NTC Program Progress Performance Report (PPPR) Information Form

Transcription

1 NTC Program Progress Performance Report (PPPR) Information Form For P.I. s Use On a semi- annual basis the NTC sponsored P.I. must report Program Progress Performance Report (PPPR) using the format specified in this PPPR Information Form. The form must be submitted electronically to the corresponding NTC Associate Director by 2/22/2016. Cover Period: 9/30/2015 3/31/2016 NTC Funded Project Information (Round/Year 3, ) University Name Project Title Principal Investigator PI Contact Information University of Maryland Stochastic Deployment of Emergency Vehicles Considering Sequence of Incidents Ali Haghani, Professor Department of Civil and Environmental Engineering 1124C Glenn Martin Hall University of Maryland College Park, MD haghani@umd.edu The form includes the following six parts: Part I Performance Indicators Part II Accomplishments: What was done? What was learned? Part III Products: What has the program produced? Part IV Participants & Collaborating Organizations: Who has been involved? Part V Impact: What is the impact of the program? How has it contributed to transportation education, research and technology transfer? Part VI Changes/Problems Supplementary documents/materials can be attached to this form with the submission. 1

2 Part I Performance Indicators Reporting Period 9/30/2015 3/31/ Transportation- related courses offered during the reporting period that were taught by faculty and/or teaching assistants who are associated with the UTC Undergraduate courses Graduate courses 2. Students supported by this grant Undergraduate students Masters students Doctoral students 3. Students participating in transportation research projects funded by this grant (but not supported by this grant) Undergraduate students Graduate students 4. Students supported by this grant who received degrees Undergraduate degrees Masters degrees Doctoral degrees ENCE 677: OR Models for Transportation Systems Analysis Hyoshin Park 2

3 Part II Accomplishments: What was done? What was learned? The information provided in this section allows the OST- R grants official to assess whether satisfactory progress has been made during the reporting period. Reporting Period 9/30/2015 3/31/ What are the major goals of the program? The National UTC aims to promote strategic transportation policies, investment, and decisions that bring lasting and equitable economic benefits to the U.S. and its citizens. The Center is concerned with the integrated operations and planning of all modes serving the nation s passenger and freight transportation system, including the institutional issues associated with their management and investments. A balanced multi- modal approach will be used that considers freight and passenger travel mobility, reliability, and sustainability, as well as system operations during periods of both recurring and non- recurring incidents, including response to major emergencies. The modes in this theme include highway, transit, rail, and inter- modal interfaces including ports, terminals and airports. In particular, the center focuses on research, education, and technology transfer activities that can lead to (1) Freight efficiency for domestic shipping and for our international land, air, and sea ports; (2) Highway congestion mitigation with multi- modal strategies; and (3) Smart investments in intercity passenger travel facilities such as high speed rail. Major center activities are as following: Advanced & Applied Research Promoting Economic Competitiveness: Our research activities are multimodal/intermodal and multidisciplinary in scope, with the aims of addressing nationally and regionally significant transportation issues pertinent to economic competitiveness and providing practice- ready solutions. Education, Workforce Development, Technology Transfer, & Diversity The consortium is committed to providing high- quality transportation education and workforce development programs for a broad and diverse audience. Center s efforts will support the development of a critical transportation knowledge base and a transportation workforce that is prepared to design, deploy, operate, and maintain the complex transportation systems of the future. 3

4 2. What was accomplished under these goals? Location and allocation of emergency response units are assumed to be interdependent in the incident management system. System costs will be excessive if delay regarding allocating decisions is ignored when locating response units. The objective of the location- allocation problem is to accurately capture the cost of multiple- stop routes within a location model. This research incorporates a realistic and stochastic process into the design of deployment of emergency response vehicles. The incident occurrence includes accumulated probabilities of secondary incidents in future steps, in which the impact of primary incidents overlaps. In general, a secondary incident may occur during the clearance or recovery of a primary incident. The project encompasses eight tasks as follows: 1. Incident probability: The focus of this task is to introduce a stochastic process of future stages of incidents. The expected probability of incidents is developed for each site and stage. Each sequence of incidents is represented in a matrix form with expected probability of each scenario. 2. Pure clearance time. In this task, potentially delayed clearance is modeled by integrating delay- type with normal clearance time following a previous study. The main goal of this task is to generate input to our optimization model that deploys required emergency vehicles to the incident site as quickly as possible. 3. Problem formulation. This task addresses deploying response units to potential sites before an incident occurs and dispatching response units to the scene after an incident occurs. To solve such a complex problem, we expect to have non- linear terms in the formulation. 4. Equivalent linear representation. In this task, we introduce new variables for linearization not to cut off the optimal solution. To prevent numerical issues, we will propose a method of selecting an appropriate parameter. 5. Scenario reduction heuristic. The goal of this research task is to develop an efficient heuristic algorithm to deal with time consuming process of a large- scale problem in real- time. 6. Preparing scenarios in a real- network. The focus of this task is to calculate conditional probabilities for each scenario and quantify a primary incident s impact on traffic conditions to detect secondary incidents on a real transportation network. 7. Model Comparison. Comparing different approaches. In this task, we make comparisons of location policies and optimization models. We will test computation time and gap of scenarios with different combination of parameters. 8. Report Preparation. A comprehensive literature review of the 4

