Thesis supervisor: Tibor Sipos
Location of studies (in Hungarian): Department of Transport Technology and Economics Abbreviation of location of studies: DTTE
Description of the research topic:
The purpose of studying Traffic Safety is to examine the many factors that
influence the likelihood of occurrence of, and the resulting harm or losses from collisions. A difference in one of those factors could have led to a different outcome.
The consequences of collisions include fatalities, various levels of injuries and property damage. Interventions to improve safety may target the two components of risk:
• Collision prevention refers to measures aimed at preventing the collision from occurring.
• Crashworthiness refers to engineering features aimed at reducing losses, given that a specific collision occurs..
Road collisions involve various types of vehicles, motorized and nonmotorized, including
bicycles and animal-powered vehicles as well as pedestrians and fixed objects; vehicles and
pedestrians will be referred to as road users. All of this factors will be considerable amount to
estimate safety models as a function of explanatory variables describing the traffic
system. The explanatory variables can be viewed as falling typically into one of the following
three categories: the road, the vehicle, and the driver. A safety model, also called a crash
prediction model (CPM) or safety performance function, typically takes the form of an equation
linking safety to a set of variables.
Traffic Safety may be considered along the time and space dimensions. Comparing safety for
different periods allows study of temporal trends and evaluation of the impact of
countermeasures in development prediction models. Prediction models in traffic safety for different
locations or spatial entities will done to identify locations that are particularly hazardous and thus have
the most promising potential for safety improvements, which will be done through rankings and prioritization.
Prediction models of the Traffic safety link a set of variables describing the road system to safety
and other collision characteristics, such as collision outcomes. Historically, these models were
developed based on collision records stored in databases. More recently, there has also been an
interest in attempting to develop collision models and perform safety analysis based on actual
observations of collisions and interactions as recorded, for example, through video and various
in-vehicle sensors, as done in naturalistic driving experiments . In this study the prediction models in traffic safety will be built using Artificial Intelligince after reviewing previous studies to see methods used in prediction models , which is based on the appropriate methodology in our study, accident data, traffic flow, and engineering characteristics will be selecte for a number of major arterial streets, and two phases will be done:
- The first stage:
1. Identify the study area, a number of major arterial streets.
2. Compilation of engineering data, and traffic flow on the chosen roads.
3. Get traffic accidents on the chosen roads.
4. Blackspot sections which are determined from the analysis of the historical accident data.
- The second phase:
1. symblolize all data that as inputs to built the prediction model.
2. Build a predictive model using Artificial Intelligence.
3. Get results, test and check
At the end, accident prediction models are usually used to monitor the effectiveness of various road safety policies that have been introduced to minimize accident ccurrences. They also give an idea to transportation planners and engineers to determine new policies and strategies about road safety.
Required language skills: English Recommended language skills (in Hungarian): German Further requirements: Coding Python/R
Number of students who can be accepted: 1
Deadline for application: 2023-01-31
2024. IV. 17. ODT ülés Az ODT következő ülésére 2024. június 14-én, pénteken 10.00 órakor kerül sor a Semmelweis Egyetem Szenátusi termében (Bp. Üllői út 26. I. emelet).
Login
