Position estimation of an outer rotor permanent magnet synchronous machine using linear hall-effect sensors and neural networks

This thesis presents and evaluates a new method for estimating the angular position for an outer rotor permanent magnet synchronous machine (PMSM). PMSMs are increasingly used as prime movers in electric vehicles such as cars and bicycles, and the precise control of these machines requires reliable feedback of the rotational position of the rotor. Conventional methods of achieving this feedback signal rely on either physically connected sensors or the implementation of sensorless methods, each of which has certain drawbacks. The proposed method uses an array of linear Hall-effect sensors located in the leakage magnetic field of the rotor. These sensors detect the rotation-dependent changing field, which is fed into a machine-learning based neural network algorithm to interpret the signals. Due to the use of machine-learning, the algorithm will first need to be trained to properly correlate the sensor signals to the rotor angle. Data sets of training signals are acquired with commercial sensors and an outer rotor PMSM, and offline training steps and results are discussed. The main objective is to design a cost-effective position estimation system that is comparable to encoders and resolvers in functionality and performance, without the limitations of sensorless position estimation methods.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Electrical & Computer Engr

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2019

Author dc:creator

• Wang, Yuyao

Contributors dc:contributor

• Haran, Kiruba S

• Permanent magnet synchronous machine, position estimation, Hall effect, neural networks

Statement dc:rights

• Copyright 2019 Yuyao Wang

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/105643

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/105643