5 state- of- the- art practices as well as description of the models and results will be provided in a final report. Three tasks of this research were completed before the project start. During this reporting period we completed tasks 4, 5, and How have the results been disseminated? The results so far have been published in archival journals and presented in refereed conferences. 4. What do you plan to do during the next reporting period to accomplish the goals? (3/31/2016 9/30/2016) Currently we are working on Task 7 and will prepare the final report for the project and submit. 5

6 Part II Products: What has the program produced? Publications are the characteristic product of research projects funded by the UTC Program. OST- R may evaluate what the publications demonstrate about the excellence and significance of the research and the efficacy with which the results are being communicated to colleagues, potential users, and the public, not the number of publications. Many research projects (though not all) develop significant products other than publications. OST- R may assess and report both publications and other products to Congress, communities of interest, and the public. Reporting Period 9/30/2015 3/31/ Journal publications: The following papers were accepted for publication in 2016: Park, H., A. Shafahi, and A. Haghani, A stochastic emergency response location model considering secondary incidents on freeways. In press in IEEE Transactions on Intelligent Transportation Systems. DOI: /TITS Park, H. and A. Haghani, Stochastic capacity adjustment considering secondary incidents. In press in IEEE Transactions on Intelligent Transportation Systems. DOI: /TITS The following papers were published based on earlier research: Park, H. and A. Haghani, Real- time prediction of secondary incident occurrences using vehicle probe data. Advance online published in Transportation Research Part C: Emerging Technologies. DOI: /j.trc Books or other non- periodical, one- time publications 3. Other publications, conference papers and presentations Park, H., A. Haghani, and X. Zhang, Interpretation of Bayesian neural networks for predicting the duration of detected incidents. Advance online published in Journal of Intelligent Transportation Systems: Technologies, Planning, and Operations. DOI: / The following papers will be presented, or were presented in 2016: Park, H. and A. Haghani. Online redeployment and routing of emergency vehicles considering sequence of incidents. The 9 th Triennial Symposium on Transportation Analysis (TRISTAN), Oranjested, Aruba, June 13-17, Accepted for presentation. 6

7 Park, H., A. Haghani., and A. Shafahi. Stochastic emergency response units allocation considering secondary incident occurrences. The 95 th Annual Meeting of Transportation Research Board, Washington, DC, January, Park, H. and A. Haghani. Stochastic capacity adjustment considering secondary incidents. The 95 th Annual Meeting of Transportation Research Board, Washington, DC, January, Park, H. and A. Haghani. Online emergency vehicle dispatching with look- ahead on a transportation network. The 95 th Annual Meeting of Transportation Research Board, Washington, DC, January, Park, H. and A. Haghani. Use of clustering model and adjusted boxplot model for identification of secondary incidents. The 95 th Annual Meeting of Transportation Research Board, Washington, DC, January, The following papers were presented based on earlier research: Park, H. and A. Haghani. Capacity adjustment considering the impact of secondary incidents. The 94 th Annual Meeting of Transportation Research Board, Washington, DC, January, Park, H. and A. Haghani. An optimal fleet allocation of emergency response teams on freeway using a two stage stochastic programming. The 20 th Conference of the International Federation of Operational Research Societies (IFORS), Barcelona, Spain, July, Park, H., A. Haghani, and Y. Aliari. A pedagogical rule extraction from Bayesian neural networks for prediction of secondary incidents. The 1 st International Conference on Engineering and Applied Sciences Optimization (OPTI), Kos Island, Greece, June, Park, H. and A. Haghani. Identification and prediction of secondary crashes using vehicle probe data. The 93 rd Annual Meeting of Transportation Research Board AFB20(2) Subcommittee, Washington, DC, January, Park, H., A. Haghani., and H. Masoud. Real- time filtering of vehicle probe data for secondary incident prediction. The 8 th Triennial Symposium on Transportation Analysis (TRISTAN) San Pedro de Atacama, Chile, June,

