Predicting Individual Trip Destinations with Artificial Potential Fields

Abstract

Abstract. This paper presents a method to model the intended destination of a subject in real time, based on a trace of position information and prior knowledge of possible destinations. In contrast to most work in this field, it does so without the need for prior analysis of habitual travel patterns. The method models the certainty of each POI by means of a virtual charge, resulting in an artificial potential field that reflects the current estimate of the subject's intentions. The virtual charges are updated as new information about the subject's position arrives. We experimentally compare a number of update rules with various parameter settings, showing that it is important to take the distance to a potential destination into account when updating the charge.

Generation of Human-Like Movements Based on Environmental Features

“…the scheme of interconnecting lines and points that represent the streets or roads existing in the real world, a virtual navigable representation is required. In this paper, we chose the Artificial Potential Field (APF) [19] idea as a base to describe the routing system. Points of Interest (POIs) need to be defined in order to codify all the environmental features of the area under analysis.…”

“…[2] employs recurrent neural networks (RNNs) in order to predict a user's destinations from their partial trajectories estimating the transition probabilities for the next time step. [19] presents a method to model the intended destination of a subject in real-time, based on a trace of position information and prior knowledge of possible destinations. The method models the certainty of each Point of Interest by means of a virtual charge, resulting in an artificial potential field that reflects the current estimate of the subject's intentions.…”

“…Given the similarity between the path planning problem and generation of trajectories task, the famous path planner algorithm A* was chosen to become the root of two novel ways to generate trajectories. A way to describe the environment was presented in [19], which suits perfectly the needs of the A* algorithm given its matrix/graph representation. Environmental features are easily downloadable from online map services and codified into a map, which is inspired by the concept of electrical charges.…”