SHORT-NOTICE BUS-BASED EVACUATION PLANNING FOR FLOOD PRONE AREAS

• 資料種別: 博士論文
• タイトル: SHORT-NOTICE BUS-BASED EVACUATION PLANNING FOR FLOOD PRONE AREAS
• 著者・編者: ASIF, NAWAZ QAZI
• 著者標目: ASIF, NAWAZ QAZI
• 出版年月日等: 2016
• 出版年（W3CDTF）: 2016
• 並列タイトル等: 直前の警告が可能な災害に対するバスによる避難計画に関する研究
• タイトル（掲載誌）: 博士論文（埼玉大学大学院理工学研究科（博士後期課程））
• 授与機関名: 埼玉大学
• 授与年月日: 2016-09-23
• 授与年月日（W3CDTF）: 2016-09-23
• 報告番号: 甲第1043号
• 学位: 博士(学術)
• 博論授与番号: 甲第1043号
• 本文の言語コード: eng
• 対象利用者: 一般
• 一般注記: type:text; The natural disasters are less frequent but more severe in terms of losses to property and humanity due to their least predictability and lower warning times, demanding some extensively planned preparation by the transportation planners for such areas. For some of the natural disasters with prior information like floods and hurricanes, the complete evacuations were not possible. Hurricanes Katrina and Rita, the two largest evacuations in the history of U.S. were the most successful and effective ones for the people with automobiles, but failed to evacuate public transport dependent citizens. Further, a huge number of life-losses are occurring quite frequently in Pakistan due to floods in different areas. Life losses caused by disasters with short-notice available prior to their occurrence can be minimized through effective evacuation planning measures and preparedness for evacuation of the people living therein before / during floods. This was the main motivation of my research discussed in details in Chapter 1.A number of studies have been proposed to present some flood evacuation models with different optimized variables, mostly dealing with private vehicle / car based evacuations. Unfortunately, bus-based evacuation planning has not been studied quite well in the existing literature. The main focus of our study is short-notice bus-based evacuation planning for flood prone areas. During the course of study about the existing evacuation models, it was observed that selection of road network data to be used in evacuation operations plays an important role. Therefore, in Chapter 2 of my research, I have carried out a study about identification of important or critical links to be used in the evacuation operations.In the literature, evaluation parameters like traffic volumes, link capacities, travel times, and population of nodes are found to be used to simulate specific scenarios. These parameters significantly affect the results for critical links. Moreover, it is quite difficult to get the data of such parameters in developing countries. A measure based on link position within the network and its connectivity, Network Path-length Index (NPI), has been proposed to evaluate geometric importance of a link. The NPI of a link “L” is defined as the ratio of sum of all possible shortest origin-destination (OD) shortest path lengths when the link “L” is removed to the sum of all shortest OD shortest path lengths when the link “L” is present. Revised shortest path lengths after removal of subject link are calculated, thereby reflecting solely the effect of network topology. To investigate the effect of including other evaluation parameters, critical links identified by three indexes proposed in the literature are also compared with those by the NPI. Applications of the proposed index to the real world explained through a case study include prioritization and ranking of links within a network and identification of isolating links, if any.The second part of the study about evacuation planning starts from Chapter 3. To develop an evacuation model, extensive literature review of existing evacuation models was carried out. The researchers have assumed either a known fixed number of evacuees or a single bus trip to shelter on specified evacuation routes, or both of the constraints. These constraints place their models away from real situations, where evacuees continuously arrive at pickup points and multiple bus trips are necessitated for economical use of a limited number of buses. A model using mixed integer linear programming, namely, model for short-notice bus-based evacuation under dynamic demand conditions (SBED Model), has been proposed to evacuate evacuees that continuously arrive at pickup points. Evacuees are sheltered through multiple trips of available buses with flexible route options. Factors affecting bus trip assignment investigated through a case study of flood evacuation planning for Kawajima Town were identified. These include: the number of evacuees at a pickup point, distance / travel time to shelter from that pickup point, and the total modeled warning time. A higher efficiency of bus capacity use in performing multiple trips by the buses and efficient use of warning time suggests the model use in bus-based evacuation planning.Different techniques used to minimize the evacuation time aim to improve either the demand or the supply side of the mechanism. Great care must be exercised to avoid negative impacts of such improvement efforts. In Chapter 4 of our study, two important factors, variation in demand and evacuation route flexibility, are discussed with respect to their effects on the optimality of model outputs. Continuous demand at all pickup point is compared with that of fixed static demand option and the flexible evacuation route option is compared with the fixed one-way evacuation route. The resulted output indicated two important findings. First, although the evacuees were assumed to arrive at pickup points quite early (within the first 15 min of total warning time of 90 min) in the continuous-demand case, better results in terms of evacuation times were observed for the fixed-demand case. Second, introducing evacuation route flexibility into the model was found to have a positive impact (i.e., fewer resources are required to evacuate the same evacuee demand), although the model run times were increased due to induced complexity in the model.Lastly, in Chapter 5 of the study, an effort is made to model the gradual flooding of pickup points and increase in the travel times between different points, either due to congestion delays or because of flooding of links connecting the points. A case study for the evacuation planning of Kawajma Town is presented and a