The Traffic Management System provides a paradigm shift in traffic optimization, addressing the vexing concerns of traffic congestion and potholes as interconnected challenges. Using Braess' paradox method, our system dynamically changes driving patterns, effectively minimizing traffic congestion. We created ‘Braess routes’ using the Frank-Wolfe approach to decrease the travel time. A Two-Way Transformer neural network was used for accurate road condition evaluation, including the detection of potholes and traffic accidents. After being thoroughly trained on a dataset from a Pune Hyperloop station, our proposed approach significantly improves traffic flow and road safety. The use of advanced image segmentation facilitates the capacity to assess road conditions (e.g., the existence of potholes, construction zones, accidents, etc.) in real-time, improving its potential to ease traffic and decrease accidents. This entire system boasts outstanding figures such as a 35% reduction in peak-hour traffic congestion, a 47% reduction in incident response time, and a 33% reduction in average trip duration. Furthermore, it has significant implications for urban planning and transportation systems since it solves the pressing concerns of potholes and traffic congestion at the same time.
Congestion to Clarity: Innovative Traffic Management with Braess Paradox and Advanced Image Segmentation
2023-12-29
1665628 byte
Conference paper
Electronic Resource
English
Braess' paradox and congestion pricing in MATSim
DataCite | 2017
|Braess' paradox in a generalised traffic network
Online Contents | 2015
|Braess Paradox of traffic networks with mixed equilibrium behaviors
Online Contents | 2016
|Braess' paradox: Some new insights
Online Contents | 1997
|