International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 01 | Jan 2024
p-ISSN: 2395-0072
www.irjet.net
Dynamic Urban Transit Optimization: A Graph Neural Network Approach for Real-Time Public Transportation Network Design and Management Mayank Nagar Prof. Atul V. Dusane, Dept. of Computer Science Engineering, MGM’s JNEC, Aurangabad, Maharashtra, India ---------------------------------------------------------------------***--------------------------------------------------------------------1. INTRODUCTION Abstract - As urban populations rise and sustainable urban development becomes more and more important, public transport networks must be designed and managed effectively. The purpose of this study is to investigate and overcome the difficulties involved in designing a real-time public transit network. Real-time data feeds and APIs from local authorities are used to dynamically portray the transport network. In the dynamic graph, transit stops or stations are represented as nodes, and connectivity and routes are represented as edges. Our GNN model's main objective is to optimize public transportation networks using data by learning in real time from features like vehicle location, arrival time, and passenger load. The objectives of this project are to increase overall urban mobility, decrease traffic, and improve efficiency. Through ongoing model training and inference, the system adjusts to the dynamic character of urban transit, giving planners and transportation authorities decision-support capabilities. To help make well-informed decisions in response to changing urban transport scenarios, visualization techniques are used.
The demand for efficient design and management of public transportation networks has reached an all-time high in the rapidly evolving landscape of urban mobility. This highlights the crucial role of sustainable development. To tackle the complex challenges associated with real-time public transportation network design and management, this research employs an advanced methodology based on Graph Neural Networks (GNNs). The aim is to explore this intricate domain and develop effective solutions. We use real-time data feeds and APIs from local transportation authorities to create a dynamic graph representation. In this representation, nodes represent transit stops or stations, and edges show the connectivity and routes that define urban transit systems. Our GNN model is designed to use real-time features such as vehicle locations, arrival times, and passenger loads, to optimize public transportation networks. We focus on improving efficiency, reducing congestion, and enhancing urban mobility, which contributes to the development of adaptive and responsive urban transit systems.
Urban planning and machine learning are coming together thanks to this research, which shows how GNNs are both practical and efficient. With the help of the suggested framework, urban transit systems can be redesigned to be more adaptable and responsive, encouraging sustainability and resilience in the face of changing mobility patterns and urban growth. The research creatively employs GNNs to address the intricacies of real-time public transport network design, in response to the growing issues posed by growing urban populations globally. The study advocates for the decrease of traffic congestion and the enhancement of overall urban mobility by utilizing real-time data feeds and APIs. The system adapts dynamically through ongoing model training and inference, providing urban planners and transportation authorities with decision-support tools. Visualization approaches are essential for providing in-the-moment insights and enabling educated decision-making in the dynamic field of urban transportation.
Our system is equipped with continuous model training and inference mechanisms that enable adaptation to the dynamic nature of urban transit. This provides decision-support tools for transportation authorities and urban planners. We have also incorporated visualization techniques that empower stakeholders with real-time insights, facilitating informed decision-making in response to evolving scenarios in urban transport. Our research intersects the fields of machine learning and urban planning. It showcases the effectiveness of GNNs in addressing the intricate demands of real-time public transportation network design. Our proposed framework establishes a foundational paradigm for creating adaptive and resilient urban transit systems. This is crucial to navigating challenges posed by urban growth and shifting mobility patterns, aligning with the overarching goals of sustainability and resilience in contemporary urban landscapes.
Key Words: Urban planning, public transit, Graph Neural Network, Decision-support, and connectivity.
To achieve these objectives, our research focuses on an indepth exploration of the dynamics of urban transit networks,
© 2024, IRJET
|
Impact Factor value: 8.226
|
ISO 9001:2008 Certified Journal
|
Page 708