Akademska digitalna zbirka SLovenije - logo
VSE knjižnice (vzajemna bibliografsko-kataložna baza podatkov COBIB.SI)
  • Machine learning methods for proposing mobility policies in urban areas using microscopic traffic simulations : master thesis = Metode strojnega učenja za predlaganje mobilnostnih strategij v urbanem okolju z uporabo mikroskopskih simulacij prometa : magistrsko delo
    Shulajkovska, Miljana
    Urbanisation and population growth have resulted in an increased demand for travel in large cities, which has led to high environmental and social costs. This has made mobility planning one of the ... biggest challenges for policy-makers in the 21st century. A highquality mobility system is crucial for improving the daily life of citizens by providing good connectivity, both between and within cities. However, it can be complex and expensive to develop, implement and evaluate mobility policies to meet the needs of people. In light of these challenges, this thesis proposes a novel machine learning (ML) approach to automatically generate mobility policies. The proposed method allows decision-makers to define potential city scenarios and a utility function, and the ML model uses this information to determine the best policy that meets the given constraints and preferences. The study focused on determining the best policy for closing different areas of the city centre at certain times and for certain durations. To evaluate the effectiveness of the proposed approach, a large amount of data were collected using a microscopic traffic simulator, specifically the Multi-Agent Transport Simulation (MATSim) software, to simulate a range of scenarios. The collected data was then used to train an ML model, which was developed to determine the best mobility policy based on key performance indicators (KPIs), such as air pollution and traffic analysis. The results of the study showed that the proposed approach was capable of effectively proposing mobility policies that met the needs of the city. This thesis makes a significant contribution towards developing a more efficient and effective mobility system in large cities. The proposed ML approach has the potential to transform the way mobility policies are developed and implemented by automating the process and ensuring that the policies are based on data-driven insights. This will help policy-makers to overcome the challenges posed by the increased demand for travel in large cities, and provide citizens with a better quality of life.
    Vrsta gradiva - magistrsko delo ; neleposlovje za odrasle
    Založništvo in izdelava - Ljubljana : [M. Shulajkovska], 2023
    Jezik - angleški
    COBISS.SI-ID - 168796419

Nobena knjižnica v sistemu COBISS.SI nima izvoda tega gradiva.
loading ...
loading ...
loading ...

Vnos na polico