Analysis and prediction of time distribution of road accidents in Karaj-Qazvin freeway

Document Type : Applied Article


1 Ph.D. Student of Meteorology, Faculty of Earth Sciences, Shahid Beheshti University

2 Professor, Department of Natural Geography, Faculty of Earth Sciences, Shahid Beheshti University

3 Assistant Professor, Department of Natural Geography, Faculty of Earth Sciences, Shahid Beheshti University

4 Assistant Professor of Climatology, Faculty of Social Sciences, Payam Noor University, Tehran



Humans face danger in some way every day, preventing environmental hazards and effectively dealing with them as traumatic events is one of the most important concerns of policymakers and executives in various areas of social issues [3]. One of the most important risks are traffic accidents and humans have always been looking for a solution to reduce or at least control the risks in their lives. Among the types of accidents, road accidents are one of the most important causes of death and serious injuries [4]. The cost of accidents on the country's roads is very high, and one of the most effective ways to reduce accidents, violations and increase safety in driving is to accurately identify the parameters that affect accidents and measure their effectiveness, the results of which can be provided to the responsible organizations in order to increase The safety level on the roads should be used to reduce accidents [1,2].
Therefore, the purpose of this research is to analyze and predict the time distribution of road accidents, of course, considering the extent and diversity of the country's roads, this research is done on one of the important roads that is the connecting bridge between Tehran and Qazin cities. So, considering the importance of this axis and of course the need to obtain correct information at the right time for better decision-making, the achievement of this research is to obtain the temporal pattern of the occurrence of road accidents on the Karaj-Qazvin freeway, so that by predicting the relative time of the occurrence of accidents, it is possible to plan to provide the correct measures to reduce the costs caused by accidents to some extent.
To achieve this goal, two basic questions are raised: What is the temporal distribution of road accidents in the Karaj-Qazvin axis? Is it possible to find a pattern between the season, day and time of occurrence of road accidents and will this pattern be able to predict the possible time of occurrence of road accidents in the studied axis?
Karaj-Qazvin freeway is one of the two connecting roads between Qazvin and Tehran, which is used by cars, especially passenger cars, more than the old road due to its freeway and relative safety conditions. And it is one of the main traffic routes in Qazvin province.
Materials and methods
 In terms of objectives, this research is classified as applied research, and from the perspective of interpretation and analysis of findings, it is descriptive-analytical, the general approach to this research is quantitative, and based on data mining and pattern finding techniques based on J48 algorithm of Weka software, that it is one of the simplest and perhaps the most widely used data mining software for preparing the decision tree, so first the statistical data of road accidents from the reports of the traffic police, at the provincial level in the desired axis, were collected and based on the process of data analysis and data mining and by CRISP technique has been analyzed in 6 stages of data.
The main reason for using this technique is the large amount of information that puts it in the category of big data, and therefore we have had to use data mining techniques to analyze this type and volume of data. Of course, before entering the data into the data mining software, in order to ensure the existence of a relationship and correlation between the time variables of this research (season, time and day of the occurrence of road accidents), these variables were nominal from the correlation test of nominal variables, lambda coefficient is used, data related to road accidents were extracted from the data contained in the traffic police report form during the years 2019-2020 on the Karaj-Qazvin highway.
Discussion and results
In this article, accident data has been analyzed based on the day, season, and time of the accident:
Fig. 1. Time of occurrence of road accidents in Karaj Qazvin Freeway during 2010-2019
Fig. 2. The occurrence of accidents in different seasons during the years 2010-2019 on the Karaj-Qazvin freeway.
Fig. 3. The occurrence of accidents based on days of the week on the Karaj-Qazvin freeway
Considering the above figures, can be concluded that most of the accidents on the Karaj-Qazvin freeway occurred at 19-21 hours, in the spring and summer seasons, and the accidents had a relatively equal distribution among the days of the week, but on Thursdays and Fridays, according to The location of this freeway and the connection of Tehran to the northern cities, and the possibility of congestion on weekends due to the increase in trips, can be justified.
In order to ensure the temporal relationship of accidents with each other on the Karaj-Qazvin freeway, the correlation coefficient test has been used, of course, considering that the main variables here (season, day, and hour) are all nominal variables, so The correlation test of nominal variables (lambda coefficient) is used here, the sig calculated for all three variables is less than 0.05 and with 95% confidence, the correlation and relationship of these three variables can be confirmed. In the following, the decision tree and prediction model of the purpose of this research is drawn by the data mining software Weka (J48 algorithm) in the form of Figure 4.
Fig. 4. Prediction tree and time pattern of road accidents in Karaj-Qazvin freeway
 In the model presented in Figure 4, the probability of road accidents on the Karaj-Qazvin freeway, the relationship between the three variables of the time of the accidents, and then the seasons and finally the days of the week are analyzed. This model will actually help in predicting the time of road accidents on the Karaj-Qazvin freeway.
The results of the analysis of the branches in Figure 4 showed that: on the Karaj-Qazvin highway, in the early morning hours, which is, between 24:00 and 4:00 in the early morning, in all seasons of the year (spring, summer, autumn and winter), the probability of road accidents on Fridays is higher than other days of the week.
But in the morning hours, between 4 and 11 am: in the spring season: Sundays, in the summer season: Fridays, in the fall season: Wednesdays, and in the winter season: Thursdays; the probability of road accidents is much higher than other days. Also, during the noon hours, that is, between 11:00 and 15:00 in the spring and autumn: the probability of road accidents on Fridays is higher than on other days of the week. In the summer season: this possibility on Thursdays and in the winter season on Mondays every week; the possibility of road accidents on Karaj-Qazvin highway will be more than other days.
In the afternoon between 15:00 and 17:00: in all seasons (summer, autumn and winter), the probability of road accidents on Saturdays will be higher than on other days of the week. In the evening hours between 17:00 and 19:00 in the summer, autumn and winter seasons (three consecutive seasons of the year): the possibility of road accidents on Fridays and in spring (the first season of the year) the possibility of road accidents on Saturdays more than others it will be the days of the week.
And also during the night hours between 19:00 and 24:00 in the spring and winter seasons: the possibility of road accidents on Fridays, in the autumn season: on Sundays and in the summer season: on Saturdays; the probability of road accidents will be higher.


