Juha-Matti Kuusinen

A study on the arrival process of lift passengers in a multi-storey office building

This article presents a study on the process of how passengers arrive at lift lobbies to travel to their destinations. Earlier studies suggest that passengers arrive at the lift lobbies individually with exponentially distributed inter-arrival times, that is, according to a Poisson process. This study was carried out in a multi-storey office building. The data was collected using a questionnaire, digital video recordings and the lift monitoring system. The results show that, in the studied building, passengers arrive in batches whose size varies with the time of day and the floor utilization. In addition, the batch arrivals follow a time-inhomogeneous Poisson process with piecewise constant arrival rates. Practical applications : This article contributes to the basic understanding of passenger behaviour, and how people move around in buildings and arrive at the lift lobbies. It is proposed that the model for the passenger arrival process should take into account that passengers do not always arrive individually but also in batches. The passenger arrival process affects the design of elevators. It will also affect the passenger generation in building traffic simulations.

The Elevator Trip Origin-Destination Matrix Estimation Problem

In this paper, we consider the problem of finding the passenger counts for the origin-destination pairs of a particular single transit route called elevator trip. Assuming that passengers first alight and then board a stopping elevator, we can define an elevator trip as successive stops in one direction of travel with passengers inside the elevator. The elevator trip origin-destination passenger counts, i.e., elevator trip origin-destination matrices, estimated for a given time interval can be combined into a building origin-destination matrix that describes the passenger flow between every pair of floors in the building during that interval. The building origin-destination matrices of successive intervals form traffic statistics that can be used to forecast passenger traffic. The forecasts model the uncertainties related to future passengers, and need to be taken into account in elevator dispatching to make robust dispatching decisions in constantly changing traffic conditions. Many methods exist for estimating an origin-destination matrix for a single transit route such as a bus line. These methods estimate average origin-destination passenger counts from observations made during a given time period on the same route. Because an elevator trip is request driven, there may not be two similar elevator trips even within a day. This means that we need to estimate a separate origin-destination matrix for each elevator trip. A natural requirement then is that the estimated origin-destination passenger counts are integer valued. We formulate the elevator trip origin-destination matrix estimation problem as a box-constrained integer least squares problem, and present branch-and-bound-based algorithms for finding all solutions to the problem. The performance of the algorithms with respect to execution time is studied based on numerical experiments. The results show that the formulation and the algorithms are fast enough for solving elevator trip origin-destination matrix estimation problems in a real elevator group control application.

Estimation of elevator passenger traffic based on the most likely elevator trip origin-destination matrices

We present a constraint programming formulation for the elevator trip origin-destination matrix estimation problem using a lexicographic bi-criteria optimization method where least squares minimization is applied to the measured counts and the minimum information or the maximum entropy approach to the whole matrix. An elevator trip consists of successive stops in one direction of travel with passengers inside the elevator. It can be defined as a directed network, where the nodes correspond to the stops on the trip and the arcs to the possible origins and destinations of the passengers. The goal is to estimate the most likely counts of passengers for the origin-destination pairs of every elevator trip occurring in a building that are consistent with the measured boarding and alighting counts and any prior information about the trip matrix. These counts are used to make passenger traffic forecasts which, in turn, are used in elevator dispatching to reduce uncertainties related to future passengers and therefore to improve passenger service level. Artificial test data was obtained by simulations of lunch hour traffic in a typical multi-story office building. This resulted in complex problem instances that enable robust performance and quality testing. The results show that the proposed approach outperforms previous alternatives in terms of quality, and can take an advantage of prior information. In addition, the proposed approach satisfies real time elevator group control requirements for estimating elevator trip origin-destination matrices. Practical application : The elevator trip origin-destination matrix estimation problem is important since it makes it possible to obtain complete information and statistics about the elevator passenger traffic. The statistics can be used to model future passengers which, when taken into account in the elevator group control, helps to improve passenger service level. Simulation experiments show that most of the problems occurring in reality can be solved within a reasonable time considering a real application, and the solving algorithms are relatively easy to implement using available constraint programming tools. Hence, this work is undoubtedly of interest to the building and elevator industry.