scenario to simulate such situation has been analyzed using our developed SBED model. The selected pickup points are assumed to become unavailable gradually after lapse of certain specified time. In addition, a scenario to take into account the effect of congestion on travel times has also been analyzed.It was observed that the number of evacuees was almost the same for the case of gradual flooding as for the normal case of a uniform warning time, mainly due to higher flooding-time intervals. However, large reductions were observed for the cases with congestion effects. This output emphasizes the need to take steps to reduce congestion delays during evacuation operations. Secondly, while determining the bus trip patterns, the SBED model did not consider any other factor to prioritize the evacuation of some particular point over the minimum travel time to the shelter and number of evacuees at a pickup point, as long as the number of buses available was less than that required for complete evacuation .i.e. the model worked similar to an evacuation model for no-notice disasters primarily focusing on maximizing the number of evacuees that can be saved within the warning time. However, once the number of buses was sufficient to complete the evacuation, the priority order of pickup points became exactly the same as the order of gradual flooding. This finding confirms the applicability of the SBED model to gradual flooding of pickup points. Moreover, by assigning higher values to the sink arcs from the points going to be flooded earlier than from the points to be flooded later, the buses could be forced to assign greater priority to the former; however, this compromised the optimality of the model output.In the last Chapter 6 of the thesis, I have summarized the remark conclusions and the limitation of our study with recommendations for further research. The limitations of the study and further research options can be summarized as below:Firstly, the NPI use is limited to link importance evaluation for single link removal at a time and within a single network only i.e. it cannot be used for comparison between the links of different networks. Also, it quantifies the importance of links out of existing links within a network, instead of identifying some absolute important links for a network. Further generalization of the NPI can be made once it is explained in details through its use in complex urban road networks.For the second part of study about bus-based evacuation planning and its extensions, the lengthy run times limit the model use to only planning stage applications and limited number of pickup points. Further research is suggested to develop an efficient algorithm based on the patterns of bus trip results observed to solve the model in shorter run times so that it can be used for a higher number of pickup points with route flexibility. Such an algorithm will widen the scope for using the proposed model for large urban areas with complex road networks. Additionally, model input modification during evacuation execution stage, if necessitated based on changed input scenarios, can only be made possible if shorter run times are achieved in the model. Furthermore, uncertainties about the availability of road links due to earlier flooding can also be included in the model. For instance, flooding can damage road links due to the overflow of local drains, and some points may become inaccessible.; Copyright iDissertation Approval iiAcknowledgment iiiExecutive Summary ivTable of Contents viiiList of Table xiList of Figures xiiCHAPTER 1: INTRODUCTION 011.1 General 011.2 Evacuation Planning 031.3 Framework of the Dissertation 031.4 Objectives of the Research Study 051.4.1 Link Importance Evaluation (Part-I) 051.4.2 Bus-based Evacuation Model for Short-notice Disasters (Part-II A) 051.4.3 Evacuation Demand and Evacuation Route Flexibility (Part-II B) 061.4.4 Modeling the Gradual Flooding and Congestion Delays in Bus-based Flood Evacuation Planning (Part-II C) 06References 07PART-I CHAPTER 2: LINK IMPORTANCE EVALUATION 082.1 General 082.2 Proposed Measure 092.3 Effect of Evaluation Parameter 102.3.1 Index by Scott et al. 2006 102.3.2 Index by Taylor et al. 2006 132.3.3 Index by Yang et al. 2012 152.4 Applications of the Network Path-length Index 172.4.1 Explanation through a Case Study 172.4.2 Limitations of the Proposed Index 222.5 Conclusions 25References 27PART-IIA CHAPTER 3: BUS-BASED EVACUATION MODEL FOR SHORT-NOTICE DISASTERS 283.1 Introduction 283.2 Problem Statement 313.3 Model Description 323.3.1 Time-Space Network 323.3.2 Model Formulation 353.4 Explanation through a Case Study 383.4.1 Collection of Data 383.4.2 Summary of Results 403.4.3 Discussion 423.5 Conclusions 46References 48PART-IIB CHAPTER 4: EVACUATION DEMAND AND EVACUATION ROUTE: TWO KEY FACTORS IN EVACUATION PLANNING 504.1 Introduction 504.2 The Model 524.3 Case Study 524.4 Results and Discussion 544.4.1 Objective Function for Evacuation Models 554.4.2 Demand Variations 564.4.3 Evacuation Route Flexibility 584.5 Conclusions 60References 62PART-IIC CHAPTER 5: MODELING THE GRADUAL FLOODING AND CONGESTION DELAYS IN BUS-BASED EVACUATIONS 645.1 Introduction 645.2 The Model 655.3 The Case Study 675.4 Results and Discussion 705.4.1 Bus Trip Pattern Analysis 725.5 Conclusions 79References 82CHAPTER 6: CONCLUSIONS AND FURTHER RESEACH 846.1 Remark Conclusions 846.2 Limitation of the Study and Recommendations for Further Research 86APPENDEXES 88Appendix 1 List of Publications and Proceedings related to the Dissertation 88Appendix 2 Curriculum Vitae of the Author 89; 主指導教員 : 久保田尚
• DOI: 10.24561/00010433; https://doi.org/10.24561/00010433
• 国立国会図書館永続的識別子: info:ndljp/pid/10952392; https://dl.ndl.go.jp/pid/10952392
• コレクション（共通）: 障害者向け資料
• コレクション（障害者向け資料：レベル1）: テキストデータ
• コレクション（個別）: 国立国会図書館デジタルコレクション > デジタル化資料 > 博士論文; https://dl.ndl.go.jp/collections/A00014
• 収集根拠: 博士論文（自動収集）
• 受理日（W3CDTF）: 2017-09-02T15:28:24+09:00
• 作成日（W3CDTF）: 2017-07-19
• 記録形式（IMT）: application/pdf
• オンライン閲覧公開範囲: 国立国会図書館内限定公開
• デジタル化資料送信: 図書館・個人送信対象外
• 遠隔複写可否（NDL）: 可
• 掲載誌（URI）: http://dx.doi.org/10.24561/00010433; http://id.nii.ac.jp/1586/00010433/
• 連携機関・データベース: 国立国会図書館 : 国立国会図書館デジタルコレクション; https://dl.ndl.go.jp

目次

2023-09-04 再収集