A Novel Approach for Route Prediction in Multimodal Transport Networks: A Monte Carlo Simulation and Long Short-Term Memory-based Model

Surya PrakashEmail

Bibhya SharmaEmail

School of Information Technology, Engineering, Mathematics and Physics, The University of the South Pacific, Suva 00679, Fiji.

Abstract

This research introduces a pioneering approach for route prediction in multimodal transport networks, leveraging the capabilities of Artificial Intelligence (AI). At the core of the proposed methodology is the integration of Monte Carlo simulations with Long Short-Term Memory (LSTM) networks, aimed at understanding and predicting complex route dynamics in multimodal transport systems. This combination aims to capture the intricate dynamics and uncertainties of multimodal transport, utilizing LSTM's strength in processing temporal sequences and recognizing data dependencies over time.Our obtained results show that the model is predicting feasible routes given starting sequences and those predicted routes are optimal and matches the data that the model was trained in. The results also show that the obtained model generates plausible predictions for new, unseen sequences thus demonstrating remarkable adaptability. This demonstrates the model's real learning potential, beyond mere data memorization. The integration of Monte Carlo simulations and LSTM networks represents a significant step forward in multimodal transport route prediction, opening avenues for future enhancements and broader applications in transportation studies, including navigating network obstacles. This research not only contributes to the field by improving route prediction accuracy but also by suggesting a framework for future developments in the domain.

A Novel Approach for Route Prediction in Multimodal Transport Networks: A Monte Carlo Simulation and Long Short-Term Memory-based Model