Please use this identifier to cite or link to this item:
Title: Adaptive incident detection algorithm for suburban expressway
Authors: Boonsiripant, S
Sumalee, A 
Jedwanna, K
Rungrittidech, W
Keywords: Incident detection
Comparative algorithm
Single station algorithm
Issue Date: 2012
Source: Proceedings of the 17th International Conference of Hong Kong Society for Transportation Studies (HKSTS), Hong Kong, 15-17 December 2012, p. 281-288 How to cite?
Abstract: Incident is one of the major causes of non-recurring congestion on the expressways. Effective incident detection and management would greatly reduce non-recurring congestion delays caused by incidents. This paper aims to implement an incident detection algorithm for a suburban expressway route in Greater Bangkok, Thailand. The overly long sensor spacing, i.e., 5 km apart presents a challenge in developing an effective detection algorithm. Two classic algorithms, namely, California and McMaster algorithms were experimented. Based on the results, it was found that the comparative approach such as California algorithm is not suitable for a long span section due to its low detection rate and extremely long detection time. A single station algorithm like McMaster algorithm showed a more promising result and was then modified. By introducing a simple rule-based algorithm with an adaptive threshold, the Rule-Based McMaster algorithm yielded an exceptionally high detection rate and low false alarm rate.
Appears in Collections:Conference Paper

Show full item record

Page view(s)

Last Week
Last month
Citations as of Feb 11, 2019

Google ScholarTM


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.