Establishing a Universal Relationship between the Traffic Mix in a Building and the Stops Data

Abstract

Previous work has used machine learning techniques to model the relationship between stops data and their locations within a building with the type of traffic mix. The methodology required knowledge of the types of floors in the building (occupant floors or entrance/exit floors), and used stops data in the up and down direction on occupant floors and entrance/exit floors to infer the mix of traffic in the building. However, previous work was based on one specific building.

This paper extends the previous work in order to show that the relationship is independent of the number of entrance floors and occupant floors. Such a universal relationship would allow the use of this derived set of equations for any building regardless of the number of floors and passenger numbers.

For this purpose, the research in this paper will employ three different sizes of buildings in terms of number of occupant floors and entrance/exit floors, as well as the variation in the car capacity (i.e., number of passengers in each round trip).

The methodology will rely on the principle of normalisation. Moreover, a Monte Carlo Simulation tool will be used to generate the data required for training the machine learning model.