مدل تحلیلی برنامه‌ریزی تولید سفر در شبکه حمل و نقل پس از بحران زلزله

با توجه به اثرات خاص وقوع بحران بر تقاضای سفر، نظیر هجوم تقاضا به شبکه در دوره زمانی کوتاه و رفتار آشوبناک استفاده‌کنندگان شبکه حمل‌و‌نقل، بررسی مساله تقاضای سفر پس از بحران امری ضروری به نظر می‌رسد. در این مطالعه بررسی شده است که سفرها بلافاصله پس از وقوع زلزله با چه اهدافی خواهد بود و چه فاکتورهایی می‌تواند در تصمیم‌گیری افراد برای سفر پس از زلزله موثر باشد. عوامل موثر بر رفتار ترافیکی افراد تحت 2 سناریو زلزله (شدید، متوسط) با استفاده از مدل لوجیت چندگانه مورد بررسی قرار گرفته است. کاربرد مدل ارائه شده در این تحقیق از نوع توصیفی می‌باشد. سفرها پس از زلزله با 4 هدف امدادرسانی، بازگشتن به خانه، تخلیه اضطراری و نداشتن سفر تعریف شده اند. داده‌های جمع‌آوری‌شده نشان داد در صورت وقوع زلزله شدید 90% از افراد ترجیح می‌دهند به منظور امدادرسانی و یا بازگشت به خانه سفر داشته باشند. در صورت وقوع زلزله متوسط نیز این آمار حدود 35% از افراد خواهد بود که برخلاف انتظار می‌تواند منجر به بحران ترافیکی در شبکه حمل‌و‌نقل گردد. نتایج مدلسازی نشان داد داشتن فرزند و همچنین فاصله زمانی از محل کار به محل امداد از مهمترین عوامل تاثیرگذار بر انتخاب سفر امدادرسانی هستند. بعلاوه متغیر تحصیلات (غیرمرتبط با زلزله) تاثیری در رفتار افراد پس از بحران ندارد.

Analytical model of trip production in transportation network after disaster

Unexpected events always occur without any alert about where or when they will happen. According to history of earthquake in Tehran, Iran, the probability of a huge earthquake occurrence (about 7 Richter) is high. The Unpredictable human behavior in disasters can affect the performance of the transportation networks. Considering the specific effects of an earthquake on the travel demand (i.e. the influx of travel demand in a short period and chaotic behavior of the users of the transport network), the issue of postearthquake travel demand needs to be investigated. Since people travel behavior would be quite different from the ordinary situation, this research proposes a method to estimate the demand based upon the interview survey. The goal of this study is to determine the trip purposes immediately after the earthquake and the factors affecting the individuals' decisions on their trip purposes. In most previous literatures, the majority of policies which have been modeled are based on unrealistic assumed demand. Many previous studies have acknowledged that more trip purposes in response to earthquake exist but few, if any, have examined it indepth. For example, Since an earthquake cannot be predicted, in a study conducted by Chang et al. to estimate the postearthquake travel demand, it is assumed that people will evacuate directly from their current locations immediately after earthquakes because under the nonotice earthquake scenarios, there is no time or considerably less time for people to return home or go to other places to pick up their relatives or friends [Chang et al. 2012] while most people will return to home to rescue their family [Hara, 2013]. This research developed discrete choice (Multinomial Logit) model to represent effective factors on travel demand behavior after 2 earthquake scenarios (Strong & weak) in a workday, with 4 trip purposes (rescue and Inquiry on Safety, returntohome, evacuation and noaction). This study investigated on travel behavior after an earthquake, based on a statistical analysis on stated preference (SP) questionnaires which were answered by 364 interviews in Tehran. The survey data indicated that, 90% of people may prefer to make trips in order to return to home or to rescue survivals after a powerful day earthquake The collected data expressed that although, it is not expected to have a problem in transportation network after a weak earthquake, the statistics from this study represented that about 35% of people will travel with different purposes because of their fear and it should be considered that despite a moderate earthquake will not destroy transport infrastructures, heavy traffic congestion will cause an emergency situation in transportation network. The goodness of fit (ρ2 statistic) of the model was obtained 0.425 that is a fairly good indicator for the discrete choice models. The model has also predicted the trip purposes in 67% of the observations correctly. The results of the model show that the most effective factors on destination choice behavior are gender, age, travel time, magnitude of earthquake, house ownership and family number. Also unrelated education to the earthquake is not effective on people travel behavior. Informing people about probable open routs after earthquake in advance would help planners for disaster management.