تهیه نقشه ریسک تصادفات رانندگی مبتنی بر اطلاعات مکانی مردم گستر با استفاده از روش های تصمیم گیری تحلیل سلسله مراتبی و تحلیل شبکه

یکی از معضلات اجتماعی در جهان و به خصوص در کشورهای در حال توسعه که هر ساله تلفات انسانی و هزینه های اجتماعی، اقتصادی و فرهنگی زیادی را به جامعه تحمیل می کند، تصادفات رانندگی است. بنابراین شناسایی نقاط حادثه خیز تصادفات امری بسیار ضروری است. هدف این مطالعه، تعیین نقاط حادثه خیز و تهیه نقشه حساسیت تصادفات رانندگی در شهر بابل با استفاده از اطلاعات مکانی مردم گستر می باشد. بدین منظور، با توجه به ویژگی های منطقه، از معیارهای فاصله از  مراکز جمعیتی، نزدیکی به میدان های شهری، فاصله از پل عابر پیاده، نزدیکی به تقاطع استفاده شده است. نقاط حادثه خیز شناسایی شده توسط اطلاعات مکانی مردم گستر به همراه معیارهای تعیین شده با به کارگیری روش های تحلیل سلسله مراتبی و تحلیل شبکه به صورت زوجی مقایسه و وزن هر یک از آنها محاسبه گردید. تقشه نهایی حساسیت تصادفات رانندگی حاصل از هر دو روش متداول و اطلاعات مردم گستر تهیه شد. به منظور ارزیابی دقت نقاط حادثه خیز شناسایی شده حاصل از اطلاعات مکانی مردم گستر و همچنین نقشه های حساسیت، از داده های رسمی پلیس با استفاده از معیارهای منحنی تشخیص عملکرد سیستم و ضریب کاپا استفاده شده است. بر اساس نتایج به دست آمده، مرکز خرید مسجد جامع، مرکز خرید یکشنبه بازار (شهاب نیا)، تقاطع فرهنگ و میدان ولایت، به عنوان حادثه خیزترین نقاط در شهر بابل شناخته شدند. همچنین از بین 4 معیار تعیین شده، فاصله از مراکز جمعیتی و فاصله از تقاطع، به ترتیب، مهمترین معیارها می باشند. نتایج حاصل از معیارهای ارزیابی نشان دادند که استفاده از اطلاعات مکانی مردم گستر با دقت خوبی می تواند در تعیین نقاط حادثه خیز تصادفات در شهر بابل موثر واقع شود. همچنین روش تحلیل شبکه نسبت به روش تحلیل سلسله مراتبی در تهیه نقشه حساسیت بهتر عمل کرده است.

Mapping the risk of road traffic accidents based onVolunteered Geographic Information usingAnalytic Hierarchy Process (AHP)and Analytic Network Process (ANP)

Extended AbstractIntroduction and ObjectiveRoad traffic accidents impose numerous social, economic, and cultural costs upon various societies, especially developing countries. Identification of accident blackspots is a method proposed to deal with car accident risks. Among various events associated with transportation network, road traffic accidents play a significant role, because of their specific features, including high frequency, high intensity and the chance of direct involvement of all members of the community.This problem is more conspicuous in developing countries such as Iran. The present study aims to identifyaccidentblackspotsand to prepare risk map for road trafficaccidents in Babol city using volunteered geographic information. Materials and methodsAccording to the characteristics of the study area, the present study takes advantage of criteria such as distance from population centers, proximity to city squares, distance from footbridges, and proximity to road intersections to identifyaccidentblackspotsand a prepare risk map for roadtraffic accidents in Babol city. Accident blackspots detected by volunteered geographic information, along with the criteria determined by applying analytic hierarchy process (AHP) and analytic network process (ANP) were compared in a pairwise manner, and their respective weight was calculated to showtheir specific level of impact. Ultimately, a risk map was produced for the risk of road traffic accidents obtained from each method. In order to evaluate the accuracy of the identified accident blackspots obtained from volunteered geographic information, as well as the accuracy of susceptibility maps, ROC curve and Kappa Coefficient were applied to police official records. Results and Discussion According to the findings, Jame Mosque shopping center, Shahabnia shopping center, intersection of Farhangstreet and Velayat square were identified as the most accidentprone areas in Babol city. Also, among the prespecified criteria, distance from population centers and distance from intersections are considered to be the most important criteria, respectively. Results obtained from the evaluation criteria indicatedhigh accuracy of volunteered geographic information, and thus it is concluded that this kind of information can be effective in determining the accident blackspotsinBabol city. Also, the ANP method works better than AHP method in preparing the risk map of accidents. Conclusion and Future worksDue to the large number of road accidents, especially in developing countries,the issue of accident blackspotsand providing a risk map for road trafficaccidents are an essential part of roads safety. In the present study, volunteered geographic information was used, along with multivariate decisionmaking methods of analytic hierarchy process (AHP) and analytic network process (ANP) to identifyaccident blackspots based on number, causes and severity of accidents and to develop a risk map for driving accidents in Babol city. Moreover, the criteria of distance from population centers, proximity to the city squares, distance from the footbridges, and adjacency to intersections were used to determine accident blackspotsand to prepare a risk map for driving accidents in Babol city. According to the results, Jame Mosque shopping center, Shahabnia shopping center, Farhang intersection and Velayat square were identified as the most accidentprone points in Babol city. Also, distance from population centers and distance from intersectionswere identified as the most important criteria, respectively. Evaluation criteria demonstrated that volunteered geographic information can be effective and accurate in determining accident blackspotsinBabol city. Also, the ANP method worked better than AHP method in preparing the risk map of driving accidents. The method proposed in this study to identify accident blackspots and preparedriving accidents risk maps can be generalized to other areas. Basedon the characteristics of specific routes, other criteria such as arc radius, longitudinal slope can alsobe used. It is also suggested that the results of other methods used for investigation ofaccidentblackspotsand production of risk maps based onvolunteered geographic information (VGI) are compared with the results of the present study.