Spelling suggestions: "subject:"4traffic incident timeline"" "subject:"ktraffic incident timeline""
1 |
Evaluation of Traffic Incident Timeline to Quantify the Performance of Incident Management StrategiesHaule, Henrick Joseph 01 January 2018 (has links)
Transportation agencies are introducing new strategies and techniques that will improve traffic incident management. Apart from other indicators, agencies measure the performance of the strategies by evaluating the incidents timeline. An effective strategy has to reduce the length of the incident timeline. An incident timeline comprises various stages in the incident management procedure, starting when the incident was detected, and ending when there is the recovery of normal traffic conditions. This thesis addresses three issues that are related to the traffic incident timeline and the incident management strategies.
First, co-location of responding agencies has not been investigated as other incident management measures. Co-location of incident responders affects the incident timeline, but there is a scarcity of literature on the magnitude of the effects. Evaluation of the co-location strategy is reflected by the response and verification durations because its effectiveness relies on improving communication between agencies. Investigation of the response and verification duration of incidents, before and after operations of a co-located Traffic Management Center (TMC) is done by using hazard-based models. Results indicate that the incident type, percentage of the lane closure, number of responders, incident severity, detection methods, and day-of-the-week influence the verification duration for both the before- and after- period. Similarly, incident type, lane closure, number of responders, incident severity, time-of-the-day, and detection method influence the response duration for both study periods. The before and after comparison shows significant improvements in the response duration due to co-location of incident response agencies.
Second, the incident clearance duration may not necessarily reflect how different types of incidents and various factors affect traffic conditions. The duration at which the incident influences traffic conditions could vary – shorter than the incident duration for some incidents and longer for others. This study introduces a performance measure called incident impact duration and demonstrates a method that was used for estimating it. Also, this study investigated the effects of using incident impact duration compared to the traditionally incident clearance duration in incident modeling. Using hazard-based models, the study analyzed factors that affect the estimated incident impact duration and the incident clearance duration. Results indicate that incident detection methods, the number of responders, Traffic Management Center (TMC) operations, traffic conditions, towing and emergency services influence the duration of an incident.
Third, elements of the incident timeline before the clearance duration have been overlooked as factors that influence the clearance duration. Incident elements before the clearance duration include verification time, dispatch duration, and the travel time of responders to the incident scene. This study investigated the influence of incident timeline elements before clearance on the extent of the clearance duration. Also, this study analyzed the impact of other spatial and temporal attributes on the clearance duration. The analysis used a Cox regression model that is estimated using the Least Absolute Shrinkage and Selection Operator (LASSO) penalization method. LASSO enables variable selection from incidents data with a high number of covariates by automatically and simultaneously selecting variables and estimating the coefficients. Results suggest that verification duration, response travel duration, the percentage of lane closure, incident type, the severity of an incident, detection method, and crash location influence the clearance duration.
|
Page generated in 0.0914 seconds