Dissertation/Thesis Abstract

In-Node, Low Power Vehicle Classification and Identification System
by Shuck, Timothy A., M.S., California State University, Long Beach, 2018, 58; 10978575
Abstract (Summary)

Traffic in major metropolitan areas is only increasing, requiring transportation authorities to mitigate traffic congestion. Intelligent Transportation Systems (ITS) offer transportation authorities the data needed to accurately measure current traffic patterns and to make accurate predictions of future patterns.

The proposed system is a wireless sensor network (WSN) incorporating small embedded processing nodes equipped with anisotropic magneto resistive (AMR) sensors which detect the magnetic field of passing vehicles for vehicle classification and identification. Machine learning algorithms are implemented to produce a Machine-Learning Vehicle Classification (MLVC) system capable of classifying vehicles without external computation—only transmitting the aggregated results of the classification to system handlers. Various machine-learning algorithms were compared to determine their viability in being deployed in a manner befitting a system with an emphasis in wireless sensor node duration. An in-node decision tree-based implementation was selected to perform the analysis as it offers low-power utilization awhile retaining a high degree of accuracy.

Indexing (document details)
Advisor: Mozumdar, Mohammad
Commitee: Ahmed, Aftab, Nazari, Masoud
School: California State University, Long Beach
Department: Electrical Engineering
School Location: United States -- California
Source: MAI 58/05M(E), Masters Abstracts International
Source Type: DISSERTATION
Subjects: Engineering, Electrical engineering
Keywords: In-node vehicle classification system, Machine learning, Wireless sensor network
Publication Number: 10978575
ISBN: 978-1-392-07389-6
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy
ProQuest