8 Park, H. and A. Haghani. Quantifying non- recurrent congestion impact on secondary incidents using probe vehicle data. The 54 th Annual Transportation Research Forum, Annapolis, MD, March, Park, H., A. Haghani, and X. Zhang. ATIS: Interpretation of Bayesian neural network for predicting the duration of detected incidents. The 92 nd Annual Meeting of Transportation Research Board, Washington, DC, January, Website(s) or other Internet site(s) 5. Technologies or techniques 6. Outreach activities 7. Courses and workshops 8. Inventions, patent applications, and/or licenses 9. Other products Several modeling approaches have been developed through the research process that are documented in the papers published in archival journals. 8

9 Part III Participants & Collaborating Organizations: Who has been involved? OST- R needs to know who has worked on the project to gauge and report performance in promoting partnerships and collaborations. Reporting Period 9/30/2015 3/31/ What organizations have been involved as partners? 2. Have other collaborators or contacts been involved? No other organization has been involved in this research. No. 9

10 Part IV Impact: What is the impact of the program? How has it contributed to transportation education, research and technology transfer? DOT uses this information to assess how the research and education programs: increase the body of knowledge and techniques; enlarge the pool of people trained to develop that knowledge and techniques or put it to use; and, improve the physical, institutional, and information resources that enable those people to get their training and perform their functions. Reporting Period 9/30/2015 3/31/ What is the impact on the development of the principal discipline(s) of the program? 2. What is the impact on other disciplines? When an incident occurs, emergency management systems initiate immediate responses to clear the event and re- open lanes as quickly as possible. For example, fatalities and loss of property can be greatly reduced by reduction of the response time to incidents. An efficient deployment of emergency response units considering sequence of future incidents has the potential to mitigate non- recurring congestion, to minimize its negative impact of incidents, and to increase road network reliability. In addition, our real- time optimal resource allocation reacts to external events effectively, and anticipates the future learning from the past to produce effective solutions. The solution to the stochastic optimization will provide prompt response to incidents and will play a crucial role in reducing delay, fuel consumption, and pollutant emission rates. Different policies for emergency vehicles (e.g., incident- to- incident response, fixed- depot) will produce different delays and ultimately contribute to the economic competitiveness. Advanced Traveler Information System (ATIS) also can provide emergency responders with accurate and timely deployment information via mobile applications. The expected waiting time omitted by previous studies can significantly impact the expected total delay compared to the relatively short travel time of response units. Allowing for flexibilities with secondary incidents decreases the expected total delay time compare to the solution without considering secondary incidents. As the number of available response vehicles increases, shorter total delay is expected. Further assignment of ERUs covering new locations occurs by using information about the most promising sites. 10

11 3. What is the impact on the development of transportation workforce development? 4. What is the impact on physical, institutional, and information resources at the university or other partner institutions? 5. What is the impact on technology transfer? The dissemination of research outcomes and their publications can help in developing a more informed and better educated workforce. The results and the methods of this project are disseminated to the research community and practitioners via publications and professional conferences. Emergency management personnel in various states can use the results of this research in developing better strategies for emergency management. 6. What is the impact on society beyond science and technology? An efficient control of emergency response and incident management greatly reduce injuries, incident induced delay, and adverse impacts on environment including fuel economy and emissions. 7. Additional impacts Nothing to report. 11

12 Part V Changes/Problems If not previously reported in writing to OST- R through other mechanisms, provide the following additional information or state, Nothing to Report, if applicable: Reporting Period 9/30/2015 3/31/ Changes in approach and reasons for change 2. Actual or anticipated problems or delays and actions or plans to resolve them 3. Changes that have a significant impact on expenditures 4. Significant changes in use or care of human subjects, vertebrate animals, and/or biohazards 5. Change of primary performance site location from that originally proposed Nothing to Report. Nothing to Report. Nothing to Report. Nothing to Report. Nothing to Report. 12