Metaheuristic Techniques for Origin-Destination Matrix Estimation: Applied to Transport and Telecommunication Networks

This project addresses the critical need for accurate estimation of the Origin-Destination (OD) matrix in transportation and telecommunication networks, where existing models fail to fully represent real travel demand, leading to low computational reliability and operational issues. Specifically, it targets the Quito public transport system (Trolleybus), where the absence of demand measurement results in inadequate unit supply, uncertain waiting times, passenger saturation, and reduced profitability for the operating company (EPMTPQ). The proposed solution is structured in two main stages: the first focuses on estimating and updating OD matrices using a trip destination choice model based on empirical probability distributions derived from historical data. The key contribution lies in the multinomial model employed for these updates. The second stage is dedicated to predicting future OD matrices using an efficient predictive engine. The methodology will first be tested with simulated data and subsequently with real operational data from the Central Trolleybus corridor.<br/><br/><b>Goal</b>: <br/>To design a model based on metaheuristic techniques to estimate the origin-destination (OD) matrix, aiming to solve complex problems in vehicular traffic and telecommunication networks by improving system planning and operation.<br/><br/><b>Research lines</b>: <br/>Data science and simulation