  • احمدی، مهدی؛ و علی‌محمدی، عباس (1394). «آنالیز زمانی و مکانی تصادفات رانندگی با استفاده از تراک پنجره‌ای فازی»، مهندسی حمل‌ونقل، دورۀ 7، شمارۀ2، ص 205-191.
  • افشاری آزاد، محمدرضا (1387). «بررسی عناصر اقلیمی بر روی تصادف‌های جاده‌ای محور رشت- بندرانزلی»، مطالعات برنامه‌ریزی سکونتگاه‌های انسانی، دورۀ 3، شمارۀ 7، ص26-9.
  • بررسی آمار تصادف‌های جاده‌ای در ایران (1399). تصادف نیوز، بازیابی‌شده در 12 آذر 1401.
  • بهبهانی، حمید؛ عفتی، میثم؛ و مرتضایی، سمانه (1399). «ارائۀ روش جهت تحلیل شدت تصادفات راه‌های برون‌شهری مبتنی بر توابع خوشه‌بندی مکانی و داده‌کاوی به روش درخت تصمیم، محور مورد مطالعه: آزادراه قزوین- لوشان»، مهندسی عمران امیرکبیر، دورۀ 52، شمارۀ 6، ص 1438-1419.
  • بهتوئی، حسن (1393). تحلیل تصادف‌های جاده‌ای با رویکرد اقلیمی در محور قزوین- رشت، پایان‌نامۀ کارشناسی ارشد، دانشکدۀ جغرافیای دانشگاه تهران.
  • بهتوئی، حسن؛ و التماسی، مهشید (1399). «داده‌کاوی تصادف‌های جاده‌ای شمال غرب تهران، مطالعات مدیریت ترافیک، دورۀ 1399، شمارۀ 59، ص 148-127.
  • پردال، حسن (1397). ارائۀ مدل تصمیم‌گیری در رابطه با ایمنی سازی راه‌های برون‌شهری در شرایط آب‌وهوایی و اقلیمی نامطلوب-مطالعه موردی جاده ایرانشهر-سرباز، پایان‌نامۀ کارشناسی ارشد، دانشگاه آزاد اسلامی، واحد زابل.
  • پیروتی، منصور (1389). بررسی تأثیر پارامترهای اقلیمی بر تصادف‌های جاده‌ای (مطالعۀ موردی: جادۀ سردشت-ارومیه)، پایان‌نامۀ کارشناسی ارشد، دانشگاه تبریز، دانشکدۀ علوم انسانی و اجتماعی.
  • خالدی، شهریار؛ بهتوئی، حسن؛ کیخسروی، قاسم؛ و التماسی، مهشید (1401). «تحلیل داده‌ها و مدل‌سازی علل تامۀ تصادف‌های جاده‌ای در شرایط نامساعد جوی: محور تهران- قزوین (جاده قدیم)، مطالعات مدیریت ترافیک، دورۀ 17، شمارۀ پیاپی 65، ص 94-63.
  • ذوالفقاری، سارا؛ رضائیان، علی؛ و شکوهیار، سجاد (1394). «خوشه‌بندی داده‌های تصادفات جاده‌ای با استفاده از فنون داده‌کاوی»، مطالعات پژوهشی راهور، دورۀ 4، شمارۀ 14، ص 47-79
  • رتبۀ ایران در آمار جهانی مرگ‌ومیر (1397). تابناک، بازیابی‌شده در 12 آذر 1401.
  • ساری صراف، بهروز؛ ولیزاده کامران، خلیل؛ و مجیدی، عثمان (1388). «بررسی اثرات عناصر اقلیمی بر تصادف‌های جاده‌ای (مطالعۀ موردی: محور ساری- رامسر)»، اولین کنفرانس ملی تصادفات و سوانح جاده‌ای و ریلی.
  • شتابزدگی برای مردن (1397). همدلی (27 آبان)، بازیابی‌شده در 20 آبان، 1401.
  • شهابی، هیمن؛ خورشیددوست، علی‌محمد؛ و حسینی، میرکامل (1390). «ارزیابی نقش عناصر اقلیمی بر تصادفات جاده‌ای (مطالعۀ محور سقز- سنندج)، تحقیقات جغرافیایی، دورۀ 26، شمارۀ 3، ص 189-212.
