Izvedba Kalmanovega filtra s sigma točkami na mikrokrmilniku

JANDRIĆ, NIKOLA (author), Mihelj, Matjaž (mentor)

Diplomska naloga opisuje razvoj sistema za merjenje orientacije v prostoru (angl. attitude and heading reference system - AHRS). Opisana je strojna oprema inercijsko-magnetne merilne enote (IME) in algoritem, ki je bil uporabljen za nadgradnjo te enote v AHRS, to je Kalmanov filter. IME vsebuje žiroskop, pospeškometer in magnetometer. Vsak senzor meri veličino v treh ortogonalnih smereh, tako da ima IME 9 prostostnih stopenj. V uvodu je opis razvoja IME, ki je sledil napredku integriranih vezij in potrebi po merjenju orientacije. Naštetih je nekaj področij uporabe in nekaj izdelkov, ki so že prisotni na trgu in uporabljajo IME za merjenje gibanja. V poglavju o Kalmanovem filtru so opisane tri verzije Kalmanovega filtra. Od prvotne verzije diskretnega Kalmanovega filtra, prek vmesnega razširjenega Kalmanovega filtra do verzije, ki smo jo uporabili za AHRS, Kalmanovega filtra s sigma točkami. Tretje poglavje opisuje uporabljeno opremo, lastnosti IME enote in princip merjenja merjenih veličin. Opisuje tudi standarda IEEE 802.15 in ZigBee, s katerima je skladen modul za komunikacijo, ter primerja ZigBee in Bluetooth tehnologiji. V četrtem poglavju so opisani potek programa na AHRS, kako z njim upravlja uporabnik, na kakšen način pošilja podatke na računalnik in kako je realiziran Kalmanov filter. V zadnjem poglavju so podani rezultati dveh različnih meritev. Najprej so opisane meritve običajnih gibov, ki jih izvajamo vsakodnevno, v nadaljevanju pa meritve sunkovitih gibov. Za vsako od meritev so podani grafi izhodov vseh senzorjev, z AHRS izračunan in z Optotrakom izmerjen kvaternion ter razlika med njima, pretvorjena v zasuk po posameznih oseh.

• Language: Slovenian
• Keywords: AHRS, IME, orientacija v prostoru, Kalmanov filter, ZigBee
• Work type: Undergraduate thesis
• Organization: FE - Faculty of Electrical Engineering
• Year: 2015
• PID: 20.500.12556/RUL-30502
• Publication date in RUL: 13.01.2015

Secondary language

• Language: English
• Title: Implementation of an Unscented Kalman filter on a microcontroller

Abstract

The thesis describes development of an attitude and heading reference system (AHRS). The thesis includes the description of used hardware of a magneto-inertial measurement unit (IMU) and description of a Kalman filter. The Kalman filter algorithm was used to upgrade IMU into AHRS. Integrated 3D gyroscope, 3D accelerometer and 3D magnetometer make IMU a 9 degree of freedom measurement device. Development of IMUs, that is driven by progress of integrated circuits and the necessity for simple measurements of orientation is described in introduction. There are also a few areas of IMU usage listed and some products already available on the market, that use IMU, for tracking motion. The second chapter gives the description of three versions of Kalman filter. From the very first version of the discrete Kalman filter to the Extended Kalman filter, and finally the Unscented Kalman filter, that was used for AHRS. Detailed description of hardware is given in the third chapter. Measuring principles of IMU sensors and their properties are listed there. Communication module is IEEE 802.15 and ZigBee compliant, so description of both standards is given. The comparison of the ZigBee and Bluetooth technologies is provided. In the fourth chapter software for AHRS is described. Description of it's user interface, Kalman filter implementation and communication is given there. Measurements of normal and fast motion are analysed in the last chapter. For each measurement there is a graph of sensors output, graph of calculated and measured quaternion and difference between them, transformed into error angle.

• Keywords: AHRS, IMU, attitude determination, Kalman filter, ZigBee