMU which comprise gyroscope and accelerometer. The estimated angle is then used by closed loop control in self-balancing scooter, for its feedback, that designed by writer’s partner. Kalman Filter is implemented on ATMEGA16 microcontroler. Result from comparing estimated angle with angle from encoder reading shows that the filter has -0,064o error average and 1,013o error standard deviation, and the filter is stable. The error standard deviation is pretty high. It happens due to hysteresis on the encoder. Even so, Kalman filter works well and successfully implemented with the closed loop control algorithm that writer’s partner design.