  • شهرداری قزوین (1401). مشخصات طبیعی و تقسیمات کشوری در استان قزوین، بازیابی‌شده در 27 خرداد 1401.
  • شناسایی و تحلیل علل و عوامل تصادف‌های جاده‌ای استان قزوین (1394). خبرگزاری ایرنا، بازیابی‌شده در 4 تیرماه 1401.
  • فرد علایی، غلامرضا؛ لطیفی، آرمان؛ پیوندی، رضا؛ افکاری، علی؛ و محمدی، زهرا (1399). «بررسی عوامل مرتبط با تصادف‌های جاده‌ای محور مراغه- هشترود (یک مطالعۀ مقطعی)، طب اورژانس ایران، دورۀ 7، شمارۀ1،  ص 6-1.
  • کلانتری، علی؛ و علیان، سحر (1401). «تحلیل تصادف‌های جاده‌ای با تأکید بر خصوصیات محیط و جاده در سیستم اطلاعات مکانی (مطالعۀ موردی: محور کرج-کندوان)»، پژوهش‌های جغرافیایی انسانی، دورۀ 54، شمارۀ 2، ص 582-563.
  • متین پارسا، محمد؛ و غلامی، نبی‌اله (1397). «مفهوم‌شناسی پاسخ به مخاطرات طبیعی براساس رویکرد جرم‌شناختیِ گذار از نظر به عمل»،مدیریت مخاطرات محیطی، دورۀ 5، شمارۀ 2، ص 142-127.
  • وطن‌پرست، مهدی؛ افشاری، علیرضا؛ رضائی عارفی، محسن؛ و نورمحمدی، علی‌محمد (1396).«ارزیابی تأثیر عناصر اقلیمی و عوامل انسانی در بروز تصادف‌های جاده‌ای با استفاده از منطق فازی (نمونۀ موردی محور مشهد- قوچان)،مجلۀ کاربردی سیستم اطاعات جغرافیایی و سنجش از دور در برنامه‌ریزی، دورۀ 8، شمارۀ 4، ص 66-52.
  • یزدانی، محمدحسن؛ پاشازاده، اصغر؛ و زادولی، فاطمه (1397). «تحلیل مکانی- زمانی تصادفات شهری شهر اردبیل و علل بروز آن»، جغرافیای اجتماعی شهری؛ دورۀ 5، شمارۀ 13، ص 147-127.
  • Abdella, M.; & Shaaban, K. (2020). “Modeling the Impact of Weather Conditions on Pedestrian Injury Counts Using LASSO-Based Poisson Model”, Arabian Journal for Science & Engineering (Springer Science & Business Media B. 46(5), pp: 4719-4730.
  • Comi, A.; Polimeni, A.; & Balsamo, (2022). “Road Accident Analysis with mining approach: evidence from Rome. Transportation Research Procedia”, 62, pp: 798-805.
  • Kim, S.; Lee, J.; & Yoon, T. (2021). “Road surface conditions forecasting in rainy weather using artificial neural networks”, Safety Scince,
  • Uddin, M.; & Huynh, N. (2020). “Injury severity analysis of truck-involved crashes under different weather conditions”, Accident Analysis & Prevention,
  • Viswanath, P. K.; Kumar, G.A.; Wamique, M.; & Kumar, I. (2021). Road Accident Prediction Model Using Data Mining Techniques,. (2021) 5th International Conference on Computing Methodologies and Communication (ICCMC), pp: 1618-1623, doi: 10.1109/ICCMC51019.2021.9418336.