Prediction of Pedestrian Speed with Artificial Neural Networks

TL;DR

This paper explores using artificial neural networks to predict pedestrian speeds in complex environments, showing they outperform classical models in differentiating geometries and improving speed estimates.

Contribution

It introduces a neural network approach for pedestrian speed prediction and demonstrates its effectiveness over classical models in mixed geometry scenarios.

Findings

Neural networks can differentiate between corridor and bottleneck geometries.

Neural networks improve pedestrian speed estimation in mixed environments.

Classical models are less effective in complex, variable geometries.

Abstract

Pedestrian behaviours tend to depend on the type of facility. Therefore accurate predictions of pedestrians movements in complex geometries (including corridor, bottleneck or intersection) are difficult to achieve for classical models with few parameters. Artificial neural networks have multiple parameters and are able to identify various types of patterns. They could be a suitable alternative for forecasts. We aim in this paper to present first steps testing this approach. We compare estimations of pedestrian speed with a classical model and a neural network for combinations of corridor and bottleneck experiments. The results show that the neural network is able to differentiate the two geometries and to improve the estimation of pedestrian speeds when the geometries are mixed.